Marc

2012
Users Guide
Corporate
MSC.Software Corporation
2 MacArthur Place
Santa Ana, CA 92707
Telephone: (800) 345-2078
FAX: (714) 784-4056
Europe
MSC.Software GmbH
Am Moosfeld 13
81829 Munich
GERMANY
Telephone: (49) (89) 43 19 87 0
Fax: (49) (89) 43 61 71 6
Asia Pacific
MSC.Software Japan Ltd.
Shinjuku First West 8F
23-7 Nishi Shinjuku
1-Chome, Shinjuku-Ku
Tokyo 160-0023, JAPAN
Telephone: (81) (3)-6911-1200
Fax: (81) (3)-6911-1201
Worldwide Web
www.mscsoftware.com
User Documentation: Copyright
2012 MSC.Software Corporation. Printed in U.S.A. All Rights Reserved.
This document, and the software described in it, are furnished under license and may be used or copied only in accordance with the
terms of such license. Any reproduction or distribution of this document, in whole or in part, without the prior written authorization of
MSC.Software Corporation is strictly prohibited.
MSC.Software Corporation reserves the right to make changes in specifications and other information contained in this document
without prior notice. The concepts, methods, and examples presented in this document are for illustrative and educational purposes
only and are not intended to be exhaustive or to apply to any particular engineering problem or design. THIS DOCUMENT IS
PROVIDED ON AN AS-IS BASIS AND ALL EXPRESS AND IMPLIED CONDITIONS, REPRESENTATIONS AND WARRANTIES,
INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE
DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY INVALID.
MSC.Software logo, MSC, MSC., MD Nastran, Adams, Dytran, Marc, Mentat, and Patran are trademarks or registered trademarks of
MSC.Software Corporation or its subsidiaries in the United States and/or other countries.
NASTRAN is a registered trademark of NASA. Python is a trademark of the Python Software Foundation. LS-DYNA is a trademark of
Livermore Software Technology Corporation. All other trademarks are the property of their respective owners.
This software may contain certain third-party software that is protected by copyright and licensed from MSC.Software suppliers.
METIS is copyrighted by the regents of the University of Minnesota. HP MPI is developed by Hewlett-Packard Development Company,
L.P. MS MPI is developed by Microsoft Corporation. PCGLSS 6.0, copyright 1992-2005 Computational Applications and System
Integration Inc. MPICH Copyright 1993, University of Chicago and Mississippi State University. MPICH2 copyright 2002, University
of Chicago.
Use, duplication, or disclosure by the U.S. Government is subject to restrictions as set forth in FAR 12.212 (Commercial Computer
Software) and DFARS 227.7202 (Commercial Computer Software and Commercial Computer Software Documentation),
as applicable.
MA*V2012*Z*Z*Z*DC-USR
Cont ent s
Marc Users Guide

Contents
Preface
Organization of this Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Part 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Part 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Part 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Documentation Conventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Section 1: Introduction
1 Introduction
Introducing Mentat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Brief Look at the Finite Element Analysis Process. . . . . . . . . . . . . . . . . . . . 50
Some Mentat Hints and Shortcuts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Mechanics of Mentat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Marc/Mentat Window Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
How Mentat Communicates with You. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
How This Manual Communicates with You . . . . . . . . . . . . . . . . . . . . . . . . . 54
How You Communicate with Mentat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Menu Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
List Specification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Identifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Menu Customization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Comprehensive Sample Session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Boundary Conditions, Initial Conditions, and Links . . . . . . . . . . . . . . . . . . 163
Material and Geometric Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Loadcases and Jobs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Results Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Marc Users Guide

4
Getting Started. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Starting the Mentat Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Procedure Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Stopping the Mentat Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Recommended Starting Chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Following a Sample Session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Detailed Session Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Answers to Frequently Asked Questions . . . . . . . . . . . . . . . . . . . . . . . . 204
Consistent Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Evaluation of Stresses in Finite Elements . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Shear Strains used in Marc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Extrapolation/Averaging Tips in Mentat . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Stress Coordinate Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Composite Shells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Material Axis Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Gauss Point Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Selective Results to the Post File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Continuum and Generalized Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Result Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Appendix A: Shape Function Interpolation. . . . . . . . . . . . . . . . . . . . . . . 260
Implication: The Evaluation of Element Displacements . . . . . . . . . . . . . . . 263
Implication: Linear Versus Quadratic Elements? . . . . . . . . . . . . . . . . . . . . 263
Implication: Nodal Temperature Loading With Temperature Dependent
Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Implication: Element Thickness Interpolation . . . . . . . . . . . . . . . . . . . . . . . 266
Appendix B: Finite Element Equilibrium. . . . . . . . . . . . . . . . . . . . . . . . . 268
Implication: Smoothed or Unsmoothed Stress Contours? . . . . . . . . . . . . . 269
Implication: Limitations of the Averaging Scheme . . . . . . . . . . . . . . . . . . . 271
Appendix C: Coordinate Transformation . . . . . . . . . . . . . . . . . . . . . . . . 273
Appendix D: Principal Stresses (Plane Stress) . . . . . . . . . . . . . . . . . . . 273
Appendix E: Python Example (Max Stress Results) . . . . . . . . . . . . . . . 274
Appendix F: Python Example (Displacements at Nodes) . . . . . . . . . . . 278
5 CONTENTS
Section 2: Recent Features
2.1 New-style Table Input
Summary of Reinforced Cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Post Buckling Analysis of a Reinforced Shell with Nonuniform Load 285
List of User Subroutines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Summary of Can Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Can Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
2.2 Thermo-Mechanical Analysis of Cylinder Head Joint
with Quadratic Contact
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Simulation of a Cylinder Head Joint . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Modeling Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
Load Steps and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
Save Model and Run Job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
2.3 RBE3 (General Rigid Body Link)
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
Soft and Rigid Connections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
Marc Users Guide

6
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
Loadcases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Submit Job and Run the Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
2.4 Arc Welding Process Simulation
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
Welding Process Simulation of Cylinder-Plate Joint . . . . . . . . . . . . . . . 374
Procedure File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
Geometric Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Weld Path Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
Weld Filler Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
Contact Body Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Initial/Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
Loadcase Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
2.5 FEM Simulation of NC Machining and PRE STATE
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Example 1: Pocket Cutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Input data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Initial Geometry and Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Local Mesh Adaptivity Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
Visualization of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Verification of Material Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Example 2: Thin Frame Cutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Input Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Initial Stress and Local Adaptive Remeshing Definition . . . . . . . . . . . . . . . 413
Loadcases and Machining Job Definition . . . . . . . . . . . . . . . . . . . . . . . . . . 414
FEA Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
7 CONTENTS
Example 3: Imported Initial Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Import with Text Data File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Import with PRE STATE Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
2.6 Parallelized Local Adaptive Meshing
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Animation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
2.7 New Magnetostatic Elements
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Magnetostatic Field Around a Coil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Inserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 444
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
2.8 Coupled Electrostatic Structural Analysis of
a Capacitor
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Capacitor Loaded with Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Mesh Adaptivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Marc Users Guide

8
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
2.9 3-D Contact and Friction Analysis using
Quadratic Elements
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Sliding Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Model Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
Loadcases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
2.10 Pin to Seal Contact with Various Friction Models
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
Friction Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
2.11 Analysis of a Manhole with Structural Zooming
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
Global Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
Local Model and Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
9 CONTENTS
2.12 Radiation Analysis
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
2.13 Application of BC on Geometry with Remeshing
Geometry and Finite Element Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544
2.14 Glass Forming of a Bottle with Global Remeshing
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Analysis with Remeshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
2.15 Marc Adams MNF Interface
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
Generation of an MNF for HDD HSA Suspension Arm . . . . . . . . . . . . . 566
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
HSA Suspension Arm Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Marc Users Guide

10
Local Model and Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
2.16 Analysis of Stiffened Plate Using Beam and Shell
Offsets
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
Analysis of Beam Reinforced Shell Structure using Offsets . . . . . . . . 574
Procedure File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
Geometric Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Loadcase Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582
2.17 3-D Tetrahedral Remeshing with Boundary Conditions
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
Simulation Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
Pressure on a Rubber Cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
Metal Compression with Prescribed Displacements . . . . . . . . . . . . . . . . . . 586
Rubber Ring Seal with Pressure Testing after Compression . . . . . . . . . . . 587
Tube Hydro-forming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Rubber Seal Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Rubber Seal and Steel Interaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Glass Forming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Rubber Bars with Prescribed Displacement on Curves. . . . . . . . . . . . . . . . 594
Rubber Seal Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Model Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
11 CONTENTS
2.18 Induction Heating of a Tube
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Heating of a Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Initial Conditions and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . 611
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 615
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
2.19 Magnetostatics with Tables
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
Nonlinear Analysis of an Electromagnet Using Tables. . . . . . . . . . . . . 620
Reading the Model and Adding Material Properties . . . . . . . . . . . . . . . . . . 620
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 627
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
2.20 Delamination and Crack Propagation
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
2.21 Progressive Failure Analysis of Lap Joint
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638
Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
Marc Users Guide

12
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
Modeling Tips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
2.22 Sheet Metal Forming With Solid Shell Elements
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
Model Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
2.23 Plastic Limit Load Analysis of a Simple Frame Structure
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
Detailed Marc Input Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Detailed Mentat Session Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
Modeling Tips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
2.24 Directional Heat Flux on a Sphere from a Distance
Source
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
Model Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
2.25 Deep Drawing of A Sheet With Global Remeshing
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
13 CONTENTS
Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
2.26 Artery Under Pressure
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
Material Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
2.27 Modeling Riveted Joint with Bushing, CFAST, or
CWELD
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
2.28 Speed and Memory Improvements
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698
Fast Integrated Composite Shells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698
Combined Multi Frontal Sparse and Iterative Solver. . . . . . . . . . . . . . . 700
Storage of Element Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
Marc Users Guide

14
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
2.29 Implicit Viscoplastic Creep Analysis of Solder
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Flow Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Requested Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
Modeling Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
Element Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
Material Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
Loading and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711
Solution Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
Result and Plots. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
2.30 Crack Propagation Capability in Shells
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Requested Solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Modeling Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
Crack Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
Loading and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Solution Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
15 CONTENTS
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
2.31 Segment-to-Segment with Friction
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Contact Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Requested Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Modeling Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Element Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Material Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Friction Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Loading Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
Solution Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
Result and Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
2.32 Directionally Dependent Friction Summary
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Requested Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Modeling Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Element Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738
Material Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Loading and Boundary Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Marc Users Guide

16
Solution Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
Result and Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751
2.33 Improved Accuracy with Remeshing of Herrmann
Elements
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Simulation of Elastomeric Seal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Contact Body Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
Contact Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757
Mesh Adaptivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Solution Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761
Video . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Section 3: Mechanical Analysis
3.1 Solid Modeling and Automatic Meshing
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
17 CONTENTS
About HexMesh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Advantages of HexMesh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Advantages of Hexahedral Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Activating the HexMesh Feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
About the HexMesh Menu in Mentat. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
About the Input for HexMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
Key Steps in the Meshing Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786
Using HexMesh Parameters and Commands. . . . . . . . . . . . . . . . . . . . . 787
Specifying Element Size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
Specifying Edge Sensitivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
How the Value of Edge Sensitivity Affects the Edge Detection Process . . 788
Specifying Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
How the Value of Gap affects the Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . 790
Specifying the Number of Shakes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Using the Runs Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Using the Coarsening Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792
How the Level of Coarsening affects the Elements . . . . . . . . . . . . . . . . . . 792
Using the Allow Wedges Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
About the Coons Patches Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
Using the Detect Edges Command. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
Selecting Edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Deselecting Edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Checklist for the HexMesh Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Applying the HexMesh Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796
About the Meshing Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796
Rectifying an Unsuccessful Hexmeshing Operation. . . . . . . . . . . . . . . . . . 796
Using HexMesh Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
About the Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
Example Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
Running the Procedure File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Preparing the Model for Surface Meshing . . . . . . . . . . . . . . . . . . . . . . . . . 798
Applying the Delaunay Tri-Mesh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
Preparing the Input List for HexMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
Applying HexMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
Marc Users Guide

18
3.2 Manhole
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
Idealization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
Requirements for a Successful Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833
3.3 Contact Modeling of Pin Connection Joints using
Higher-Order Elements
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Pin Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
Contact Bodies and Contact Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
Loadcases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
3.4 Beam Contact Analysis of an Overhead Power Wire of
a Train
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
Pantograph of a Train Touching the Overhead Power Wire . . . . . . . . . 852
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856
Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
Geometry Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
19 CONTENTS
Loadcases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861
Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 865
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868
3.5 Gas Filled Cavities
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Simulation of an Airspring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Axisymmetric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884
3.6 Tube Flaring
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Requirements for a Successful Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909
3.7 Punch
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 912
Requirements for a Successful Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 913
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
Marc Users Guide

20
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932
3.8 Torque Controlled dies with Twist Transfer
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Belt and Pulley Assembly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
3.9 Break Forming
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946
Detailed Session Description of Break Forming . . . . . . . . . . . . . . . . . . 948
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 954
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 958
Modeling Tips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 960
Animation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962
3.10 Hertz Contact Problem
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 964
Run Jobs and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 970
FEA versus Theoretical Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972
21 CONTENTS
3.11 Anisotropic Sheet Drawing using Reduced Integration
Shell Elements
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
Simulation of Earing for Sheet Forming with Planar Anisotropy. . . . . 975
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
Load Steps and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 989
Advanced Topic: Drawbead Modeling using Nonlinear Spring . . . . . . 991
Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . 994
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
3.12 Chaboche Model
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Blade on a Fan of a Turbine Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1008
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1008
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . 1010
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1013
Marc Users Guide

22
3.13 Modeling of a Shape Memory Alloy Orthodontic
Archwire
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1016
Simulation of an Archwire with Shape Memory Alloy Models . . . . . . 1016
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1017
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
Load Steps and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1029
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
3.14 Implicit Creep Analysis of Solder Connection between
Microprocessor and PCB
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036
Microprocessor Soldered to a PCB . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1040
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1052
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
3.15 Continuum Composite Elements
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
Model Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
23 CONTENTS
Boundary Conditions and Loads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059
Composite Layer Property Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1060
Composite Layer Orientation Definition . . . . . . . . . . . . . . . . . . . . . . . . . . 1062
Define Job Parameters, Save Model, and Run Job . . . . . . . . . . . . . . . . . 1063
View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064
Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1065
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1065
3.16 Super Plastic Forming (SPF)
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
SPF Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1070
Preprocessing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1070
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1084
Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087
SPF with Adaptive Remeshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089
Discussion of Adaptive Meshing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1092
3.17 Gaskets
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094
Simulation of a Cylinder Head Joint . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Tyings and Servo Links. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1101
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1104
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1104
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1111
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1112
Load Steps and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1113
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . 1117
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122
Marc Users Guide

24
3.18 Cantilever Beam
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124
Detailed Session Description of Cantilever Beam . . . . . . . . . . . . . . . . 1126
Add Plasticity to Cantilever Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135
3.19 Creep of a Tube
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1138
Detailed Session Description of Oval Tube . . . . . . . . . . . . . . . . . . . . . 1140
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
What can improve the results? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154
3.20 Tensile Specimen
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1156
Detailed Description Session . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1157
Tensile Specimen Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1157
Overlay Technique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1160
Advancing Front Technique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1161
Mapped Meshing Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1161
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164
Tensile Specimen Uniform Gage Section . . . . . . . . . . . . . . . . . . . . . . . . . 1170
Tensile Specimen Composite Material . . . . . . . . . . . . . . . . . . . . . . . . . . . 1173
Modeling Tips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1179
25 CONTENTS
3.21 Rubber Elements and Material Models
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1182
Lower-order Triangular Rubber Elements . . . . . . . . . . . . . . . . . . . . . . 1183
Using Quadrilateral Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1183
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1193
Using Triangular Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1194
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1195
Tube with Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1198
Cavity Pressure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1200
Buckling of an Elastomeric Arch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1204
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1204
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1209
Comparison of Curve Fitting of Different Rubber Models . . . . . . . . . 1212
Mooney . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215
Arruda-Boyce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1217
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219
3.22 Modeling of General Rigid Body Links using RBE2/
RBE3
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1222
Cylindrical Shell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1222
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1222
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1224
Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1226
Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1226
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1226
Geometric Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1227
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1228
Save Job, and Run the Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1229
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1230
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1231
Marc Users Guide

26
3.23 Cyclic Symmetry
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1234
Pure Torsion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1235
Mechanical Analysis of Friction Clutch. . . . . . . . . . . . . . . . . . . . . . . . . 1238
Coupled Analysis of Friction Clutch . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
3.24 Axisymmetric to 3-D Analysis
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
Simulation of a Rubber Bushing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
Description of Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
Axisymmetric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249
3-D Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256
Automobile Tire Modeling with Rebar Elements . . . . . . . . . . . . . . . . . 1261
Description of Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
Axisymmetric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262
3-D Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1263
Analysis of a Rubber Cylinder using Remeshing. . . . . . . . . . . . . . . . . 1268
Description of Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1268
Axisymmetric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1268
3-D Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1270
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1276
3.25 Interference Fit
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1278
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1283
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1285
27 CONTENTS
3.26 3-D Remeshing with Tetrahedral Elements
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1288
Why Remeshing with Tetrahedral Elements? . . . . . . . . . . . . . . . . . . . 1288
Tetrahedral Element Type 157. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1288
Tetrahedral Remeshing Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1290
Tetrahedral Remeshing Controls and Meshing Parameters. . . . . . . . 1291
Tetrahedral Remeshing Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1293
Elastomeric Seal Simulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1300
Model Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1302
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1303
Contact Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1304
Mesh Adaptivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1307
Loadcases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1308
Jobs and Run Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1309
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1311
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1313
3.27 Rubber Remeshing and Radial Expansion of Rigid
Surfaces
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1316
Model Highlights. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1316
Results Highlights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1320
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1321
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1321
3.28 Automatic Remeshing/ Rezoning
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1324
Elastomeric Seal Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1325
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326
Marc Users Guide

28
Tape Peeling Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1347
3.29 Multibody Contact and Remeshing
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
Squeezing of a Rubber Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
Background information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1370
3.30 Container
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1372
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1372
Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1372
Idealization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1373
Requirements for a Successful Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1375
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401
3.31 Analyses of a Tire
Steady State Rolling Analysis1404
Simulation of a Tire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1404
Run Job and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1417
More Results on Contact Friction Stresses . . . . . . . . . . . . . . . . . . . . . . . . 1420
Tire Bead Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1423
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1423
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1423
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1424
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1425
Detailed Session Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1425
29 CONTENTS
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1448
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1448
3.32 Transmission Tower
Chapter Overview1450
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1450
Tower Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1450
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1451
Requirements for a Successful Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . 1451
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1451
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1452
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1452
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1500
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1500
3.33 Bracket
Chapter Overview1502
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1502
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1502
Idealization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1502
Requirements for a Successful Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . 1502
Full Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1503
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1504
Detailed Session Description of the Linear Static Case . . . . . . . . . . . 1504
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1521
Dynamic Modal Shape Analysis1522
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1522
Detailed Session Description of the Modal Shape Analysis. . . . . . . . 1522
Dynamic Transient Analysis1526
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1526
Detailed Session Description of Dynamic Transient Analysis. . . . . . 1526
Marc Users Guide

30
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1533
Pressure Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1533
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1534
3.34 Single Step Houbolt Dynamic Operator
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
Impact of a Ball on a Plate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
Eigenvalue Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
Transient Analysis1546
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1553
3.35 Dynamic Analyses of a Cantilever Beam
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556
Modal Analysis1557
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1557
Modal Analysis1558
Harmonic Analysis1560
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560
Harmonic Analysis and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560
Transient Analysis1563
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1563
Analysis and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1563
Damping Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565
Over Hanging Beam Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1567
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1571
3.36 Plastic Spur Gear Pair Failure
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1574
Gear Geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1575
31 CONTENTS
Material Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1576
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1576
Failure Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1577
Model Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1579
Experimental Test Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1580
Results & Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1582
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1583
Animation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1584
3.37 Girkmann Verification Problem
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1586
Detailed Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1587
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1589
Modeling Tips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1591
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1596
Section 4: Heat Transfer Analysis
4.1 Thermal Contact Analysis of a Pipe
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
Pipe in a House. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1601
Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1602
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1603
Marc Users Guide

32
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1604
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1605
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1607
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1612
4.2 Dynamics with Friction Heating
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1614
Friction Heat Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1616
Run Jobs and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626
4.3 Radiation with Viewfactors
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1628
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1630
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1633
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1635
4.4 Cooling Fin Analyses
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1638
Steady State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1639
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1639
Overview of Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1640
Detailed Session Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1640
Transient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649
Detailed Session Description with Fin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649
Detailed Session Description without Fin . . . . . . . . . . . . . . . . . . . . . . . . . 1652
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1655
33 CONTENTS
Section 5: Coupled Analysis
5.1 Coupled Structural Acoustic Analysis
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1660
Two Spherical Rooms Separated by a Membrane. . . . . . . . . . . . . . . . 1660
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1660
Harmonic Analysis with Stress-free Membrane. . . . . . . . . . . . . . . . . . 1661
Harmonic Analysis with Pre-stressed Membrane . . . . . . . . . . . . . . . . 1670
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1673
5.2 Coupled Electrical-Thermal-Mechanical Analysis of a
Micro Actuator
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1676
Simulation of a Microelectrothermal Actuator . . . . . . . . . . . . . . . . . . . 1676
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1676
Actuator Model1677
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1682
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683
5.3 Coupled Transient Cooling Fin
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1687
Run Jobs and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1688
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1690
5.4 Temperature Dependent Orthotropic Thermal Strains
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
Run Jobs and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
Marc Users Guide

34
Thermal Expansion Data Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1698
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1701
Section 6: Miscellaneous Analysis
6.1 Magnetostatics: Analysis of a Transformer
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1706
3-D Analysis of a Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1706
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1715
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1718
6.2 Fracture Mechanics Analysis with the J-integral
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1720
Specimen with an Elliptic Crack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1720
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1720
Modeling Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1721
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1722
Crack Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1730
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1733
Contact Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1733
Run Job and View Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1739
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1741
6.3 FEM Simulation of NC Machining Process
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1744
Input Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1744
Model Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1745
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1745
35 CONTENTS
Residual Stresses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1746
Procedure Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1747
Machining Process Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1748
Loadcase1 (cut the top part of the workpiece) . . . . . . . . . . . . . . . . . . . . . 1749
Loadcase2 (release the bottom boundary condition and apply to the top face)
1750
Loadcase3 (cut the pocket from the lower face part) . . . . . . . . . . . . . . . . 1751
Loadcase4 (final release springback) . . . . . . . . . . . . . . . . . . . . . . . . . . 1752
Job Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1753
Visualization of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1754
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1758
6.4 Piezoelectric Analysis of an Ultrasonic Motor
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1760
Eigenvalue Analysis of the Stator of an Ultrasonic Motor . . . . . . . . . 1760
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1762
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1763
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1771
Save Model, Run Job, and View results . . . . . . . . . . . . . . . . . . . . . . . . . . 1772
Harmonic Analysis of the Stator of an Ultrasonic Motor . . . . . . . . . . 1773
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1773
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1774
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . 1775
Transient Analysis of the Stator of an Ultrasonic Motor. . . . . . . . . . . 1779
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1780
Save Model, Run Job, and View Results . . . . . . . . . . . . . . . . . . . . . . . . . 1781
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1783
6.5 Analysis Performance Improvements
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1786
Speed and Memory Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1786
Case 1: Rigid-Deformable Body Contact . . . . . . . . . . . . . . . . . . . . . . . . . 1786
Case 2: Deformable-Deformable Contact. . . . . . . . . . . . . . . . . . . . . . . . . 1787
Marc Users Guide

36
Case 3: Model with Solid and Shell Elements . . . . . . . . . . . . . . . . . . . . . . 1788
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1790
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1790
6.6 Robustness of Automatic Load Stepping Schemes
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1792
Usage of the Auto Step Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1792
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1802
6.7 Marc Running in Network Parallel Mode
Run CONTACT WITH DDM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1804
Run CONTACT WITH DDM on a Network . . . . . . . . . . . . . . . . . . . . . . . 1804
UNIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1804
Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1805
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1809
6.8 Convergence Automation and Energy Calculations
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1812
Convergence Automation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1812
AUTO SWITCH Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1813
Energy Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1819
Usage of the Energy Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1820
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1829
6.9 Capacitors
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1832
Capacitance Computation in Symmetric Multiconductor Systems . . 1833
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1834
Element and Node Set Selection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1834
37 CONTENTS
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1838
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1839
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1841
Loadcase and Job Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1841
Save Model, Run Job and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1845
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1849
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1850
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1850
6.10 Inductance Between Two Long Conductors
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1852
Inductance Computation in Two Infinitely Long Rectangular Conductors
1854
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1855
Element Selection as Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1856
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1858
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1859
Modeling Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1860
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1862
Loadcase and Job Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1864
Save Model, Run Job, and View results . . . . . . . . . . . . . . . . . . . . . . . . . . 1868
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1872
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1872
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1873
6.11 Lamination Loss in Magnetostatic-Thermal coupling
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1876
Lamination Loss Computation and ohmic Winding Loss in a C Core
Cylindrical Inductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1877
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1881
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1884
Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
Modeling Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1888
Initial and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
Marc Users Guide

38
Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1891
Loadcase and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1892
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1895
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899
6.12 Magnetic Levitation of a Ferromagnetic Sphere
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902
Magnetic Levitation of a Ferromagnetic Sphere. . . . . . . . . . . . . . . . . . 1903
Mesh Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1903
Material Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1904
Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1905
Modeling Tools and Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1906
Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1908
Loadcases and Job Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1909
Save Model, Run Job, and View Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1912
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1912
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1913
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1913
New Mentat Menus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1914
Section 7: Mentat Features and Enhancements
7.1 Past Enhancements in Marc and Mentat
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1918
Preprocessing Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1918
New Attach Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1918
Boundary Conditions on Geometric Entities . . . . . . . . . . . . . . . . . . . . . . . 1924
Combined Mesh Generation Commands . . . . . . . . . . . . . . . . . . . . . . . . . 1925
Change Class . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1927
Improved Links Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1927
Patran Tetrahedral Mesher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1929
New Select Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1931
New Domain Decomposition Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1934
Multi-Dimensional Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1935
39 CONTENTS
User-defined Text Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1943
64-bit Version of Mentat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1943
Python . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1943
Postprocessing Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1944
MPEG and AVI Animations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1944
Creating a Movie. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946
Postprocessing in 3-D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1948
Animation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1949
7.2 Importing a Model
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1964
7.3 HyperMesh

Results Interface
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1966
About Postprocessing of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1966
Interfacing Analysis and Postprocessing . . . . . . . . . . . . . . . . . . . . . . . . . 1966
Data Written into the HyperMesh Results File . . . . . . . . . . . . . . . . . . . . . 1966
About Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1967
Mentat Preprocessing for HyperMesh . . . . . . . . . . . . . . . . . . . . . . . . . 1967
Important Data Preparation Considerations Regarding Eigenmodes. . . . 1974
Relation to other Types of Results Files . . . . . . . . . . . . . . . . . . . . . . . . . . 1974
Postprocessing using HyperMesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1976
Marc Users Guide

40
7.4 Translators
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
New Mentat Writers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
dxfout: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
stlout: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
vdaout: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
vrmlout: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1988
New Mentat Readers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1989
c-mold: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1989
stl: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1991
Improved Readers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
acis: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
dxf: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
ideas: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
nastran: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
patran: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
7.5 Sweep Nodes on Outlines
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1994
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1994
Overview Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1994
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1994
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1998
7.6 Transition Parameter for Meshing
Chapter Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000
Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000
Overview Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2000
Input Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2004
41 CONTENTS
7.7 Mentat Features 2001 and 2003
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2006
2001 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2006
Optimized Element Graphics Generation . . . . . . . . . . . . . . . . . . . . . . . . . 2006
Optimized Entity Recoloring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2006
Post Reader Optimization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2006
Flowline Plotting2007
Particle Tracking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2008
PostScript Thin Lines Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2009
Curve Direction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2010
New Viewing Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011
2003 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013
User Defined Variable Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013
Status File Information2014
DCOM Server Support for Windows NT. . . . . . . . . . . . . . . . . . . . . . . . . . 2014
User-defined NUMERIC Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2015
Previous and Last Increment Buttons2016
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2016
7.8 Generalized XY Plotter
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2018
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2018
Overview Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2018
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2018
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2023
7.9 Beam Diagrams Example
Chapter Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2026
Background Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2026
Overview of Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2026
Detailed Session Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2027
Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2034
Marc Users Guide

42
Preface
Preface

Organization of this Manual 44

Documentation Conventions 45
44 Marc Users Guide: Part I
Preface
Organization of this Manual
This manual introduces the Iirst-time user to the Mentat program. The User`s Guide covers the basics oI the program
and helps the novice user in becoming comIortable with Mentat through a number oI examples.
The manual is divided into several parts
a basic introduction
sample problems
an archive oI past new Ieatures highlighted during each release oI the product.
Part 1
Section I Introduction
The Iirst section introduces the basics oI the program and provides inIormation that helps user interaction with Mentat.
It also consists oI a sample session that provides the user with hands-on experience with the Iunctionality oI the
Mentat program:
Section 2 Recent Features
This consists oI examples oI recent Ieatures Ior Marc and Mentat.
Part 2
Section 3 Sample Session
This section demonstrates how to set up the basic requirements Ior a linear elastic stress analysis.
Introduction provides inIormation on the basic steps oI the Iinite element analysis
cycle and on how Mentat is used as a tool to accomplish these steps.
Mechanics of Mentat describes the user interIace aspects oI the program
Background Information expands on the common Ieatures oI Mentat and describes some oI the
underlying philosophies oI the program.
Getting Started introduces you to Mentat with a simple example oI how to create a Iinite
element model.
A Simple Example introduces you to use Mentat and Marc to perIorm a complete linear
elastic analysis oI a rectangular strip with a hole subjected to tensile
loading. Both the preprocessing, analysis, and postprocessing steps will
be demonstrated.
45
Preface
Part 3
Section 4 Example Features 2000 - 2003
Section 4 consists oI prior releases oI the new Ieature examples Ior Mentat versions 2000 through 2010.
Documentation Conventions
Listed below are some Iont and syntax conventions that are used to make the inIormation in this manual easier to
understand:
Cross-reIerences (links) are highlighted in Blue.
Names oI buttons that appear on the Mentat screen are UPPER CASE in the Arial Iont.
Literal user input and program prompts are in courier Iont.
Names oI processors are indicated in BOLD UPPER CASE.
A carriage return keystroke is indicated by <CR>.
The leIt mouse button is indicated by <ML>.
The middle mouse button is indicated by <MM>.
The right mouse button is indicated by <MR>.
The mouse cursor is indicated by .
A Iilename implies a concatenation oI pathname and Iilename. The pathname may be omitted iI the Iilename
is in your current directory.
~
46 Marc Users Guide: Part I
Preface
Section 1: Introduction
Section 1: Introduction
MSC.Marc Users Guide 48
Chapter 1: Introduction
1
Introduction

Introducing Mentat 50

Some Mentat Hints and Shortcuts 51

Mechanics of Mentat 52

Comprehensive Sample Session 109

Background Information 153

Getting Started 170

A Simple Example 186

Input Files 203

Answers to Frequently Asked Questions 204

Appendix A: Shape Function Interpolation 260

Appendix B: Finite Element Equilibrium 268

Appendix C: Coordinate Transformation 273

Appendix D: Principal Stresses (Plane Stress) 273

Appendix E: Python Example (Max Stress Results) 274

Appendix F: Python Example (Displacements at Nodes) 278

50 Marc Users Guide: Part I
Introduction
Introducing Mentat
Welcome to Mentat a graphical user interIace program that allows you to execute a Iinite element analysis process
Irom start to Iinish.
Brief Look at the Finite Element Analysis Process
In order to enhance your understanding oI Mentat, we will review the Iinite element analysis cycle beIore introducing
the mechanics oI the program. The Iinite element analysis cycle involves Iive distinct steps as is shown in Figure 1.1-1.
This process may be traversed more than once Ior a particular design; that is, iI the results do no meet the design
criteria, you can return to either the conceptualization (Step 1) or modeling (Step 2) phase to redeIine or modiIy
the process.
Figure 1.1-1 The Analysis Cycle
The Iive distinct steps oI the Iinite element analysis cycle provide the Ioundation Ior this guide. In order to improve
your productivity, this guide has been designed to Iocus your attention on two steps oI the Iinite element analysis cycle:
Step 2: the model generation phase, and Step 4: results interpretation phase.
CONCEPTUALIZATION
MODELING
ANALYSIS
INTERPRETATION
ACCEPTANCE
Step 1
Step 2
Step 3
Step 4
Step 5
Preprocessing
Postprocessing
Determine what
to include in analysis
Meet
criteria?
51
CHAPTER 1
Introduction
Typical engineering problems are used in this guide as a vehicle to demonstrate the key Ieatures oI Mentat. Steps 2
and 4 oI the analysis cycle, modeling and results interpretation, were the pacing parameters Ior the selection oI the
example problems. The engineering problems were Iurther selected to meet two criteria:
1. to introduce you to the intricacies oI generating a model in a variety oI ways,
2. to demonstrate a diversity oI analysis types so that you become Iamiliar with as many oI the diIIerent
capabilities oI Mentat as possible. The analysis will be perIormed with the general purpose Iinite element
program Marc. Within the graphical user interIace, the complete input requirements can be speciIied.
The dimensionality oI each object prescribes the technique used to generate a model and display the results.
Accordingly, the Iinite element models have been grouped by dimensionality and, in many cases, the complexity oI
the model corresponds to dimensionality. From this you could conclude that once you know how to solve
three-dimensional problems, you also know how to solve one-dimensional problems. However, the unique Ieatures oI
one-dimensional objects require that we cover examples oI that particular topology. For example, it is diIIicult to
contour a quantity on line elements. With this in mind, the geometry and analysis types, such as heat transIer, statics,
or dynamics, have been selected to minimize duplicity in this guide.
The material covered in this tutorial is very basic and should be easy to access and understanding Ior the Iirst time user.
Once you have worked through the sample sessions in Comprehensive Sample Session, you should Ieel comIortable
enough to do a complete analysis simply by extrapolating Irom what you have learned Irom the example problems.
Some Mentat Hints and Shortcuts
1. Enter Mentat to begin, Quit to stop
2. Mouse in Graphics: LeIt to pick, Right to accept pick
3. Mouse in Menu: LeIt to pick another menu or Iunction, Middle (or F1) Ior help, Right to return to previous
menu. <CR> means keyboard return.
4. Save your work Irequently. Go to FILES and select SAVE AS and speciIy a Iile name.This will save the current
Mentat database to disk. Use SAVE Irom then on.
5. Dialog region at the lower leIt oI screen displays current activity and prompts Ior input. Check this region
Irequently to see iI input is required.
6. Dynamic Viewing can be used to position the model in the graphics area. When activated, the mouse buttons,
LeIt translates the model, Right zooms in/out, Middle rotates in 3-D. Use RESET VIEW and FILL to return to
original view. Be sure to turn oII DYNAMIC VIEW beIore picking in the graphics area.
7. CTRL P/N recall previous/next commands entered.
8. All oI the workshop problems have Mentat procedure and data Iiles. They are located in a marc.ug directory
under Mentat`s main directory. The directory/Iile structure looks like:
~mentat/examples/marc.ug/s3/c3.9/ Ior Section 3, Chapter 3.9.
Furthermore, you can click on the Iilename listed in the input Iiles table to download the Iiles via the web.
52 Marc Users Guide: Part I
Introduction
Mechanics of Mentat
BeIore you get started with Mentat, you need to know how to communicate with the program. The goal oI this section
is to give you an overview oI how Mentat works and to provide you with the basic inIormation to interact comIortably
with the program. Upon completion oI this, you should have a clearer understanding oI the Iollowing areas:
The basic window layout
How Mentat communicates with you
How you communicate with Mentat
The menu system
Where, say in directory
/marc_ug/s3/c3.9, there is a
procedure Iile called s4.proc. It
automatically runs Mentat to build,
run Marc, and process the results.
These directories can be copied to
your local disk area to work on during
the workshop.
Check out the HELP & RUN A DEMO
menus.
53
CHAPTER 1
Introduction
Marc/Mentat Window Layout
The starting point Ior all communication with Mentat is the window shown below that appears at the start oI
the program.
Figure 1.1-2 Basic Mentat Window
The Mentat Window is divided into three major areas:
Graphics area is used to display the current state oI the database. When you start Mentat, the graphics area is
blank to indicate that the database is empty.
Menu area is reserved to show the selectable menu-items and is divided into two submenus,
Static always present and contains items that are applicable and selectable at all times
Dynamic contents oI the dynamic menu area change as the menu items are selected.
54 Marc Users Guide: Part I
Introduction
How Mentat Communicates with You
Mentat communicates with you via prompts and messages and other visual queues. Mentat's prompts urge you to take
action through the input oI data or commands. These prompts have three types oI trailing punctuation marks to indicate
the required type oI input:
II you misspell a keyword or enter an incorrect response, Mentat warns you through a message posted in the dialogue
area. Mentat does not require that you complete every action you initiate. For example, iI you are prompted Ior a
Iilename, and you change your mind, entering a <CR> instead oI typing in the Iilename tells Mentat to abort the action.
II the program is waiting Ior a list oI items to operate on, and instead you enter a command that also requires a list oI
items or any additional data, Mentat ignores your original request and process the command. II the command you enter
does not request additional data, you are returned to the original data request Irom beIore the interrupt.
The program assumes at all times that you want to repeat the previous operation on a new set oI items and prompts
you Ior a new list to operate on. This process repeats itselI until you indicate otherwise by entering a new command
or a <CR>.
How This Manual Communicates with You
Mentat menus suggest the next step or steps to complete Ior the simulation; commonly the menus are abstracted into
a button name sequence leaving room Ior more meaningIul Iigures. For example, to save the current model in Mentat,
the button sequence:
is used to represent the actual menus as shown in Figure 1.1-2 where clicking on FILES in the static menu (1) brings
up the FILE I/O menu. Clicking on the button SAVE AS (2), brings up the SAVE AS FILE Iile browser, and it is implied
that you enter the Iile name, spf, (3) and accept the entry by clicking the OK button (4). Placing the mouse in the menu
Dialogue area is a scrollable area oI about Iive visible lines where all program prompts, warnings, and responses
appear, and where the user can input data or commands.
Status Area within the dialogue area and reserved to communicate the state oI the program to
the user. Either working or ready appears in the status area to reIlect the current
state oI the program. For intensive operations, an additional progress widget will
appear.
: enter numeric data, e.g., .283/384;
~ enter a character string, typically a command, Iile name or set name;
? enter a YES or NO answer.
MAIN MENU (panel title)
FILES (pulldown)
SAVE AS spf (button that launches a file browser)
<CR> or Cancel (submenu button)
55
CHAPTER 1
Introduction
area and clicking the right mouse button will RETURN (5) you to your previous menu (MAIN MENU). This button
sequence has colored the buttons as they appear in the menus and a Iew introductory chapters use the colored button
names whereas others use black Ionts assuming you have understood how the button sequence represents the menus.
In summary, green buttons take you to another menu, where as blue (cyan) buttons perIorm an action.
56 Marc Users Guide: Part I
Introduction
Figure 1.1-3 Mentat Menu
How You Communicate with Mentat
All interaction with Mentat is done through the mouse, keyboard, or a combination oI both. This section Iirst discusses
the usage oI the mouse, Iollowed by a discussion on how to use the keyboard as a means to enter commands and data.
The Mouse
The mouse is used to select items Irom the menu area or to point at items in the graphics area. It is important to make
a distinction between using the mouse in the menu area versus the graphics area because the three mouse buttons have
very diIIerent Iunctions in each area. Figure 1.1-4 shows a graphical representation oI the mouse, mouse buttons, and
corresponding cursor.
57
CHAPTER 1
Introduction
Figure 1.1-4 The Mouse, Mouse Buttons, and Corresponding Cursor
The leIt button is represented by <ML>, the middle button by <MM>, and the right button by <MR>. For a two button
mouse <MM> = <ML> + <MR> depressed at the same time. Click reIers to a quick single depress-release action.
Using the Mouse to Select a Menu Item
To select a menu item with the mouse, move the over the item that you want to select and click the <ML>. To
return to the previous menu, move the over the menu area, and click the <MR>. Alternatively, you can click on
the RETURN button in the menu area using <ML>. Clicking on the MAIN button takes you to the main menu.
On-line Help
Using the Classic version oI Mentat, each menu item has a help panel with a short description and explanation oI the
Iunction oI that menu item. To activate the help Ieature, position the over the menu item on which you require
help, Iollowed by a click oI the <MM>. The help panel disappears the moment you select another menu item. You may
also use the F1 Iunction key to activate the HELP Ieature.
Figure 1.1-5 Using the Mouse in the Menu Area
Cursor, <^>
<MM>
<MR>
<ML>
< >
< >
< >
<ML> for
Use this button,
command selection
<MR>
for RETURN
Use this button,
<MM>
for on-line help
Use this button,
58 Marc Users Guide: Part I
Introduction
Using the Mouse to Point
The mouse is used in two ways to operate in the graphics area: to point to or pick existing items, or to point to or pick
the location oI yet to be created items.
1. To pick, the mouse is used by moving the over the item to be identiIied Iollowed by a click oI the <ML>.
HenceIorth called by clicking on an item. You can undo that action by clicking the <MM> anywhere in the
graphics area. At times, you will need to identiIy more than a single item. A list oI items must be terminated
by a click oI the <MR> with the positioned anywhere in the graphics area. Alternatively, you can click on
the END LIST button in the menu area using <ML>.
2. To locate a position in Mentat, it is possible to deIine a grid that is positioned in space and where the grid
consists oI points that can be pointed to. II you click in the vicinity oI a grid point, the coordinates oI the item
that you created are snapped to that grid point. In addition, you can also pick an existing node, point, or
surIace-grid-point to speciIy a location.
Figure 1.1-6 Using the Mouse in the Graphics Area
Keyboard Input
Not all data can be entered through the mouse; numerical and literal data must be entered via the keyboard. The
program mode prescribes the speciIic requirements Ior proper entry oI each type oI data. The program can be in data
mode or in command/literal data mode and is described under the Iollowing two headings.
Numerical Data
You must use the keyboard Ior numerical data entry. The program interprets the data entry according to the context in
which it is used. II the program expects a real number and you enter an integer, Mentat automatically converts the
number to its Iloating point value. Conversely, iI a Iloating point Iormat number is entered where an integer is
expected, the program converts the real number to an integer.
ScientiIic notation Ior real numbers is allowed in the Iollowing Iormats:
.12345e01
.12345e01
-0.12345e-01
< >
< >
<ML>
Use this button,
to pick
<MR>
for end of item list
Use this button,
<MM>
to undo last pick
Use this button,
59
CHAPTER 1
Introduction
The interpreter does not allow imbedded blanks in the Iormat. Whenever the program encounters an illegal Iormat, the
message bad float! appears in the dialogue area. The prompt Ior numerical data is a colon (:).
Literal Data
Literal data is used Ior Iile, set, and macro names. A literal data string may not be abbreviated. Commands as
introduced in the beginning are considered string data (as opposed to literal string data) and can be abbreviated as long
as the character string is unique within the Mentat command library. For example, *add_elements cannot be
abbreviated to *add because oI the other commands that start with the same characters such as *add_nodes and
*add_curves. The program checks the input Ior validity against the internal library oI valid responses. For example,
iI you enter an ambiguous or misspelled command, Mentat responds by listing all the valid entries that start with the
same Iirst letter oI the command. The prompt Ior literal data is a greater-than symbol (~).
II the program is in data mode which is identiIied by the : prompt, you must enter a command preceded by an asterisk
(*) to instruct the program that you are entering a command.
For example: Enter node (1): *add_nodes
II you enter a command without the asterisk when the program is in data mode, Mentat responds with an error message
in the dialogue area.
The asterisk can be omitted when the program is in command or literal data mode which is indicated by the
greater-than symbol (~).
For example: Command > add_nodes
Editing the Input Line
The experienced user can enter a sequence oI commands or requests in a single 160-column input line. Note that
anything typed beyond the input line limit is lost! Use <CR> to avoid this. You must use a blank space to separate
entries when you are entering multiple responses on a single input line. All entries in the buIIer are processed
sequentially.
Mentat maintains a history oI lines that are entered and oIIers limited recall and editing capabilities Ior the command
line. The arrow keys . and v on the keyboard can be used to scroll up and down in the dialogue area to make these
lines visible. Use CTRL-p (that is, hold down the CTRL key and press the p key) to recall a previously entered input
line. Repeat the CTRL-p sequence to recall as many lines as you need. Use CTRL-n to move to the next line in the
history oI command lines. ( p and n stand Ior previous and next, respectively, in these control sequences.)
Edit Iunctions Ior the current line are: backspace Ior character delete and CTRL-u Ior line delete. The leIt and right
arrow keys are used to position the cursor at the desired location to overwrite or insert characters. The TAB key is used
as a toggle to switch Irom insert to overwrite mode and vice versa. For example, iI you type *view_viewpont 0.0p
0.0 1.0, the program responds with the message unknown command in the dialogue area. To correct the entry,
recall the line using CTRL-p, use the leIt arrow key to move the cursor to the letter n oI view_viewpont, press the
TAB key, type i, and press <CR> to enter the line. The command will now be *view_viewpoint.
60 Marc Users Guide: Part I
Introduction
Menu Structure
This section Iocuses on the menu system as a means to communicate with Mentat. The Iirst subsection discusses the
structure oI menus that constitute the program. The second subsection analyzes the components oI each menu.
The Mentat 2012 release unites the capabilities oI the existing capabilities with a modern user interIace. Most oI the
capabilities are in assessable as beIore, but there are a Iew signiIicant changes that are reviewed here.
Upon starting Mentat, one will see the Iollowing menu.
61
CHAPTER 1
Introduction
The display is divided into several areas that will be reviewed here.
Mentat is driven by icons such as: . II the main menu is not hidden by any other menu, then hovering the mouse
over the icon will result in a short description oI the icon, such as:
62 Marc Users Guide: Part I
Introduction
Menu Bar
The Iirst layer is the Menu bar which consists oI seven Icons:
- Program Icon activates Pull-down
Icon activates Pull-down to control the setting oI the home directories, opening and closing Iiles,
and saving the data base model. The Import and Export pull-downs are critical Ior interoperability with other
systems.
63
CHAPTER 1
Introduction
Navigator Menu - This menu reports to the user the number oI entities in the model and the amount oI
memory that is being used. By selecting an entity, one automatically jumps to the appropriate menu where
that entity is deIined. Note iI this menu is leIt open, the values dynamically change as one is building the
model.
64 Marc Users Guide: Part I
Introduction
In particular, the Import Button activates a pull-down that allows the import oI CAD models and other Iinite
element models into the system. AIter speciIying the type, the Iile browser will be activated.
The General CAD with Repair is the new button that activates the new CAD import.
65
CHAPTER 1
Introduction
- The Select Icon activated menus that allow the user to gather a subset oI the model to be
either put into a set, control the model being viewed or used in a subsequent command. The Menus are similar
to the previous release, but slightly reorganized.
Note that the Selection Control and Set Control menus are Iloating menus and it is oIten advantageous to have both oI
them open at the same time. In Iact, it is oIten useIul to have the Selection Control menu open throughout the session.
Selects a Group of Entities Creates and Manipulates Sets
66 Marc Users Guide: Part I
Introduction
- The View Icon activates multiple pull-downs and pop-up menus to control what is displayed in
the graphics area. Initially, the Iollowing pop-up is seen.
Lead View Pull-down - activates a series oI menus
67
CHAPTER 1
Introduction
Wireframe vs. Solid Model Entity Type Identify
Selects which type of entities Selects which entity type will be
used for displaying grous will be displayed
68 Marc Users Guide: Part I
Introduction
The Plot Setting and Visibility menus - Iloating menus and it is oIten useIul to have the Plot Setting menu
open throughout the session.

Select which type of entities will be displaced.
The settings allow control of the plot.
Select which type of entities are visible.
69
CHAPTER 1
Introduction
IdentiIy Menu
Selects which entity type will be used to control the displayed groups.
70 Marc Users Guide: Part I
Introduction
Viewing Menu - Controls the viewing orientation and rate at which the display will be updated.
71
CHAPTER 1
Introduction
Avtivate Light Source Shading
72 Marc Users Guide: Part I
Introduction
Light Setting
Controls the image of the model when
light source shading is used.
Rendering
Controls the image of the model when
photo-realistic rendering is used.
73
CHAPTER 1
Introduction
Activate Lighting Control
Define Colors used in Display
74 Marc Users Guide: Part I
Introduction
Define or Reset Graphic Font
Graphic Font Selection
75
CHAPTER 1
Introduction
Image using Default Graphic Font
Image using Size 14 Bold Graphic Font
76 Marc Users Guide: Part I
Introduction
- The Tools Icon is used to allow the user to select utilities that may be used to assist in creating
the model. The Iollowing pull-down is activated.
77
CHAPTER 1
Introduction
Controls the creation and playback Controls Python
of procedure files
Allows the user to define Parameters that will be used later or repetitively in the session.
If the expression contains other parameters, one has the choice of evaluating the value
of the parameter now or when it is used.
78 Marc Users Guide: Part I
Introduction
Allows the user to rename a command within Mentat
79
CHAPTER 1
Introduction
Allows user to combind multiple X-Y type plots Controls the creation of movies
80 Marc Users Guide: Part I
Introduction
Allows user to add
annotations to the image
Controls the use of a postscript display
Allows the user to obtain
geometric parameters or
subregions of the model
81
CHAPTER 1
Introduction
VRML - Controls the creation oI VRML Iiles.
82 Marc Users Guide: Part I
Introduction
By turning oII UNDO, one can improve perIormance but user error cannot be reversed.
Controls Drawing Process, Picking Process, and Save Undo Process
Controls Fonts in the Menu
83
CHAPTER 1
Introduction
- This Icon is used to control the layout oI the graphic windows. This initiates the Iollowing
pull-down.

Note the types oI plots to be displayed will be dependent on what has been previously requested.
Show/Hide controls which
model repreentation is
going to be displayed
Allows the quick creation
of window captures
Snapshot
84 Marc Users Guide: Part I
Introduction
- This icon initiates the Marc documentation set. The Iollowing menu appears to a allow you to
select the particular manual.

Manual Description
Release Guide Summarizes the New Capabilities oI the Release
Whats New PDF that shows images oI problems that can be solved in the current
release
Installation Guide Instructions oI how to install and customize the soItware
Users Guide Describes Usage oI Program through Mentat
Volume A: Theory and User Information Theoretical review oI the mechanics available in Marc
Volume B: Element Library Description oI the diIIerent element types.
Volume C: Program Input Description oI the Marc input Iormat
Volume D: User Subroutine Templates Ior the user subroutines
Volume E: Demonstration Problems Example oI techniques used Ior engineering analysis
Python Manual Describes the Mentat API
MAR 101 Introduction Course Introductory Class Material
MAR 103 Experimental Elastomer Analysis Class notes Ior elastomer material testing and parameter
identiIication
MAR 104 Electromagnetic Analysis Class notes Ior electrostatic, magnetostatic, electromagnetic, Piezo
electric, and Joule heating.
In the new version oI Mentat, clicking Mentat Help activates a
HTML based help system. This helps system can be downloaded
separately containing the latest inIormation about the Mentat
produce.
85
CHAPTER 1
Introduction
About Marc Mentat will provide the customer identiIier required to obtain customer support.
Tool Bar
The TOOL BAR contains the Iollowing icons. Moving your cursor over the icon, results in a short deIinition its
Iunction.
Icon Description
Deletes the current model and clears the database
Open a new directory
Save current data base
Undo Mentat has only one level oI undo
Reset View
Regenerate graphics area
Expand or contract image to Iill the graphics area

Activate / Deactivate the interactive model manipulation
86 Marc Users Guide: Part I
Introduction
Tabs
The Tabs are the major control to the Mentat Iunctionality. One can move Irom one tab to another tab at any time as
long as a locked popup is not displayed
The tab that is currently selected is a slightly lower color. Once the tab is selected, one or more panels appears in the
'Main Menu Area
Geometry & Mesh
Tables and Coordinate Systems
Geometric Properties
Zoom in on a region identiIied by a polygon
Zoom in () ; Zoom out (-)
Translate in x, -x, y, -y, z, -z directions respectively
Rotate about x , -x, y, -y, z, -z axis
Switch between wireIrame element and solid display oI elements
Switch between wireIrame element and solid display oI surIaces
Switch between wireIrame element and solid display oI solids
Icon Description
87
CHAPTER 1
Introduction
Material Properties
Contact
Toolbox
Links
Initial Conditions
Boundary Conditions
Mesh Adaptivity

Loadcases
88 Marc Users Guide: Part I
Introduction
Jobs
Results
Side Bar
The Side Bar has two sets oI icons, the Iirst are short-cuts to control the selecting oI entities. The second are used Ior
controlling the positioning oI the post Iile. These icons can be repositioned around the graphics window.
Selection control
The menu on the leIt is what appears when the user actually needs to select a list oI entities. II this is not currently
required, then the menu is deactivated and has the appearance oI the Iigure on the right.
89
CHAPTER 1
Introduction
Post File control
When the post Iile is open Ior postprocessing. the post Iile control menu will appear.
Windows Controls
The Mentat product uses standard window control which allows you to resize the main menu. One can also resize the
display image. The pop-up menus come in either two Iorms Iloating/undockable and Iloating/dockable.
In the Iloating/undockable style, such as the Identify Menu shown above, the menu can be repositioned anywhere on
the screen and will remain open until it is closed using the OK command. Any other command may be executed beIore
this pop-up is closed.
In the Iloating/dockable menu as shown below, the Plasticity Property menu can repositioned anywhere, but only the
commands in the Plasticity Property menu can be executed. The command in the parent Structural Properties pop-up
cannot be executed and the parent menu cannot be moved until the child pop-up is closed. This restriction applies to
all parent windows including the Main window. To execute any oI these commands, the child must be closed using the
OK button.
90 Marc Users Guide: Part I
Introduction
Figure 1.1-7 Main Window with Structural Properties Pop-up and Child Plasticity Properties Pop-up.
91
CHAPTER 1
Introduction
Figure 1.1-8 Repositioning Child Pop-up, Parent Pop-up cannot be Repositioned
The Main, Dynamic and Dialog Menus may also be undocked Irom the Graphics Area. This is oIten useIul Ior two-
terminal type displays. In this case, these menus can be shiIted to the second monitor, allowing a larger graphical
display region Ior the model.
92 Marc Users Guide: Part I
Introduction
To achieve this click on the Window Icon associated with these menus. To rejoin these menus to the Iull image
double click on the menu header
Drag and Drop
The icons in the tool bar can be customized using drag and drop. Click on the little dots with the leIt mouse button and
move the icons. This allows you to customize the behavior oI the interIace.
93
CHAPTER 1
Introduction
The post Iile control icons have been moved to bottom.
The user can see all the commands that the program is generating in the Dialog menu.
These commands are also put into a Iile called Mentat.proc that is started when the user invokes the program. They
are also put into the Procedure File Ior later use iI a procedure Iile has been created. Note that the user can scroll in the
Dialog area. Additional use oI the dialog area is discussed later.
The command line area can be used to manually enter commands or literal strings such as Set Names or Boundary
Condition Names or Numerical Data such as the Young`s modulus.
94 Marc Users Guide: Part I
Introduction
Scrollable Widgets.
Mentat makes signiIicant use oI scrollable widgets such as when selecting the scalar value to be displayed.
One can either use the scroll bar to control
Or one can expand the scroller menu or do both.
95
CHAPTER 1
Introduction
II all items are visible, then the scroller disappears.
96 Marc Users Guide: Part I
Introduction
Menu System
The kernel oI the Mentat program consists oI a set oI processors in a parallel conIiguration that operate on the data
base. The data base is the most compact, yet complete, description oI the current state oI the model you are analyzing.
Typical examples oI processors are SUBDIVIDE and PATH PLOT.
Every processor may depend on a number oI parameters that inIluence the process. The combined number oI
processors and parameters in Mentat is too large to show in one menu. To help you in the scheduling oI tasks, we have
structured menus around the processors that lead you through the steps Irom top down. Figure 1.1-9 shows you the
organization oI the main menu that appears when you start Mentat and how it corresponds to the main tasks oI the
analysis cycle depicted in Figure 1.1-1.
For your convenience, the menu items have been grouped in panels by the Iour main tasks: preprocessing, analysis,
postprocessing, and conIiguration. The menu items and subtasks on each oI these panels represent yet another group
oI corresponding tasks. It is important to realize that most oI the menus Ior the global tasks do not contain processors;
these menus are Ior navigation purposes only and are not part oI the kernel oI the program.
Figure 1.1-9 Organization of Main Menu
A task and corresponding subtask is selected by clicking on a menu item oI that menu. AIter the (sub)task is
accomplished, it is necessary to traverse the menus in the opposite direction. There are two ways to do this:
97
CHAPTER 1
Introduction
1. Click on the RETURN or MAIN menu items in the static menu area.
RETURN takes you to the previous menu and MAIN takes you to the main menu.
2. Move the over the menu area and click the <MR>. The result oI this sequence is equivalent to clicking on
the RETURN menu item.
Menu Components
This section describes the anatomy or diIIerent components oI the Mentat menu and the meaning oI each component.
Understanding the deIinition oI components will help you to use the menu system in a proIicient manner. We have
already mentioned that the menu items are grouped into panels or submenus by task and related subtasks. Each menu
has a title that describes its task.
Positioned on the panel are Ilat and raised rectangles. The raised rectangles in the released state suggest a light shining
directly Irom above. The task is printed on the raised rectangle and is selectable by clicking on it with the <ML>. Flat
rectangles are not selectable; they convey the setting oI parameters. The program does not respond to clicking the
<ML> or <MM> on the Ilat rectangles.
Mentat contains Iive types oI raised rectangles. Throughout the remainder oI this document, the term button is used
Ior a raised rectangle. Below is a list oI the diIIerent types oI buttons and their Iunctions.
Cycle Button
A cycle button is used to set a parameter to a value when there is a choice oI three or more alternatives. The parameter
is set to the value that is currently displayed on the button. Clicking on this button changes the displayed value to the
next consecutive value in the list oI alternatives. II the list is exhausted, the process starts over again with the Iirst
alternative. This button is identiIied by a V symbol. Note that the symbol is indicative oI the unidirectional way the
list oI alternatives is traversed.
Toggle Button
A special type oI cycle button is the toggle button where the number oI alternatives is limited to two. It is a switch that
denotes a state oI on or off; a button is depressed to Ilag on or active, and released (or raised) to Ilag that the listed
parameter is oII or inactive. This button is identiIied by a o symbol.
Tabular Button
A tabular button represents a combination oI a parameter button and a Ilat rectangle. They show one or more numerical
or alpha-numerical values that are associated with the parameter represented by the button. Clicking on this button
type usually implies that you have to enter data through the keyboard, which is then displayed in the rectangular Iields
aIter the keyboard input is completed.
Tabular buttons may contain a large number oI numerical data Iields. There are instances where the tabular buttons
pop-up over the graphics area. II this is the case, you need to conIirm that all entries have been completed by clicking
on the OK button. BeIore returning to regular menu selection, you can clear all entries by clicking on the RESET button
which usually appears in the lower leIt-hand side oI the panel. The pop-up table then disappears Irom the graphics area
and the original graphics area is restored. Typical examples oI these compounded tabular buttons can be Iound in the
boundary conditions and material properties menus.
< >
98 Marc Users Guide: Part I
Introduction
ACTIVATE/DEACTIVATE
A button that changes the behavior oI the program that looks like a square. When not activated, it looks like . When
activated, it looks like .
One-Only Button Group
The alternative values oI cycle buttons are also represented as individual toggles under a one-only button group. In a
cycle, only one value can be selected; hence iI a button in a one-only group oI buttons is depressed, another is released.
The one-only button sequence is identiIied by an symbol that is picked resulting in .
As a typical example oI a menu, the Coordinate System panel oI the Mesh Generation menu as shown in Figure 1.1-10
will be discussed. These buttons are also summarized in Figure 1.1-14.
99
CHAPTER 1
Introduction
Figure 1.1-10 Coordinate System Panel
The GRID button is a toggle; it can be switched on or off. The deIault position Ior this button is the raised or released
state which means that the grid is oII. Clicking it will turn the grid on and leave the button in a depressed state.
100 Marc Users Guide: Part I
Introduction
Figure 1.1-11 Released and Depressed States of a Toggle Button
The button next to it displays RECTANGULAR and has the symbol which implies one only. In this example, the
button is an adjective to grid and speciIies the type oI grid to be used.
Figure 1.1-12 gives you examples oI tabular buttons that are Iound in the SET submenu.
Grid Off Grid On
101
CHAPTER 1
Introduction
Figure 1.1-12 Example of Simple Tabular Buttons
Examples of
Simple Tabular
Buttons
102 Marc Users Guide: Part I
Introduction
The Figure 1.1-13 summarizes the diIIerent types oI buttons Iound in the Mentat menu.
Figure 1.1-13 Summary of Mentat Menu Buttons
List Specification
How This Manual Communicates with You discussed the diIIerence between menu buttons that are used to navigate
through the menus and buttons that represent processors. Processors generally require two types oI data:
Parameters associated with the process
A list oI items to operate on.
II the list to operate on consists oI only one item, you can use the mouse to point to that item on the graphics screen
(see Using the Mouse to Point). II the list oI items contains twenty items, pointing to each item individually becomes
a cumbersome task; iI the list contains a hundred items, pointing becomes an impossible task. This section
concentrates on the capabilities in Mentat to speciIy a list oI items.
The Mentat program recognizes the Iollowing items:
Points
Curves
One-Only
Group of
Buttons
103
CHAPTER 1
Introduction
SurIaces
Solids
Vertices oI solids
Edges oI solids
Faces oI solids
Nodes
Elements
Edges oI elements
Faces oI elements.
A simple example oI how to generate a list Iollows. You can then extrapolate Irom what you have learned in this
section to do more intricate examples. Assume you want to subdivide an existing element that is already displayed in
the graphics area oI the Mentat window. The processor to use is SUBDIVIDE, and it operates on elements only.
Figure 1.1-14 Locating the SUBDIVIDE Processor in the Mesh Generation Menu
Note: The examples in the remainder oI this chapter show menus using an icon style that was elimated Irom the
Mentat 2012 release.
104 Marc Users Guide: Part I
Introduction
AIter you activate the SUBDIVIDE processor and click on the ELEMENTS button in the subdivide submenu, the
Iollowing program prompt appears in the dialogue area:
Enter subdivide element list:
Chances are that you don't know the element number (nor should you care at this point). For this reason, answering
this question by typing a number in the dialogue area may be possible but is not necessarily a viable option. Instead,
move the mouse over the graphics area and use the <ML> to click on the center oI the element. You have now entered
the Iirst element in the list. The program keeps prompting you Ior more elements; iI this is the only one you want to
subdivide, you must let the program know that this is the end oI the list. This can be done in one oI three ways:
1. Press the END LIST button in the menu area,
2. Type a '#' sign in the dialogue area, or
3. Click <MR> with anywhere over the graphics area.
The most convenient way oI ending the list is oI course to click <MR> since the is most likely already over the
graphics area and saves you a keystroke Irom the keyboard.
Using a Box to Specify a List
Suppose the number oI subdivisions was set to 20 by 20, creating 400 elements. Assume you want to enter the leIt 200
elements in a list by creating a rectangle to Ience in those elements. Position the at one oI the corners oI the box.
Depress the <ML> and move the to the opposite corner oI the box you want to create. The rectangle that appears
tells you exactly which elements are included in the box. Once you have reached the desired position, release <ML>
(see Figure 1.1-15). For Every element that is completely inside the box is included in the list speciIication.
There are times when you may need the guidance oI cross hairs to help you determine what is to be included in your
selection. To activate the cross hairs, press the SHIFT key while moving the in the graphics area.
Using a Polygon to Specify a List (CTRL Key + <ML>)
An alternative to using the box pick Ior list speciIication is to use a polygon around the elements (Figure 1.1-16) that
you want to include in the list. As with the box pick method, only those elements that are completely inside the polygon
are entered into the list. To use the polygon pick, move the to the Iirst corner point oI the polygon. Click <ML>
while holding down the CTRL key on the keyboard. Move to the next vertex oI the polygon and click <ML> again,
continue to hold the CTRL key down. Repeat this process until a closed loop is Iormed. The last point needs to be in
the vicinity oI the starting point and must be clicked on to end the selection. A variation on this polygon pick is the
lasso pick This is done by holding down the CTRL key and the <ML> down simultaneously while slowly moving the
mouse, until the elements to be selected are surrounded by the lasso (Figure 1.1-17). With either approach, a Iinal click
on <ML> is required at the position near the beginning oI the polygon or lasso.
Note: You can relax the completely inside constraint mentioned previously by using the PARTIAL button on the
picking panel under DEVICE.
Note: The PARTIAL and COMPLETE buttons mentioned under the VIEW > DEVICE menu Pick Method also apply
to the Polygon Pick Method.
< >
< >
< >
< >
< >
< >
105
CHAPTER 1
Introduction
Table 1.1-1 at the end oI this chapter summarizes the mouse selection options in both the graphics and menu areas.
Figure 1.1-15 Selecting an Element Using the Box Pick Method
Step 1: Position the cursor
Step 2: Hold down the <ML> and drag the cursor to the desired position
Step 3: Release the <ML> button
106 Marc Users Guide: Part I
Introduction
Figure 1.1-16 Selecting an Element Using the Polygon Pick Method
Figure 1.1-17 Selecting an Element Using the Lasso Pick Method
LIST Buttons
For your ease oI use, we have preprogrammed some oI the more common list options and assigned them buttons which
are located in the lower leIt-hand side oI the static menu.
Press Ctrl key, and click <ML>
at each vertex of
the polygon
Polygon pick
Press Ctrl key, and hold <ML> down
while dragging the < >
until a closed loop is
Lasso
formed around the
elements to be selected
107
CHAPTER 1
Introduction
The LIST buttons are:
Figure 1.1-18 Location of LIST Buttons
All: EXISTING
Perhaps the most used list button is all EXISTING. It speciIies all existing elements, nodes, curves, points, or surIaces
(whichever is applicable), to be operated on by the processor that requested the list.
The contents oI the selected/unselected, visible/invisible list are determined by the two operators: SELECT and
VISIBLE. The meaning oI each and their connection is explained in the next paragraphs.
All: SELECTED/UNSELECTED
The SELECT operator is a very powerIul way to separate speciIic items Irom others. The methods by which items are
selected range Irom a single item to a path oI nodes, a box oI items, or all items on a plane, and are connected by
Boolean operators such as and, except, invert, and intersect. An example oI this syntax is:
(use) single |items| and (a) box (oI) |items| except single |items|
where the words use, a, of, and item are implied because they do not appear as buttons. A powerIul Ieature oI the
SELECT processor is the ability to name a group oI items, and reIer to them by that name in list speciIications. The
STORE command Iacilitates this process.
All: VISIBLE/INVISIBLE
all: EXISTING
all: SELECTED
all: UNSELECTED
all: VISIBLE
all: INVISIBLE
all: OUTLINE (all nodes and edges on the outline)
all: SURFACE (all nodes, edges and faces on the outer surface)
all: TOP
all: BOTTOM
SET
END OF LIST
108 Marc Users Guide: Part I
Introduction
Sometimes the model may be so complex that it takes an unacceptably long time to update the graphics screen every
time the database changes. It is advantageous to Iocus on the items that you are working on. By activating and
deactivating items Irom the display list, you can minimize the items that are displayed. Note that activating or
deactivating does not imply that they are removed Irom the database.
The VIEW>VISIBILITY processor Iacilitates this activation and deactivation process by using the VISIBLE and
INVISIBLE commands.
Tables 1.1-1 and 1.1-2 summarize the Iunctions oI the three-mouse buttons in the graphics and menu areas.

Identifiers
In many applications, an identiIier is associated with a group oI data. These applications include material properties,
link properties, geometric properties, boundary conditions, initial conditions, tables, transIormations, beam sections,
loadcases, and jobs. The identiIier can be any name, iI none is given a deIault name is given. These ID names are then
reIerenced in other commands. The use oI IDs is detailed below.
When using many oI the menus, the Iollowing buttons appear.
Creates a new entry in the list oI applications and makes it the current application.
Table 1.1-1 Mouse Button Functions in Graphics Area
<ML> <MM> <MR>
single pick or box pick unpick end oI list
single pick or box pick with cross hairs unpick end oI list
polygon pick or lasso pick unpick end oI list
Table 1.1-2 Mouse Button Functions in Menu Area
<ML> <MM> <MR>
command selection on-line help return
SHIFT
CTRL
109
CHAPTER 1
Introduction
Removes the current application ID and the associated data.
Allows you to provide a name to the current application.
Creates a new entry in the list oI applications by copying the current application ID; the new entry becomes the current
application.
Allows you to position to either the previous or next item.
Displays a list oI the IDs and allows you to select a particular ID. The selected ID becomes the current one.
Menu Customization
The menu system oI Mentat may be customized to suit special end user requirements. The SHORTCUTS menu button
in the lower leIt-hand corner provides some useIul shortcuts; howeve,r these can be changed to be any list oI quickly
accessible commands. See the Mentat Menu Customization Guide under the Mentat directory oI
doc/menu/MenuGuide.html. You may open a menu Iile Ior editing by moving the cursor over a displayed menu
and pressing the F2 Iunction key.
Comprehensive Sample Session
In this hands-on session, you will create a simple 3-D mesh and add all appropriate boundary conditions, material
properties, etc. You will run the analysis and view the results.
A linear elastic analysis oI the Iollowing 3-D structure will be perIormed:
110 Marc Users Guide: Part I
Introduction
Boundary conditions:
Iace 1: clamped
Iace 2: loaded by a uniIormly distributed shear load
(Iorce per unit area), magnitude 40, direction
Material properties:
Young`s modulus
Poisson`s ratio
20 20
20
30
200
60
face 1
face 2
x
y
z
0 1 1
T
E 4
5
10
v 0.3
111
CHAPTER 1
Introduction
Start up window Mentat:
112 Marc Users Guide: Part I
Introduction
Mouse buttons
ML MM MR
menu area
graphic area
select
command
help on
command
return to
previous menu
pick
entity
undo last
pick
end of list
single pick
box pick
polygon pick
(CTRL)
113
CHAPTER 1
Introduction
Mesh generation: top menu
114 Marc Users Guide: Part I
Introduction
Mesh generation (continued): set and display grid for easy input of coordinates and fill view
SET: COORDINATE SYSTEM
GRID ON
U DOMAIN
0 200
U SPACING
20
V DOMAIN
0 80
V SPACING
10
FILL
RETURN
115
CHAPTER 1
Introduction
Mesh generation (continued): create points (geometric entities), switch off grid and fill view
PTS: ADD (Add the following points with mouse clicks)
(0,80,0)
(20,80,0)
(40,80,0)
(0,60,0)
(20,60,0)
(40,60,0)
(20,0,0)
(40,0,0)
(200,80,0)
(200,60,0)
FILL
116 Marc Users Guide: Part I
Introduction
Mesh generation (continued): create quad surfaces (geometric entities)
SRFS: ADD
Pick corner points for quad surfaces with mouse clicks to obtain four surfaces as shown.
A half-arrowhead is used to indicate the first side of the surface.
117
CHAPTER 1
Introduction
Mesh generation (continued): convert surfaces to elements (mesh entities)
CONVERT
DIVISIONS
6 2
BIAS FACTORS
-0.3 0
SURFACES TO ELEMENTS (pick the rightmost surface)
DIVISIONS
2 2
BIAS FACTORS
0 0
SURFACES TO ELEMENTS (pick the two small surfaces)
DIVISIONS
2 3
SURFACES TO ELEMENTS (pick the lower surface)
118 Marc Users Guide: Part I
Introduction
Mesh generation (continued): modify sweep tolerance and use sweep option to merge coincident nodes
SWEEP
TOLERANCE
0.001
ALL
119
CHAPTER 1
Introduction
Mesh generation (continued): use renumber option to obtain consecutive numbering
RENUMBER
ALL
120 Marc Users Guide: Part I
Introduction
Mesh generation (continued): use expand option to expand the mesh in z-direction
EXPAND
TRANSLATIONS
0 0 15
REPETITIONS
2
MODE: REMOVE (no action required, this is the default)
ELEMENTS
ALL: EXISTING
121
CHAPTER 1
Introduction
Mesh generation (continued): remove unused nodes and repeat renumber command
SWEEP
REMOVE UNUSED: NODES
ALL
RETURN
RENUMBER
ALL
122 Marc Users Guide: Part I
Introduction
Mesh generation (continued): show view 4 and fill view
VIEW
SHOW VIEW
4
FILL
123
CHAPTER 1
Introduction
Mesh generation (continued): define increment of rotation for the model
VIEW SETTINGS
MODEL INCREMENTS: ROTATE
90
124 Marc Users Guide: Part I
Introduction
Mesh generation (continued): rotate model in positive direction around model x- and y-axis and
fill view
MANIPULATE MODEL
ROTATE IN MODEL SPACE:
X+
ROTATE IN MODEL SPACE
Y+
FILL
125
CHAPTER 1
Introduction
Mesh generation (continued): plot elements in solid mode, switch off plotting geometric entities
VIEW
DRAW SETTING
turn off POINTS and SURFACES
ELEMENTS: SOLID
REDRAW
OK
126 Marc Users Guide: Part I
Introduction
Boundary conditions: top menu
BOUNDARY CONDITIONS
MECHANICAL
127
CHAPTER 1
Introduction
Boundary conditions (continued): switch to view 1 and select appropriate nodes
VIEW
SHOW VIEW
1
FILL
OK
NODES: ADD (add as shown by the box pick method)
END LIST (for end list use button or use right mouse
click in graphics area)
128 Marc Users Guide: Part I
Introduction
Boundary conditions (continued): switch to view 4, define mechanical boundary conditions, face loads
VIEW
SHOW VIEW
4
OK
NEW
NAME shear
FACE LOAD
U SHEAR
28.2843
V SHEAR
28.2843
OK
129
CHAPTER 1
Introduction
Boundary conditions (continued): zoom in locally and select appropriate element faces
130 Marc Users Guide: Part I
Introduction
ZOOM BOX (zoom in on the right end of structure)
FACES: ADD (add appropriate element faces with mouse)
END LIST
131
CHAPTER 1
Introduction
Boundary conditions (continued): overview of boundary conditions
ID BOUNDARY CONDITIONS (on)
FILL
ARROW PLOT SETTINGS
SOLID (on)
DRAW
OK
SAVE
132 Marc Users Guide: Part I
Introduction
Material properties: top menu
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NEW
STANDARD
NAME
linear_elastic
133
CHAPTER 1
Introduction
Material properties (continued): mechanical material type, isotropic properties, apply to all elements
STRUCTURAL
YOUNGs MODULUS = 400000
POISSONS RATIO = 0.3
OK
ELEMENTS: ADD
ALL: EXISTING
134 Marc Users Guide: Part I
Introduction
Geometric properties: top menu
GEOMETRIC PROPERTIES
135
CHAPTER 1
Introduction
Geometric properties (continued): select assumed strain formulation for all existing elements to
improve bending behavior
NEW
NAME
assumed_strain
3-D
SOLID
ASSUMED STRAIN (on)
OK
ELEMENTS
ADD
ALL: EXISTING
136 Marc Users Guide: Part I
Introduction
Jobs: define mechanical analysis; for a single linear analysis no loadcases are necessary and the default
analysis options can be used
JOBS
NEW (TYPE)
MECHANICAL
NAME
example_3d
PROPERTIES
137
CHAPTER 1
Introduction
Jobs (continued): select post file quantities
JOB RESULTS
AVAILABLE ELEMENT TENSORS
Stress
OK
138 Marc Users Guide: Part I
Introduction
Jobs (continued): check if boundary conditions are selected as initial loads
INITIAL LOADS (check to see if they are on)
OK
OK
139
CHAPTER 1
Introduction
Jobs (continued): select mechanical 3-D solid element type 7, save model
ELEMENT TYPES
ANALYSIS DIMENSION
3-D (select element type 7)
OK
ALL: EXISTING
SAVE
OK
140 Marc Users Guide: Part I
Introduction
Jobs (continued): save Mentat database and submit job model1_example_3d
RUN
STYLE
OLD (old style of table input)
SUBMIT 1 (look for EXIT NUMBER 3004)
141
CHAPTER 1
Introduction
Submitting a job:
Marc data file:
submit1: Mentat-directory/bin/submit1
marc -j model1_example _3d -q b -v no
* use model1_example_3d.dat as data file
* run job in background
* dont wait for confirmation of correct input
outputfile: model1_example_3d.out
log file: model1_example_3d.log
post file: model1_example_3d.t19 (formatted)
model1_example_3d.t16 (binary)
Parameter
options
Model
definition
options
History
definition
options
sufficient
for a
single
linear
analysis
title
sizing
elements
...
...
end
connectivity
coordinates
isotropic
geometry
fixed disp
dist loads
point load
...
...
end option
...
continue
...
...
continue
dist loads
point load
disp change
...
142 Marc Users Guide: Part I
Introduction
Jobs (continued): use monitor to observe current status
MONITOR
OPEN POST FILE (RESULTS MENU) (opens defalut post file and goes to results menu)
143
CHAPTER 1
Introduction
Marc post file:
Header
Increment 0
coordinates
connectivity
nodal
quantities;
element
quantities
if selected
Increment 1
nodal
quantities;
element
quantities
if selected
open post file
next increment;
skip to increment 0
next increment;
skip to increment 1
144 Marc Users Guide: Part I
Introduction
Postprocessing (continued): skip to increment 0 and select equivalent von Mises
stress to be displayed
SCALAR
EQUIVALENT OF STRESS
OK
145
CHAPTER 1
Introduction
Postprocessing (continued): plot deformed and undeformed structure for increment 0 using
contour bands
DEFORMED AND ORIGINAL
CONTOUR BANDS
146 Marc Users Guide: Part I
Introduction
Postprocessing (continued): deformed shape settings. You may magnify the displacements.
DEFORMED SHAPE: SETTINGS
MANUAL FACTOR
4
MANUAL FACTOR
1
RETURN
147
CHAPTER 1
Introduction
Postprocessing (continued): define plotting style settings for cutting planes
SCALAR PLOT: SETTINGS
POINT
00 00
NORMAL
1 0 0
PLANES
12
SPACING
18
CUTTING PLANES
SCALAR PLOT: SETTINGS
148 Marc Users Guide: Part I
Introduction
Postprocessing (continued): switch back to contour bands plotting and define a node path for a
path plot
CONTOUR BANDS
PATH PLOT
NODE PATH (pick the nodes shown to define the path - N1, N2, N3)
END LIST
149
CHAPTER 1
Introduction
Postprocessing (continued): add path plot curve and scale the plot axes
ADD CURVES
ADD CURVE
ARC LENGTH (X variable)
EQUIVALENT OF STRESS (Scroll Down to Y variable)
FIT
150 Marc Users Guide: Part I
Introduction
Postprocessing (continued): vector plot of displacements
Workshop tasks:
Perform the discussed 3-D analysis and store the Mentat commands in a procedure Iile, which can be
created in the TOOLS Menu Bar
Analyze the same 3-D structure, but now subjected to a distributed shear load with a magnitude 40 and a
direction oI (bending load)
Analyze the structure subjected to the bending load using 4-node plane strain elements (select Marc element
type 11) and compare the results with the 3-D solution
MORE
VECTOR PLOT: ON
0 1 0
T
151
CHAPTER 1
Introduction
Additional workshop: linear elastic analysis of an infinitely long pressurized thick-walled cylinder
Dimensions: , ,
Apply Iixed displacements in axial direction
Internal pressure:
Material: ,
Workshop tasks:
Determine the radial stress as a Iunction oI the radial coordinate using:
A: axisymmetric element 10:
B: plane strain element 11 (model one quarter oI the cross section):
p
r R
L
section to be considered
L 4 r 5 R 12
p 15
E 2.1
5
10 v 0.3
152 Marc Users Guide: Part I
Introduction
C: brick element 7 (model one quarter oI the section to be considered):
Apply the correct boundary conditions and compare the results.
Infinitely long pressurized thick-walled cylinder
0.0
-5.0
-10.0
-15.0
4.0
6/0 8.0 10.0 12.0
radial coordinate
r
a
d
i
a
l

s
t
r
e
s
s
Analytical solution
10 axisymmetric elements; radial bias -0.5
80 plane strain element; radial bias -0.5
80 brick elements; radial bias -0.5
153
CHAPTER 1
Introduction
Background Information
The purpose oI this section is to give you an overview oI the most common Mentat Ieatures. These Ieatures recur
throughout the sample sessions in the last part oI this guide. For example, all sample sessions contain a mesh
generation step.
You may Iind it helpIul to use the inIormation in this section as you work through the example problems. The order in
which the common Mentat Ieatures are described in this section is based on the preprocessing, analysis, and
postprocessing sequence oI the Iinite element analysis process.
The key Ieatures discussed in this chapter are listed below.
Mesh generation
mesh entities
geometric entities
direct meshing technique
geometric meshing technique
Boundary conditions, initial conditions and links
Material and geometric properties
Contact
Loadcases and jobs
Results interpretation
Mesh Generation
The preprocessing task is considered a signiIicant part oI the Iinite element analysis process. In Iact, at times it may
be the most complex and time consuming part oI the entire job. For this reason it is important that you use the
conceptualization phase as indicated in Figure 1.1-1 to determine in advance what the objective oI your analysis is and
what answers you are seeking since both Iactors strongly inIluence the choice oI your model.
Mentat distinguishes two techniques to build a mesh. The Iirst is the direct or manual approach where you generate
Iinite elements Irom bottom up. The second is the geometric approach where the model is Iirst generated using
geometric entities Iollowed by a conversion step in which these entities are converted to Iinite elements. The two
techniques are by no means mutually exclusive and oIten the best results are obtained by alternating between the two.
The Iollowing guidelines will simpliIy the task oI generating a mesh using either one oI the available methods.
Plan your model
Plan your model careIully and take the time to Iormulate a strategy. This will save you time and resources.
154 Marc Users Guide: Part I
Introduction
Look for symmetry and duplication
Many structures exhibit some Iorm oI symmetry. Look Ior the simplest component oI your model. Also look Ior
duplicates (or close duplicates) oI another portion, and use the SYMMETRY or DUPLICATE processor.
Select a logical origin
The lower-leIt corner oI your model or drawing is not necessarily the best location Ior the origin. Take some time to
examine the model Ior an origin that makes the creation process easier. For example, iI a model is symmetrical about
a hole in a plate, consider placing the origin at the center oI the hole. You may change the location oI the origin within
the same session while generating your model.
Choose a logical coordinate system
The deIault coordinate system in Mentat is rectangular Cartesian. A cylindrical or spherical coordinate system may be
more suitable Ior a particular model.
Create 1-D and 2-D before 3-D
For many structures, oIten the best mesh generation strategy is to create Iirst a 1-D and/or 2-D mesh and to drag it into
a 3-D mesh using the EXPAND processor.
Mesh Entities
Figure 1.1-17 Element Classes
X
Y
Z
2 nodes 3 nodes
3 nodes 6 nodes
6 nodes 4 nodes
8 nodes
9 nodes
The Stick Family
The Triangular Family
The Quadrilateral Family
1
LINE (2) LINE (3)
TRIA (3)
TRIA (6)
QUAD (4)
QUAD (6)
QUAD (8) QUAD (9)
155
CHAPTER 1
Introduction
Two types oI mesh entities can be distinguished: nodes and elements.
Nodes
Nodes are characterized by three coordinates and symbolized by a on the screen. When nodes are attached to a
geometric entity, the symbol is used instead.
Elements
Elements consist oI element edges and element Iaces and are deIined by a sequence oI nodes. The number oI edges,
Iaces and nodes depend on the element class. Mentat employs a wide variety oI element classes which are identiIied
in Figure 1.1-17 and Figure 1.1-18. The CHANGE CLASS processor allows you to change the class oI existing
elements. The CLEAR MESH button in the MESH GENERATION~MANUAL menu enables you to remove all mesh
entities Irom the database. When elements are drawn in wireIrame mode, the Iaces are indicated with a cross and the
Iirst edge is indicated with a halI-arrowhead.
Figure 1.1-18 Element Classes
During the mesh generation phase, it is not required to make a decision on the element type to be used. Only the
element class is important at this stage. For instance, planar elements can be used to model a planar or an axisymmetric
structure. In the phase oI the analysis type deIinition, it has to be decided iI the element is axisymmetric, plane stress
or plane strain.
Y
X
Z
The Tetrahedral Family
The Pentahedral Family
The Hexahedral Family
4 nodes
10 nodes
6 nodes
15 nodes
8 nodes
12 nodes
20 nodes
27 nodes 27 nodes
1
TETRA (4)
PENTA (6)
HEX (8)
HEX (20)
TETRA (10)
PENTA (15)
HEX (12)
HEX (27)
156 Marc Users Guide: Part I
Introduction
Geometric Entities
The building blocks oI the geometric mesh technique are points, curves, surIaces, and solids.
Points
Points are characterized by three coordinates and symbolized by a '' on the screen.
Curves
The Iollowing curve types can be used to deIine curves: line, polyline, tangent, arc, Iillet, circle, cubic spline,
interpolate, Bezier curve, NURB curve, composite curve, and sampled.
The line curve is a straight line segment between two points, the polyline is a concatenation oI linear line segments
through a series oI points. The tangent is a straight line tangent to an existing curve. The arc curve is part oI a circle
and Iive methods are available to speciIy this circle segment. The Iillet curve creates an arc between two curves. The
circle curve is a complete circle and can be speciIied in two ways. The cubic spline and the interpolate curve pass
through a series oI points (a curve Iitting technique). The Bezier curve and NURB curve can be used to deIine more
general curves. The composite curve enables joining several previously deIined curves.
Surfaces
Currently, Mentat recognizes the Iollowing surIace types which can be speciIied directly: quad, ruled surIace, driven
surIace, cylinder, Bezier surIace, NURB surIace, sphere, swept surIace, interpolate, coons, skin surIace, and sampled.
The quad surIace is the most simple surIace deIinition as it is deIined by 4 (non-collinear) points. The cylinder surIace
is a cone which is deIined by the coordinates oI two points on the axis and two radii. The sphere surIace is deIined by
the center and the radius. The ruled surIace is spanned by a Iamily oI straight lines between two curves. Both the driven
and the swept surIace are generated by dragging a curve along another curve. The interpolate, Bezier and NURB
surIaces are logical extensions oI the interpolate, Bezier and NURB curves. The coons surIace is created Irom a closed
boundary consisting oI Iour curves. The skin surIace is created through a list oI curves.
You can also generate axisymmetric surIaces by revolving a curve about the local y-axis using the
REVOLVE processor.
In addition, the CAD interIaces allow Mentat to read in trimmed surIaces.
When you display a curve or surIace, you oIten see a crude representation oI that entity on your screen. We emphasize
the word representation here. By deIault, the resolution oI a curve is set to 10. The curve is represented by 10 straight
line segments. For small curves, this may be suIIicient to give the impression oI a smooth curve. For larger curves,
however, 10 subdivisions may not be suIIicient. You can change the resolution by increasing the number oI divisions
in the DIVISIONS submenu oI the continued part oI the NEW>PLOT SETTING>SETTINGS menu. Be aware that
increased resolution requires more time Ior the program to draw the curve.
Solids
A solid is a volume which is bounded by a number oI Iaces. Solid Iaces are bounded by edges and solid edges are
bounded by solid vertices. Mentat oIIers Iive basic solid types: block, cylinder, sphere, torus, and prism.
157
CHAPTER 1
Introduction
The block entity is a rectangular block which is deIined by the coordinates oI a corner point and three dimensions. The
cylinder entity is a solid cone which is deIined by two points on the axis oI revolution and two radii. The sphere is
deIined by the center point and the radius and the torus is deIined by the coordinates oI the center and the two radii.
The prism is deIined by two axis points, a radius, and the number or prism Iaces.
The basic solids can be manipulated through the SOLIDS processor. First a series oI boolean operations such as
UNITE, SUBTRACT, and INTERSECT can be used to modiIy the basic solid entities. In addition, the BLEND and
CHAMFER operations exist to make smooth transitions between various Iaces.
Geometric entities can be converted to other geometric entities using the CONVERT or SOLIDS processor. This
allows the Iollowing conversions:
The CLEAR GEOM button in the mesh generation menu allows you to remove all geometric entities Irom the database.
The Direct Meshing Technique
Elements are used as the basic building blocks to generate a coarse mesh that can be reIined later using the tools
provided by Mentat speciIically Ior this purpose. This approach is particularly suitable Ior a domain with a simple
geometry. The direct meshing technique is not based on an algorithm but consists oI the enumeration, by you, oI the
most coarse mesh that still represents the desired geometry. Use the ADD button oI the element and node panels in the
mesh generation menu to deIine the building blocks.
Once you have generated a coarse model (Figure 1.1-19), you can reIine it (locally) to the desired level using the
SUBDIVIDE processor. You can expand the model to a higher dimension using the EXPAND processors. The
DUPLICATE, SYMMETRY, and MOVE processors allow scaling, translation, rotation, and duplication oI part oI
the model. The RELAX and STRETCH processors are available to relocate nodal points based on a given element
connectivity. Removal oI duplicate nodes or elements is achieved through the SWEEP processor and renumbering oI
the mesh can be perIormed with the RENUMBER option. CHECK allows speciIic checks on the correct deIinition oI
the mesh.
curve polyline curve
curve interpolated curve
surIace polyquad surIace
surIace interpolated surIace
solid vertex point
solid edge curve
solid Iace trimmed surIace
trimmed surIace solid Iace
158 Marc Users Guide: Part I
Introduction
Figure 1.1-19 Example of a Coarse Mesh
The direct mesh generation process in Mentat is a three step procedure:
The Geometric Meshing Technique
The basic building blocks Ior this technique are geometric entities rather than mesh entities. The geometric entities
available in Mentat are points, curves, surIaces, and solids. They may be converted to mesh entities aIter you have
completed the geometric model. This approach is more complex than the direct meshing technique as it involves the
extra layer oI geometric entities. However, the advantage oI the geometric meshing technique is that increased
complexity is oIIset by increased Ilexibility in generating geometries oI complex shape. It is important to diIIerentiate
mesh entities Irom geometric entities; Ior example, a two-noded line element is not the same as a line curve, and a
node is not the same as a point.
Convert
To change the geometric model to a Iinite element mesh, you may convert the geometric entities to Iinite elements.
For instance, curves can be converted into line elements and surIaces into quadrilateral elements. The Iollowing
conversions are possible:
Step 1 Generate nearly correct coordinates and Iully correct connectivity. Subdivide and reIine the initially
speciIied elements where necessary.
Step 2 ModiIy the boundary nodes Ior exact boundary coordinates.
Step 3 Redistribute the internal coordinates to create reasonably shaped or relaxed elements.
point node
curve line elements
surIace quadrilateral elements
X Y
Z
4
159
CHAPTER 1
Introduction
The ATTACH processor is a very powerIul tool to put nodes on a curve or surIace. Please note that aIter a CONVERT
operation, the resulting nodes have been attached to the geometric entity.
Automesh
Mentat contains as optional products automatic mesh generators which generate Iinite element meshes on solids, on
trimmed surIaces, and within curves in a plane.
Typically, three steps can be considered.
1. Clean up and repair oI the curves (coming Irom a CAD tool)
2. Set the curve divisions which basically controls the mesh density
3. Automatic generation oI the mesh
Three classes can be distinguished Ior the automatic meshers:
1. 2-D PLANAR MESHING
2. SURFACE MESHING
3. SOLID MESHING
Both the 2-D Planar and the SurIace meshing have several alternatives Ior creation oI a mesh. All meshers with
exception oI the OVERLAY mesher use the seed points deIined in the CURVE DIVISIONS menu.
The OVERLAY processor allows you to describe the geometry by its boundary instead oI a surIace. This can be
applied either to a planar structure or a trimmed surIace. You may use any curve type available in Mentat to speciIy
the boundary. (This also implies that a combination oI curve types is permitted.)
The TRI MESH! mesh generator creates triangular elements, on a trimmed surIace or within curves in a plane. Either
the Delaunay technique or the Advancing Front technique can be used.
The QUAD MESH! mesh generator creates quadrilateral elements on the Iaces oI a solid, on a trimmed surIace, or
within curves in a plane based on an Advancing Front technique.
Triangular Mesh
160 Marc Users Guide: Part I
Introduction
Figure 1.1-20 Overlay Mesh
In addition, mixed triangular and quad elements can be generated using the Advancing Front technique.
The two solid meshers use the surIace mesh created with the above mentioned meshers as input. The TET MESH! mesh
generator creates tetrahedral elements in a solid volume or within a volume bounded by triangular elements. The
HEXMESH! generator creates hexahedral solid elements in a volume spanned by the surIace mesher.
Y
X
Z
Boundary Description
Meshed Interior using Overlay
1
Quadrilateral Mesh
161
CHAPTER 1
Introduction
What Constitutes a Good Mesh?
UnIortunately, this question can only be answered a posteriori. Only when the analysis is complete, and a convergence
study conducted, is it possible to quantiIy the answer to this question. A priori qualiIications, although oIten necessary,
are generally not suIIicient.
Elements have ideal shapes when there is little or no error in the numerical computation oI individual stiIIness
matrices. It would be convenient iI triangles were always equilateral, quadrilaterals always squares, and hexahedra
always cubes. However, it is almost impossible to model complex systems with a mesh oI ideally shaped elements.
ThereIore, it is advisable to match the mesh density to stress gradients and deIormation patterns which imply that
elements vary in size, have unequal side lengths, and are warped or tapered.
With the above in mind, the remainder oI this section concentrates on a Iew guidelines you can use to determine the
quality oI a mesh. These guidelines are aspect ratio, distortions, and transitioning.
Aspect Ratio
The element aspect ratio is the quotient between the longest and the shortest element dimensions. This ratio is by
deIinition greater than or equal to one. II the aspect ratio is 1, the element is considered to be ideal with respect to this
measure. Acceptable ranges Ior the aspect ratio are element and problem dependent, but a rule oI thumb is:
AR 3 Ior linear elements
AR 10 Ior quadratic elements.
Elements with higher-order displacement Iunctions and higher-order numerical quadrature Ior a given displacement
Iunction are less sensitive to large aspect ratios than linear elements. Elements in regions oI material nonlinearities are
more sensitive to changes in the aspect ratio than those in linear regions. II a problem has a deIlection or stress gradient
dominant in a single direction, elements may have relatively large (10) aspect ratios, provided that the shortest element
dimension is in the direction oI the maximum gradient.
Tetrahedral Mesh
162 Marc Users Guide: Part I
Introduction
Distortions
Skewing oI elements and their out-oI-plane warping are important considerations. Skewness is deIined as the variation
oI element vertex angles Irom 90 Ior quadrilaterals and Irom 60 Ior triangles. Warping occurs when all the nodes oI
three-dimensional plates or shells do not lie on the same plane, or when the nodes on a single Iace oI a solid deviate
Irom a single plane.
Transitioning
Two types oI transitioning exist. The Iirst type is the change in element density in the direction oI the stress gradient.
The greatest reIinement is then in the region with the highest gradient. A good tool to apply to this type oI transitioning
is biased subdivision.
Figure 1.1-21 A Biased Mesh, Bias = -0.4 in X-direction
The second type is transverse transitioning, which is used between element patterns with diIIerent densities across a
transverse plane.
163
CHAPTER 1
Introduction
Figure 1.1-22 A Transition Mesh
II a model requires transverse transition regions, they should only be used in low-stress gradient regions, never near
regions oI maximum stress, deIlection, or other regions oI interest. The REFINE option in Mentat allows you to create
a transition region.
Boundary Conditions, Initial Conditions, and Links
The BOUNDARY CONDITIONS processor is used to deIine the boundary conditions applied to the model in order
to perIorm the analysis.
Mentat distinguishes the Iollowing groups oI boundary conditions based upon the Analysis Class:
Structural
Thermal
Joule
Acoustic
Bearing
Electrostatic
Magnetostatic
Magnetodynamic
Note: Within the Iramework oI the CONTACT option, Marc and Mentat allow the automatic connection oI two diIIerent parts
which do not have common nodal points. Thus, various parts in the structure can be modelled with diIIerent mesh
densities without the need Ior transition regions.
164 Marc Users Guide: Part I
Introduction
Depending on the analysis class, boundary conditions must be taken Irom one oI these groups. An exception to this
rule is the coupled analysis, Ior which both mechanical and thermal boundary conditions may be deIined. The
Iollowing boundary condition types can be Iound in the Mechanical submenu:
Fixed displacement
Fixed acceleration
Point load
Edge load
Face load
Global load
Gravity load
CentriIugal load
Fluid drag
Edge Ioundation
Face Ioundation
Cavity pressure load
Cavity mass load
Degree oI Ireedom - set nodes
State variable
Nodal temperature
Release nodes
The speciIications oI the boundary conditions and associated parameters, along with the location, are grouped in one
menu. The application oI boundary conditions can best be thought oI as an answer to the question: 'Apply what,
where, and when.
Every what requires a list speciIication Ior where and possibly when. It will be clear that Iixed displacements are
applied to nodes as are point loads. Edge loads are applied to edges oI elements, while Iace loads to Iaces oI elements,
etc. For speciIication oI the where part we reIer back to the beginning oI Chapter 2 on List Specification. II nodes have
been attached to a curve or surIace, it is also possible to apply the boundary conditions to the curve or surIace. The
associated nodes, element edges or element Iaces will inherit this boundary condition.
An important consideration oI the when part is that one is deIining potential boundary conditions, based upon a unique
boundary conditions id. The boundary conditions are not applied in an analysis, unless they are selected in the
LOADCASE processor, and the loadcase is selected in the JOBS processor or unless they are selected as INITIAL
LOADS in the JOBS processor. Note that boundary conditions can also be speciIied as a Iunction oI time through the
TABLE option.
Note: It is important to apply the correct number oI boundary conditions. Too many causes the system oI equations to become
over constrained; too Iew causes a rigid body mode.
165
CHAPTER 1
Introduction
In addition to the boundary conditions, oIten a set oI initial conditions can be present. Examples oI these are the initial
velocity in a dynamic analysis, and the initial temperature in a heat transIer analysis. The initial conditions can be
deIined in the INITIAL CONDITIONS processor.
Similar to boundary conditions, one deIines here only potential initial conditions. They become active only iI they are
selected as INITIAL LOADS in the JOBS processor.
For speciIic analyses, it can be required to set up constraint equations between various components oI the boundary
conditions. Also springs can be present between two nodes. The LINKS processor allows the deIinition oI constraint
equations and links or dashpots. (Note that springs are not associated with element behavior.)
Material and Geometric Properties
Virtually all oI the required material data Ior an analysis with Marc may be entered through Mentat. The program
recognizes the Iollowing material data:
Elastic-Plastic Isotropic
Elastic-Plastic Orthotropic
Elastic-Plastic Anisotropic
Rigid Plastic
Hypoelastic
Mooney
Ogden
Gent
Arruda-Boyce
Foam
NLELAST
Shape Memory
Composite
Gasket
Soil
Powder
Heat transIer
Joule heating
Acoustic
Bearing
Electrostatic
Magnetostatic
166 Marc Users Guide: Part I
Introduction
Magnetodynamic
Note that Ior a coupled analysis the heat transIer material type must be combined with one oI the mechanical
material types.
In List Specification, it explains how to apply material data to elements. The MATERIAL PROPERTIES processor
in the main menu Iacilitates the application oI material constants and Iunctions to elements.
Both in the Orthotropic and the Anisotropic material type, direction dependent material constants have to be deIined.
These material properties are usually deIined in a local material axis system. The ORIENTATION processor allows
speciIication oI the material axis system. In addition, the COMPOSITE processor is available to deIine layered shell
structures with diIIerent (direction dependent) properties and thicknesses.
Truss, beam, plane stress, plane strain, axisymmetric, membrane, plate, and shell elements are based on theories that
are limiting cases oI the general continuum theory. Shell theory, Ior instance, requires the shell element to have a
thickness. This thickness (although strictly speaking a part oI the geometry) does not enter into the mesh generation
phase. This data is entered through the GEOMETRIC PROPERTIES processor. Other element types have similar
properties such as area Ior truss elements and moments oI inertia and local axis systems Ior beam elements.
For some element types, special options may be Ilagged in order to get more accurate results. For instance, the classical
4-node plane strain element is known to give a too stiII behavior iI the element is subjected to bending. By selecting
the assumed strain Iormulation, the element type is modiIied into a description with improved bending behavior. II,
Ior the same element, the material behavior is nearly incompressible, the constant dilatation Iormulation has to also be
selected. These special options are also deIined in the GEOMETRIC PROPERTIES processor.
Furthermore, the data Ior the Marc gap/Iriction elements can be entered here.
Contact
The automatic contact analysis is a very powerIul capability in the Marc program. The boundary nodes and segments
Ior a given set oI elements are determined, and when the analysis requires it, the boundary conditions to be applied are
automatically adapted. Mentat supports this analysis capability completely. It allows deIinition oI both deIormable and
rigid bodies, Iriction and thermal contact.
A deIormable contact body is deIined by a list oI elements. A rigid contact body is deIined by curves Ior 2-D
applications and surIaces Ior 3-D applications.
167
CHAPTER 1
Introduction
The CONTACT processor in the MAIN menu allows the deIinition oI the Iollowing tasks:
Note that similar to Boundary and Initial Conditions, both the CONTACT TABLES and CONTACT AREAS only deIine
potential diIIerent applications oI these options. They are only applied iI they are selected in the LOADCASE or the
JOBS processor.
Loadcases and Jobs
Linear Iinite element analysis is characterized by a Iorce-displacement relationship that only contains linear terms. The
system oI equations always produces a unique solution. In contrast, nonlinear analysis does not guarantee a unique
solution. In Iact, there may be multiple solutions or no solution at all. The task oI providing analysis directives (i.e.,
controls by which the program will come to a solution) is Iar Irom simple. Solving nonlinear equations is an
incremental and iterative process.
A linear static mechanical analysis with a known external load can be perIormed in one step. II nonlinearities are
expected, it may be necessary to apply the load in increments and let each load increment iterate to the equilibrium
state, within a speciIied tolerance, using a particular iteration scheme such as Newton-Raphson. Also the complete
load history might consider oI a number oI load vectors, each applied at a speciIic time in the load history. Each (set
oI) loads to be applied in a speciIic time period can be considered as a loadcase. A job is then the subsequent
perIormance oI various loadcases. In this way, the complete loading history can be deIined. Note that a loadcase is not
necessarily identical to a load step. A loadcase may consist oI 10 load steps to reach the total load oI the loadcase. In
a loadcase, multiple boundary condition IDs can be present.
CONTACT BODIES: deIining the contact bodies, the properties oI the contact body and allowing a
graphical veriIication iI the bodies are deIined correctly.
CONTACT TABLES: deIining Ior which bodies contact will be checked, local Iriction coeIIicients, local
separation Iorces, and heat transIer coeIIicients. Also here it can be speciIied that the
so-called glued contact will occur, which implies automatic coupling oI diIIerent
parts.
CONTACT AREAS: deIining Ior which subset oI the nodes in a contact body contact will be checked.
168 Marc Users Guide: Part I
Introduction
A dynamic transient analysis oI a beam structure with pre-load P1 and dynamic load P2 using the modal superposition
technique consists oI the Iollowing loadcases:
Depending on the analysis type (e.g. mechanical, heat transIer), the LOADCASE processor on the analysis panel oI
the main menu allows you to speciIy the Iollowing:
Load incrementation i.e selecting the boundary conditions, the number oI steps,
automatic versus Iixed stepping, and the controls Ior this loadcase.
The JOBS processor is used to control the overall Ilow oI the analysis process. This includes the analysis class, the
selection oI the loadcases, the analysis options, the results which are required, the initial loads, contact control, and
other parameters. Also the element type speciIication, the check on integrity oI the job, and the actual submitting oI
the job is done in this processor.
Typically, the Iinite element analysis produces an enormous volume oI numerical data. BeIore you submit the job Ior
analysis, use the JOB RESULTS processor to control which variables are to appear in the results Iile beyond the
deIault parameters associated with the analysis type.
BeIore you submit the job, it is advisable to perIorm an integrity veriIication to check Ior inconsistencies in
geometrical and material properties. The program automatically veriIies the determinant oI the Jacobian Ior all
elements in the mesh. Errors Iound during this process are reported and corrective action should be taken beIore the
job is submitted.
Once the data is veriIied by the program and passes the validity test, the job may be submitted. The SUBMIT button
initiates the job in the background and leaves the terminal Iree to do other tasks. Use the UPDATE or MONITOR button
to monitor the progress oI the job during execution.
Results Interpretation
Once you have completed the analysis, you need to analyze the results and veriIy the criteria Ior acceptance. For each
increment, the requested results are stored in a sequential Iile. Use the Iollowing thre basic steps to gain access to
the results.
The RESULTS processor on the postprocessing panel gives you access to the various plot options available in Mentat.
Loadcase 1: Apply pre-load P1.
Loadcase 2: PerIorm eigenIrequency analysis based on pre-stressed structure
Loadcase 3: PerIorm transient analysis using superposition oI eigen modes. The load P2 is deIined as a Iunction
oI time through the TABLE option. Each loadcase can have diIIerent control values Ior the iterative
processes used.
Step 1 Open the results Iile.
Step 2 Select the desired inIormation.
Step 3 Select an appropriate display technique and display the results.
169
CHAPTER 1
Introduction
As we have already mentioned, a typical nonlinear Iinite element analysis consists oI several steps called increments.
The results Ior an increment can be accessed through the OPEN, NEXT, or SKIP sequence oI commands. OPEN
accesses the Iile and opens it Ior reading. The results Iile name is a concatenation oI the job name and the suIIix .t19
or .t16. NEXT Iorwards the Iile pointer to the next increment. The results data Ior the increment that was read by the
NEXT command is available Ior processing.
The solution oI the Iinite element analysis involves a geometrical discretization oI the object and, iI applicable, a
temporal discretization. The geometrical discretization is obtained by creating the Iinite element mesh that consists
primarily oI nodes and elements. The results (depending on their nature) are supplied at either the nodes or the
integration points oI the elements. We make the distinction by reIerring to one as data at nodes, and the other as data
from elements at integration points.
Data at nodes is a vector where the number oI degrees oI Ireedom oI the quantity indicates the number oI components
in the vector. Data Irom elements at integration points is either scalar, vector, or tensor data.
The data Irom elements at integration points are not in a Iorm that can be used directly in a graphics program. Data
Irom elements at integration points is extrapolated to the nodes thus creating data at nodes from elements. The values
at the nodes are calculated by a linear extrapolation oI the average centroidal value and the integration point closest to
the node.
A node may be shared by several elements. Each element contributes a potentially diIIerent value to that shared node.
The values are summed and averaged by the number oI contributing elements.
II a node is shared by elements oI diIIerent materials, the averaging process may not be appropriate. To prevent the
program Irom averaging values, use the ISOLATE option.
Scalar Plots
Scalar data may be represented graphically by means oI contour bands, contour lines, symbols, numerics, iso-surIaces,
cutting planes, beam contours, or beam values. A legend to the leIt oI the drawing shows the correspondence between
the colors used and the numeric interval they represent. Contour plots are lines or bands oI equal value drawn over the
elements. This display technique is applicable to two-dimensional elements, such as shells and plates, or to Iaces oI
three-dimensional elements, such as bricks. The three-dimensional counterpart to contour plot is the iso-surfaces plot,
where the surIaces oI constant value are displayed.
Vector Plots
Vector data may be represented graphically by arrows that are displayed at the nodes.
Tensor Plots
Tensor data may be represented graphically by arrows that are displayed at the centroid oI the elements.
Deformed Shape Plots
The deIormations Iound in a mechanical analysis can be shown in what is known to Mentat as a deformed shape plot.
The mesh is deIormed by an amount that is proportional to the actual displacement at the node.
170 Marc Users Guide: Part I
Introduction
Beam Diagrams
Display a shear diagram or a moment diagram Ior beam structures is with this option.
Path Plots
Path plots are snapshots created by Ireezing time or an increment. The variables Ior the abscissa and ordinate are
selected Irom the list oI available variable names. For path plots, the position where the quantity is evaluated is the
most likely candidate Ior the abscissa.
History Plots
As the name indicates, history plots capture phenomena over time or increments. The abscissa variable is very likely
to be time or an increment number. As Mentat keeps only one increment oI data in memory, it is necessary to collect
data by scanning over the range oI increments or time that is oI interest beIore the history can be displayed.
Getting Started
This section describes the routine interactions with Mentat listed below.
Starting Mentat,
Using the PROCEDURE option,
Stopping a Mentat session,
Recommended Starting Chapters.
This section concludes with a simple example to acquaint you with the program. It is best to Iocus on the overall
session and not to dwell on the details. Once you have mastered the basic steps described in this chapter, you should
Ieel comIortable enough with Mentat to venture on to the sample sessions in the remaining oI this manual.
Starting the Mentat Program
Before you start the Mentat program. . .
1. You will need an account so you can use Mentat on your system.
2. II you don't know how to invoke Mentat, ask a current Mentat user or call MSC.SoItware customer support.
Although the starting command is system dependent, it most likely is mentat.
On machines supporting OpenGL graphics, one would type: mentat -ogl.
3. The Mentat program is based on X-Windows; you must start the program in a window environment.
Assuming you are already logged in on your computer, type mentat at the prompt oI your operating system. Provided
your version has been installed correctly, once Mentat is loaded into memory, the program should start by opening a
window on your X-terminal. This displays the basic Mentat screen which consists oI a main menu, a blank graphics,
and a dialogue area. Figure 1.1-23 shows you the initial Mentat display.
171
CHAPTER 1
Introduction
II the Mentat script does not invoke the program or does not invoke it correctly, ask your system administrator or call
your nearest MSC.SoItware oIIice Ior support. Our telephone numbers are on the back cover oI this manual.
Figure 1.1-23 Initial Mentat Display
Procedure Files
A procedure Iile is a record oI all commands issued during a session and is useIul Ior the tasks listed below.
Protecting your work.
PerIorming repetitive operations.
Doing parametric design.
Demonstrating your work.
Reporting errors.
The PROCEDURES command has two modes oI operation:
1. Record mode: creates a new procedure Iile or appends an existing procedure Iile
2. Playback mode: partial or complete execution oI a procedure Iile
172 Marc Users Guide: Part I
Introduction
In both modes oI operation, a choice can be made iI the procedure Iile should reIlect the changes in the menus. II the
MENU RECORD button is activated, all changes in the menu are recorded while creating procedure Iiles. II the MENU
EXECUTE button is activated and iI a procedure Iile is used in which the menu changes have been recorded, the menus
modiIy while playing back the procedure Iile. Upon clicking on any oI the CREATE, APPEND, LOAD, or EXECUTE
buttons, a Iile browser appears. The FILTER block indicates the Iile extensions Ior the Iile type being used in the
current application (here a.proc extension). Either click an existing Iile in the FILES block, or type a new Iile name in
the SELECTION blocks Iollowed by an OK.
In playback mode, the LOAD button Iollowed by the STEP mode allows stepwise playback oI the procedure Iile.
(Observe that the Mentat Procedure Control window can be moved to any position oI the screen). START/CONT will
start the execution oI the procedure Iile until the STOP button is clicked.
All sessions listed in Section II through IV oI this manual are procedure Iiles that are included on the Mentat
installation CD in the examples/marc_ug directory.
You can play these sessions back by executing the procedure Iile using the Iollowing button sequence:
Remember to use <ML> to click on a button. AIter you click on the LOAD button, enter the Iile name oI the procedure
Iile you want to execute. Once you have done this STEP, observe the changes as the inIormation stored in the procedure
Iile is executed. START/CONT automatically continues until the either the STOP button is activated or until in the
procedure Iile the *stop_procedure command is present. Continue with the remaining inIormation with either
STEP or START/CONT.
An excellent way to learn more about the program is to make changes to the procedure Iile or to mimic it and to try to
predict the results.
On the previous page, you were introduced to the concept oI a button sequence diagram. A button sequence diagram
is a way oI prescribing a sequence oI mouse clicks and corresponding data entry. An indent indicates a new menu.
Aligned options indicate they are available Irom the same menu. A button sequence diagram starts at the main menu
and works its way to the desired option. Buttons in the static menu do not require you to start with the main menu.
The button sequence diagram is used Irequently throughout the remainder oI this chapter and in the sample sessions.
II there is any ambiguity as to which button you must click on, the button will be preceded by the speciIic panel or
menu title. For example, iI elems ADD appears in a button sequence, the idea is to click on the ADD button next to
'elems rather than nodes or curves on a particular panel. Another example is all: EXIST. which indicates that you
should click on the EXIST. button oI the 'all: panel.
UTILS (located at the bottom of the static menu)
PROCEDURES
LOAD (located on the PROCEDURE window)
path/filename
OK
STEP or START/CONT
173
CHAPTER 1
Introduction
II you are not at the main level beIore you execute the button sequence diagram, you can click on the MR~ with
anywhere over the menu area until you reach the main menu. You can also click on the MAIN button in the lower leIt
hand corner oI the menu area to return immediately to the main menu.
The initial state oI the program prescribes, wherever possible, a deIault Ior every setting. These settings are chosen
because they are applicable to most cases. For example, the deIault number oI divisions Ior SUBDIVIDE is set to 2, 2,
2. You can return to this deIault state at any time during the execution oI the program by clicking on the RESET
PROGRAM button. Use the Iollowing button sequence Irom the TOOLS tab:
When you create a procedure Iile, you are only recording commands that are issued Irom the time the procedure
was started. The procedure Iile does not contain inIormation on the state, or settings oI the program at the time it
was started.
Stopping the Mentat Program
Always make sure to save your work beIore you stop the Mentat program. Use the SAVE button to write a copy oI the
database in Mentat Iormat. The SAVE button is located in the static menu directly under the graphics area. This way
you are assured all data is saved. Using other Iormats such as the Marc Iormat does not guarantee all inIormation
is saved.
Normal Stop
Use the EXIT button located on the FILE pop-up to end a Mentat session. Be sure you save your model beIor exiting
the program. Alternatively, you can type *quit in the dialogue area Iollowed by a y Ior yes at the Exit program?
prompt at any time during a session.
Emergency Stop
An emergency stop can be made at any time by using CTRL-C (that is, hold down the CTRL key and press C) Irom
the parent window. Typing CTRL-C in the dialogue area does not stop the program. A host-induced stop usually does
not oIIer you much oI an option as you lose some or all oI the data in memory.
PROGRAM SETTINGS
RESET PROGRAM
< >
174 Marc Users Guide: Part I
Introduction
Recommended Starting Chapters
The subsequent chapters in this User`s Guide perIorm a variety oI simulations with varying levels oI diIIiculty; iI you
are just starting with Mentat, the table below recommends some oI the simpler starting problems keeping the model
complexity to a minimum while covering the basic Iunctionality oI Mentat. You might consider taking these chapters
in the order below since the later ones assume you have gained conIidence with changing views, turning node displays
oII, or rotating the model and the button sequences to perIorm simple manipulations are omitted.
175
CHAPTER 1
Introduction
Following a Sample Session
At this point, you may begin duplicating the Iirst sample session on your computer. Do not try to understand everything
at once; all concepts are explained as you progress through the subsequent chapters. For now, concentrate only on
becoming comIortable with the Mentat user interIace.
The structure you are going to model has the dimensions shown in Figure 1.1-24.
Analysis Type User Guide Chapter Problems: Recommended Starting Chapters
Static Chapter 3.20: Tensile Specimen
Static Chapter 3.18: Cantilever Beam
Static Chapter 3.21: Rubber Elements and Material Models
Static Chapter 3.9: Break Forming
Static Chapter 3.16: Super Plastic Forming (SPF)
Static Chapter 3.19: Creep of a Tube
Static Chapter 3.25: Interference Fit
Static Chapter 3.10: Hertz Contact Problem
Heat TransIer Chapter 4.4: Cooling Fin Analyses
Coupled Thermal Mechanical Chapter 5.3: Coupled Transient Cooling Fin
Coupled Thermal Mechanical Chapter 4.2: Dynamics with Friction Heating
Heat TransIer Chapter 4.3: Radiation with Viewfactors
Dynamics Chapter 3.35: Dynamic Analyses of a Cantilever Beam
176 Marc Users Guide: Part I
Introduction
Figure 1.1-24 Dimensions of Structure to be Modeled
The Iirst step is to type mentat. The MSC logo appears on your screen and is immediately replaced with a window
that displays the main menu.
Use the <ML> to click on the MESH GENERATION tab.
The next step is to establish an input grid to help you speciIy the nodes oI your model. Click on the EDIT button oI the
COORDINATE SYSTEM panel. The dynamic portion oI the menu is replaced by the set coordinate system menu where
the grid settings are located. In Mechanics of Mentat, we mentioned that the 'GRID button was a toggle button that
can be switched on or off. Click on the GRID button to turn the grid on.
The object you want to model has maximum dimensions 8 x 6 units. Click on the U DOMAIN button and use the
keyboard to enter 0 10 to set the grid size in x-direction.
The updated coordinate system menu with the numerals 0 and 10 appearing in the Ilat Iields next to the U DOMAIN
button will appear.
The spacing between the grid points does not need to be Iiner than 1 unit since all the corner points are at integer
distances Irom the origin.
Click the U SPACING button and type in 1. The program updates the menu accordingly as is shown in. Repeat the
steps Ior the V DOMAIN and V SPACING to set the values in y-direction.
Click on the FILL button to scale the picture to Iit the screen. The FILL button is located in the Tool Bar directly under
the graphics area.
The window will appear as shown in Figure 1.1-25.
Y
8 units
6 units
2 units
4 units
y
x
z
177
CHAPTER 1
Introduction
Figure 1.1-25 Scaling the Picture
Below Iollows a button sequence diagram oI the steps required to set the coordinate system that we discussed in the
previous pages. A comparison oI this button sequence to a detailed step description shows you that the button sequence
step Iormat is condensed and easy to Iollow.
MESH GENERATION
Grid Activate
Edit
U DOMAIN
0 10
U SPACING
1
V DOMAIN
0 10
V SPACING
178 Marc Users Guide: Part I
Introduction
To help you keep track oI the elements and nodes that you are going to create, you must label them. Click on the Fix
View Draw Setting pull down to control the plots.
The VIEW DRAW Setting panel determines whether or not an entity is drawn. With the LABEL panel, it can be indicated
iI the entity will be labeled. Click on the NODES button oI the LABEL panel. The NODES button is a toggle button; as
long as it is depressed, every node you create is labeled by its respective node number.
Similarly, click on the ELEMENTS button oI the LABEL panel. The ELEMENTS button is also a toggle and stays
depressed indicating that every element you create is labeled by the corresponding element number.
Turn Off Faces
Return to the mesh generation menu by clicking on the RETURN button located in the bottom leIt corner oI the menu
area or by clicking <MR> with the over the menu area.
To enter an element, click on the ADD button oI the ELEMS panel. The deIault element class is QUAD(4), a 4-noded
quadrilateral, which is the element type you are going to use Ior your model.
The program prompts you to enter Iour nodes. Look Ior the prompt in the dialogue area.
Pick the grid point at the origin oI the u-v system shown on the screen Ior the Iirst node oI a quadrilateral element.
Click the <ML> with the close to that grid point. The program conIirms the location oI the Iirst node with a small
square at the grid point and the node number 1 in the graphics area . The entry is conIirmed in the dialogue area with
node(0,0,0) at the Enter element node (1) : prompt.
To create Node 2, repeat the steps Ior Node 1 six units, or grid points, to the right oI the Iirst node.
Repeat this Ior Node 3 at a location two units above Node 2.
Finally, pick a location two units above Node 1 Ior the Iourth node.
The program draws the entire element. It includes a cross at the center oI the element and a halI-arrowhead on the Iirst
side oI the element in the direction oI the connectivity. The cross in the middle oI the element is the handle oI the Iace
oI the element; in 2-D, the Iace is the element itselI. II you need to pick this element, click the in the center oI the
element.
1
FILL
< >
< >
< >
179
CHAPTER 1
Introduction
Figure 1.1-26 Element 1 Completed
You do not need to click on the ADD button on the ELEMS panel again to enter a second element. As discussed in How
Mentat Communicates with You, until you explicitly instruct it otherwise, the program assumes you want to continue
the previous action: in this case, adding elements.
Pick Node 2 oI Element 1 Ior the Iirst node oI the second element. You will see this node light up on your screen.
The second node oI Element 2 is positioned two units to the right oI the Iirst node. Again, the program conIirms this
by drawing a square and the node number, Node 5, at that location.
The third node is positioned Iour units above Node 5. Click the <ML> on that particular grid point to create Node 6.
Pick the last node oI Element 2 so that it coincides with the third node oI Element 1. The program conIirms this pick
by highlighting the existing node. The connectivity Ior this element is complete and is conIirmed by the display oI the
entire element.
180 Marc Users Guide: Part I
Introduction
Figure 1.1-27 Element 2 Completed
Add the third element by repeating the same sequence oI steps. Pick Node 1 oI Element 3 to coincide with Node 4 oI
Element 1.
To create the second node oI Element 3, pick Node 3 oI Element 1 which also coincides with Node 3 oI Element 2.
Once again, the node will light up to conIirm it has been picked.
The third node oI Element 3 is positioned above the second node (Node 3). Use <ML> to pick this node by clicking on
the grid point that is two units above the previous one.
Complete the element by picking a grid point two units above the Iirst node oI this element (Figure 1.1-28).
181
CHAPTER 1
Introduction
Figure 1.1-28 Element 3 Completed
Turn the grid oII by clicking on the GRID button located on the Tool bar (Figure 1.1-29). The toggle returns to the
deIault released state.
182 Marc Users Guide: Part I
Introduction
Figure 1.1-29 Turning Off the Input Grid
Click on the FILL button in the Tool bar to scale the picture to Iit the screen. The picture was previously scaled by the
object and the grid (see Figure 1.1-29). Now that the grid is turned oII, the object occupies the entire graphics area.
Assume you want to subdivide the elements. Click on the SUBDIVIDE button to update the dynamic portion oI the
menu.
The DIVISIONS button shows the deIault values Ior subdivisions in the Iirst, second, and third direction. The Iirst
direction is deIined by the halI-arrowhead on the Iirst side oI each individual element. The element type you are using
in this model is a two-dimensional QUAD(4) element. Even though a two-dimensional element does not have a third
direction, its coordinate must still be entered. Click on the DIVISIONS button (using the <ML>) and type in 2 2 1.
Click on the ELEMENTS button to indicate that you are ready to subdivide elements using the current settings. The
program prompts you Ior the element list to be subdivided with the Iollowing string:
Enter subdivide element list:
183
CHAPTER 1
Introduction
Figure 1.1-30 Activating the Subdivide Processor
Use the <ML> to click on the handle oI each element to indicate that you want to subdivide it. Each element that you
click on will light up.
Once you have picked every element on the screen, click the <MR> with anywhere over the graphics area to
indicate an end of list to the program. Alternatively, you can click on the END LIST (#) button icon. All elements are
now subdivided. Instead oI picking all the individual elements, you can click on the all: EXIST. button to subdivide all
elements.
Notice the double nodes at the corners oI the original elements. The SWEEP processor eliminates duplicate nodes.
< >
184 Marc Users Guide: Part I
Introduction
Figure 1.1-31 After Elements were Subdivided
Note that there are 35 nodes, some oI which are duplicate nodes. Click on the NODES button on the SWEEP panel to
eliminate the duplicate nodes. Use the deIault value 0.0001, Ior the tolerance.
The program prompts you Ior the list oI nodes to sweep with the Iollowing string:
Enter sweep node list:
Click on the all: EXIST. icon to indicate that you want to sweep all nodes (Figure 1.1-32). The program removes the
duplicate nodes and displays the mesh with only 21 nodes remaining.
185
CHAPTER 1
Introduction
Figure 1.1-32 Identifying the Nodes to Sweep
Save the mesh by clicking on the SAVE icon. The mesh is saved in Mentat Iormat in a binary Iile called model1.mud.
If model1.mud already exists on the disk, the numeral in the Iile name is automatically incremented by one, and the
Iile name is thus called model2.mud.
Stop the session with the Iollowing button sequence:
You should now Ieel comIortable interacting with Mentat. We encourage you to practice with the example detailed in
this chapter, and to build on the experience you gained through this session.
FILE
EXIT
186 Marc Users Guide: Part I
Introduction
A Simple Example
In this section, it will be demonstrated how to set up the basic requirements Ior a linear elastic stress analysis. For this
purpose, a Ilat square plate with a circular hole subjected to a tensile load will be analyzed. It is generally known that
a stress concentration exists around the hole. Both the deIormed structure and the stress distribution need to be
determined. The goal oI the analysis is to demonstrate:
a simple mesh generation technique, using the geometric meshing approach
how to apply boundary conditions
how to set material properties
how to set geometric properties
selecting quantities to be calculated in the analysis Ior subsequent postprocessing
how to submit a job using the Marc Iinite element program
how to generate deIormed structure plots, contour plots, and path plots
Background Information
A square plate with dimensions 20 * 20 mm and a thickness oI 1 mm contains a circular hole with radius 1 mm at the
center oI the plate. The material behavior is assumed to be linear elastic with Young`s modulus E 200000 N/mm
2

and Poisson`s ratio v 0.3. A tensile load with magnitude p 10 N/mm
2
will be applied both at the top and the bottom
oI the plate.
Calculate the deIormed structure and determine the yy-component oI stress along the cross-section near the hole.
Figure 1.1-33 Plate with a Hole Subjected to Tension
20
2
0
t = 1
X
Y
1
10 N/mm
2
10 N/mm
2
187
CHAPTER 1
Introduction
Due to the symmetry oI the problem, it is suIIicient to analyze only a quarter oI the problem. At the line x 0 and y 0
symmetry boundary conditions have to be applied.
Figure 1.1-34 Quarter of the Plate with Symmetry Conditions and Tensile Load
Overview of Steps
Step 1: Mesh generation
Step 2: Boundary conditions
Step 3: Material behavior
Step 4: Geometric properties
Step 5: Job definition
Step 6: Postprocessing
Detailed Session Description
Step 1: Mesh generation
The applied approach Ior generating the model is to use the geometrical technique to speciIy the boundary curves and
the surIace spanned by these curves. Subsequently, the surIace will be converted into Iinite elements.
As in the sample session in Following a Sample Session, the Iirst step Ior building the mesh is to establish an input
grid. Click on the MESH GENERATION tab.
Next click on the SET button to access the coordinate system menu where the grid settings are located. Use the
Iollowing button sequence to set the horizontal and vertical grid spacing to 1 and both the horizontal and vertical grid
dimensions to 10.
10 N/mm
2
Y
X
188 Marc Users Guide: Part I
Introduction
Two geometrical entities are used to describe the boundary contour. First, set the curve type to a circular arc and deIine
the arc segment.
In the graphics window, a circular arc will now be visible. Change the curve type subsequently to a polyline.
MAIN MENU
MESH GENERATION
SET
U DOMAIN
0 10
U SPACING
1
V DOMAIN
0 10
V SPACING
1
GRID ON (on)
FILL
MAIN MENU
MESH GENERATION
CURVE TYPE
CENTER/POINT/POINT
RETURN
CRVS ADD (pick the following points from the grid)
0 0 0 (center point)
1 0 0 (starting point)
0 1 0 (ending point)
MESH GENERATION
CURVE TYPE
POLY LINE
RETURN
CRVS ADD (pick the following points from the grid)
point(10,0,0) (or type in POINT(x,y,z))
point(10,10,0)
189
CHAPTER 1
Introduction
Figure 1.1-35 Boundary Curves of Quarter of the Plate
The two basic curves will now be used to describe a ruled surIace. Set the surIace type to ruled and speciIy the
both curves.
point(0,10,0)
END LIST (#)
FILL
MESH GENERATION
SURFACE TYPE
RULED
RETURN
190 Marc Users Guide: Part I
Introduction
Figure 1.1-36 Surface Definition
With the CONVERT processor, the surIace are converted to Iinite elements. In the CONVERT menu, it can be
observed that by deIault the mesh division are set to 10 by 10 elements. The BIAS FACTORS will be used to ensure
that the mesh is more reIined in the direction oI the hole. The Iirst surIace direction is along the arc; the second surIace
direction runs Irom the arc to the polyline. A negative bias Iactor will be speciIied here, indicating that the reIinement
must be near the hole. Now convert the surIace to a Iinite element mesh.
SRFS ADD
1 (pick the arc)
2 (pick the polyline)
MESH GENERATION
CONVERT
191
CHAPTER 1
Introduction
Figure 1.1-37 Generated Element Mesh
The arrows near the element edges indicate that the elements are numbered clockwise. Marc requires that planar
elements are numbered counter-clockwise. The numbering can be changed using the UPSIDE DOWN and FLIP
ELEMENTS options in the CHECK menu.
BIAS FACTORS
0 -0.5
SURFACES TO ELEMENTS
1 (pick the surface)
END LIST (#)
DEACTIVATE GRID (off)
FILL
192 Marc Users Guide: Part I
Introduction
While checking, all elements with incorrect numbering are put in the temporary selection buIIer, which is graphically
shown by a change oI color. ThereIore, the list oI elements that need to be Ilipped can easily be speciIied using the
ALL: SELECTED icon. Repeating the check will show that no upside-down elements are Iound anymore so that the
temporary selection buIIer will be empty again.
Step 2: Boundary conditions
The symmetry conditions can be applied using the Iollowing button sequence:
MESH GENERATION
CHECK
UPSIDE DOWN (this will automatically select upside down elements)
FLIP ELEMENTS
ALL: SELECTED
UPSIDE DOWN (no elements should be selected - they have all been fixed)
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
OK
NODES ADD (box pick the nodes on the line x=0)
END LIST (#)
NEW
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
OK
NODES ADD (box pick the nodes on the line y=0)
END LIST (#)
193
CHAPTER 1
Introduction
Figure 1.1-38 Applied Boundary Conditions at the Lines x=0 and y=0
The applied loading is a tensile edge load with magnitude 10. Mentat allows to prescribe a distributed pressure on
element edges. The Iollowing button sequence will give the prescribed loading.
BOUNDARY CONDITIONS
MECHANICAL
NEW
EDGE LOAD
PRESSURE
-10 (specify pressure)
OK
EDGES ADD (box pick the edges on the line y = 10)
END LIST (#)
194 Marc Users Guide: Part I
Introduction
A graphical veriIication oI the applied edge loading is now obtained.
Figure 1.1-39 Boundary Conditions and Edge Loading
Step 3: Material behavior
The material behavior is identical Ior all elements. Elastic behavior with Young`s modulus and Poisson`s ratio must
be speciIied Ior this material. The Iollowing button sequence IulIills the requested task. Note that aIter entering the
Young`s modulus, Mentat automatically requests Ior the Poisson`s ratio, which can subsequently be entered. AIter
entering this value, the mass density is requested. By entering a <CR>, this sequence may be stopped.
AIter speciIying the list oI elements, the material description is complete.
MAIN MENU
MATERIAL PROPERTIES
MATERIAL PROPERTIES
195
CHAPTER 1
Introduction
Figure 1.1-40 Material Properties
NEW
STANDARD
STRUCTURAL
YOUNGS MODULUS 200000.
POISSONS RATION 0.3
OK
ELEMENTS ADD
ALL: EXIST.
196 Marc Users Guide: Part I
Introduction
Step 4: Geometric properties
Many elements require geometrical properties such as cross-sectional areas Ior beams and thickness Ior plate and shell
elements. For this plane stress analysis, the thickness must be speciIied Ior all elements.
Step 5: Job deIinition
All ingredients Ior a linear elastic static analysis are now created. No incremental steps are required nor does the
loading consist oI various load vectors. ThereIore, entering the LOADCASES menu is not necessary.
In the JOBS menu, Iirst set the analysis class to MECHANICAL, indicating that a stress analysis will be perIormed. In
the pop-up menu Iirst select JOB RESULTS. Here, the analyst has to speciIy which element quantities have to be
written to the post Iile. For simplicity, the Iull stress tensor is selected. Alternatively, all requested components oI the
stress tensor can be selected. (Note that the stress tensor writes 6 components to the post Iile; three oI them are zero
Ior a plane stress element).
In the INITIAL LOADS menu, it can be veriIied iI all boundary conditions (symmetry conditions and edge load) are
active as initial loads. The initial loading is the complete loading Ior a linear elastic analysis. Loading histories or
diIIerent loading steps require the use oI the LOADCASE option. The INITIAL LOADS screen must contain the
Iollowing below:
GEOMETRIC PROPERTIES
PLANAR
PLANE STRESS
THICKNESS
1
OK
ELEMENTS ADD
ALL: EXIST.
JOBS
NEW
NEW JOB: MECHANICAL
PROPERTIES
JOB RESULTS
AVAILABLE ELEMENT TENSORS
Stress (select stress)
OK
INITIAL LOADS
OK (twice)
197
CHAPTER 1
Introduction
Figure 1.1-41 Initial Loads
Next, the element type will be set. In this analysis, the 4-noded plane stress element type 3 with Iull integration will
be used Ior all elements.
JOBS
ELEMENT TYPES
ANALYSIS CLASS
USE CURRENT JOB ON (MECHANICAL), OK
ANALYSIS DIMENSION
PLANAR
SOLID
3 (full integration, QUAD(4))
198 Marc Users Guide: Part I
Introduction
Figure 1.1-42 Selecting the Element Type
The job is now ready Ior submission. The RUN command controls the submission oI the job. One oI the SUBMIT
commands starts the analysis, and with the MONITOR option, the current status oI the job can be observed.
OK
ALL: EXIST.
JOBS
RUN
SAVE MODEL
SUBMIT (1) ( this will automaticall monitor the job)
199
CHAPTER 1
Introduction
Figure 1.1-43 Screen After Job Completion
Step 6: Postprocessing
Once the job is complete, we are able to do postprocessing. The postprocessing tasks are perIormed on the Marc post
Iile which does not contain all the inIormation Irom the Mentat database. ThereIore, it is always recommended to save
the database beIore doing any postprocessing.
The Marc post Iile contains an analysis header (containing the mesh topology inIormation) and the results oI the
various increments. ThereIore, NEXT INC must be used aIter opening the post Iile to view the results oI the Iirst
increment (usually increment 0). Clicking the DEF & ORIG button only does not seem to show any deIormation on the
screen. The magnitude oI the displacement is simply too small Ior any visual eIIect. (By deIault the displacement will
be added with multiplication Iactor 1 to the original coordinates to get the deIormed structure). Automatic scaling will
show the requested deIormed structure.
FILE
OPEN DEFAULT
200 Marc Users Guide: Part I
Introduction
Figure 1.1-44 Deformed Structure Plot
Contour plots oI stress or displacement components can be made with the continuous contours or the contour bands
option. First, the quantity to be contoured has to be selected Iollowed by clicking the CONTOUR BANDS option.
RESULTS
MODEL PLOT
DEFORMED SHAPE
STYLE: DEFORMED & ORIGINAL
FILL
RESULTS
SCALAR
Comp 22 of Stress
201
CHAPTER 1
Introduction
Figure 1.1-45 Contour Plot of yy Component of Stress
A plot oI the yy component oI stress along the cross-section near the hole can be obtained with the PATH PLOT option.
First, a node path has to be selected (Iollowed by an end list) and then the curve to be plotted can be speciIied. With
the SHOW MODEL option, the screen can be changed to display the model again.
OK
CONTOUR BANDS
RESULTS
PATH PLOT
NODE PATH (pick the first and last node at the line y = 0)
END LIST (#)
ADD CURVES
202 Marc Users Guide: Part I
Introduction
Figure 1.1-46 Path Plot
The results show that the plate has been analyzed correctly. Around the hole, a stress concentration Iactor oI about
three is present and the deIormed structure plot shows an to be expected deIormation Iield.
Close the post Iile and leave Mentat.
ADD CURVE
Arc Length
Comp 22 of Stress
FIT
SHOW PATH PLOT
SHOW MODEL (select SHOW MODEL to return to model view)
203
CHAPTER 1
Introduction
Input Files
The Iile ~/mentat/examples/marc_ug/s1/c1.1/linear_elastic_stress.proc is on your delivery media
or it can be downloaded by your web browser by clicking the link (Iile name) below.
File Description
linear_elastic_stress.proc Mentat procedure Iile to run the above example
204 Marc Users Guide: Part I
Introduction
Answers to Frequently Asked Questions
This section is an ongoing collection oI answers to Irequently asked questions that relate to the output oI results Irom
Marc and Mentat. A general description oI some oI the derived results available Ior each element is given in Marc
Volume A: Theory And User Information under the heading "Element InIormation". The speciIic results available Ior
each element are indicated in the Marc Volume B: Element Library under the heading Output of Stresses and Output
Of Strains. Details on obtaining speciIic results in the output Iile are available in the section entitled "Selective
Printout" in Marc Volume A: Theory And User Information. Further inIormation oI beams and bars may be Iound in
Chapter 4 oI the Marc Volume A: Theory And User Information.
Appendix A: Shape Function Interpolation presents a simpliIied description oI the deIinition and use oI shape Iunctions
in Iinite element analysis, as well as some important implications oI their use.
Appendix B: Finite Element Equilibrium discusses issues such as averaging and unaveraging oI element stresses in more
detail.
Appendix C: Coordinate Transformation) presents the equations necessary to transIorm stresses Ior the plane stress
case.
Appendix D: Principal Stresses (Plane Stress) presents the equations necessary to evaluate principal stresses Irom
component stresses.
Appendix E: Python Example (Max Stress Results) gives an example oI extracting the maximum stress Iound using a
Python script.
Appendix F: Python Example (Displacements at Nodes) gives an example oI extracting nodal displacements using a
Python script.
Consistent Units
MARC is unit independent and any system oI units may be used. It is essential, however, that all data input
quantities are speciIied using a consistent unit system. For example, iI N, m, and kg are used, then the model
dimensions must be speciIied in meters, the elastic modulus in N/m
2
and density as kg/m
3
. With respect to the
output oI results - iI N, m, kg have been used in the data input, the results will provide stresses in N/m
2
, Iorces
and reaction in N, displacements in m etc.
Rotations are output in radians. This includes angular velocity and acceleration results.
The units oI temperature being used in an analysis are dependent on what is speciIied in Jobs> Heat Transfer>
Job Parameters> Units & Constants . The deIault is Celsius.
When choosing a set oI units, an appropriate system should be used to avoid problems with numerical round-
oII. For instance using the unit oI meters to model a component that is only a Iraction oI a millimeter in
overall size would require that the distance between element nodes would be oI the order oI 1e-6. Similarly,
the use oI Mega-Tonnes would be preIerable when analysing a massive civil engineering structure compared
to kg.
205
CHAPTER 1
Introduction
Using a dimensionally 'consistent set oI units is essential in dynamic analyses to obtain meaningIul results.
For a static analysis without any body Iorce loading (using density), the requirement Ior consistency is
relaxed. In general, iI units are chosen Ior mass (m), length (l) and time (t), the measure used Ior Iorce must
dimensionally correspond with F m a ml/t/t.
The standard SI unit system is normally recommended; i.e., Newton, Metres, Kilograms. This complies by
deIinition with the equation FMa (N kg m/s) and is termed a consistent set oI units. II both sides are
divided by 1000, a Iurther set oI consistent units becomes apparent as (kN Tonnes m/s). Similarly,
multiplying both sides gives (N T mm/s).
Any consistent system can be used; Ior example:
N, Millimetres, Tonnes
KN, Metres, Tonnes
MN, Metres, kTonnes
Dyne, Gram, Centimetre
Poundal, Pound, Feet
An example oI nonstandard sets oI units would be (kN, Millimetres, Tonnes), (N, Millimetres, kg).
Evaluation of Stresses in Finite Elements
In a Iinite element analysis, the stresses are evaluated at positions in each element called 'Gauss points. These points
are usually not located at the element nodes but some distance into the element. These points are used to numerically
integrate system matrices like stiIIness, mass, surIace and volumetric loads. See the section entitled 'Marc Element
ClassiIications in Marc Volume B: Element Library Ior details on the exact positions oI these points Ior each element
class). Gauss point values are the most accurate results Irom a Iinite element analysis.
For general ease oI use, however, nodal results are required and in order to obtain them it is necessary to extrapolate
the Gauss point stresses to the nodes. This is perIormed by using the 'assumed element displacement (shape)
functions. These shape Iunctions are one oI the basic building blocks oI the Iinite element method and vary Ior each
element type. In the general case, there are linear and quadratic shape Iunctions. The Iormer is typically used Ior
corner-noded elements (e.g., 4-node quadrilaterals, 8-node bricks) and are oIten called 'linear or 'lower order
elements. The latter is used Ior midside noded elements (e.g., 8-node quadrilaterals or 20-node bricks) and oIten
termed 'quadratic or 'higher order elements. For the elements in Marc, see the section entitled 'Marc Element
ClassiIications in Marc Volume B: Element Library Ior details on the shape Iunctions used Ior individual element
classes. See Appendix A: Shape Function Interpolation Ior a brieI explanation oI both shape Iunctions and
extrapolation.
Shear Strains used in Marc
Marc uses Iull tensor components Ior the stress matrices when all components are available (e.g. continuum elements)
- except Ior the shear strains, where Marc uses a modiIied deIinition oI shear strain namely , instead
oI the actual tensor components . Since the strain tensor is symmetric, then . This nontensor Iorm

xy
c
xt
c
yx

c
xy

xy
2c
xy

206 Marc Users Guide: Part I

Introduction
comes Irom the equaiton

where is the shear modulus oI the material. Herein, we shall reIer to
as the shear angle. Shear angles are used through out Marc and are always given in the user subroutines.
Stress and strain measures must be considered when displacements become large, Ior example strain measures used in:
Large Displacement Green Lagrange
Updated Lagrangian Logarithmic
Nothing Engineering
To obtain a correct energy evaluation using the standard equation (area under the stress-strain curve), it is
necessary to multiply the correct stress and strain matrices together, otherwise they are not energy-conjugate and will
not give the correct result. The tensor stress and strain matrices are not energy-conjugate in this regard. The
incremental energy is evaluated as:
sum_i { sig(i)*deps(i) + dsig(i)*deps(i) }
where sig is the beginning increment stress, dsig is the incremental stress and deps is the incremental strain. i
ranges Irom 1 to the number oI stress/strain components.
Extrapolation/Averaging Tips in Mentat
There are three element extrapolation methods in Mentat (RESULTS> SCALAR PLOT SETTINGS>
EXTRAPOLATION) where:
Linear - the average oI the integration point values is calculated and placed at the centroid. Then Mentat
perIorms a linear extrapolation oI the values Irom the centroid through the integration point to the nodes.
Translate - the values at the integration points are simply copied to the nearest nodes. II there are Iewer
integration points than nodes, Mentat averages the values oI neighbouring integration points. When combined
with the Isolate Ieature (RESULTS>ISOLATE) to isolate just one element, this enables you to see the exact
integration point values produced by Mentat in history plots. No extrapolation is used.
Average - Mentat computes the average oI all the values at the integration points and assigns an equal value to
all the nodes. No extrapolation is used.
The accuracy oI the extrapolation procedure is dependent on both the presence oI a reasonably uniIorm stress Iield and
the type oI shape Iunction used in the element chosen. For instance, a high stress gradient across an element would be
more likely to extrapolate incorrectly; particularly iI a linear shape Iunction element is being used. Thus, although
equilibrium conditions will always be met, the stress distribution can be wholly inaccurate Ior an insuIIiciently reIined
mesh.
In a typical analysis, each node point will be connected to more than one element and, hence, an averaging process is
required to obtain a single value oI stress at this node (simply adding all the component stresses and dividing by the
number oI components) - this is the output obtained in Mentat when choosing 'nodal averaging on in SCALAR
PLOT> SETTINGS> EXTRAPOLATION. Setting 'nodal averaging oII will not perIorm this process. See Appendix
B: Finite Element Equilibrium Ior a more detailed description oI the nodal averaging process.
o
xy
G
xy
G
xy
1 2oc
207
CHAPTER 1
Introduction
In areas oI interest where stresses will be used as input to the design process, a contour plot oI averaged nodal stress
should very similar to one using unaveraged nodal results (a smooth transition across element boundaries). This would
indicate that the stress distribution in the structure is being simulated suIIiciently accurately. For other sections oI the
model which are not oI interest, a coarser mesh would normally be used and a comparison in these areas would
typically give signiIicantly diIIerent contour plots - the unaveraged contours appearing more like a patchwork quilt.
The diIIerence in contour values at the element boundaries indicating either that the mesh is too coarse or the element
type insuIIicient to simulate the stress variation adequately.
To turn on/oII nodal averaging, select RESULTS> SCALAR PLOT SETTINGS> EXTRAPOLATION
NUMERICS in Mentat always displays nodal averaged values, unless the ISOLATE Ieature is used
(RESULTS> ISOLATE)
The maximum and minimum values on the legend Ior a NUMERICS plot are obtained by looping over all the
nodes. This means that values at nodes that do not belong to any element (contact control nodes, Ior example)
do inIluence the min/max Ior NUMERICS plots
Contact control nodes can be excluded Irom the nodes that are being considered in the NUMERICS plot by
using
RESULTS> POST NODES REM
.the list oI nodes to remove can be generated easily:
The maximum and minimum values on the legend Ior a CONTOUR PLOT are obtained by looping over all the
nodes oI each element. This means that values at nodes that do not belong to any element (contact control
nodes, Ior example) do not inIluence the min/max Ior CONTOUR plots.
II all values in a CONTOUR plot are identical (say equal to v), the automatic range is set to (v, 1.01*v).
Nodal averaging in Mentat is independent oI what is currently visible on the screen. This means that when
contouring selected parts oI a structure, the elements that have been removed Irom the active list will still
make a contribution to the averaging process. Depending on the relative size oI the stress contributions Irom
the elements that are missing, the contours around the edge could have changed signiIicantly iI this was not
the case.
To suppress averaging Ior speciIic elements select RESULTS> SCALAR PLOT SETTINGS>
EXTRAPOLATION> ISOLATE ELEMENTS. This means that when contouring selected parts oI a structure, the
elements that have been removed Irom the active list will no longer make a contribution to the averaging
process. Depending on the relative size oI the stress contributions that are now missing, the contours around
the edge could change signiIicantly.
SELECT
SELECT BY
NODES BY ELEMENTS
ALL: EXIST.
POST NODES REM
ALL:UNSELEC.
208 Marc Users Guide: Part I
Introduction
Nodal averaging oI element results to the output file is carried out using a weighted averaging scheme based
on the angle spanned at a node oI an element. This should not have an eIIect Ior a perIectly regular mesh.
The CONTOUR CENT method in Mentat (RESULTS) is a Iurther way to obtain contour plots. SpeciIying this
command sets the results plotting style to element centroid contours. The element centroid value is
determined by the extrapolation method being used. II it is set to linear, the Gauss point stresses are
extrapolated to the centroid, averaged and displayed. II it is set to translate, the Gauss point values are directly
averaged and then displayed.
Both Mentat and Patran use the same Mentat libraries (when the attach option is used) to extract results. This
means that a comparison oI extrapolated results will be identical. When reading results Irom a post (t16) Iile
using the Import option, then Patran will use its own libraries and a slightly diIIerent algorithm to extrapolate
the Gauss point results.
It can look like gravity and volumetric loads are applying a moment on the mid-side nodes oI shell elements
49 and 72. This is a post processing issue, since Mentat uses the average oI the corner node values Ior the
midside node results - not only Ior displacements, but also Ior Iorces. When running the analysis without no
print, the midside nodes oI elements 49 and 72 have zero external Iorces
Stress Coordinate Systems
Search Ior 'Material PreIerred Direction or 'Material Property (Element) Coordinate Systems in Marc in
Marc Volume C: Program Input Ior Iurther details oI the coordinate systems available in Marc.
Every element type in Marc has a deIault orientation (that is, a deIault coordinate system or reIerence axis)
within which element stress-strain calculations take place. There are three coordinate systems used by Marc:
1. Global Coordinate System
For 2-D or 3-D solid continuum and composite solid elements the coordinate system is aligned, by deIault, with
the global XYZ coordinate system. This is not normally a problem Ior isotropic materials since every direction
is then a principal direction. For orthotropic materials, however, the material principal coordinates are seldom
aligned with the global coordinates. For this reason the Orientation command is used to deIine a 'preIerred or
'material coordinate system.
2. Local (or Marc) Coordinate System
For truss, beam, and shell elements, 'local element dependent coordinate systems are used by deIault. For
example, Ior shell element 72, the local coordinates are the surIace coordinates, where is
deIined by the Iirst edge direction (node 12). is in the normal shell surIace direction.
correspond to the local result directions. As in continuum problems, the Orientation command may
be used to deIine a preIerred coordinate system iI required.
3. User-deIined (or PreIerred/Material) Coordinate System
II the stress/strain results are required in coordinate axes that cannot be achieved through the Global or Local
systems, then the user-deIined PreIerred or Material coordinate system should be used. This coordinate system
is usually organized such that it coincides with the material principal axes and is deIined through the
Orientation command on an element basis. This is especially important Ior non-isotropic materials
v
1
v
2
v
3
, , ( ) v
1
v
3
v
1
v
2
v
3
, , ( )
o
x
o
y
o
z
, ,
209
CHAPTER 1
Introduction
(orthotropic, anisotropic, nlelast, or composite materials). Thus, the Orientation command determines the
relationship between the material or element coordinate system and the global coordinate system (or the
ply angle direction, iI composite). Essentially, the Orientation command deIines the axis system into which the
stress and strain results will be transIormed. For isotropic materials, this could be carried out as a
postprocessing calculation using SCALAR PLOT> SETTINGS> RESULTS COORDINATE SYSTEM (in
Mentat). For orthotropic materials assigned to a regular, quadrilateral mesh, this would also be the case. For
orthotropic materials assigned to non-regular shell meshes (quadrilateral or triangular), the orientation oI the
0
o
ply will not be consistent, and require the Orientation command to be used.
Consider the Iollowing examples, in which a 2-D and 3-D mesh is shown together with arrows deIining the element
x-axis direction:
Figure 1.1-47 panels 1 and 2 show a 3-D shell mesh, in which the deIault coordinate axes are, respectively the local
system (as deIined by the Iirst edge direction) and the global system; panels 3 and 4 show a 2-D solid mesh, in which
the deIault coordinate axes are, respectively the local system (as deIined by the Iirst edge direction) and the global
system.
One oI the main issues to consider Ior both isotropic and orthotropic materials is the diIIiculties in interpreting
results iI a local element axis is used and the local element axes are not aligned consistently. In Figure 1.1-47
panels 1 and 3, a in one element is in a diIIerent direction to the oI an adjacent element in many cases.
The remedy is to deIine a preIerred system Ior all elements. Contouring using global result values will not be
aIIected by the local axis deIinitions and, hence, quadrilateral and triangular elements may be mixed without
diIIiculty, as shown in panel 4. However, in the case panel 2 in which the shell surIace is curved, the stresses
will be diIIicult to interpret since the chosen stress component 'intersects the element plane. In such cases
direction independent stress measures such as principal or von Mises would be easier to interpret. In addition,
Ior thin shell elements where the stress output does not include out oI plane stress terms , Marc
will transIorm the stress state Ior panel 2 Irom the deIault local stress state to a global state which, Ior an
arbitrarily positioned element in space, will give the appearance oI through-thickness shear.
0
o
x
o
x
o
yz
o
zx
, ( )
210 Marc Users Guide: Part I
Introduction
Figure 1.1-47 Local (First Edge) and Global Orientations for 3D and 2D Elements
Composite Shells
In composite shells, the orientation oI the materials in each shell layer can vary Irom layer to layer. In this case, the
Orientation command is used to locate the ply angle direction in the shell element surIace. II the Orientation
command is omitted, the ply angle direction coincides with the axis. For each layer, additional ply angle
oIIsets Irom this ply angle direction are given in the COMPOSITE command. This allows the arbitrary orientation
oI a composite layup in shells with local coordinates. Consider a Iour element composite shell model in which a
1. First Edge Orientation 3D 2. Global Orientation 3D
3. First Edge Orientation 2D 4. Global Orientation 2D
0
0 v
1
0
211
CHAPTER 1
Introduction
uniIorm in-plane Iorce load is applied to one end. One oI the elements has been split into Iour triangles. The boundary
conditions are minimal to allow Iree movement. The red arrows show the direction oI the Iirst Orientation vector that
has been assigned the preIerred x-axis:
The material used is a Iour-ply layup oI a single orthotropic material with an angle to the preIerred x-axis oI 45
o
Ior
each ply, where layer one is at the bottom oI the stack shown below and the material properties Ior each ply are shown
on the right.
Red Arrows = 1
st
orientation vector
F
x
S =
x
A
x
------- 83
1
3
-- - =
F
y
A
y
------- 0 = S =
y
212 Marc Users Guide: Part I
Introduction
Clearly, this represents a isotropic material and will simpliIy the discussion oI the results to the uniaxial case. Plotting
the numerical results Ior the non-preIerred stresses in layer one presents a nonuniIorm picture because oI mixing oI
components since shells use element dependent coordinate systems, namely:
Four ply stack
Layer one
83
1
3
---
83
1
3
-- -
0
=
83
1
3
-- - 83
1
3
-- - 0
3
+ +
213
CHAPTER 1
Introduction
Whereas, iI the preIerred directions are used, the nodal averaging proceeds correctly by adding components in the
correct directions as shown below.
As shown below, the principal values oI the preIerred stress components are drawn in the principal direction.
See the sections entitled 'Output oI Strains or 'Output oI Stresses in the Marc Volume B: Element Library
to determine the deIault element orientation axis Ior speciIic elements.
83.3333
83.3333 83.3333 83.3333
83.3333 83.3333 83.3333
83.3333 83.3333
83.3333
job1
Comp 11 of Stress in Preferred Sys Layer 1
X Y
Z
83.333
83.417
83.500
83.583
83.667
83.750
83.833
83.917
84.000
84.083
84.167
job1
Principal Stress in Preferred Sys Layer 1 Major
Inc: 0
Time: 0.000e+000
X Y
Z
214 Marc Users Guide: Part I
Introduction
All results quantities can be transIormed in Mentat using the RESULTS> SCALAR PLOT> SETTINGS>
RESULTS COORDINATE SYSTEM. See Appendix C: Coordinate Transformation Ior an example oI Plane
Stress transIormation.
The preIerred coordinate system oI an element can be requested by selecting the element quantities 1st Element
Orientation Vector, 2nd Element Orientation Vector, and Ply Angle in the JOB RESULTS menu. The
latter should be selected Ior all layers oI interest.
II the element orientation vectors and the ply angle are available on the post Iile, then Mentat will display the preIerred
coordinate system as an orientation in postprocessing.
The stress in preIerred system is now transIormed to the global system and the principal directions are displayed
correctly. Furthermore, iI the preIerred system inIormation is not available on the post Iile, then the user will not be
able to plot the principal stresses in preIerred system anymore as they will be wrong.
A similar problem exists Ior the regular stress tensor Ior shell elements. This tensor is deIined in the local coordinate
system oI the element (which may be diIIerent Irom the preIerred system). At this point, Marc does not output the
directions oI this coordinate system to the post Iile, so Mentat cannot compute the principal directions correctly Ior
these elements. The principal direction will still be displayed incorrectly.
Material Axis Definition
In the presence oI either orthotropic or anisotropic material behaviour, the material axes may be deIined
accordingly. In general this transIormation is termed 'orientation
'Orientation is diIIerent Irom 'TransIormation (Iound in BOUNDARY CONDITIONS> TRANSFORMS).
The latter permits the transIormation oI individual nodal degrees oI Ireedom Irom the global direction to a
local direction through an orthogonal transIormation that Iacilitates the application oI boundary conditions
and the tying together oI shell and solid elements.
TransIormations are assumed to be orthogonal. Once you invoke a transIormation on a node, you must input all
loads and kinematic conditions Ior the node in the transIormed system. Nodal output is in the transIormed
system. This option is invoked using the TRANSFORMATION option. The UTRANFORM option allows
transIormations to be entered via the UTRANS user subroutine to change; Ior example, the transIormation with
each increment. When you invoke this option, the nodal output is in both the local and the global system. Note
that, when a node oI a deIormable body contacts a deIormable body, a multipoint constraint (called tying) is
automatically imposed. These can produce conIlicts with any other nodal transIormations and is not
recommended. A preIerred method to apply transIormations in this case would be to use a rigid line or surIace.
See the ORIENTATION option in Marc Volume C: Program Input Ior Iurther details on the data Iile syntax to
orientate the material axes.
Various user speciIied orientations may be speciIied Irom within Mentat using MATERIAL PROPERTIES
>ORIENTATIONS.
With the ORIENTATION option, it is possible to speciIy the orientation oI the material axes oI symmetry (or
the 0-ply angle line, iI composite) in one oI Iive diIIerent ways:
215
CHAPTER 1
Introduction
1. A speciIic angle oIIset Irom an element edge (in Mentat: edge12, edge23, edge34, edge31, edge41) Edge
types Ior 2-D continuum and shell elements. The direction vector along the speciIied edge is projected onto
the surIace tangent plane (x-y plane iI continuum, V1-V2 plane iI shell) at each integration point. The Iirst
preIerred direction is given by a rotation about the surIace normal (z axis iI continuum, V3 axis iI shell)
equal to the orientation angle. The third preIerred direction is given by the surIace normal, and the second
preIerred direction is given by the cross product oI the third and Iirst directions.
2. A speciIic angle oIIset Irom the line created by two intersecting planes (in Mentat: xy plane, zx plane, yz
plane). Global intersecting plane types Ior 2-D elements. The speciIied global coordinate plane is
intersected with the surIace tangent plane. The Iirst preIerred direction is given by a rotation about the
surIace normal Irom this intersection equal to the orientation angle. The third preIerred direction is given
by the surIace normal, and the second preIerred direction is given by the cross product oI the third and Iirst
directions.
3. A speciIic angle oIIset Irom the line created by two intersecting planes (in Mentat: xu plane, yu plane, zu
plane, uu plane). User-deIined intersecting plane types Ior 2-D elements. For types xu, yu, and zu plane, the
plane determined by the coordinate direction and a user-deIined vector is intersected with the surIace
tangent plane. For type uu plane, the plane determined by two user-deIined vectors is intersected with the
surIace tangent plane. The Iirst preIerred direction is given by a rotation about the surIace normal Irom this
intersection equal to the orientation angle. The third preIerred direction is given by the surIace normal, and
the second preIerred direction is given by the cross product oI the third and Iirst directions
4. A particular coordinate system speciIied by user-supplied unit vectors (in Mentat: 3d_aniso) 3-D type Ior
3-D elements. The Iirst and second preIerred directions are as given by the user. The third preIerred
direction is given by the cross product oI the Iirst and second directions.
5. Direct speciIication oI a local coordinate system. This is consistent with the CP identiIication number on
the MD Nastran GRID Bulk Data Entry. The coordinate systems are similar to the MD Nastran CORD1R,
CORD1C, CORD1S, CORD2R, CORD2C, and CORD2S options. Note that the data entered here should not
be changed upon restart (COORDINATE SYSTEM)
6. For hexahedral elements, a local element system deIined by element nodal connectivity may be used. This
system can be rotated around the three local axes. The Iirst preIerred direction joins the centroids oI Iaces
4-1-5-8 and 3-2-6-7. A second vector joins Iaces 1-2-6-5 and 4-3-7-8. The third preIerred direction is given
by the cross product oI the Iirst preIerred direction and this vector. This system is then rotated around the
three local axes by the three given angles (3D LOCAL)
7. One or more NURBS curves may be used to deIine the preIerred system. These curves must be deIined with
the CURVES model deIinition option and only the NURBS variant is allowed. Using the centroid oI the
element, the closest point on any oI the given curves is Iound. The Iirst preIerred direction is given by the
tangent vector at this point. For 2-D elements, the second preIerred direction is given by the cross product
oI the global z direction and the Iirst preIerred direction. For 3-D elements, this option is only supported
Ior solid shell elements and solid composite elements. The third preIerred direction is given by the thickness
direction and the second preIerred direction by the cross product between the third and Iirst preIerred
direction. The Iirst preIerred direction is recalculated as the cross product between the second and third
preIerred directions to insure that we have an orthogonal system (CURVES).
8. As speciIied by the ORIENT user subroutine (in Mentat: Usub Orient), the UORIENT user subroutine type
Ior all types oI elements. Marc will call UORIENT to obtain the transIormation matrix between global
coordinates (Ior continuum elements) or local coordinates (Ior beams, plates, or shells)
216 Marc Users Guide: Part I
Introduction
In addition to the preceding orientation options, the Iollowing Iacilities are available in Mentat:
1. Align: This command sets the type oI the current orientation to 3d_aniso and determines the user-deIined
vectors Irom the three speciIied points representing a cartesian coordinate system. This command should
only be used Ior orientations that are applied to 3-D elements
2. Rotate: This command rotates the current orientation by the speciIied rotations and should only be used Ior
orientations oI type 3d_aniso
3. Local: This command creates a separate orientation oI type 3d_aniso Ior each oI the speciIied elements and
determines the user-deIined vectors Irom the two speciIied points representing the axis oI a local coordinate
system. The resulting orientations are initialised to be aligned with the local coordinate system oI each
element, and then rotated about the local X, Y, and Z axes by the given angles. The speciIied elements
should be 3-D elements.
4. Cylindrical: This command creates a separate orientation oI type 3d_aniso Ior each oI the speciIied elements
and determines the user-deIined vectors Irom the two speciIied points representing the axis oI a cylindrical
coordinate system. The speciIied elements should be 3-D elements.
Gauss Point Results
To obtain Gauss point results in Mentat, select the TRANSLATE extrapolation method and switch oII nodal
averaging. When combined with the Isolate Ieature (RESULTS>ISOLATE) to isolate just one element, this
enables the exact integration point values to be displayed by MARC in history plots.
User subroutines such as PLOTV and ELEVAR can also be used to provide such results.
Selecting JOBS> JOB RESULTS> OUTPUT FILE> Full Element and Node Print in Mentat will give Gauss
point element results directly to the output Iile.
PRINT ELEM can be used to obtain a selective set oI elements.
Temperature results using post codes 9/180 (elements) are given at integration points. It is possible to select
Gauss point results Irom Patran also.
By deIault, iI the t16 Iile is attached, the standard extrapolation to the nodes as done by Mentat is available. II
on the translation parameter Iorm Gauss points are requested, then on attachment, the gauss point results are
accessed. UnIortunately, the limitation in Patran is that you can't swap back and Iorth between these without
detaching and re-attaching the results.
Selective Results to the Post File
In an analysis with large models, the post (results/t16/t19) Iile oIten becomes large, involves signiIicant
I/O and utilizes a large amount oI disk space. Post Iile version 13 has been implemented in Marc to reduce the
size oI the post Iile. Post version 13 was introduced in 2005r3. The deIault remains as version 12. This allows
Patran to be able to read the deIault post Iile. To select the version 13 post Iile Iormat in Mentat:
Jobs> Job Results> POST FILE = 2006 Style
217
CHAPTER 1
Introduction
To select the version 13 post Iile Iormat directly in the Marc data Iile, speciIy '13 in the 11th Iield oI the 2nd
data block oI the POST option. A zero in this Iield will give the deIault version 12 Iormat post Iile. Note that
only those elements or nodes selected will be available/visible when post processing, and care should be taken
to ensure all entities oI interest are selected
Using version 13 provides the Iollowing beneIits:
The option to select a subset oI elements, contact bodies and nodes Ior output to the post Iile via additional
options select element, select body and select node on the POST option. For example, the user can select only
the exterior elements oI contact bodies to be on the post Iile. It would also be possible to change the elements,
nodes, and postcodes Ior each loadcase independently.
II the post Iiles are used as input Ior Iurther analysis such as in jobs with PRE STATE, INITIAL STATE,
AXITO3D, and GLOBAL LOCAL options, the Iull model must be present in the post Iile. In Mentat, this may be
speciIied using both LOADCASE> LOADCASE RESULTS and JOB> JOB RESULTS.
The version 13 Iormat oI the post Iile is diIIerent Irom earlier versions since the element variables are stored
based on the element types. In version 12 and earlier, the maximum number oI integration points is used Ior
all elements, and postcodes are applied to all element types. In version 13, the true number oI integration
points Ior that element type is used. Post codes with layers apply only to layered elements and postcodes Ior
beams apply only to beam elements. This leads to signiIicant post Iile size reduction iI there are mixed
element types, postcodes with layers and postcodes Ior beams in the analysis
The reduction oI post Iile size depends on the element types and postcodes in the analysis. In one example using 2400
brick elements and 200 shell elements, stress tensors oI two layers are requested Ior the post Iile. In version 12, post
Iile size is 56 Mbytes. In version 13, it reduces to 36 Mbytes. II only exterior elements oI the contact bodies are
selected, the post Iile size Iurther reduces to 21 Mbytes.
Continuum and Generalized Stresses
There are two main types oI stress output Ior elements in Marc:
Continuum Stresses
Also known as 'physical stresses in Marc, these are named aIter the stress deIinitions used in classical continuum
mechanics theory and are deIined in the usual manner as Iorce per unit area and are denoted by the Iollowing symbol
types
The Iirst subscript Ior the shear stress component stresses represents the Iace on which the stress acts, the second is
the normal direction oI this Iace. Hence, the stress is acting on the x-Iace in the global y-direction. The Marc
deIinition is given as Iollows:
o
x
o
y
o
z
o
xy
o
yz
o
zx
, , , , ,
o
x
o
y
o
z
t
xy
t
yz
t
zx
, , , , ,
S
x
S
y
S
z
S
xy
S
yz
S
zx
, , , , ,
o
xy
218 Marc Users Guide: Part I
Introduction
Normal Stress values:
The shear stress in two perpendicular cutting directions at a point have the same value, i.e.
The stress tensor comprises the 6 independent stress values Irom 3 perpendicular cutting planes at a point
and describes the stress behavior completely: where .
Further to the in-plane Gauss point evaluation, shell elements also evaluate stresses at a number oI points
through their thickness. This is carried out in order to obtain bending and membrane eIIects as a result oI
increasing distance Irom the neutral axis. These through-thickness stress evaluation points evaluate what are
commonly termed 'surIace, 'layer or 'Iibre stress results.
Y
X

XY

XY

X,

Y
+ve tension

XY
+ve into XY quadrant

Y
X
Z

XZ

YZ

XY

YZ

XZ

XY

X
Arrows indicate +ve
stress directions
o
xy
o
x

o
yy
o
y

o
zz
o
z

o
xy
o
yx

o
xz
o
zx

o
yz
o
zy

o
ij
o
ij
o
xx
o
xy
o
xz
o
yx
o
yy
o
yz
o
zx
o
zy
o
zz
o
ij
o
ji

219
CHAPTER 1
Introduction
Numerical integration through the shell thickness is perIormed using Simpson`s rule Ior homogeneous shells
and the trapezoidal rule Ior a composite shell. The number oI equally spaced integration points (layers) can be
deIined Ior homogeneous shells via the data Iile SHELL SECT parameter or using JOBS> JOB
PARAMETERS> # SHELL/BEAM LAYERS in Mentat. For nonhomogeneous materials, the number oI layers is
deIined through the COMPOSITE data Iile parameter. This number must be odd. Seven points are enough Ior
simple plasticity or creep analysis. Eleven points are enough Ior complex plasticity or creep (Ior example,
thermal plasticity). For linear material behavior, three points are suIIicient. The deIault is 11 points
For shell and beam elements, the layers at which stress results are required may be speciIied in JOBS> JOB
RESULTS> SELECTED ELEMENT QUANTITIES> LAYERS. The options Irom within Mentat are:
1. All: This command activates all element layers Ior a selected post tensor. Values Ior this tensor will be
written at all layers.
2. Out & Mid: This command activates the outer and middle element layers (top, middle, and bottom) Ior a
selected post tensor.
3. List: User speciIied layer numbers at which the selected post tensor should be output
4. Default: This sets the deIault layer Ior the selected post tensor - which, Ior shells, is the mid-plane (see the
Mentat User Guide command post_tensor_default_layer). Furthermore, the results will be generalised
stress/strain measures.
Shell Layer Convention: The layer number convention is such that layer one lies on the side oI the positive
normal to the shell, and the last layer is on the side oI the negative normal. The normal to the element is based
upon both the coordinates oI the nodal positions and upon the connectivity oI the element:
The local element directions are deIined Irom the element numbering with the local X-axis deIined Irom the
Iirst element node to the second. The element numbering can be obtained Irom the output Iile or Irom the
MESH GENERATION> ELEMENT> SHOW command (it will print the element node numbering to the Mentat
command window). The local Y-axis is at right angles in the plane oI the element surIace and the local Z-axis
is the normal direction according to the right hand rule.
X
Z
Y
Top Layer
(layer 1)
Bottom Layer
(layer "n")
Local
Element Axis
220 Marc Users Guide: Part I
Introduction
The orientation oI the local z-axis can be displayed by using MESH GENERATION > CHECK > ID
BACKFACES. This presents a contoured display indicating the orientation oI this element normal. II both the
surIaces and their underlying elements are displayed, the surIace contours will be visualised hence turn oI
the surIace display (PLOT> SURFACES.) to see the element orientation.
Back/Outside reIers to the bottom surIace, Front/Inside reIers to the top surIace and Front reIers to the top
surIace. The top surIace corresponds to layer 1 (most positive normal direction). The bottom surIace to layer n
(most negative normal direction). The middle layer Ior a seven layer shell will be layer 4, whilst Ior a Iive
layer shell, layer 3. For Nastran shell elements, the bottom surIace is layer 1.
Middle surIace stresses represent membrane action only. For Example top and bottom surIace stresses
additionally include the eIIects oI bending stress.
II you have a mixture oI elements on a shell that may not have been created at the same time, then there is a
possibility that some local z-directions will point in opposite directions. In this case, use the MESH
GENERATION > CHECK > ALIGN SHELLS (pick the reference element and Mentat will make all adjoining
elements to have the same outward normal by changing the node orientation) to get them all pointing in the
same direction. Aligning shells does not align the in-plane x/y-axes. Aligning shells does not take account oI
visible groups and will always work on all the analysis elements.
This choice is not available iI the shell biIurcates, but it is possible subsequently to Ilip elements` Ior groups
that may be wanted in diIIerent directions. The flip element command works by spinning the element through
180 about a line parallel to the element node 1-2 line that passes through the element centroid.
With composite elements, the layer numbering Iollows the same rules; however, the stress recovery points are
at midlayer. II a uniIorm material is represented as a composite material, the surIace stresses will not be
recovered.
For plate elements, not having the additional membrane stress terms, this layered approach is not necessary and
continuum stress output is computed at the neutral axis only - giving a single set oI stresses.
Generalised Stresses
Generalised stresses are also known as 'average membrane stresses and are available Ior shell elements.
They are dimensionally deIined as a stress. They are not to be conIused with 'stress resultant output, deIined
as Iorce or moment per unit width
Generalised stresses are obtained at each Gauss point by integrating the shell 'layer continuum stresses over
the thickness oI the shell. In the same way, generalised strain are their work-conjugate counterparts and
examples include curvature and rotation
More inIormation on how the generalised stresses and strains are evaluated can be obtained Irom the 'Shell
Elements section oI Marc Volume A: Theory And User Information (search Ior 'Element InIormation). The
equation given Ior the generalised stress is:
o
G
o y d
t 2
t 2
}

221
CHAPTER 1
Introduction
Where is the generalised stress and corresponds to the through-thickness direction. To see what this
means a little more clearly:
Where , and are the stress, thickness and Iorce associated with layer 1. is the shell width. The
quantity is the Iorce per unit width Ior the layer i.
Generalised stresses and strains are only available Irom the Marc output Iile.
Result Types
Search Ior 'Element Post Codes and 'Nodal Post Codes in Marc Volume C: Program Input Ior details oI the
element/node-based results available Irom a Marc analysis.
Stress (post code 311)
Element-based results.
II an updated Lagrangian analysis is being carried out, then this will be the Cauchy stress. II a total
Lagrangian analysis, this will be the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Post Codes 11-16 give generalised stress quantities iI no layer number is speciIied Ior a shell analysis. In this
case, the Mentat display can be conIusing, because Mentat still uses comp11, comp22 etc. II a layer number is
given, these are physical layer quantities (i.e. local continuum stresses).
This stress measure may be local or global, depending on the element type. For shells and beams, it will be based on
the local element coordinate system whilst Ior continuum it will be global.
Cauchy Stress (post code 341)
Element-based result. Also known as the 'true stress.
The stress is thus given in terms oI current area and current deIormed geometry (Iorce per unit deIormed
area). As a result, it is the most naturally understood stress measure, that is, it most naturally describes the
material response.
II a total Lagrangian analysis is being carried out, then this will be evaluated accurately Irom the 2
nd
Piola-
KirchhoII stresses used by preIerence in a total Lagrangian analysis. See later section on Updated and Total
Lagrangian solutions.
This stress measure may be local or global, depending on the element type. For shells and beams, it will
usually be based on the local element coordinate system whilst Ior continuum it will be global.
Its corresponding True/Log strain measure is also available using post code 681.
o
G
dy
o
G
o y d
t 2
t 2
}
1
t
---
f
i
lt
i
-----
\ .
| |
t
i
i

~
o
1
t
1
, f
1
l
fi
l
---
222 Marc Users Guide: Part I
Introduction
Stress in Preferred System (post code 391)
Element-based result.
Components oI stress in the user 'preIerred coordinate system deIined by the ORIENTATION option
(MATERIAL> ORIENTATIONS in Mentat). This is a material characteristic and may be modiIied on this basis
throughout the mesh.
II an Updated Lagrangian analysis is being carried out, then this will be the Cauchy stress. II a Total
Lagrangian analysis, this will be the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
II no ORIENTATIONS are deIined, the stress measure may then be local or global, depending on the element
type. For shells and beams, it will be based on the local element coordinate system; while Ior continuum, it
will be global.
Global Stress (411)
Element-based result.
Stresses in global (X,Y,Z) directions.
II an Updated Lagrangian analysis is being carried out, then this will be the Cauchy stress. II a Total
Lagrangian analysis, this will be the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
For post codes 411 (and 421, 431, 441), global quantities Ior shell elements are reported Ior as many layers as
requested and the same layer numbering system is used as Ior regular shell quantities. Layer 1 is the top
surIace; layer 2 is the next surIace, etc. This convention is Iollowed Irom Marc 2000 on.
UseIul measure Ior shells that are not regularly aligned particularly in the case oI triangular elements.
See 'Interpretation oI results Ior issues regarding shell global stresses.
Default Stress
Element-based result.
Stresses in the element directions (i.e. Ior shells their local directions and global directions Ior continuum
elements).
Tresca Intensity
Element-based result.
This is deIined in Chapter 12 oI Marc Volume A: Theory And User Information and is the maximum shear
stress obtained Irom Mohr's circle. There is no post code Ior this quantity and is not available in the post Iile at
present. The PLOTV user subroutine would be needed to obtain this quantity in the post Iile. It is evaluated
and printed to the output Iile when stresses are requested.
Note: When the LARGE STRAIN parameter is used, Cauchy stress in the PreIerred system is automatically shown
in the Post File Ior all cases Ior post code 391 - i.e., Ior both Total Lagrangian and Updated Lagrangian.
223
CHAPTER 1
Introduction
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Normal Stress
Element-based result.
At the outer surIace oI the structure the normal stress can be evaluated Irom the stress tensor using the normal
vector to the surIace .
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Shear Stress
Element-based result.
At the outer surIace oI the structure the shear stress can be evaluated Irom the stress tensor using the
tangential vector to the surIace
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Normal Total Strain
Element-based result.
At the outer surIace oI the structure the normal strain can be evaluated Irom the strain tensor using the normal
vector to the surIace .
II an Updated Lagrangian analysis is being carried out, then this will be based on the log (true) strain. II a
Total Lagrangian analysis, it will be based on the Green-Lagrange strain. See later section on Updated and
Total Lagrangian solutions.
Shear Total Strain
Element-based result.
At the outer surIace oI the structure the shear strain can be evaluated Irom the strain tensor using the
tangential vector to the surIace .
II an Updated Lagrangian analysis is being carried out, then this will be based on the log (true) strain. II a
Total Lagrangian analysis, it will be based on the Green-Lagrange strain. See later section on Updated and
Total Lagrangian solutions.
o
ij
*
n
j
( )
o
ij
*
t
j
( )
c
ij
*
n
j
( )
c
ij
*
t
j
( )
224 Marc Users Guide: Part I
Introduction
Mean Normal Intensity
Element-based result.
This is deIined in Chapter 12 oI Marc Volume A: Theory And User Information and is otherwise known as
'mean normal stress. It is the trace oI the stress tensor divided by three (i.e. the hydrostatic pressure part oI
the stress).
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Equivalent Mises Stress/Equivalent von Mises stress (post code 17)
Element-based result.
This and other related issues are discussed in a separate technical note 'Equivalent Stresses.
This is alternatively called the 'von mises intensity and corresponds to the 'mises intensity obtained in the
output Iile. It is described in the section entitled 'Element InIormation in Marc Volume A: Theory And User
Information.
The 'equivalent von Mises stress is evaluated in Marc during the solution and passed to Mentat as Gauss
point results, where the values are linearly extrapolated to the nodes and then averaged (usually). This
extrapolation can result in negative values and is a sign oI poor mesh quality in that area.
Mentat also provides an 'equivalent stress measure. This uses the same equations as Ior post code 17, but is
computed directly Irom the stress tensor during post processing by extrapolating the Gauss point stress tensor,
averaging to obtain a nodal value and then evaluating the equivalent stress in the usual manner. Small
diIIerences may be seen between this measure and postcode 17 in some circumstances.the diIIerence is
broadly that the equivalent von Mises stress (postcode 17) is a derived-extrapolated result whilst the
'equivalent stress is an extrapolated-derived result.
The diIIerences should be minimal when the 'translate extrapolation method is used, as expected.
Equivalent Stress/yield (post code 59)
Element-based result.
The initial yield stress is used, not the current. The values will be in the range 0-1 Ior un-yielded material and
greater than 1 Ior yielded material. For temperature dependency (where so is normally speciIied as unity, this
will give the equivalent stress.
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Equivalent Stress/yield at Current Temperature (post code 59)
Element-based result.
The initial yield stress is used, not the current. In this case, account is taken oI the temperature dependency.
The values will be in the range 0-1 Ior un-yielded material and greater than 1 Ior yielded material.
225
CHAPTER 1
Introduction
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Effective Plastic Strain
Element-based result.
There are two ways oI using the equivalent (von Mises) plastic strain.
a. 'Total Equivalent Plastic Strain (post code 7): The integral oI equivalent plastic strain rate
b. 'Current Equivalent Plastic Strain (post code 27)
Both are evaluated using the same equation .
a. represents the integration oI the incremental equivalent plastic strain values oI the analysis time period. This
is what the plasticity routines use since plasticity is, by deIinition, a history-dependent process as seen in
the incremental nature oI the numerical Ilow rules. This measure may increase (or remain constant)
throughout the analysis. It will not decrease. Broadly, it captures the maximum value oI the von Mises
plastic strain measure provided by the continuously changing combination oI the direct plastic strain
components at each increment.
b. is the equivalent value oI the plastic strain tensor at a particular increment. This may increase or decrease
over the analysis - especially so in the case oI cyclic loading. The reason Ior this is due to the negative cross-
terms in the equation. It is hard to think oI a useIul application Ior this result quantity. Note that the plastic
strain components themselves will not decrease - only the scalar result Irom this von Mises equation.
Nastran does not output this measure.
For example:
where is the 'total equivalent plastic strain and , the 'equivalent plastic strain. Also,
, the incremental equivalent plastic strain.
As is always positive, will always increase and never reduce, whilst EPS can increase
and decrease. II a simple bar tension and compression test is carried out, will be seen to go Irom 0
to 1 and then to 2, whilst goes Irom 0 to 1 and back to 0. measures the accumulated total
plastic deIormation regardless iI it is in compression or tension.
In general it can be said that the 'current measure will never rise above the 'total measure.
The eIIective plastic strain output Irom Nastran is a 'total quantity. It is an accumulative scalar that accounts
Ior the plastic Ilow during the analysis. Thus, to compare the 'current scalar result oI the von Mises strain
equation using total strain components with that Irom the 'accumulated von Mises strain equation with
plastic components is not appropriate.
dc
p
SQRT 2 3 ( )
*
dc
ij
p
:dc
ij
p
( )
1 2
( )
TEPS
t
TEPS
t 1
d TEPS ( )
EPS EPS
t

TEPS EPS
d TEPS ( ) d EPS ( )
d TEPS ( ) TEPS
TEPS
EPS TEPS
226 Marc Users Guide: Part I
Introduction
An additional point is that when evaluating the Total Equivalent Strain (using the total strain components), the
use oI the 'current plastic strain components means that the result is some sort oI combination oI the elastic
and the 'current plastic strains, i.e. not the total plastic strains - hence the plastic strain component being
used in this evaluation will necessarily be less than the Total Equivalent Plastic Strain that is evaluated
separately. The diIIerence can be signiIicant.
It should also be born in mind that the summation oI the 'current equivalent plastic strain and the elastic
equivalent strain will rarely (and even then, coincidentally) equal the total equivalent strain, simply because it
is not a strain in the usual sense and the additive assumption made does not apply. The summation only
applies to the direct strain components:
Apart Irom the von Mises strain measure being an unusual one to use in general, it is not recommended to use
the total eIIective strain measure to interpret structural response when both plastic and elastic strains are
present. These measures need to be investigated separately.
The easiest way to get plastic strain tensor components Ior elements Irom Marc is to select Full Element &
Node Print in the JOBS> JOB RESULTS > OUTPUT FILE menu oI Mentat. This will print the inIormation to
the output Iile. II the output needs to be minimised, then the PRINT CHOICE command may be manually
added to the data Iile to speciIy only those elements Ior which plastic strain results are required. NUMERICS
can also be used in Mentat to obtain the plastic strain component values individually - note that the shear
strains in Marc are engineering values and not tensor values.
Principal Stress
Element-based result.
The principal stresses associated with either the generalised stress or continuum stresses are available Ior
output. The axis system will be the same as that discussed above.
II an Updated Lagrangian analysis is being carried out, then this will be based on the Cauchy stress. II a Total
Lagrangian analysis, it will be based on the 2
nd
Piola-KirchhoII stress. See later section on Updated and Total
Lagrangian solutions.
Principal stresses are assumed to act along planes in which shear stress is zero and are
computed using the standard equations developed Irom the Mohr's circle stress diagram. The minimum and
maximum principal stresses are always orthogonal and are used Irequently Ior problems involving crack
progression.
Principal stresses are not stored in the post Iile, but are calculated directly Irom the stress tensor in Mentat.
This means that they are not calculated in the element loops during the incremental solution (unless locally
required Ior one or two oI the material models). Principal stresses are seen in the output because they have
been requested speciIically - and they are calculated in a separate loop during the writing oI the output Iile.
The principal values are calculated Irom the physical components. Marc/Mentat solve the eigenvalue problem
Ior the principal values using the Jacobi transIormation method. Note that this is an iterative procedure and
may give slightly diIIerent results Irom those obtained by solving the cubic equation exactly.
Mentat uses an extrapolate-derive approach to evaluate the Principal nodal stress in Mentat.
c
elastic i ( )
c
plastic i ( )
c
creep i ( )
c
damage i ( )
c
total i ( )

o
min
o
max
o
int
, , ( )
227
CHAPTER 1
Introduction
There are Iour Principal results quantities available in Mentat:
1. Principal Stress Min
2. Principal Stress Int
3. Principal Stress Max
4. Principal Stress Major
The Iirst three are based on both the sign and the magnitude oI this derived quantity that is, the minimum
principal stress will be the most negative value obtained and the maximum will be the most positive obtained.
The principal major stress is simply the maximum value obtained without taking the sign into account
Principal directions in Mentat
From a stress tensor, Mentat computes principal values and these can be displayed as a triad (with their
directions) as a tensor plot. At present, that is all that is available, that is, it is not possible to obtain the
principal angles Irom Mentat or the Marc output Iile.
The principal stress directions need not be tangential to the plane oI the element. This is the case Ior thin shell
elements, but not Ior thick shells. Due to the presence oI the xz and yz stress components, the principal
directions may have a component normal to the plane oI the element.
There are certain circumstances when the maximum principal stress value is not the same as the major
principal stress value. II you consider a node where the Iollowing principal stresses exists:
Node 1: Pmax = 5.0 Pint = -2.0 Pmin = -9.0
Then maximum principal will be 5.0, while the principal major will show -9.0 (the largest numerical modulus
value).
The maximum principal shear stress may be important iI the shear strength oI a material is signiIicantly lower
than the direct strength. 'Luders bands, Ior instance, are shear lines along which Iailure can occur and are
typically seen when the yield stress is just attained.
At present Mentat does not evaluate the maximum shear stress or strain. It is necessary to write a user subroutine or a
python script (py_post module) to evaluate the maximum principal shear values iI required.
Damage:
Damage post code 80.
This is actually a damage indicator Ior CockroIt-Latham, Oyane, Principal stress and one deIined in
UDAMAGE_INDICATOR.F.
For Lemaitre damage criterion, use 178 (Damage Iactor) or 179 (relative damage).
228 Marc Users Guide: Part I
Introduction
Thermal/Temperature / Flux:
Common result post codes are as Iollows:
When using shells in heat transIer, it is important to enter a code Ior each layer in chronological order iI post
Iile is to be correctly read by the INITIAL STATE or CHANGE STATE options.
It is not recommended to use nodal temperatures obtained by extrapolating the Gauss point temperatures
(postcodes 9/180) when post processing in Mentat.
This is because such nodal temperatures will have Iirstly been interpolated back to the Gauss point locations
to obtain the derived temperatures and then a Iurther extrapolation will have been carried out to obtain the
nodal values.
It is Iar better to use the nodal temperatures Irom the solution directly (post code 14)
For the same reason, it is not reasonable to compare the nodal temperatures Irom the output Iile to the nodal
temperatures obtained in the post Iile by extrapolating the Gauss point results to the nodes.
Post
Code Element Results Notes
9 Total temperature For heat transIer, this is used Ior all heat transIer
elements. Known in Mentat (pre/post processing), as
'Temperature (Integration Point). Available Ior use
with structural analyses to check that any
temperatures have been applied correctly.
10 Increment oI temperature Available Ior use with both thermal and structural
analyses to check that any temperatures have been
applied correctly
180 Total temperature Available Ior use with thermal/coupled only
371 Thermal strain tensor Known in Mentat as the 'Thermal strain element
tensor. Available Ior use with structural analyses (not
sensible Ior thermal analyses)
181-183 Components oI temperature gradient T Available Ior use with thermal / coupled only
184-186 Components oI Ilux Available Ior use with thermal / coupled only
Provides heat Ilux per unit area
Nodal Results
14 Temperature Available Ior use with thermal / coupled only
15 External Heat Flux Available Ior use with thermal / coupled only
External Heat Flux is a total nodal value
16 Reaction Heat Flux Thermal / coupled only
Reaction Heat Flux is a total nodal value
229
CHAPTER 1
Introduction
Any diIIerences will be less signiIicant as the mesh is reIined and the temperature gradient across an element
is reduced.
An example oI the data Iile syntax Iollows that saves nodal displacements (1) and temperatures (14) as well as
the global element stress tensor (411):
POST
2, 16, 17, 0, 0, 19, 20, 0, 1, 0, 0, 0, 1,
411,
9,
NODAL, 1
NODAL, 14
The Marc 'q that is Iound in Marc Volume A: Theory And User Information is total Ilux (Q) per unit area and
is consistent with the generally accepted deIinition.
Energies
The undocumented Marc data command:
POST,,5
permits all energy variables to be written to the post Iile although this is the deIault anyway.
Marc Users Guide, Chapter 6.8 demonstrates the use oI the energy results
The Iollowing energies are available:
a. Total strain energy density
Available (total or group-based) in OUT via PRINT ELEM/VMAS, and POST via postcode 48 or Global
Variables.
Total over the whole model is always in the output Iile.
b. Elastic strain energy density
Available (total or group-based) in OUT via PRINT ELEM/VMASS and POST via postcode 58 or Global
Variables.
Total over the whole model is always in the output Iile.
c. Plastic strain energy density
Available (total or group-based) in OUT via PRINT ELEM/VMASS and POST via postcode 68 or Global
Variables.
Total over the whole model is always in the output Iile (where applicable).
d. Creep strain energy
Available (total only) in POST via Global Variables.
Total over the whole model is always in the output Iile (where applicable).
e. Work done by applied Iorce or displacement
Available (total only) in POST via Global Variables. Not available on an element or group basis (not
sensible).
Total over the whole model is always in the output Iile.
I. Work done by Irictional Iorces
230 Marc Users Guide: Part I
Introduction
Available (total only) in POST via Global Variables. Not available on an element or group basis (not
sensible).
Total over the whole model is always in the output Iile.
g. Work done by contact Iorces
Available (total only) in POST via Global Variables.
Total over the whole model is always in the output Iile.
h. Kinetic energy (dynamic analyses)
Available (total only) in POST via Global Variables. Not available Ior groups since element mass matrices
would need to be recomputed costly.
Total over the whole model is always in the output Iile (where applicable).
DeIining dummy contact bodies Ior the elements oI interest would cause Marc to provide contact body
velocities as a global output. Getting kinematic energies Irom these would then be possible (mass is
available on a contact body basis).
i. Damping energy (dynamic analyses)
Available (total only) in POST via Global Variables. Not available Ior groups since element damping
matrices would need to be recomputed costly.
Total over the whole model is always in the output Iile (where applicable).
j. Thermal energy (thermal analyses)
Available (total only) in POST via Global Variables.
Total over the whole model is always in the output Iile (where applicable).
Damping Energy
Damping energy and total work done by Iriction Iorces can have negative values
Damping energy is calculated Ior mass dampers
The way to view the energy balance is dependent on the analysis type
For analysis with dynamics, the energy is balanced between the change oI kinetic energy and the work done
by external Iorces, excluding the energies dissipated by plastic/creep strain and dampers
Energy loss might be observed Ior dynamic analysis because oI numerical dissipation
Energy Balance
(1-1)
The total work done by external Iorces should be viewed as:
(1-2)
For static analysis, the energy balance can be calculated by Equation 1-3.
(1-3)
SE CSE KE DE WE KE
initial

WE WC WA WF
WE SE CSE ES EF
231
CHAPTER 1
Introduction
From Equations (2) and (3), the energy balance can be calculated by Equati1-4
(1-4)
where:
The strain energy output Irom Marc is obtained Irom:
where and are the stress and strain at the integration point. is the element volume. This integration is
numerically 'exact Ior both Iully and reduced integration elements but the constant stress Iields associated
with the single integration points oI reduced integration elements will have the expected tendency to produce
less accurate stress and, thereIore, strain energy results. For a coarse mesh, this lack oI accuracy gets more
proIound.
This means that the Work Done value is much better than the value, in particular when the mesh
is coarse, since the stiIIness and internal Iorces oI a reduced integration element do vary across the element
and, hence the energy evaluation can be more accurate
Total Strain Energy
Total Elastic Strain Energy
Total Plastic Strain Energy
Total Creep Strain Energy
Thermal Energy (Available Ior heat transIer or coupled stress/thermal analysis)
Total Kinetic Energy
Initial Kinetic Energy
Total Energy Dissipated By Dampers
Total Energy Contributed By Springs
Total Energy Contributed By
Foundations
Total Work By All External Forces
within which various contributions are also calculated as:
Total Work By Contact Forces
Total Work By Applied Forces
Total Work By Friction Forces
WC WA WF ES EF SE CSE
SE
ESE
PSE
CSE
ME
KE
KE
initial
DE
ES
EF
WE
WC
WA
WF
SE ( )
o
*
c
*
V
o c V
WD ( ) SE
232 Marc Users Guide: Part I
Introduction
A Iiner reduced integration mesh will improve the value and brings it closer to the WD value. Using Iull
integration also improves the values, even Ior a coarse mesh (more integration points in the element) and
the agreement oI and is good Ior coarse and Iine meshes
Re: DiIIerences between total strain energy` and the area under the load-deIlection curve
This may be noticed when carrying out a crack growth analysis in which the area under the load-deIlection
curve is being evaluated using a user subroutine. The diIIerence is related to an initial hydrostatic pressure
load being applied since this means that work was being done against the pressure Iield as the crack grew. II
this work is calculated Irom the volume change oI the model when the crack is introduced and subtracted Irom
the strain energy reduction, the results Ior the energy release rate will be consistent the Iorce-deIlection curve.
There may still be a very small diIIerence in the absolute values oI the energy, which will be attributable to
approximations in the numerical integration
ConIusion can arise because when calculating the STRAIN energy release rate, since this is only accurate iI
no work is done against external Iorces. This is usually achieved by preventing movement oI the model
boundaries.
Re: Negative or positive contact work quantities
The work done by external contact Iorces usually means the work done 'to deIormable bodies. So, iI a
deIormable body receives work Irom an external or contact/Iriction Iorce, the work contributed to this body
will be positive. However, iI the deIormable body does work to the environment, e.g. other bodies contacting
it, the work contributed to this deIormable body will be negative.
Especially in the case oI Iriction, the sign oI Irictional work is merely determined by the Iriction Iorce and
relative motion between the contact bodies.
Assume that a body is moving on a contact surIace and let denote the Irictional Iorce, the velocity oI
the deIormable body and the velocity oI the contact body. Based on our method Ior work calculation, the
total work done by the Iriction Iorce will be .
Please note that at this moment will take the sign determined according to the relative velocity oI the two
contact bodies, say, , That is:
From the equation above, we will know that iI is positive, will be positive. Otherwise,
will be negative. Here, it is assumed that the coordinate system is deIined so that is positive.
In Marc, the values are accumulated Irom increment to increment, so it is thereIore possible to observe such
phenomena that the total Irictional work may go Irom positive to negative or maybe the opposite
Because oI the incremental accumulation required in a nonlinear analysis, the energy results are, to some
degree, dependent on the number oI steps oI applied loading. The extent oI this variation will depend on the
degree and type oI nonlinearity in the system.
SE
SE
SE WD
F1 V1
V2
Wf F1
*
V1
*
sign F1 ( )
F1
V2 V1
Wf F1
*
sign V2 V1 ( )
sign V2 V1 ( ) Wf
Wf V1
233
CHAPTER 1
Introduction
Herrmann Variable (Nodal post code 40):
This is only available via ELEVAR and PLOTV user subroutines Ior linear elements
It is available in the post Iile Ior quadratic elements however
The additional Herrmann degree oI Ireedom is:
(mean pressure variable Ior Herrmann)
(negative hydrostatic pressure Ior Mooney, Ogden, or Soil)
Reaction Forces (Nodal post codes 5 and 6)
Reaction Iorces may also be obtained in the output Iile by selecting a 'Full Element & Node Print (Summary
only gives max/min values). The nodal results obtained are reactions at Iixed boundary conditions and
residual load correction values elsewhere. The residual load correction is the diIIerence between the internal
Iorces and the externally applied loads and its magnitude is controlled by the residual convergence criterion
mainly. In theory, the residual load correction values should be zero. In practice they should simply be
negligible compared to the reactions. The component and resultant reaction values can be made available in
Mentat.
Total reaction Iorces can be obtained most easily by using rigid bodies and then viewing the results Ior the
rigid body directly.
To Iind the total Iorce acting along a section: The easiest way to do this is via the PATH PLOT Iacility in
Mentat. In this way, the beginning and end nodes are speciIied between along the section oI interest. It is then
possible to speciIy a variable to evaluate over this speciIied length. II is speciIied (Ior example), a graph
oI the variation oI along this length is plotted (this can also be done in Patran). Note that the NODE PATH
should be selected in the order in which the nodal values are expected in the graph.
In Mentat, this variation can be readily converted to a table (CONVERT > TABLE) - where there is also an
integration tool. Simply click on Integrate and the Iinal graph point that is calculated represents the area under
the curve.
II you do not have much experience oI Mentat then, Ior a third party application like Excel, MathCAD, etc.
would be able to do similar.
Note that the Iree body Iacility Iound in Patran under the Nastran preIerence is only Ior Nastran use. It uses
Iorces that are evaluated during the Nastran solution.
Shell elements print stress and strain to the post Iile but can give Iorces and moments to the output Iile iI
requested..
Electric Current (Nodal post code 88)
In an electric analysis, the results 'Electric Current (I) and 'Current Density (J) can be conIused. Mentat
displays the Current Density (J).
Here is an example oI a 3-D solid bar:
L=20, l=2, h=0.1, Resistivity = 1.7e-7, U=5
R = Rho * L / S
o
kk
E
p
c
xx
c
yy
c
zz
( ) 1 2v ( )
o
X
o
X
234 Marc Users Guide: Part I
Introduction
U = R * I and
I = J * S = [ U / ( Rho * L) ] * S
Giving
J= 1.47e6 and I= 2.94e5
The electric current as mentioned above (Ohmic) is a global quantity and cannot be used in a Iem analysis.
The same is true Ior applied nodal currents, these are also diIIerent Irom the Ohmic current.
The same issue exists in magnetodynamic analyses. This would be an enhancement, although this is not
straightIorward, since this current is the summation oI the integration point current densities divided by the
local areas, in which the direction oI the current and the normal oI the area needs to be taken into account.
Composite Layer Results
II there are diIIerent numbers oI layers across the composite materials used, it is possible to still get the top,
middle and bottom layer results easily:
Layer 1 will be the Iirst (or top) layer iI no global IDs are speciIied.
Layer 15,000 will give the Iirst (or top) layer iI global layer IDs are used and ID1 is not the top layer.
Layer 5,000 will give the midsurIace layer results based upon (1number oI layers)/2.
Exact Ior an odd number oI layers.
Close Ior an even number oI layers.
Layer 10,000 will give the last (or bottom) layer.
Composite Failure Indices
Element-based result.
Search Ior 'Failure Index in Marc Volume A: Theory And User Information Ior details on the Iailure models
available.
It is only possible to speciIy a maximum oI three oI the available Iailure models in any analysis.
There are 13 Iailure indices available as post codes (91 to 103) Ior use in the post Iile. The meaning oI these
post codes is determined by the Iailure models speciIied and the order in which they are speciIied. From the
theory manual it can be seen that the number oI Iailure indices varies between the Iailure modes as Iollows:
Maximum stress (6 Iailure indices)
Maximum strain (6 Iailure indices)
Tsai-Wu (1 Iailure index)
HoIIman (1 Iailure index)
Hill (1 Iailure index)
235
CHAPTER 1
Introduction
This means that the maximum number oI Iailure indices possible would be iI the maximum stress, maximum
strain and one oI the other single term models were used together creating 13 Iailure indices. In the case oI
this example, the Iirst 6 terms would be related to the maximum stress model, terms 7-12 with the maximum
strain and the last term to the single term model. There are also post codes Ior the corresponding Strength
Ratios available, where Strength Ratio Margin oI saIety 1.
F (in the Iailure criteria deIinitions oI the theory manual) is the Iailure index that is the result Irom these
material models. II it is equal to or greater than unity Iailure has occurred. The mathematical Iorm (Irom the
Maximum Stress Criterion) oI is not an uncommon Iorm to state a Iailure criterion, but
actually means where is the maximum allowable stress in the 1-direction in tension and
is the actual stress calculated in the 1-direction Irom the analysis.
The use oI the user Iailure model would require that the user post codes be used and speciIied separately in
PLOTV.
Iterative Post File Result
Post Iile results Ior a contact analysis are available Ior individual iterations within an increment. The
Iollowing inIormation can be displayed iI a number is put in the 12
th
Iield oI the POST option:
The Mentat commands to activate and deactivate this Ieature are:
*job_option post_trial:on
*job_option post_trial:off
This can generate a large post Iile.
Effect of Updated or Total Lagrangian Solution
For all 'stress quantities (apart Irom the 'Cauchy stresses), the stress that is output is dependent on the type
oI analysis solution undertaken, namely, whether Updated or Total Lagrangian has been used. Total
Lagrangian is naturally Iormulated in terms oI Green-Lagrange strains and second Piola KirchhoII stress and
is based on the initial element geometry. Updated Lagrangian is naturally Iormulated in terms oI Cauchy
stress and logarithmic strain since the current conIiguration is the reIerence conIiguration.
For example, stress components such as 'Stress, 'Stress in preIerred system, 'Global stress, 'Shear stress
etc. will be based on Cauchy stress Ior Updated Lagrangian solutions and 2
nd
Piola-KirchhoII stress Ior Total
Lagrangian.
Number Information
1 displacements
2 displacements reaction Iorces
3 displacements reaction Iorces contact inIormation
o
1
X
t
| | F
F o
1
X
1
X
t
o
1
236 Marc Users Guide: Part I
Introduction
The 2
nd
Piola-KirchhoII stress is given in terms oI the initial area and current deIormed geometry
(transIormed current Iorce per unit undeIormed area) and is work conjugate to the Green-Lagrange strain
measure. For small strain, the 2
nd
Piola-KirchhoII stress can be interpreted as the Cauchy stress related to
(local) axes that rotate with the material. Without additional knowledge concerning the deIormations, these
stresses are diIIicult to interpret. 2P-K stresses are not uncommonly transIormed into Cauchy stress to give a
'true stress oI use to engineers. Marc supports this transIormation and 'Cauchy Stress may be selected
within a total Lagrangian solution.
All stress and strain measures will produce the same response as the small strains engineering stress/strain.
To tell whether an updated analysis is being used or not, the Iollowing notes may help:
1. II none oI LARGE DISP, UPDATE, or FINITE are used, Marc uses and prints Engineering stress and strain
measures.
2. Using only the LARGE DISP parameter, Marc uses the Total Lagrangian method. The program uses and
prints 2
nd
Piola-KirchhoII stress and Green-Lagrange strain.
3. With the combination oI LARGE DISP and UPDATE, Marc uses the Updated Lagrangian method. The
program uses and prints Cauchy stresses and true strains.
4. The combination oI LARGE DISP, UPDATE, and FINITE (with constant dilatation also invoked), results in
a complete large strain plasticity Iormulation using the Updated Lagrange procedure. The program uses and
prints Cauchy stresses and true strains.
5. The use oI PLASTICITY (option 3) is equivalent to the above combination and also results in a complete
large strain plasticity Iormulation using the Updated Lagrange procedure. The program uses Cauchy stress
and rotation neutralized strains (in Marc`s case, the Jaumann rate oI stress is used see Marc Volume A:
Theory And User Information Ior more inIormation).
6. The use oI PLASTICITY (option 5) results in a complete large strain plasticity Iormulation in a mixed
Iramework using the Updated Lagrange method. The results are given in Cauchy stress and logarithmic
strains.
7. The use oI PLASTICITY (option 1) results in a large displacement, small strain Iormulation using the Total
Lagrange scheme. The results are given in 2
nd
Piola-KirchhoII stress and Green-Lagrange strains.
8. Large strain rubber elasticity can be modelled in either Total Lagrange (ELASTICITY, option 1) or Updated
Lagrange (ELASTICITY, option 2). The Iormer uses 2
nd
Piola-KirchhoII stress with Green-Lagrange
strains, the latter, Cauchy stress with Logarithmic strain (Mooney or Ogden). Note that a Total Lagrangian
solution is also perIormed iI Elasticity> Small Strain is selected in Mentat (hence, Green-Lagrange and 2
nd

P-K stresses).
where:
LARGE DISP corresponds to JOBS> ANALYSIS OPTIONS> LARGE DISPLACEMENT.
FINITE does not have a direct correspondence.
UPDATE corresponds to JOBS> ANALYSIS OPTIONS> ADVANCED> Updated Lagrangian in Mentat.
PLASTICITY (option 1) corresponds to JOBS> ANALYSIS OPTIONS> PLASTICITY PROCEDURE> Small
Strain.
237
CHAPTER 1
Introduction
PLASTICITY (option 3) corresponds to JOBS> ANALYSIS OPTIONS> PLASTICITY PROCEDURE> Large
Strain Additive.
PLASTICITY (option 5) corresponds to JOBS> ANALYSIS OPTIONS> PLASTICITY PROCEDURE> Large
Strain Multiplicative.
ELASTICITY (option 1) corresponds to JOBS> ANALYSIS OPTIONS> RUBBER ELASTICITY
PROCEDURE> Large Strain Total Lagrange.
ELASTICITY (option 2) corresponds to JOBS> ANALYSIS OPTIONS> RUBBER ELASTICITY
PROCEDURE> Large Strain Updated Lagrange.
Selective Results to the Output (.out) File
For details on results output to the .OUT Iile, search Ior 'Selective Printout in Marc Volume A: Theory And
User Information.
To obtain stress and strain results in the output Iile, select Full Element & Node Print in the JOBS> JOB
RESULTS > OUTPUT FILE menu oI Mentat.
This prints the inIormation to the output Iile. Selecting this option removes the 'No Print command Irom the
data Iile.
This also includes temperature results (where appropriate) at nodal locations. In this case, temperatures are
obtained directly Irom the solution as the primary unknown and are the most accurate values
The Iollowing types oI result are then available Ior a continuum element (mechanical analysis)....
tresca mises mean p r i n c i p a l v a l u e s p h y s i c a l c o m p o n e n t s
intensity intensity normal minimum intermediate maximum 1 2 3 4 5 6
intensity
element 2 point 1 integration pt. coordinate= 0.254E-01 0.431E-03 0.431E-03
Cauchy 3.309E+05 3.309E+05 1.103E+05-7.990E+00-7.990E+00 3.309E+05 3.309E+05-7.990E+00-7.990E+00-5.738E-05-1.092E-05-5.741E-05
Logstn 2.687E+00 1.792E+00 9.452E-05-8.957E-01-8.957E-01 1.792E+00 1.792E+00-8.957E-01-8.957E-01-4.020E-10-1.881E-11-4.022E-10
plas.st 2.475E+00 1.650E+00-1.850E-16-8.250E-01-8.250E-01 1.650E+00 1.650E+00-8.250E-01-8.250E-01-3.283E-10-4.775E-12-3.284E-10
where
a. 'Element is the element number. In the example above, this is 2.
b. 'Point is the Gauss point number Ior the Iollowing section oI results. In the example, this is 1.
c. 'Integration pt. Coordinate is, as implied, the coordinates oI the current Gauss point.
d. For more inIormation on 'Tresca intensity, 'Mises intensity, 'Mean normal intensity and 'Principal
values search Ior 'tresca intensity in Marc Volume A: Theory And User Information.
e. The individual stress and strain components are given in six columns under the heading 'Physical
Components. The correspondence oI the column number with the actual stress/strain values is given at the
top oI the output Iile in a section that looks similar to:
key to stress, strain and displacement output
element type7
8-node isoparametric brick
stresses and strains in global directions
1=xx
2=yy
3=zz
4=xy
5=yz
238 Marc Users Guide: Part I
Introduction
6=xz
I. In the example above, Ior each oI the columns 1,2,3. the Cauchy stress (cauchy), the true strain (Logstn)
and the plastic strain (plas.st) components are given.
In the same way as above, the Iollowing types oI result are available Ior shell elements (mechanical
analysis)....
tresca mises mean p r i n c i p a l v a l u e s p h y s i c a l c o m p o n e n t s
intensity intensity normal minimum intermediate maximum 1 2 3 4 5 6
intensity
element 25 point 2 integration pt. coordinate= 0.677E+00 0.842E+00 -0.677E+00
section thickness = 0.100E+00
average membrane
PK2str 9.999E+00 9.999E+00 3.333E+00-2.747E-04 3.690E-04 9.999E+00 9.999E+00 9.435E-05-8.699E-05-3.184E-04 4.284E-04
moment 1.666E-02 1.662E-02 5.507E-03-6.889E-05 0.000E+00 1.659E-02 1.659E-02-6.842E-05 8.878E-05 2.647E-22 0.000E+00
Grnstch 9.999E-05 1.155E-04 0.000E+00 0.000E+00 0.000E+00 9.999E-05 9.999E-05 9.435E-10-1.740E-09-6.368E-09 8.567E-09
curvatr 1.999E-04 2.294E-04 0.000E+00-8.267E-07 0.000E+00 1.991E-04 1.991E-04-8.210E-07 2.131E-06 0.000E+00 0.000E+00
layer 1
PK2str 1.100E+01 1.100E+01 3.663E+00-4.038E-03 2.512E-05 1.099E+01 1.099E+01-4.011E-03 5.240E-03-3.184E-04 4.284E-04
Cauchy 1.100E+01 1.100E+01 3.664E+00-4.038E-03 2.512E-05 1.100E+01 1.100E+01-4.011E-03 5.240E-03-3.184E-04 4.284E-04
Grnstn 2.198E-04 1.269E-04 0.000E+00-1.099E-04-4.013E-08 1.099E-04 1.099E-04-4.011E-08 1.048E-07-6.368E-09 8.567E-09
layer 2
where
a. 'Section thickness is the shell thickness.
b. The values under the heading 'average membrane are based on the 'generalised stresses.
c. The values under the headings 'Layer 1, 'Layer 2, etc are based on the layer continuum stresses in the
local shell element directions.
There is a SUMMARY command in the Marc data Iile which prints a summary oI the results obtained in the
analysis.
This option prints the maximum and minimum quantities in tabular Iorm. The table is designed Ior direct
placement into reports. The increment Irequency oI summary inIormation and the Iile unit to which the
inIormation is written can be controlled Irom within Mentat using JOBS> > JOB RESULTS> OUTPUT
FILE> SUMMARY. Note that in the summary output the incremental and total displacements are given, but any
prescribed displacement boundary conditions are Iiltered out to give the 'real maximum and minimum
displacements.
'Selective nodal and element output to the .OUT Iile can be obtained using the PRINT NODE and PRINT
ELEMENT commands. This option allows you to choose which elements, and what quantities associated with
an element are to be printed.
The results can only be printed on an individual node/Gauss point basis not as a total Ior the speciIied
elements/nodes. For total quantities over a group oI elements, see PRINT VMASS.
PRINT NODE supports the Iollowing results:
INCR: Incremental displacement or potentials
TOTA: Total displacement or potentials
VELO: Velocity
ACCE: Acceleration
LOAD: Total applied load
REAC: Reaction / Residual Iorce
TEMP: Temperature
239
CHAPTER 1
Introduction
PRINT ELEMENT supports the Iollowing results:
An example Iollows that prints nodal displacements and temperatures Ior all nodes as well as the stress tensor
only Ior element 1. It should be placed beIore the POST command:
print node
FLUX: Flux (only available iI the HEAT, 0, 0,2 parameter is used)
MODE: Eigenvector (modal or buckle)
STRESS: Average generalized stresses at nodes
VOLT: Voltage (Joule analysis)
PRES: Pressure (bearing analysis)
COOR: Coordinates (Ior rezoning)
INER: Inertia relieI load (Ior inertia relieI analysis)
ALL: All relevant quantities
STRAIN: Total strain
STRESS: Total stress
PLASTIC: Plastic strain
CREEP: Creep, swelling and viscoelastic strain
THERMAL: Thermal strain
ENERGY: Strain energy densities:
Total strain energy
Incremental total strain energy
Total elastic strain energy
Incremental elastic strain energy
Plastic strain energy
Incremental plastic strain energy
CRACK: Cracking strain
CAUCHY: Cauchy stress
STATE: State variables
PREFER: Stresses in preIerred system
ELECTRIC: Electric Iield and electric Ilux
MAGNETIC: Magnetic Iield and magnetic Ilux
CURRENT: Current
ALL: All oI the above
240 Marc Users Guide: Part I
Introduction
1 1
tota temp
1 to 1890
print element
1 1
stress
1 to 1
1
Here is an example oI the data Iile commands needed to obtain the energies Ior two sets oI elements (1-10, 20-
30) at GPs 1-4:
print elem
2,1
energy
1,2,3,4,5,6,7,8,9,10
1,2,3,4
energy
20,21,22,23,24,25,26,27,28,29,30
1,2,3,4
Spring Iorces have an independent control Ior their output via the PRINT SPRING command.
One can also visualize spring Iorces when one end is Iixed via reactions.
Spring Iorces are written to the post Iile as well, and can be visualized Irom the Global Variables list when
graphing results.
Another way oI obtaining spring Iorces visually is to use bush elements instead oI springs. The results are
then obtainable through the Beam Axial Force variable on the post Iile.
'Selective mass, costs, volume, 2
nd
moment oI inertia about origin and energy (strain and plastic) results
may be output to the .OUT Iile using the PRINT VMASS command.
Options are provided to print:
a. Total quantities Ior each group oI elements and the quantities Ior each element in the group or
b. Total quantities Ior each group oI elements (or element SETS) only.
The Iollowing will print the summed values only Ior the two sets oI elements (1-10, 11-21):
print vmass
2,1,
1 to 10
11 to 21
In order to have correct mass computations, mass density Ior each element must be entered through the
ISOTROPIC/ORTHOTROPIC option.
In order to have the correct cost, the cost per unit mass or the cost per unit volume must be deIined through the
ISOTROPIC/ORTHOTROPIC option.
Note that volumes and masses Ior some special elements (Ior example, gap element, semi-inIinite element,
etc.) are not computed. Similarly, the lumped mass initial conditions are not included. These quantities can be
written on either standard output Iile unit 6, or a speciIied unit.
Currently, creep, kinetic, damping and thermal energies are not available Ior output on a group basis.
241
CHAPTER 1
Introduction
The Marc PRINT CHOICE command permits the selection oI how much oI the element and nodal inIormation
is to be printed, Ior example, group oI elements, group oI nodes, which shell/beam layers, which integration
points etc. Mentat does not support this. An example to print results at the Iive layers oI each oI the Iour
integration points Ior three shell element sets is as Iollows:
print choice
3,0,4,5,1
2649,2650,3090,3090,3154,3154
1,2,3,4
1,2,3,4,5
Mentat supports these model deIinition options: PRINT ELEMENT, PRINT NODE, PRINT SPRING, PRINT
VMASS, PRINT CONTACT, ELEM SORT, and NODE SORT via:
JOBS> PROPERTIES> JOB RESULTS> OUTPUT FILE
In addition, support has been added Ior the history deIinition options: PRINT ELEMENT, PRINT NODE,
PRINT SPRING, PRINT VMASS, PRINT CONTACT, ELEM SORT, and NODE SORT via:
LOADCASES> Loadcase Results
The old job option 'noprint is no longer used
The case oI job option 'noprintoII is now covered by 'resultelementoutput, 'Iull and
'resultnodeoutput, 'Iull
Compatibility in reading oI old model Iiles has been achieved. Compatibility in writing to old model Iiles
remains to be done. Old procedure Iiles that set the "noprint" job option can be made compatible by adding the
line:
*prog_option compatibility:prog_version:ment2010
It is also possible to make use oI the IMPD and ELEVAR user subroutines to process and print the required
results. This would give the most Ilexibility iI experience in Fortran is available. There is a simple example oI
the use oI these subroutines in Marc Volume D: User Subroutines and Special Routines.
Error Estimates
This can be requested Ior printing to the Marc output Iile in Mentat via JOBS> MECHANICAL> JOB
RESULTS> OUTPUT FILE> STRESS DISCONTINUITY/GEOMETRIC DISTORTION. The corresponding
command in the Marc data input Iile is Error Estimate.
There are two measures available:
1. The stress discontinuity between elements in which Marc calculates a nodal stress based upon the
extrapolated integration point values. These nodal values are compared between adjacent elements and
reported.
2. The geometric distortion in the model in which the aspect ratios and warpage oI the elements are monitored
- subsequent increments indicating how much these ratios change.
More details are given in Marc Volume A: Theory And User Information under the heading Error
Estimates. The output obtained in the output Iile is as Iollows:
worst current aspect ratio is 2.044 at element 391
worst current warpage ratio is 1.252 at element 1
largest change in aspect ratio is 1.250 at element 21
largest change in warpage ratio is 1.252 at element 1
largest normalized stress jump is 3.661E+03 at node 371 component 5 mean value is 9.11E-7
largest stress jump is 5.773E-01 at node 22 component 2 mean value is -8.399E-01
242 Marc Users Guide: Part I
Introduction
The term 'warpage used here actually means the ratio oI the largest and the smallest diagonal in quad/hex
elements. This would not have any signiIicant meaning Ior triangles/tets. The aspect ratio used is, as usual,
related to the largest and smallest element edge lengths.
The evaluation oI the stress error measure is moderately expensive. The evaluation oI the geometric error
measure is very inexpensive.
The ERROR ESTIMATE option can be used Ior either linear or nonlinear analysis.
The ADAPTIVE option can be used to ensure that a speciIied level oI accuracy is achieved. The elastic
analysis is repeated with a new mesh until the level oI accuracy requested. This is detailed Iurther in Marc
Volume A: Theory And User Information under the heading 'Adaptive Meshing.
Sort node and element quantities by magnitude
This is invoked using the NODE SORT or ELEMENT SORT commands and allows results to be sorted, with
the output given in report Iormat. NODE SORT allows either an ascending or descending sort order. In
addition, either real numeric value or absolute value can be used. A range can also be given over which to sort
This option is in eIIect until a NO ELEM SORT or NO NODE SORT command is encountered
The element sort codes (through which ordering is controlled) are as Iollows:
Code Description Code Description
1 First stress 28 Fourth plastic strain
2 Second stress 29 FiIth plastic strain
3 Third stress 30 Sixth plastic strain
4 Fourth stress 31 Equivalent plastic strain
5 FiIth stress 32 Mean plastic strain
6 Sixth stress 33 Tresca plastic strain
7 Equivalent stress 34 First principal plastic strain
8 Mean stress 35 Second principal plastic strain
9 Tresca stress 36 Third principal plastic strain
10 First principal stress 37 First creep strain
11 Second principal stress 38 Second creep strain
12 Third principal stress 39 Third creep strain
13 First strain 40 Fourth creep strain
14 Second strain 41 FiIth creep strain
15 Third strain 42 Sixth creep strain
16 Fourth strain 43 Equivalent creep strain
17 FiIth strain 44 Mean creep strain
18 Sixth strain 45 Tresca creep strain
19
Equivalent strain
46
First principal creep strain
20
Mean strain
47
Second principal creep strain
243
CHAPTER 1
Introduction
Similar codes are available Ior NODE SORT (see Marc Volume C: Program Input).
Nodal Force Output for Continuum Elements
It is possible to obtain a node point Iorce balance similar to Nastran, using the GRID FORCE data command.
This option controls the output oI the contribution to the nodal Iorce at either an element level or a nodal level.
This is useIul when constructing a Iree body diagram oI part oI the structure. The grid Iorce balance is with
respect to the global coordinate system.
On an element level, the grid Iorce balance is based upon:
Internal Iorces
Distributed Loads
Foundation Forces
Reaction Force
On a nodal basis, it is much more complete and includes:
Internal Forces Distributed Point Forces
Foundation Forces Spring Forces
Contact Normal Forces Contact Friction Forces
Tying/MPC Forces Inertia Forces
Damping Forces DMIG Forces
Reaction Force
Using nodal stress and an associated area is not recommended because oI its inherent inaccuracy.
Eigenvalue Output File Results
The Iollowing three results are provided:
a. Frequency: The magnitude oI the Irequency oI vibration Ior each mode. The relationships between the
eigenvalue, , circular Irequency, , and Irequency, , are

21
Tresca strain
48
Third principal creep strain
22
First principal strain
49
Temperature
23
Second principal strain
61
Voltage
24
Third principal strain
73
First gradient
25
First plastic strain
74
Second gradient
26
Second plastic strain
75
Third gradient
27
Third plastic strain
Code Description Code Description
e |
e 2tf e
2
2tf ( )
2
f
1 2
2t
244 Marc Users Guide: Part I
Introduction
b. (th)trans*m*th: This represents (the diagonal modal/generalised mass) as described in the Iinite
element equations given in Marc Volume A: Theory And User Information. The magnitude is equal to unity
when mass normalisation has been requested.
Generalised mass or stiIIness do not have any helpIul physical meaning. They are mathematical concepts
that enable the use oI the 'real stiIIness and mass oI a component in the modal domain. Unlike
static analyses, the modal domain is Irequency dependent, so that the eIIective values oI the generalized
stiIIness and mass will changes according to the Irequency oI excitation. That is why there are terms like
and - the corresponds to the natural Irequency shapes oI the structure - these are used
to 'Iactor the static and .
The value oI the generalized mass can be made any value simply by choosing a diIIerent normalization
method. Moreover, the relative size oI the modal mass between modes do not have any signiIicance a low
value oI modal mass in mode A and a high value in mode B cannot be interpreted to imply that mode A is
unimportant with respect to mode B or that more oI the structural mass is associated with mode B. The
modal mass is given so that an analyst can perIorm subsequent modal response calculations.
Generalized mass is most oIten used to normalize the eigenvector results so that, together with the modal
stiIIness, they can be later used in a post-eigensolution analysis such as a harmonic/Iorced Irequency
response. Again, this normalization with mass is a mathematical requirement.
What you can say is that the ratio oI the modal stiIIness (i) to modal mass (i) is the eigenvalue (i).
Alternatively, it can be seen as an indication oI the amount oI mass participating in a particular mode
compared to the mass participating in rigid body motion but this is only Ior an individual mode.
c. (th)trans*k*th/w*w: This represents (the diagonal modal/generalised stiIIness, divided by the
eigenvalue) as described in the Iinite element equations given in Marc Volume A: Theory And User
Information. Similar to the modal mass, this has no particular signiIicance. The ratio oI the modal stiIIness
to the modal mass is always the eigenvalue however. For mass normalization, the modal stiIIness becomes
the eigenvalue.
What are oI use are the participation and mass participation Iactors. But these are only available Irom a
Irequency response calculation where there is some Iorm oI Irequency dependent loading input to compare
with the actual response oI the structure.
|
T
M|
K ( ) M ( )
|
T
M| |
T
K| |
M K
|
T
K| e
2

245
CHAPTER 1
Introduction
Output from Contact Analyses
Detailed contact inIormation to the output Iile:
When the debug printout PRINT parameter is used in a contact analysis (value oI 5 or 8), it produces
inIormation on when any node on the boundary comes into contact or separates Irom any surIace. It also
produces inIormation on whether a contact node is Iixed to a surIace or is Iree to slide along it. For example:
node 101 of body 1 is touching body 2 segment 24
the retained nodes are 59 5
the normal vector is 0.00000 -1.00000
contact body = 1
number of nodes in contact = 0
contact body = 2
number of nodes in contact = 5
total friction force change = 0.43415E+00
current total friction force = 0.43415E+00
current total normal force = 0.10007E+03
friction convergence ratio = 0.10000E+01
maximum friction force change = 0.20546E+00
current maximum friction force = 0.20546E+00
In addition to the inIormation printed with IDEV = 5, when IDEV = 8 is entered (IDEV is an internal
variable name), the incremental displacement and the reaction Iorces Ior those nodes in contact with rigid
surIaces are printed in a local coordinate system.
incremental displacements in transformed system
nodes in contact: tangential,normal
node incremental displacements
1 1.914E-21 1.388E-17
32 9.715E-17 1.388E-17
reaction forces/residuals at transformed shell nodes in transformed system
node residuals and reactions
1 -2.474E+00 -4.127E+00
32 -4.036E+00 -5.458E+00
PRINT, 5 can also be speciIied Irom Mentat via Jobs~ Mechanical~ Job Results~ Output File~ Contact.
PRINT CONTACT:
Controls the printing oI the contact summary inIormation at the end oI each increment in a more granular
manner. Previously, when NO PRINT was speciIied, this suppressed the contact summary.
This option ensures that the summary oI contact inIormation Ior each body is printed to the output Iile even iI
the NO PRINT option is activated.
NO PRINT CONTACT: This option deactivates the output oI the summary oI contact inIormation.
Selecting CONTACT NORMAL FORCE X/Y when a post Iile is loaded in Mentat will give the contact
Iorces directly. The Iorces are given in global directions. To have these Iorces rotated into normal and
shear components automatically, select the CONTACT NORMAL FORCE and CONTACT SHEAR FORCE
results variables.
246 Marc Users Guide: Part I
Introduction
For analytical contact, the Iorce direction will have been evaluated according the spline directions at each
node. For discrete contact, the Iorce direction will be an average oI all the element Iace directions attached to
each node in question.
The global contact Iorces may also be obtained in the output Iile via JOBS> MECHANICAL> JOB RESULTS>
OUTPUT FILE> Full Element & Node Print. The amount oI result data output may be controlled via the data
Iile - this is not possible using Mentat at present.
Contact Status (node post code 38):
Marc allows you to select the contact status as a post Iile variable.
A value oI 0 means that a node is not in contact.
A value oI 0.5 means the node is in near thermal contact.
A value oI 1 means that a node is in contact.
A value oI 2 means the node is on a cyclic symmetry boundary.
The image below shows a contact status values oI 0.5 on one node that is part oI an adaptively split element
that is not in contact but should be zero. This is a known Ieature oI Mentat in conjunction with local
adaptive meshing. For reIinement tying it will average the values at the corner nodes. This works in the
majority oI the time, but is not appropriate Ior discrete quantities such as contact status.
Contact Touched Body (node post code 39):
a. The 'contact touched body nodal result will give the body number oI the contact body being contacted by
the node. In Iact, it is an array oI length three; thus, it can include up to two or three bodies (depending on
whether the problem is 2-D or 3-D) iI a node is touching more than one.
b. II all entries oI the array are zero, it means the node is not touching any contact body.

247
CHAPTER 1
Introduction
Contact Stress:
a. 'Contact Normal Stress (nodal post code 34) is the underlying element stress tensor components
transIormed normal to the surIace.
b. 'Contact Friction Stress (nodal post code 36) is computed in a similar way and can be interpreted as the
Iriction generated shear stress.
Contact stresses Ior quadratic elements (with true quadratic contact) are derived Irom the extrapolated
element integration point stresses (rotated normal and tangential to the element surIace at the contact location
points) and are less accurate than the contact Iorces. For linear elements (or quadratic elements and linearised
contact), the contact stress is derived Irom the contact Iorce divided by the area. The contact Iorces are
obtained directly Irom the FE solution at the same time as the reactions and displacements. The contact Iorces
are the most accurate. Contact Iorces are available in the output Iile (choose Full Element & Node Print Irom
Mentat). It is not possible to obtain contact stresses directly in the output Iile however these are evaluated
during post-processing only. You can only control this via the nodal post code 34 ( Contact Normal Stress) -
but this writes the inIormation to the post Iile only. PLDUMP could be used to extract this inIormation Irom
the post Iile, or a suitable user subroutine during the analysis iI this suits better (using ELMVAR).
Prior to 2009r,1 the Iriction stress (nodal post code 36) on the nodes oI quadratic elements (using true
quadratic contact) may be nonzero even iI there is no Iriction between the contact bodies involved. Since the
Iriction stress is only calculated Ior post processing, this did not aIIect the other analysis results.
Note that Ior contacting nodes, we know iI there is Iriction or not, so then the Iriction stress vector can be set
to zero. But Ior nodes oI contacting segments, we don't know iI they are involved in Iriction (or glue) so,
there, the Iriction stress is always calculated as the extrapolated/averaged nodal shear stress.
Only contacting nodes are given a value Ior the result CONTACT STATUS. Contacted nodes have a value oI
zero.
Nodes that are in contact, but are considered to have 'slid oII do not get marked as in contact in the
CONTACT STATUS.
Global results variables are available Ior all contact bodies ('body variables) when in History Plot. These are:
POS X/Y/Z <body name>: The displacement oI the contact body in the component X/Y/Z directions. This
is not available Ior deIormable bodies.
POS <body name>: The resultant displacement oI the contact body. This is not available Ior deIormable
bodies.
Angle POS <body name>: The rotation (radians) oI the contact body. This is not available Ior deIormable
bodies.
VEL X/Y/Z <body name>: The velocity oI the contact body in the component X/Y/Z directions.
VEL <body name>: The resultant velocity oI the contact body.
Angle VEL <body name>: The rotational velocity (radians/second) oI the contact body.
FORCE X/Y/Z: <body name>: The Iorce on the contact body in the component X/Y/Z directions. Based on
the contact Iorces created during the solution.
FORCE <body name>: The resultant Iorce on the contact body. Based on the contact Iorces created during
the solution.
248 Marc Users Guide: Part I
Introduction
MOMENT X/Y/Z: <body name>: The moment on the contact body in the component X/Y/Z directions.
Based on the contact Iorces created during the solution. This would not be available Ior a deIormable body
based on continuum elements.
Contact Area
The contact normal stress on the post Iile depends on whether or not we have quadratic elements and true
quadratic contact. For linear elements (or quadratic elements and linearised contact), we get the Iorce divided
by the area. For quadratic elements (and true quadratic contact), we get the extrapolated stresses.
For linear elements, thereIore, it is possible to use the Iact that the contact stress results are based on the
contact Iorce divided by the contact area around the node. Marc uses shape Iunctions to get this area, so
simply extract the contact Iorce and contact stress, divide one by the other to leave a Iairly accurate contact
area. This could be automated with a Python script iI using Mentat or PCL iI using Patran.
For quadratic elements, it would be possible to edit a Marc routine to print the contact area (oarea) but this is
only called Ior stress-based extrapolation.
Iterative Solver Output
The iterative sparse solver prints out its measure oI convergence every 50 iterations. Three numbers are listed.
They are all related to the convergence behaviour oI the solver. They are not at all related to the global
convergence control Irom the (say) displacement and residual criteria.
Conceptually, one sequence oI iterations would correspond to a single N-R solution. As Ior the three numbers,
they signiIy checks on diIIerent quantities, to make sure satisIaction on all Ironts, they are as Iollows:
Term 1 checks on the Euclidean norm oI the residual Iorce vector vs. the norm oI the right-hand side.
Term 2 checks on the relative change in the maximum displacement component vs. the maximum value in the
updated solution vector.
Term 3 checks on the maximum residual Iorce component vs. the maximum value in the right-hand side.
Interpretation of Results
Search Ior 'Results Interpretation in the Marc Users Guide Ior general inIormation.
A negative stress is generally taken as compressive, positive as tensile.
Bending moment distributions Irom the use oI lower-order shell elements can predict nonzero values at a
'pinned boundary where, theoretically, no moment should exist. This is normally due to the Iact that a
parabolic moment distribution has been developed in the solution, while these elements can only sustain a
constant or linear moment distribution. In this case, the moment distribution will be increasingly better
approximated as the mesh is reIined. Note, however, that the values oI the bending moment at the boundary
will normally be signiIicantly smaller than the maximum values obtained.
When perIorming simple tests on a shell or plate element, it should be noted that the and results are
coupled through the Poisson's ratio material parameter (standard shell and plate theory). Hence, this value
may need setting to zero to establish correlation with simpler theories. The eIIect oI this coupling may be seen
iI the deIormation is exaggerated - in which case there will be either hogging or sagging across the width
(depending on the orientation oI the load).
M
x
M
y
249
CHAPTER 1
Introduction
All plate elements suIIer a theoretical singularity in the vicinity oI the support around an obtuse skew angle.
This limitation is documented in various academic papers. This may be mostly overcome through the use oI
the thick shell elements. Also the eIIect oI the singularity can be minimised by thickening the plates locally.
Isolating groups oI planar surIaces Irom a model containing a mixture oI planar and curves shells will permit
the use oI averaged contour plotting without any averaging errors. Generally, however, unaveraged results
should be displayed since large stress gradients can be revealed which might otherwise have been hidden by
the averaging process.
Averaged or unaveraged stress results may be used when contouring global variables since averaging will
occur on a global basis, independent oI the local axis or orientation oI the shell surIace.
Contouring using local values will be aIIected by the local axis deIinitions and, hence, quadrilateral and
triangular elements may not be mixed without diIIiculty. This is because the element local axes may not be
aligned in the same direction Ior all elements. The use oI unaveraged output is recommended, particularly Ior
any non-planar shell geometries such as curves and intersections. In the case oI shell elements in which a shell
surIace is curved or intersecting geometries are present, an averaged contour plot will not be correct at the
lines oI intersection since averaging Ior each node will be perIormed over diIIerent planes. In such cases,
direction independent stress measures such as principal or von Mises may be easiest to interpret.
Stress Concentrations
Stress concentrations are an 'expected behavior oI a Iinite element to a localised stress discontinuity. The
more you reIine the mesh, the higher the stress will get, until the element dimensions become too small Ior the
computer to handle due to precision problems.
Point loads and point boundary conditions will always give this eIIect (as will sudden changes in geometry or
material properties).
Such behavior is particularly a problem in material nonlinearity, since these can give local Iailure and cause
no end oI problems with convergence.
One way oI explaining this is in terms oI degrees oI Ireedom (see the Iigure below). It is a series oI square
blocks each with double the mesh density and each having exactly the same loading conditions, that is,
glued to the rigid body at the bottom and a point displacement in the y-direction only at the top centre. It is
clear Irom the coarse mesh that there are simply not enough degrees oI Ireedoms present in the loaded area to
adequately capture the deIormation required. As the reIinement increases, so the localised deIormation that
would be expected is achieved. In other words, the coarser mesh behaves in a 'stiIIer manner than the reIined
mesh another common trait with the Iinite elements.
250 Marc Users Guide: Part I
Introduction
The problem is partly due to the numerical model not replicating what is happening in reality. In real liIe,
there are very Iew times when a 'point load is actually applied. It is almost always distributed (albeit over a
small area). Even iI it is a very localized load in real liIe, iI the results oI interest are in that area, a suIIiciently
reIined mesh is required over which a distributed load can be applied. This will make a lot oI diIIerence to the
stresses.
This problem is the one that people dealing with Iracture have to deal with. The stresses near a point oI
Iracture rise in an exponential manner. Various ways are used to get a better distribution oI stress at this point.
One oI them is via the 'crack tip elements.
II the results near the point load/boundary condition are not oI interest, then it is quite acceptable to ignore
those high stresses - as long as there have been checks made that determine that there is nothing 'real
happening that may be oI consequence. In a nonlinear analysis with slack convergence criteria thresholds, it is
common to see stress concentrations in unexpected locations. These are eliminated by an appropriate choice
oI convergence threshold values.
The graph below shows the increase in equivalent stress underneath the loaded nodes as the mesh density
increases Ior the meshes above:
Mentat Results (General)
To show the value oI a single node:
RESULTS> TOOLS> SHOW NODE
.and then click on the node in question or type the node number into the dialog box as requested.
Display the values oI multiple nodes:
RESULTS> POST NODES|REMOVE SELECT ALL EXISTING
Since all nodes have been added by deIault, to view speciIic nodes use the REM button Iirst to remove the
nodes that are not to be displayed.
RESULTS> POST NODES| ADD SELECT NODES REQUIRED
Equivalent Cauchy Stress VS. Node Position
0.00E+00
2.00E+10
4.00E+10
6.00E+10
8.00E+10
1.00E+11
1.20E+11
1.40E+11
1.60E+11
1.80E+11
2.00E+11
0 1 2 3 4 5 6 7 8 9
Loaded Node Position
E
q
u
i
v
a
l
e
n
t

C
a
u
c
h
y

S
t
r
e
s
s
251
CHAPTER 1
Introduction
To evaluate the distance oI nodes Irom geometric entities such as curves and surIaces, use:
RESULTS> GEOMETRY DISTANCE>
The results oI this computation are made available as scalar and vector plots. Controls are available to control
the accuracy and cost oI the computation:
Mentat has the capability to reIlect your results Irom a 3-D section (say) into a Iull 3-D structure by simply
carrying out a symmetry copy (in mesh generation) aIter plotting the results.
Alternatively:
a. Write a PLDUMP application that would produce the Iull 360 post Iile Irom the current one.
b. Write a PLDUMP application that would produce an axisymmetric elements post Iile and then use
AXITO3D.
c. Redo the analysis with axisymmetric elements and then use AXITO3D.
To more clearly visualise the diIIerence between results oI diIIerent increments, the Iollowing option may be
oI use:
RESULTS>DELTA
This command toggles the diIIerence results plotting Ieature, which, when turned on, plots the diIIerence
between the current increment with the previously plotted increment.
Keyboard command:
*set_post_delta <on/off>
Mass and Volume Results:
In Mentat under UTILS, there are commands that calculate the ELEMENT MASS or ELEMENT VOLUME. The
quantities will be the current values iI an updated Lagrangian solution has been speciIied.
In Marc, the PRINT VMASS command (Model DeIinition section) evaluates element volumes, masses, costs,
strain energies, and second moment oI inertia about origin. See separate section on VMASS.
252 Marc Users Guide: Part I
Introduction
Flowline Plotting:
a. Flowlines are computed by Marc and displayed in Mentat to visualise how material Ilows during an
analysis. They are to be used in conjunction with global remeshing, since the mesh is not 'attached to the
material in this case.
Typically, the original mesh is used below to Iorm the Ilowlines. Marc then evaluates the location oI the
original mesh at each increment. This can then be superimposed in post-processing onto the
deIormed/remeshed mesh to indicate material Ilow. It large displacement analyses, it is not possible to
superimpose the original, undeIormed mesh.
b. This is invoked using the FLOW LINE command in the Marc data Iile or via Mentat:
JOBS> JOB RESULTS> FLOWLINES> BODY
This will turn on the calculations oI the Ilowlines that are attached to the material/
c. The Ilowlines are automatically plotted until turned oII. Controls are available Ior selecting which Ilowline
edges are plotted, and whether or not to restrict them to the model outline or surIace. Use the Iollowing
button sequence to get to the FLOWLINES submenu to change the plot controls:
RESULTS> FLOWLINES
d. A little more detail is given in the Marc Users Guide (search Ior 'Flowline Plotting).
e. Changing the options in the FLOWLINE Iorm oI Mentat requires the DRAW button to be clicked to activate
the changes.
Particle Plotting:
a. This Iacility displays position oI a particle as a Iunction oI time by means oI a curve. The color oI the curve
indicates the value oI the equivalent stress oI a particle as a Iunction oI time. The particles must be identiIied
during preprocessing by means oI a node set. This button is located in:
RESULTS> PARTICLE TRACKING
b. A little more detail is given in the Marc Users Guide (search Ior 'Particle Tracking).
Generalized XY Plotter:
An example oI the use oI this Mentat Iacility can be Iound by searching Ior 'Generalised XY Plotter in the
Marc Users Guide.
Rezoning:
a. There is a capability in the Post` part oI Mentat called Rezone`. It can be Iound in RESULTS> TOOLS>
REZONE MESH. It adds the current displacements oI each node to its original coordinates to enable a new
mesh to be created Irom the current results increment that has been selected.
In addition, the post Iile is closed and all post plotting is turned oII. The model existing beIore the post Iile
is opened, is replaced by the model Irom the post Iile.
To use it, go to the time step that has the necessary displacement vector, and with the deIormed shape
switched oII, run the rezone option (do it once only). II the displacements are large, you will see the change
in shape happening. Then SAVE AS another Mentat Iile.
II the deIormed shape is leIt on, the displacement will be doubled and the results will, thereIore, look rather
odd. It won`t make any diIIerence to the actual mesh created.
253
CHAPTER 1
Introduction
b. It is possible to obtain the deIormed mesh in this way, and then to export it as geometry (IGS Iormat, Ior
example). This is most easily done by deIining the structure with one or more contact bodies, and speciIying
that these bodies are deIined using an analytic deIinition (CONTACT> CONTACT BODIES>
DEFORMABLE> ANALYTICAL). Assuming that analytic surIace output has been requested (JOBS>
MECHANICAL> JOB RESULTS> CONTACT MODEL FILES), then it is a simple case oI reading in the
analytic surIace deIinition Iile corresponding to the increment required (e.g.,
filename_job1_spline_2.mfd Ior increment 2), and then exporting via FILES> EXPORT.
c. II the spline inIormation is not available Ior some reason, then it is possible to convert the element Iaces to
surIaces and then export these surIaces.
To do this mesh 'skinning:
Display the solid mesh in minimum edges (outline) and Iaces (SurIace). This is in the PLOT>
ELEMENTS > SETTINGS menu.
Then convert Iaces to elements (GEOMETRY> CONVERT) by picking all the Iaces in the view using a
'box pick aIter REGEN to make sure that the display is reIreshed completely so you get them all.
Select by class hex8 and delete ALL SELECTED. Remove unused nodes.
Reducing the post Iile size:
a. The pst_reader program in the Marc bin directory can be used to remove increments Irom a post Iile.
b. Alternatively, the deIault behaviour oI the PLDUMP program can do the same.
c. Use post Iile version 13 (available in 2005r3 and later). This provides signiIicant savings in Iile size by a
new method oI grouping element types together.
The deIault post Iile version is 12. II a 0 is entered on the post option (11th Iield), a version 12 post Iile is
created. The DEFAULT Iile version in Mentat writes a 0 on this Iield, so in Mentat 2005r3/2007/2008,
DEFAULT selects post Iile version 12.
TransIorming Nodal Results
a. Commands are available to control the coordinate system used to decompose tensors and vectors into scalar
components Ior scalar plotting. Normally tensors and vectors are decomposed in a rectangular coordinate
system aligned with the global axes.
Alternatively, the user may desire to decompose the tensors and vectors in another coordinate system, such
as a cylindrical one, not aligned with the global axes, and having a diIIerent origin.
b. To create and activate a local coordinate system Ior post processing:
RESULTS> SCALAR PLOT SETTINGS> RESULTS COORDINATE SYSTEM
c. The easiest way to create the coordinate system is using Align. In this way, one can select three existing
nodes or points. For example, pressing Align and then points (1), (2), (3) in order will create a local
coordinate system as shown.
254 Marc Users Guide: Part I
Introduction
d. The eIIect oI the current local coordinate system may be switched on or oII using the Active button.
e. TransIormations or coordinate systems are not stored in the post Iile, and must be created as needed.
I. Commonly used coordinate systems may be stored using the Save button and restored using the Load
button.
Python Scripting
The Mentat scripting language, Python, may be used very eIIectively to extract the required results directly
Irom a post Iile. There are a number oI reIerences to the Python API that are available:
Python Reference Manual (in ..\examples\python\tutorial\python_ref.pdf): This document
describes the Iunctions available to Python scripts that use the PyMentat or PyPost interIace modules.
Python 1.5 Documentation (in ..\Python\Doc\index.html): Contains Tutorial, Library reIerence,
Language ReIerence, Extending and Embedding (tutorial Ior C/C programmers) and Python/C API
(reIerence Ior C/C programmers).
Tutorial and ReIerence Manual (in ..\examples\python\tutorial\python_manual.pdf):
Introduces the user to the Python modules through examples. The examples cover the basics oI the modules
and display some typical uses oI creating and processing a model at various stages.
A recommended Python programming manual would be Programming Python by Mark Lutz or Learning
Python by Mark Lutz and David Ascher.
Visit the Python web site at http://www.python.org.
An example oI extracting scalar results and contact body inIormation is given at the end oI this document in
Appendix E: Python Example (Max Stress Results).
An example oI extracting nodal displacement results is given at the end oI this document in Appendix F:
Python Example (Displacements at Nodes).
255
CHAPTER 1
Introduction
Saving Results Directly From Mentat
A Report Writer is available in RESULTS> REPORT
WRITER.
There is an undocumented Mentat command, *post_dump,
which one typed manually speciIying an external Iile and
a list oI nodes. Mentat then writes the results, line per line,
as the node number Iollowed by the value oI the actual
plotted variable (as deIined in the SCALAR PLOT menu) Ior
the list oI nodes. An example procedure Iile to demonstrate
this is as Iollows:
*post_contour_bands
*post_value total strain energy
density
*post_skip_to 10
*post_dump
results.txt
yes
1
all_existing
#
The results are listed in value order and not node order.
Mentat will always display and post_dump by averaging
across all elements (including those that are not visible).
The general remedy to this provided in Mentat is to
ISOLATE (RESULTS> ISOLATE) the elements on the
screen. In this way, you deIine over which elements the
averaging will occur. So, no matter what is selected on the
screen, the results Ior the isolated elements will not change.
UnIortunately, the *post_dump does not handle the isolate command properly. It seems to take the nodal
isolated value Irom the associated element, and not the element you have isolated. This means that you should
always use contours in the Mentat display when using *post_dump.
With averaging on, the legend in Mentat will show the same max and min values whether NUMERICS or
CONTOUR is chosen. With averaging oII, the legend in Mentat will show diIIerent max and min values when
NUMERICS or CONTOUR is chosen. The reason is that NUMERICS will always average (unless isolate is
used), whereas CONTOUR BAND does not. What you see on the screen should be the same as what you Iind
in the post_dump Iiles, but there will be diIIerences depending on whether you have contours or numerics
turned on in the display.
It is possible to copy certain results (tabular data) to the Windows NT clipboard only. This is available Ior
results produced using PATH PLOT, HISTORY PLOT, TABLES, or GENERALIZED XY PLOT.
This data can then be pasted into a Word or Excel document.
For non-NT platIorms it is possible to save results directly to a speciIied Iile when using PATH PLOT and
HISTORY PLOT.
256 Marc Users Guide: Part I
Introduction
Visualizing Analytic Contact Surfaces
1. Create a Iile named file.proc text Iile (Ior example) with the Iollowing commands:
*reset_view
*view_model_angles
-45 45 45
*dynamic_model_on
*fill_view
*set_lighting 1 on
*set_light 1 4 on
*surfaces_solid
*set_surface_lines off
*redraw
2. Double click on the mfd Iile (that Marc has created containing the spline inIormation) to open up Mentat with
the Iile loaded (or FILE> OPEN Irom Mentat directly).
3. Bottom Menu Bar: UTILS> PROCEDURES> LOAD> ....,select the proc Iile (Irom 1 above) - then press
START/CONT.
Press OK when done.
4. Each oI the contact body surIace deIinitions has a group: UTILS> SELECT> SELECT SET.
Pick the sets you want, and then press MAKE VISIBLE.
To get all back, just press CLEAR SELECT, and then MAKE INVISIBLE.
With the DYN. MODEL lit up, one can pan (leIt mouse), rotate (middle mouse) or zoom (right mouse) RESET
VIEW, Iollowed by FILL brings everything back in the display.
257
CHAPTER 1
Introduction
Making Movies
The quick method to dynamically animate the results is via
RESULTS> MONITOR (preceded by a Rewind). This is Ior
a 'live view and does not save the animation.
Once the mouse is used to move/rotate the model, the
Monitor will stop and the Monitor button pressed again to
restart it 'animating. The Animation Iorm is Iound in
RESULTS> ANIMATION.
This is used to create movies and save them as a Iile Ior
presentations.
It is possible to create MPEG, AVI and GIF movies.
From the Help menu you can access the New Features and
in the section entitled 'MPEG and AVI Animations.
An example oI the use oI animation Irom a nonlinear load
incrementation analysis can be Iound in the 'Tube Flaring,
'Container or 'Tire examples herein.
See the 'Transmission Tower example Ior an example oI
the use oI animation with modal analysis types.
The recommended method is the GIF movie. These are good
quality, and can be embedded into presentations (not linked,
and needing to copy the avi / mpeg with the presentation).
GIF MOVIE shares many oI the settings with the
MPEG/AVI movie generation commands. It will use the
values displayed under INCREMENT SETTINGS, FIRST
(movieIirstincrement), LAST (movielastincrement),
STEP (moviestepincrement), VIEW (movieview), and
DELAY (animationpause).
The view will be that speciIied in the VIEW display and must be the current view.
258 Marc Users Guide: Part I
Introduction
The GIF animation is generated in a three step process:
1. The animation Iiles are automatically generated when
the 'Make GiI Movie is pressed.
2. Once the animation Iiles are created, the animation is
played and screen images are captured into a sequence oI
.giI Iiles. These Iiles are required Ior the GIF animation
Iile encoder.
3. The sequence oI .gif Iiles will be read and written to
the GIF animation Iile in the current working directory.
Mentat grabs the screen image Irom the graphics
window. This means that they have to be visible and
not obscured in any way Ior the images to be saved
properly. Mentat must also be pointing to an active
display without the screen saver on.
Note that this command will remove all animation display list Iiles and .giI Iiles that begin with the base Iile
name (*animation_name) beIore it starts unless the GENERATE ANIMATION FILES option
(movie_gen_files command) has been turned oII.
One can interrupt the process oI making the GIF animation Iile by pressing the Escape key, however, you will
be returned to the animation play mode. In this case, you will need to run the showmodel command in the
ANIMATION menu to display the model again.
II the movie is not being created correctly, check that the OpenGL version oI Mentat is being used (see the
title bar oI Mentat, it should have '(OpenGL) in it - and not GDI).
Creating AVI Animations
Mentat stores AVI animations as a series oI RGB Iiles. There is a utility (normally in
\mentat\bin\marc_movie.exe) on the PC to convert these multiple Iiles to a single AVI Iormatted Iile.
This can subsequently be imported to a Power Point presentation. Working on PC machines you may use this
program directly to create animation Iiles.
SGI machines have a similar utility program called Media Convert or movieconvert - there is no such utility
on other unix platIorms. II a PC machine is available, then the PC marcmovie application may be obtained
Irom MSC to convert the series oI RGB Iiles created on a unix machine to a PC Iormat AVI Iile.
The steps to use the PC-based marcmovie application are as Iollows:
a. Create the RGB Iiles on the unix system. This done by creating the on-screen animation as usual
(RESULTS> ANIMATION ). The Iiles created Irom creating the animation do not have the extension RGB
and are simply display lists basically binary snapshots oI the screen that are only readable by Mentat. To
generate RGB images that can be manipulated to create individual images or an AVI animation, it is
necessary to select MAKE MOVIE. This will take the animation images previously created and produce the
appropriate RGB images. Selecting Make Movie will ask Enter View to Create Movie From this reIers to
the current view that is active, either 1, 2, 3, or 4, as selected Irom the View button.
b. FTP the *.RGB Iiles to the PC (iI running Marc on a Unix box).
259
CHAPTER 1
Introduction
c. Run marc.movie.exe. On FILE> OPEN, select all the Iiles that are required to be part oI the animation. On
FILE> SAVE AS... save the animation in avi Iormat.
d. In Powerpoint, use the Insert option to attach this Iile to the desired slide
BeIore creating the movie Iiles, it is suggested that changes are made to the colormap. This is done under
VISUALIZATION> COLORS. Either use colormap 2 or change the colour sliders as:
a. BACKGROUND to white.
b. TEXT & WINDOW BORDERS to black.
c. EDGES to black.
d. POSTPROCESSING TEXT to black.
e. ANNOTATIONS to black.
A proc Iile can be used to save this conIiguration Ior later use.
Assembling RGB files for Animation
To create an animation Irom a series oI RGB display Iiles:
a. Assemble the RGB Iiles into a directory together.
b. Find the program Marc_Movie.exe - normally in ..\mentat\bin. Then double click on it and run it.
c. In Marc_Movie, do FILES>OPEN>... and browse to the directory containing the RGB Iiles, select them all
at the same time (holding down the control key) and press Open
Press the play button on the right-hand side oI the Marc_Movie screen. File save as will save the animation as
an AVI Iile. The best compression scheme Ior the is MicrosoIt 1.
Note: The numbering oI the RGB Iiles is critical to showing the proper sequence. This is due to a quirk
in how Windows saves and displays Iile names. II there are less than 10 RGB Iiles, they should be
numbered ...01.rgb, ...02.rgb, ~ ...09.rgb.
260 Marc Users Guide: Part I
Introduction
Appendix A: Shape Function Interpolation
Displacement shape or interpolation Iunctions are a central Ieature oI the displacement-based Iinite element method.
They primarily characterise the assumptions regarding the variation oI displacements within each element. Because
oI their relationship with displacements, the variation oI both strains and stresses is also consequently deIined.
The basic assumption oI the Iinite element method is that the subdivision oI a complex physical structure into the
assembly oI a number oI simple 'elements will approximate the behaviour oI the structure. Because oI this
subdivision, each Iinite element need not attempt to simulate the complex behaviour oI the whole structure but,
rather, assumes a relatively simple displacement variation so that the sum oI the individual Iinite element responses
approximates the response oI the whole structure.
Shape Iunctions are polynomial expressions. Any order oI polynomial can theoretically be used but, in general, linear
and quadratic variations are most common. It is Irom the order oI the shape Iunction polynomial that the terms linear
and quadratic elements originate.
A consequence oI these assumed displacement variations enables the Iinite element method to be able to solve the
equilibrium equations at discrete points, thus transIorming a continuous 'physical system (having inIinite degree oI
Ireedom) into something manageable Ior numerical procedures.
Typically, Marc uses Lagrangian shape Iunctions which provide C(0) continuity between elements (primary
variables only, and not their derivatives, are continuous across element boundaries). Shape Iunctions are deIined in
terms oI the natural coordinate system Ior line elements (bars, beams), Ior surIace elements (shells,
plates, plane membranes,) and Ior volume elements (solids).
For many two-noded line elements a linear variation is assumed as Iollows:
where and are the shape Iunctions at nodes 1 and 2 oI the element respectively (the order being dependent
on the element node numbering). Diagrammatically, their variation is as Iollows:
( ) q , ( )
q , , , ( )
N
1
1
2
--- 1 ( )
N
2
1
2
--- 1 ( )
N
1
N
2
1
1
Variation of N
1
1
1
1 1
Variation of N
2
261
CHAPTER 1
Introduction
Linear variations are also used on Iour-noded surIace elements as Iollows:
Three-noded line elements typically assume a quadratic variation as Iollows:
where , , and are the shape Iunctions at nodes 1, 2 and 3 oI the element respectively. Diagrammatically,
their variation is as Iollows:
Quadratic variations are also used on eight-noded surIace elements. The Iollowing diagrams show the variations oI
the shape Iunctions at both corner and midside nodes.
N
1

2
-- - 1 ( )
N
2
1 ( ) 1 ( )
N
3

2
-- - 1 ( )
N
1
N
2
N
3
1 0 1
1
Variation on N
1
1
Variation on N
1
1 0 1
Variation on N
3
1 0 1
1
Midside Nodes Corner Nodes
262 Marc Users Guide: Part I
Introduction
Shape Iunctions need to have the Iollowing characteristics:
(Ior )
This means that the value oI each shape Iunction evaluated at its nodal position must be unity. For example,
(Ior )
Requires that the values oI each shape Iunction, evaluated at the other nodes must be zero. That is,
The sum oI all the shape Iunctions, evaluated at any point must be unity. That is
Furthermore, to ensure that a Iinite element convergences to the correct result, certain requirements need to be satisIied
by the shape Iunctions, as Iollows
The displacement Iunction should be such that it does not permit straining oI an element to occur when the
nodal displacements are caused by rigid body displacement. This is selI evident, since an unsupported
structure in space will be subject to no restraining Iorces.
The displacement Iunction should be oI such a Iorm that iI nodal displacements produce a constant strain
condition, such constant strain will be obtained. This is essential since a signiIicant mesh reIinement will
cause near-constant strain conditions to occur in elements and they must be able to handle this condition
correctly.
N
i

j
q
j
,
j
1 i j
N
1
1

2
--- 1 ( ) 1
N
2 ( 0)
1 ( ) 1 ( ) 1
N
3 x ( 1)

2
--- 1 ( ) 1
N
i

j
q
j
( ) 0 i j =
N
1 ( 0)

2
--- 1 ( ) 0
N
1 ( 1)

2
--- 1 ( ) 0
N
i
q , ( )
i

1
N
1 ( 1 2)

2
--- 1 ( )
1
8
---
N
2 ( 1 2)
1 ( ) 1 ( )
6
8
---
N
3 ( 1 2)

2
--- 1 ( )
3
8
---
263
CHAPTER 1
Introduction
The displacement Iunction should ensure that the strains at the interIace between elements are Iinite (even
though indeterminate). By this, the element boundaries will have no 'gaps appear between them and, hence,
will show a continuous mesh.
The Iollowing sections deal with some oI the more Irequently encountered practical implications that are related to the
use oI shape Iunctions.
Implication: The Evaluation of Element Displacements
The isoparametric element Iormulation assumes that
where are the displacements at any point within an element and are the displacements at the nodes oI an
element.
This equation relates the displacements at any point within an element to the nodal displacements according to the
element shape Iunction . ThereIore, the displacement at any point in a two-noded line element can be
obtained Irom the nodal values using the Iollowing equation
II this element is Iully Iixed at one end and sustains a displacement oI 2 at the other end , the
displacement at the centre oI this element (0) would be given thus
i.e., halI the end displacement as expected.
The same can be done with any quantity that varies across an element, Ior example, coordinates, strain, stress, and
thickness.
Implication: Linear Versus Quadratic Elements?
Consider a 3-noded element that uses a quadratic shape Iunction variation oI the Iorm
The quadratic terms in thus giving a corresponding quadratic variation oI displacement over the element. The strain
variation can be deIined as
u { } N | | d { }
u { } d { }
N | | ( )
u ( ) N
1
( )d
1
N
2
( )d
2

1
2
--- 1 ( )d
1
1
2
--- 1 ( )d
2

d
1
0 ( ) d
2
2 ( )
u 0 ( )
1
2
-- - 0 ( )
1
2
--- 2 ( ) 1
N | |

2

2
-------------- 1
2

2

2
--------------
264 Marc Users Guide: Part I
Introduction
where is that strain-displacement matrix and are the three element axial displacements. It can be seen Irom
the terms that the strain is now a linear variation as will be the stress variation. In a similar manner, Ior a linear
element, the strain and stress variation will be constant.
This has a direct bearing on the type oI element to be chosen Ior an analysis. For instance, consider a bar element under
the action oI a constant uniIormly distributed load along the length oI the element. The resulting axial Iorce variation
will be theoretically linear as in the topmost picture oI the Iollowing diagram.
II this bar is modeled using linear elements (i.e., linear terms in the shape Iunction), the axial Iorce will be
approximated by a constant, 'stepped response in each element, since the shape Iunction derivatives only contain
constant terms. A quadratic element (i.e., quadratic terms in the shape Iunction) will, however, support a linear
response and provide the correct answer directly, since the shape Iunction derivatives contain linear terms. Thus, the
exact solution can be obtained with a relatively small number oI elements (or even with one element only) iI the actual
strain Iield can be matched by the shape Iunctions oI the element that is being used. In the above example, the shape
Iunction derivative terms did indeed match the linear strain oI the actual analysis.
A Irequent observation when inspecting Iorce output at a simply supported section oI a structure is to Iind
(unexpectedly) non-zero values. Depending on the degree oI mesh reIinement, these values can be signiIicant
compared to the peak values. The reason is directly related to the above discussion. For example, iI the Iorce
distribution is at least quadratic in Iorm and linear elements are used (typically supporting a constant Iorce
distribution), a stepped response will be seen hence the nonzero values these constant values represent an average
c
x
B | | d { }
d
dx
------
1
2
--- 2
1
2
--- d { }
1
J
-- -
1
2
--- 2
1
2
--- d { }
B | | d { }

Four Element
Mesh
Theoretical
Solution
Axial Force
Linear Elements
(constant 'stepped' solution)
One Element
Mesh
Quadratic Elements
(linear solution)
265
CHAPTER 1
Introduction
oI the Iorce distribution and, iI summed across the structure would be Iound to be equilibrium. The use oI quadratic
elements will improve the situation, but even these will not be able to match third order or higher Iorce distributions
without a measure oI mesh reIinement perIormed.
In spite oI this sort oI discrepancy, it should be noted that, during the solution stage, the equilibrium equation is used
to ensure that the product oI the stiIIness matrix and the computed displacements exactly
balances the externally applied Iorces. This means that, unless there are pertinent warnings or errors output during the
solution, static equilibrium will have been Iully achieved. Moreover, the derived quantities oI strain and stress will
also be Iound to be in equilibrium but not necessarily according to an expected distribution as noted in preceding
paragraphs.
Similar diIIiculties can be observed when attempting to compare the reactions at a location in a structure with the
element Iorce output at the same location. The explanation in most cases is, again, related to the order oI shape Iunction
that has been used to Iormulate the element.
The remedies are to either increase the number oI linear elements used (and reduce the size oI the 'step change
between each element) or change to quadratic elements (to more closely match the actual variation). The speciIic
element notes section in Marc Volume B: Element Library will typically give details on the variation oI Iorce that is
supported by each element.
Apart Irom the consideration oI element selection related to the order oI shape Iunction, quadratic elements would be
recommended in the presence oI high degrees oI plastic strain since they are less susceptible to 'locking. Linear
elements, however, would be recommended when the stress distributions anticipated are constant or linear. Such
elements are computationally cheaper and, in such circumstances, render the use oI higher order elements unnecessary.
Implication: Nodal Temperature Loading With Temperature
Dependent Materials
Although the temperature loading is deIined at element nodes, it is actually used by Marc at a Gauss point level. The
nodal temperature loading is interpolated Irom the nodes to the Gauss points using the element shape Iunctions.
The presence oI signiIicant temperature loading distributions over higher order elements can cause negative
temperature loading to be applied at the Gauss points even though the applied temperature Iield is entirely positive
in magnitude. Such negative temperatures can be unexpectedly out oI the user-speciIied temperature dependent
material property table.
As an example, consider the situation described in the Iirst oI the Iollowing diagrams. The temperature loading is
applied at the nodes as shown.
f { } K | | d { } ( )
266 Marc Users Guide: Part I
Introduction
As a result oI the quadratic displacement assumption used in higher order elements, the interpolation to the Gauss
points yields the variation oI temperature across the length oI the element as shown.
This variation will ensure that the applied temperature loading is applied correctly to the structure, but Ior the Gauss
points nearest to the zero temperature speciIication this may not be so.
For most cases the negative value is insigniIicant compared to the temperature loading speciIied and the variation in
the temperature dependency oI the material properties. Mesh reIinement in the area oI the greatest temperature
variations is the most appropriate remedy.
Implication: Element Thickness Interpolation
Although the thickness Ior an element is deIined at element nodes, it is actually used by Marc at a Gauss point level.
The thickness is interpolated Irom the nodes to the Gauss points using the element shape Iunctions.
For a constant thickness element, the interpolation will always produce the same constant value at the Gauss points.
For a varying thickness over an element, the actual thickness used will not be that speciIied at the nodes, but rather an
interpolated value. See the top diagram below.
x
x x
x
0
0 100
100
Gauss Points
Nodal
Temperatures
Temperature Variation
Across Element
T = 100 T = 100 T = 0
20
20
0
100
267
CHAPTER 1
Introduction
When using the quadratic displacement assumption used in higher order elements, the interpolation to the Gauss points
yields the variation oI thickness across the element as shown in the second picture (above). The eIIect oI a signiIicant
variation oI thickness over a single element may, thus, cause a zero or negative thickness value at a Gauss point.
The remedy is to check that the thickness variation applied to the speciIied element is applied correctly. II so, then the
mesh should be reIined to reduce the severity oI the thickness variation over the element.
x
x x
x
1
1 100
100
Gauss Points
Nodal
Thicknesses
Thickness Variation Across
Element
t = 100 t = 1
20
20
1
100
268 Marc Users Guide: Part I
Introduction
Appendix B: Finite Element Equilibrium
In terms oI Iinite element equilibrium, there are two important properties that are always satisIied by the Iinite element
solution using either a coarse or a Iine mesh. To describe these properties consider the Iollowing portion oI a mesh
under the application oI an arbitrary Iorce, the Iour elements (1,2,3,4) share the same node (i).
The Iollowing diagram, representing an exploded view oI these Iour elements, shows the Iorces obtained at the nodal
position (i) and those on element (2).
The two properties may now be deIined as
Nodal point equilibrium: At any node, the sum oI the internal element point Iorces is in equilibrium with any
external loads that are applied to the node. The internal Iorces include the eIIects due to body Iorces, surIace
tractions, initial stresses, concentrated loads, inertia and damping Iorces, and reactions. Thus, Ior an externally
non-loaded node in a linear static analysis, such a summation will be zero.
Element equilibrium: Each element is in equilibrium under its internal Iorces
1 2
3 4
i
1 2
3 4
Element Iorces
are in equilibrium
Sum oI nodal Iorces
are in equilibrium
with external Iorces
269
CHAPTER 1
Introduction
Nonlinear analyses may produce out-oI-balance residual Iorces at a node, depending on the degree oI convergence
obtained during the solution. For a well-converged solution, however, these are insigniIicant. See nonlinear iterative
strategy Ior more inIormation.
Although nodal and element equilibrium is achieved as described above, in a general Iinite element analysis,
diIIerential equilibrium (e.g. stress equilibrium) is not necessarily achieved at all points oI the continuum considered
most notably at the shared boundaries oI elements. The reason is as Iollows.
In the displacement-based Iinite element method, a continuous approximation Ior the displacements is assumed
within each element. This means that the displacements at any point in a mesh will be continuous and ensures that no
gaps appear between elements. The element stresses are calculated using derivatives oI the displacement, which means
that they will not necessarily be continuous and give rise to inter-element discontinuities or 'jumps in stress between
adjacent elements. This is particularly the case Ior coarse element meshes. The discontinuities at adjacent element
boundaries are reduced with mesh reIinement and the rate at which mesh reIinement reduces such discontinuities is
determined by the order oI the elements in the mesh higher order elements converging Iaster than low order.
For the same reasons that element stresses are not continuous across element boundaries, the element stresses at the
surIace oI a Iinite element model are, in general, not in equilibrium with the external applied tractions. Again, this
eIIect is minimised with mesh reIinement.
Experience has shown that the most accurate locations Ior stress output are the Gauss points. Nodal points, which are
the most accessible, are actually the worst output location Ior stresses. Reasons have been given above, but include
the Iact that shape Iunctions tend to behave badly at element extremities and it is reasonable to expect that the shape
Iunction derivatives (i.e. strains/stresses) sampled in the interior oI the element would be more accurate than those
sampled at the periphery oI the element.
This evokes the question oI how to obtain accurate stress results Irom a Iinite element model?
Implication: Smoothed or Unsmoothed Stress Contours?
One method Ior obtaining reasonable nodal stress output is by extrapolating the 'exact stresses at Gauss points to the
nodal positions using the element shape Iunctions. Consider the Iollowing diagram, representing the same exploded
view oI the Iour elements shown earlier.
C
0
270 Marc Users Guide: Part I
Introduction
For element 2, the nodal stresses at nodes (1,2,3,4) are obtained by
DeIining a Iictitious element (shown by the dashed lines) with nodes at the element Gauss points (a,b,c,d)
Extrapolating the Gauss point stresses to the nodal points oI the real element (1,2,3,4) using the displacement
shape Iunctions oI the Iictitious element, i.e.
where is the number oI Gauss points, and subscripts and denote nodal and Gauss point values,
respectively.
The accuracy oI the extrapolation procedure is dependent on both the presence oI a reasonably uniIorm
stress Iield and the type oI shape Iunction used in the element chosen. For instance, a high stress gradient
across an element would be more likely to extrapolate incorrectly, particularly iI a linear shape Iunction
element is being used.
This procedure is carried out Ior the other elements and the nodal stresses at the common node are obtained as
, , etc. As pointed out above, these stresses are not usually equal and a single 'averaged or 'smoothed
nodal stress value is obtained using
Gauss
Points
4 3
2 1
1 2
3 4


b a
d c
(oi)
3
(oi)
2
(oi)
1
(oi)
4
o
i
N
I

i
q
i
, ( )o
I
I I
N

N i I
i ( )
o
i
( )
1
o
i
( )
2
271
CHAPTER 1
Introduction
When this procedure is carried out Ior all nodes in an element assembly, the ensuing averaged stress values provide a
reasonable approximation to a continuous stress Iield. This is a straightIorward and economic solution and works well
on the whole. See later Ior more details on the circumstances that smoothing should not be used. This is the deIault
method used in Mentat when smoothed results are selected in the contour layer properties. II smoothed results are not
selected then the extrapolation procedure is still perIormed, but the averaging process is omitted.
Note that, Ior shell elements, the local Gauss point stresses and strains are transIormed to global stresses and strains
beIore extrapolation to the nodes. The mean global stresses are then transIormed to the local shell system at the nodal
point beIore evaluation oI the nodal stress resultants.
Other methods are available, based on a least squares Iit over the integration point stress values oI the elements. The
least squares procedure might be applied over the patches oI adjacent elements or even globally over a whole mesh.
However, iI the domain over which the least squares Iit is applied involves many stress points, the solution will be
expensive and, in addition, a large error in one part oI the domain may aIIect rather strongly the least squares prediction
in the other parts.
In general, it is recommended to display unsmoothed stress contours at an early point during the processing oI results.
In this way, severe stress discontinuities between elements will be apparent and the possible requirement oI mesh
reIinement and/or the use oI higher order elements may be considered.
In areas oI interest where the stress results will be used in the design process, smoothed contours would ideally be
similar to unsmoothed contours. The inIerence Irom this being that a smooth stress transition across the element
boundaries indicates that the stress distribution in the structure is being simulated suIIiciently accurately. For sections
oI the model that are not oI interest, a coarser mesh would normally be used and such a comparison in these areas
would typically give signiIicantly diIIerent contours - smoothed contours appearing more like a patchwork quilt!
The nodal averaging technique is suIIiciently robust that such stress values will tend to be pretty much those that would
be obtained at the same location with mesh reIinement as long as the element mesh is reasonably uniIorm.
At all times, it is imperative to remember that the Iinite element method is an approximate numerical technique
(albeit a good one) and that smoothed stress results can give good results but need careIul attention.
Implication: Limitations of the Averaging Scheme
In addition to taking no account oI the size oI the adjacent elements, the averaging method must not be used:
At mesh locations in which geometric or material properties change
For local or global stress output Ior shell elements that are nonplanar
o
i
o
i
( )
1
o
i
( )
2
o
i
( )
3
o
i
( )
4

4
--------------------------------------------------------------
272 Marc Users Guide: Part I
Introduction
Interconnecting BEAM elements. It is necessary to extract results Ior longitudinal members and transverse
members separately.
Consider the situation in which results have been selected Ior display, and both transverse and longitudinal
members are active. The averaged value displayed at the central node number (1) will be comprised oI the local
values Irom the two longitudinal and two transverse members that connect to this node. It must be noted that the
values are local to the elements (as shown in the diagram), so that the values Ior the longitudinal members act at
90 to the results in the transverse members. This means that the averaged values will be meaningless since the
results Irom the longitudinal members will be averaged with the values oI the transverse members that are
acting in a completely diIIerent direction.
o
x
o
y
o
x
o
y
o
x
o
y
1
2
y x
z
y x
z
y x
z
y x
z
y x
z
y x
z
x
y
z
x
y
z
x
y
z
x
y
z
x
y
z
x
y
z
Longitudinal
Members
Transverse
Members
1
M
x
M
x
M
x
M
x
M
x
M
x
M
x
273
CHAPTER 1
Introduction
Appendix C: Coordinate Transformation
The stress transIormation equations Ior plane stress are as Iollows
is the angle between the and the rotated -axis:
Appendix D: Principal Stresses (Plane Stress)
The principal stresses are the maximum values oI normal stresses under the condition:
and
and , , and are deIined in the diagram in Appendix C: Coordinate Transformation:
Giving, Ior the plane stress condition (using the transIormation equations Irom Appendix C: Coordinate
Transformation):
And leading to
|
where and are two perpendicular cutting directions, called Principal Directions.
o
x
1 2 o
x
o
y
( ) 1 2 o
x
o
y
( ) 2 t
xy
2 sin cos
o
q
1 2 o
x
o
y
( ) 1 2 o
x
o
y
( ) 2 t
xy
2 sin cos
t
q
1 2 o
x
o
y
( ) 2 t
xy
2 cos sin
x
y
x
q

do
x
d 0
do
q
d 0
q
o
x
o
y
( ) 2 ( ) 2t
xy
2 ( ) cos sin 0
2* ( ) tan 2 * t 2 ( ) | | tan 2t
xy
| | o oy | |
* * t 2 ( )
274 Marc Users Guide: Part I
Introduction
Appendix E: Python Example (Max Stress Results)
# --------------------------------------------------------------------
# Purpose:
# PyPost example
# Find the max nodal scalar values
#
# Usage:
# python <file>.py
#
# Dependencies
# Uses PyPost methods:
# node_scalars
# node_scalar_labels
# moveto
# Notes
# index = node/element number
# id = internal node/element index number (pointer)
# Scalar stress/strain results values are:
# Comp 11 of Cauchy Stress
# Comp 22 of Cauchy Stress
# Comp 33 of Cauchy Stress
# Comp 12 of Cauchy Stress
# Comp 23 of Cauchy Stress
# Comp 31 of Cauchy Stress
#
# Comp 11 of Total Strain
# Comp 22 of Total Strain
# Comp 33 of Total Strain
# Comp 12 of Total Strain
# Comp 23 of Total Strain
# Comp 31 of Total Strain
#
# Equivalent Cauchy Stress
# Total Strain Energy Density
# --------------------------------------------------------------------
from py_post import *
import sys
# specify the post file to read
def main(fname):
# open post file and define as object
p = post_open(fname)
# select loadcase of interest - don't forget that increment 0 is counted. this means
# that the last increment for a 10 increment analysis will be increment number 11
p.moveto(11)
#-------------------------------initialisation
# ...arrays
element_list = []
celements = []
max_scalars = []
max_nodes = []
total_volume = []
# ...variables
stress_threshold = 1100
#-------------------------------extract global variables
# ...number of increments in post file
nincrements = p.increments()
# ...number of contact bodies
n_contact_bodies = p.cbodies()
print 'number of contact bodies = ', n_contact_bodies
# ...time at this increment
inc_time = p.time
# ...increment number
inc_number = p.increment
# ...extrapolation method
p.extrapolation('translate')
print 'Extrapolation method is ', p.extrapolate
#-------------------------------loop over the contact bodies
for i in range(0, n_contact_bodies):
# contact body id number
275
CHAPTER 1
Introduction
cid = p.cbody(i).id
# contact type
ctype = p.cbody(i).type
# contact body type
cbodytype = p.cbody(i).bodytype
# number of elements in contact body
cnelements = p.cbody(i).nelements
# element id list for contact body
celements = p.cbody(i).elements
# contact body name
cname = p.cbody_name(i)
# contact body volume
cvolume = p.cbody_volume(i)
# print contact information
print ' contact body id: ', cid
print ' body name = ', cname
print ' body volume = ', cvolume
print ' body type = ', ctype
print ' contact body type = ', cbodytype
print ' elements in contact body = ', cnelements
print ' increment time = ', inc_time
print ' increment number = ', inc_number
#-------------------------------loop over the elements in each contact body
for j in range(0, cnelements):
# extract element number
element_number = celements[j]
# extract element id
element_id = p.element_sequence(element_number)
# print element number/id
# print ' element id: ', element_id
# print ' element number ', element_number
# extract number of nodes on this element
nnodes = p.element(element_id).len
# extract number of element scalars available
nelement_scalars = p.element_scalars(element_id)
# print ' number of element scalars: ', nelement_scalars
# print ' number of nodes: ', nnodes, ' ', p.element(j).items
# initialise current element volume
element_volume = 0
#-------------------------------loop over the scalars to find the current element volume
for k in range(0, nelement_scalars):
# extract the element scalars data label
scalar_label = p.element_scalar_label(k)
# look for volume of this element
if (scalar_label == 'Current Volume'):
# extract the element volume data (nodal-based)
tlist = p.element_scalar(element_id, k)
# extract the number of bits of information in the scalar
length_scalar = len(tlist)
# print the scalar label
# print ' scalar: ', scalar_label, ' (size: ',length_scalar,')'
#-------------------------------loop over the nodal volume data
for m in range(0, length_scalar):
# ...extract next scalar component
scalar_data = tlist[m].value
# ...print the scalar values
# print ' ',scalar_data
# check for a specific scalar result for further processing
element_volume = element_volume + scalar_data
# store the total volume for each element here for later use
total_volume.append(element_volume)
# print the total volume for this element
# print ' total volume: ',total_volume[j]
#-------------------------------loop over the contact bodies again for the stress/strain results
for ii in range(0, n_contact_bodies):
# contact body id number
cid = p.cbody(ii).id
# number of elements in contact body
cnelements = p.cbody(ii).nelements
# element id list for contact body
celements = p.cbody(ii).elements
#-------------------------------loop over the elements in each contact body
276 Marc Users Guide: Part I
Introduction
for jj in range(0, cnelements):
# extract element number
element_number = celements[jj]
# extract element id
element_id = p.element_sequence(element_number)
# print element number/id
# print ' element id: ', element_id
# print ' element number ', element_number
# extract number of nodes on this element
nnodes = p.element(element_id).len
# extract number of element scalars available
nelement_scalars = p.element_scalars(element_id)
#-------------------------------loop over the number of element scalars
for nn in range(0, nelement_scalars):
# print ' element number ', element_number
# extract the element scalars data label
scalar_label = p.element_scalar_label(nn)
# extract the element scalars data
tlist = p.element_scalar(element_id, nn)
# extract the number of bits of information in the scalar
length_scalar = len(tlist)
# print the scalar label
# print ' scalar: ', scalar_label, ' (size: ',length_scalar,')'
iloop_counter = 0
#-------------------------------loop over the data within the scalar
for ip in range(0, length_scalar):
# ...extract next scalar component
scalar_data = tlist[ip].value
# ...print the scalar values
# print ' ',scalar_data
# check for a specific scalar result for further processing
if (scalar_label == 'Comp 11 of Cauchy Stress'):
if (iloop_counter == 0):
# print ' scalar data: ', scalar_data
if (abs(scalar_data) > stress_threshold):
# print ' scalar: ', scalar_label, ' (size: ',length_scalar,')'
print 'element ', element_number, 'exceeded threshold ',scalar_data,' volume ', total_volume[j]
# set the loop counter since we are searching for the FIRST occurrence of a stress above the threshold value
iloop_counter = 1
return 1
if __name__ == '__main__':
main("python_result_extract_3d_job1.t16")
# element_list.sort()
# print element_list()
# print '-------------------- '
The results Irom this script will look as Iollows
number of contact bodies = 2
Extrapolation method is translate
contact body id: 1
body name = rubber
body volume = 0.0499988384545
body type = 0
contact body type = 2
elements in contact body = 8
increment time = 1.0
increment number = 10
contact body id: 2
body name = metal
body volume = 0.0400002449751
body type = 0
277
CHAPTER 1
Introduction
contact body type = 2
elements in contact body = 8
increment time = 1.0
increment number = 10
element 17 exceeded threshold -1189.01721191 volume 0.00624981324654
element 18 exceeded threshold -1173.24584961 volume 0.00624981324654
element 19 exceeded threshold -1105.20239258 volume 0.00624981324654
element 21 exceeded threshold -1179.69567871 volume 0.00624981324654
element 22 exceeded threshold -1199.10046387 volume 0.00624981324654
element 23 exceeded threshold -1118.07836914 volume 0.00624981324654
element 26 exceeded threshold -1592.05969238 volume 0.00624981324654
element 28 exceeded threshold -1529.88928223 volume 0.00624981324654
element 29 exceeded threshold -1208.35742188 volume 0.00624981324654
element 30 exceeded threshold -2732.16430664 volume 0.00624981324654
element 31 exceeded threshold -1222.24694824 volume 0.00624981324654
element 32 exceeded threshold -2766.19677734 volume 0.00624981324654
278 Marc Users Guide: Part I
Introduction
Appendix F: Python Example (Displacements at Nodes)
#---------------------------------------------------------------------
#!/usr/bin/env python
#*description
# PyPost example to find the time history of a specified node
#
#*useage
# to obtain all the results from this script:
# python node_loop.py filename.t16 1
#
# to obtain only the results ids without scanning the post file fully:
# python node_loop.py filename.t16 0
#
# The results are in columns:
# inc number - time - dispx - dispy - dispz - rotx - roty - rotz
#
#---------------------------------------------------------------------
from py_post import *
import sys
#-----------------------------extract command line parameters
# ...post file name (without t16 extension)
mno = str(sys.argv[1])
# ...flag to indicate whether full scan or
# just post file summary required
nns_only = int(sys.argv[2])
def main(fname):
try:
# open post file
p = post_open(fname)
# vectors for user node numbers
nodes = []
# set pointer to first loadstep
p.moveto(1)
# extract number of nodal scalars available in
# the post file
nns = p.node_scalars()
#-----------------------------User Defined Variables Start:
# number of nodes for which results are needed
nnodes = 2
# node numbers of interest
nodes.append(18)
nodes.append(15)
#-----------------------------User Defined Variables End:
print " "
print " "
print " Found", nns, "Node Scalars in POST file",fname
279
CHAPTER 1
Introduction
print " "
# extract number of increments in analysis
ninc = p.increments()
print " Found", ninc-1, "Increments in POST File"
print " "
print " Scalar Results Found:"
print " ---------------------"
for i in range(0,nns):
print " ...",p.node_scalar_label(i)," (result number ",i+1,")"
print ""
# check to see if we need to skip to the
# end if a scan-only is requested
if nns_only > 0 :
print "\n"
print "\t\t Extracting Displacement (x,y,z) and Rotation (x,y,z)"
print "\t\t For Specified Nodes and All Load Increments... "
print "\t\t ----------------------------------------------------"
print "\n"

# initialise node loop counter value
kk = 0
#-----------------------------loop over number of nodes
while kk < nnodes:
# print header
print " "
print " History for NODE ",nodes[kk]
# initialise increment loop counter value
k = 0
#-----------------------------loop over number of increments
while k < ninc-1:
# point to next set of increment results
p.moveto(k+1)
# extract results for current node
# ...disp-x
dispx = p.node_scalar(p.node_sequence(nodes[kk]),0)
# ...disp-x
dispy = p.node_scalar(p.node_sequence(nodes[kk]),1)
# ...disp-x
dispz = p.node_scalar(p.node_sequence(nodes[kk]),2)
# ...disp-x
rotx = p.node_scalar(p.node_sequence(nodes[kk]),3)
# ...disp-x
roty = p.node_scalar(p.node_sequence(nodes[kk]),4)
# ...disp-x
rotz = p.node_scalar(p.node_sequence(nodes[kk]),5)
str = " %.3i\t%.5e\t%.5e\t%.5e\t%.5e\t%.5e\t%.5e\t%.5e " %
(k,p.time,dispx,dispy,dispz,rotx,roty,rotz)
print str
# increment increment counter
280 Marc Users Guide: Part I
Introduction
k = k + 1
# increment nodal loop counter
kk = kk + 1
except AttributeError:
print ""
print " Post File is not accessible: ", fname
return
if __name__ == '__main__':
main(mno)
Section 2: Recent Features
Section 2: Recent Features
Marc Users Guide 282
Chapter 2.1: New-style Tables Input
2.1
New-style Table Input

Summary of Reinforced Cylinder 284

Post Buckling Analysis of a Reinforced Shell with

Nonuniform Load 285

Input Files 304

Summary of Can Analysis 305

Can Analysis 306

Input Files 325

284 Marc Users Guide: Part I
CHAPTER 2.1
Summary of Reinforced Cylinder
Title Post buckling analysis oI a reinIorced cylinder with nonuniIorm load
Problem Ieatures New style oI table input Ior pressure load and user subroutine Ior post
Geometry
Material properties E 30x10
6
Psi, 0.3
Analysis type Static with Riks-Ramm arc-length load stepping method
Boundary conditions
Clamped at z 0 with external pressure oI
Element type Brick element type 75 and 98
FE results Pressure history as cylindrical shell buckles
C
L
X
Y Z
X
Y Z
360 in
r = 120 in
2x2 Beam rectangular
cross section
Pressure
Shell thickness
0.5 in
v
30* cos u ( ) 1
Z 180
180
------------------

\ .
| |

Loadcase Percentage Completion (x100)

0 4.46
0
1
Increment (x100)
Buckling of Reinforced Shell with Nonuniform Load
420
120
140160
180
200
220
240
260
280
300
320
340
360
380
400
440
100
80
40
60
0
20
1
285
CHAPTER 2.1
New-style Table Input
Post Buckling Analysis of a Reinforced Shell with
Nonuniform Load
This problem demonstrates the use oI applying a nonuniIorm load by deIining an equation to prescribe the pressure.
The load is placed on the geometric surIace. The thin shell is reinIorced with beam elements at the top. The
Riks-Ramm arc-length method is used to control the applied load.
Figure 2.1-1 Geometry of Tank
The cylindrical shell as shown in Figure 2.1-1 has a diameter oI 20 Ieet 240 inches, and a height oI 30 Ieet 360
inches. The shell thickness is 0.5 inch. The material is steel with Young`s modulus 30x10
6
psi, and Poisson`s
ratio 0.3. The steel beams have a square solid section with a 2 inch width, where the shell is at the midsection oI the
beam. The pressure magnitude has a cosine like distribution with a bilinear axial variation. The magnitude may be
expressed as
This distributed load is applied on only halI oI the surIace. This problem also demonstrates the use oI user subroutine
plotv.
The Iollowing steps are used to perIorm the analysis:
Step 1: Create Model
Step 2: Define Geometrical Properties for the Shells and Beams
Step 3: Define Material Properties
Step 4: Attach element edges to curves
Step 5: Apply Boundary Conditions
Step 6: Create loadcase
Step 7: Write user subroutine
Step 8: Create Job and submit model
Step 9: Review results
h
r
r = 120
h = 360
30* cos u ( ) 1
Z 180
180
------------------

\ .
| |

286 Marc Users Guide: Part I

CHAPTER 2.1
Step 1: Create Model
The cylindrical model (Figure 2.1-2) will be created Iirst by creating two ruled surIaces representing each halI oI the
cylinder and converting these to elements. Two surIaces are used as opposed to the conventional single surIace,
because the load is to be applied on only halI oI the surIace. The curves on the top are then converted to beam elements.
Figure 2.1-2 Definition of Cylinder Geometry
FILES
SAVE AS
reinforced (create model file reinforced)
OK
RETURN
MESH GENERATION
COORDINATE SYSTEM
SET (create grid)
U DOMAIN
-120 120
U SPACING
10
V DOMAIN
-120 120
V SPACING
10
287
CHAPTER 2.1
New-style Table Input
GRID (on)
FILL
RETURN
CURVE TYPE (create curves used to make ruled surface)
ARCS
CENTER/RADIUS/ANGLE/ANGLE
RETURN
CRVS ADD
0 0 0 120
-90 90
0 0 0 120
90 -90
PLOT
CURVES SETTINGS
PREDEFINED SETTINGS
HIGH
DRAW
DIRECTION (on)
REGEN
RETURN
SURFACES SETTING
HIGH
REGEN
RETURN (twice)
DUPLICATE
TRANSLATIONS
0 0 360
CURVES
1 2
# END LIST
RETURN
SURFACE TYPE (create surfaces)
SURFACES
RULED
TRIM (on)
TRIM NEW SURFACE
288 Marc Users Guide: Part I
CHAPTER 2.1
RETURN
SRFS ADD
1 3
SRFS ADD
2 4
CONVERT (convert surfaces to elements)
DIVISIONS
18 30 (as 18 divisions are used for each shell,
the shell elements will be 10 wide.
GEOMETRY/MESH The result is shown in Figure 2.1-3)
SURFACES TO ELEMENTS
1 2
# END LIST
RETURN
SWEEP
SWEEP
ALL
RETURN
COORDINATE SYSTEM
GRID (off)
CONVERT (convert curves to line elements)
GEOMETRY/MESH
CURVES TO ELEMENTS
3 4
# END LIST
RETURN
SWEEP:
ALL
RETURN
RENUMBER
ALL
RETURN
MAIN
289
CHAPTER 2.1
New-style Table Input
Figure 2.1-3 Mesh Converted from Surface and Automatically Attached (shown as yellow)
Step 2: DeIine Geometrical Properties Ior the Shells and Beams
The shell thickness, the beam cross section and orientation are deIined. The thickness is applied to the geometric
surIace. The beam is located at the top oI the shell.
MAIN MENU
PREPROCESSING
GEOMETRIC PROPERTIES
STRUCTURAL ELEMENTS
3-D
GEOMETRIC PROPERTY TYPE
SHELL
STRUCTURAL 3-D SHELL PROPERTIES
THICKNESS
0.5
OK
SURFACES ADD
1 2
# END LIST
RETURN
Y
X
Z
4
290 Marc Users Guide: Part I
CHAPTER 2.1
GEOMETRIC PROPERTIES
NEW
STRUCTURAL ELEMENTS
3-D
GEOMETRIC PROPERTY TYPE
SOLID SECTION BEAM
STRUCTURAL 3-D SHELL PROPERTIES
CROSS SECTION
PROPERTIES
ENTERED
AREA
4
Ixx
1.33333
Iyy
1.33333
VECTOR DEFINING LOCAL ZX-PLANE
Z
1
OK
SELECT
CLEAR SELECT
SELECT BY
ELEMENTS BY
CLASS
LINE (2)
OK
RETURN
MAKE VISIBLE
RETURN
ELEMENTS ADD (associate beam elements with the geometry)
ALL: VISIB.
RETURN
PLOT SETTINGS (display beam element orientation
BEAM as shown in Figure 2.1-4)
BEAM ORIENTATION
291
CHAPTER 2.1
New-style Table Input
Figure 2.1-4 Orientation of Reinforcement Beams at Top of Shell
DRAW LOCAL AXES (on)
REDRAW
RETURN
MAIN MENU
PREPROCESSING
MESH GENERATION
SELECT
ELEMENTS
ALL: EXIST.
MAKE VISIBLE
PLOT
ELEMENT (display shell orientation and thickness)
SETTINGS
RELATED PLOT SETTING
SHELL
PLOT EXPANDED (on)
DEFAULT THICKNESS
0.5
REDRAW
MAIN
Y
Y
Y
Y
X
X
Y
X
X
Y
X
Y
X
X
Y
Y
X
X
Y
Y
X
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
X
X
X Y
Y
Y
Y
Z
X
292 Marc Users Guide: Part I
CHAPTER 2.1
Step 3: DeIine Material Properties
Both the shell and the beams are made oI standard steel, and will remain elastic.
Step 4: Attach element edges to curves
The CONVERT option attached the shell elements to the surIace and attached the beam elements to the curve. To
Iacilitate the application oI boundary conditions, it is also useIul to attach the edges oI the shell elements to the curve.
This will be demonstrated in this step.
MAIN MENU
PREPROCESSING
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NEW
NEW MATERIAL
STANDARD
STRUCTURAL
STRUCTURAL PROPERTIES
YOUNGS MODULUS
30e6
POISSONS RATIO
0.3
OK
ELEMENTS ADD
ALL: EXIST.
MAIN MENU
PREPROCESSING
MESH GENERATION
ATTACH
EDGES -> CURVE
3
EDGES -> CURVE
4
EDGES -> CURVE
1
EDGES -> CURVE
2
MAIN
293
CHAPTER 2.1
New-style Table Input
Step 5: Apply Boundary Conditions
This problem has two boundary conditions: the base oI the shell is clamped, and a nonuniIorm pressure is applied. The
displacement boundary condition is applied to the curves at the base. The pressure is applied on a surIace by giving a
reIerence value oI 30 psi and reIerencing a table. This table deIines a mathematical equation. Then Ior each element
attached to the surIace, it will, Ior each integration point, determine the integration point coordinates and evaluate the
table. Later in this analysis, we will activate the FOLLOWER FORCE option. When applying distributed load type
boundary conditions to curves or surIaces, it is important to indicate iI the load is to the top or bottom part oI the
surIace. The SELECT option is used to Iilter the surIace.
MAIN MENU
PREPROCESSING
BOUNDARY CONDITIONS
STRUCTURAL
NAME (clamped boundary condition)
fixed_base
BOUNDARY CONDITION TYPE
FIXED DISPLACEMENT
X, Y, Z, RX, RY
OK
CURVES ADD
1 2
# END LIST
RETURN
DRAW BOUNDARY COND ON MESH
NEW (pressure boundary condition)
NAME
pressure
STRUCTURAL
FACE LOAD
30 (enter the reference value)
OK
SELECT (select the top side of surface 1)
FILTER
TOP
RETURN
CLEAR SELECT
SURFACES
1
294 Marc Users Guide: Part I
CHAPTER 2.1
The pressure load is to look like a cosine Iunction multiplied by a bilinear Iunction, such that the load is maximum at
Z180 and linearly decreases as one approaches the edge. A working sketch Ior deIining the load is shown in
Figure 2.1-5.
Figure 2.1-5 Definition of Function
Note that initially this is , but as deIormation occurs, this would no longer be true.
As the independent variables are given in the order oI 1x, 2y, 3z, when entering the equation, the variable names
are replaced with the generic names v
1
, v
2
, v
3
. The equation used is then:
(v1/sqrt (v1 * v1 + v2 * v2)) * (1 - (abs (v3 - 180)/180))
# END LIST
RETURN
SURFACES ADD
ALL SELEC. (associate the selected surface
with this boundary condition)
TABLES
NEW
3 INDEPENDENT VARIABLES (toggle)
INDEPENDENT VARIABLE V1
TYPE
X0_coordinate
OK
MAX
120
VARIABLES
MORE
X
1

Y Z
p =
{ } 30 cos ( ) 1
Z 180
180
------------------

u cos x l x x
2
y
2

x r
295
CHAPTER 2.1
New-style Table Input
LABEL
X0_coordinate
PREVIOUS
INDEPENDENT VARIABLE V2
TYPE
Y0_coordinate
MIN
-120
MAX
120
VARIABLES
MORE
LABEL
Y0_coordinate
PREVIOUS
INDEPENDENT VARIABLE 3
TYPE
Z0_coordinate
MIN
0
MAX
360
MORE
LABEL
Z0_coordinate
PREVIOUS
FORMULA (define function with a formula)
ENTER
(v1/sqrt (v1 * v1 + v2 * v2)) * (1 - abs (v3 - 180)/180)
FIX V3 (evaluate the table at the 5th value of V3 which
5 is located at z=180 or midway up the cylinder)
Use DYN. MODEL to rotate model (the Mentat evaluated table is shown in Figure 2.1-6)
NAME
load_factor
toggle SHOW TABLE to SHOW MODEL
RETURN
296 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-6 Table Describing Load Evaluated at Z=180, X-direction along Axis,
Y-direction is Out-of Plane. Complete surface, p = p(u,z), at lower right.
FACE LOAD
TABLE
load_factor
OK
MAIN
Note: When evaluating the Iunction, Mentat indicates numerical errors because it evaluates the Iunction at v
1
0, v
2
0, which
results in a divide by zero. In the analysis program, the Iunction is evaluated at the element integration points which are
not at this position.
1:11:5
2:11:5
3:11:5
4:11:5
5:11:5
1:10:5
2:10:5
6:11:5
3:10:5
4:10:5
7:11:5
5:10:5
8:11:5
6:10:5
1:9:5
2:9:5
3:9:5
4:9:5
9:11:5
7:10:5
5:9:5
2:6:5
10:11:5
8:10:5
6:9:5
3:8:5
4:8:5
2:8:5
3:7:5 5:8:5 7:9:5 9:10:5 11:11:5
1:8:5
4:7:5
3:6:5
2:7:5
6:8:5
8:9:5
10:10:5
5:7:5
4:6:5
7:8:5
9:9:5
11:10:5
6:7:5
8:8:5
1:7:5
5:6:5
10:9:5
7:7:5
9:8:5
11:9:5
6:6:5
8:7:5
10:8:5
3:5:5
1:6:5
4:5:5
7:6:5
9:7:5
2:5:5
11:8:5
5:5:5
8:6:5
10:7:5
F
6:5:5
1
1:5:5
9:6:5
11:7:5
7:5:5
10:6:5
3:4:5
8:5:5
4:4:5
2:4:5
5:4:5
1:4:5
11:6:5
9:5:5
6:4:5
7:4:5
10:5:5
8:4:5
1:3:5
2:3:5
3:3:5
table1
11:5:5
4:3:5
5:3:5
9:4:5
6:3:5
10:4:5
7:3:5
1:2:5
2:2:5
3:2:5
8:3:5
11:4:5
4:2:5
0
9:3:5
5:2:5
6:2:5
1:1:5
10:3:5
0
2:1:5
7:2:5
3:1:5
11:3:5
8:2:5
4:1:5
5:1:5
9:2:5
6:1:5
10:2:5
7:1:5
11:2:5
8:1:5
X0_coordinate (x100)
9:1:5
10:1:5
11:1:5
1.2
Y
0
_
c
o
o
r
d
in
a
t
e
z

-1.5
-1
-0.5
0
0.5
1
1.5
0
50
100
150
200
250
300
350
0
0.2
0.4
0.6
0.8
1


cos() [1-abs(z-180)/180]
297
CHAPTER 2.1
New-style Table Input
Step 6: Create loadcase
This demonstration problem contains one loadcase. It is anticipated that this thin shell will buckle, so the Arc Length
procedure is invoked (Figure 2.1-7).
For more inIormation about continuation methods in buckling analyses, see Marc Volume A: Theory and User
Information. In most analyses oI this type, it would be necessary to adjust the convergence tolerances. In this
simulation, this was not required.
Figure 2.1-7 Loadcase Menu
MAIN MENU
ANALYSIS
LOADCASES
ANALYSIS CLASS
STRUCTURAL
NEW
STATIC
PROPERTIES
ADAPTIVE
ARC LENGTH
PARAMETERS
MAX # INCREMENTS IN LOADCASE
500
OK (twice)
MAIN MENU
298 Marc Users Guide: Part I
CHAPTER 2.1
Step 7: Write user subroutine
In this problem, it is worthwhile displaying the actual applied pressure on the surIace oI the element associated with
the applied boundary condition. Marc, by deIault, places the total equivalent nodal load associated with all boundary
conditions on the post Iile. This may be displayed as a contour or a vector plot.
Here, additionally, we would like to see the pressure which is based upon the reIerence magnitude, the evaluation oI
the equation, and the Iraction oI the load applied. As this is not normally available, PLOTV user subroutine is invoked
based upon a user deIined post code. This subroutine will be called Ior every element oI the model. As the load is only
applied on the shell elements when , ignore all other elements.
There are Iive steps to achieve this:
1. Begin with a skeleton plotv.f routine obtained Irom the /user subdirectory or Irom Marc Volume D: User
Subroutines and Special Routines.
2. IdentiIy elements oI interest.
3. Obtain the integration point coordinates and store them in the appropriate place.
4. Evaluate the Iunction and scale with the a reIerence value.
5. Scale with the Iraction oI the load applied in this loadcase.
Subroutine tabva2 may be used to obtain the current value oI a table or equation by the user. It is documented in Marc
Volume D: User Subroutines and Special Routines. Here, the key parameters are:
List of User Subroutines
subroutine plotv(v,s,sp,etot,eplas,ecreep,t,m,nn,layer,ndix,
* nshearx,jpltcd)
c* * * * * *
c
c select a variable contour plotting (user subroutine).
c
c v variable
c s (idss) stress array
c sp stresses in preferred direction
c etot total strain (generalized)
c eplas total plastic strain
c ecreep total creep strain
c t current temperature
c m(1) user element number
c m(2) internal element number
c nn integration point number
c layer layer number
c ndi (3) number of direct stress components
refval the reIerence value; here 30 psi
prxyz the calculated pressure
idtab the table ID; here 1.
X 0 >
299
CHAPTER 2.1
New-style Table Input
c nshear (3) number of shear stress components
c
c* * * * * *
include '../common/implicit.cmn'
dimension s(*),etot(*),eplas(*),ecreep(*),sp(*),m(2)
include '../common/elmcom.cmn'
include '../common/ctable.cmn'
include '../common/array4.cmn'
include '../common/heat.cmn'
include '../common/space.cmn'
include '../common/autoin.cmn'
jcrxpt=icrxpt+lofr+(nn-1)*ncrdel
c
c obtain coordinates of integration point
c for distributed load on shell face, integration point location
c is the same as element stiffness integration point location
c if x-coordinate is less than zero, skip as load was only applied
c to half of cylinder
c
xyz0(1)=varselem(jcrxpt)
if(xyz0(1).gt.0.0.and.ndix.ge.2) then
xyz0(2)=varselem(jcrxpt+1)
xyz0(3)=varselem(jcrxpt+2)
c
c refval is reference value of applied pressure
c idtab is the table id
c prxyz is the value of the table/function after evaluation
c the original coordinates in xyz0 are passed into the
c evaluator via common/ctable/
c
refval=100.0
idtab=1
call tabva2(refval,prxyz,idtab,0,1)
else
prxyz=0.0
endif
c
c scale by the total percentage of load applied (autacc)
c
v=prxyz*autacc
c
return
end
Step 8: Create Job and submit model
The job will be created and submitted Ior analysis. A large displacement elastic analysis will be perIormed. In this
model, the Iour-node shell element, type 75 and the two-node elastic beam element, type 52 will be used. These are
deIault element types.
The output to be written to the post Iile is selected, and the inclusion oI the user subroutine is invoked.
300 Marc Users Guide: Part I
CHAPTER 2.1
MAIN MENU
ANALYSIS
JOBS
NEW
STRUCTURAL
PROPERTIES
AVAILABLE
Select lcase1
INITIAL LOADS
pressure (deactivate pressure from the initial conditions)
OK
ANALYSIS OPTIONS
FOLLOWER FORCE (toggle) (invoke follower force)
LARGE STRAIN
OK
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress (select post variable)
LAYERS
toggle DEFAULT to OUT & MID
AVAILABLE ELEMENT SCALARS
User Defined Var #1 (select post variable change label
Rename it pressure of user-defined post variable)
POST FILE
FREQUENCY
2
OK
OK
RUN
USER SUBROUTINE FILE
shellcos_buckle.f (select user subroutine)
COMPILE / SAVE
STYLE
TABLE-DRIVEN (toggle) (invoke table driven input)
ADVANCED JOB SUBMISSION
WRITE INPUT FILE
301
CHAPTER 2.1
New-style Table Input
Step 9: Review results
In this type oI analysis, it is interesting to examine how the load increased such that it reached the Iinal magnitude.
This is dependent on the accuracy requirements and in this model the buckling phenomena. The other areas oI interest
are the applied distributed load and the deIormations.
EDIT INPUT FILE
OK
SUBMIT
MONITOR
Note: This analysis will run Ior 5 - 15 minutes depending upon the computer.
MAIN MENU
POSTPROCESSING
RESULTS
OPEN
reinforced_job1.t16
PLOT
NODES (off)
POINTS (off)
SURFACES (off)
ELEMENT SETTINGS
SOLID
RETURN (twice)
HISTORY PLOT
COLLECT DATA
ALL INCS
SHOW IDS
2
ADD CURVES
GLOBAL
Increment
Loadcase Percent Completion
FIT (This is shown in Figure 2.1-8)
RETURN
302 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-8 Loadcase Percentage Completion versus Increment Number
SHOW MODEL
RETURN
RESET VIEW
FILL
rotate model
LAST
RETURN
CONTOUR BANDS
SCALAR
pressure
DEF ON (toggle) (examine pressure on deformedconfiguration,
see Figure 2.1-9)
Loadcase Percentage Completion (x100)
0 4.46
0
1
Increment (x100)
Buckling of Reinforced Shell with Nonuniform Load
420
120
140160
180
200
220
240
260
280
300
320
340
360
380
400
440
100
80
40
60
0
20
303
CHAPTER 2.1
New-style Table Input
Figure 2.1-9 Applied Pressure on Deformed Structure
You can observe that the load has a cosine-like distribution along the circumIerence and increases, then decreases
along the height. The maximum value is at (0,0,180).
The Iinal stress on the deIormed shell is shown in Figure 2.1-10.
SCALAR
Equivalent Von Mises Stress Top Layer
OK
SCALAR PLOT
SETTINGS
SET LIMITS
0 2.E5
MANUAL
RETURN
REWIND
MONITOR
-5.673e-002
9.969e+000
2.000e+001
3.002e+001
4.005e+001
5.007e+001
6.010e+001
7.013e+001
8.015e+001
9.018e+001
1.002e+002
lcase1
Pressure
Inc: 446
Y
X
Z
304 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-10 Equivalent Stress on Deformed Structure
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
reinforced.proc Mentat procedure Iile to run the above example
shellcos_buckle.f Associated user subroutine
2.278e+002
1.967e+004
3.912e+004
5.856e+004
7.801e+004
9.745e+004
1.169e+005
1.363e+005
1.558e+005
1.752e+005
1.947e+005
lcase1
Equivalent Von Mises Stress Top Layer
Inc: 446
Y
X
Z
305
CHAPTER 2.1
New-style Table Input
Summary of Can Analysis
Title Capped Cylindrical Shell
Problem Ieatures New style oI table input Ior material properties and boundary conditions
Geometry
Material properties E 10x10
6
Psi, 0.3, with workhardening
Analysis type Static with variable load stepping
Boundary conditions Fixed at z 0 with pressure in hole and on dome
Element type Brick element type 138
FE results Stress contours at end oI load history
X
Y
Z
Cylinder
r = 4 in
h = 12 in
Blend radius
r = 0.2 in
Hole
r = 1 in
center @ z = 6
Spherical Cap
v
4.048e-006
2.583e+003
5.166e+003
7.749e+003
1.033e+004
1.291e+004
1.550e+004
1.808e+004
2.066e+004
2.325e+004
2.583e+004
lcase1
Equivalent Von Mises Stress Top Layer
Inc: 61
Time: 1.000e+000
X
Y
Z
306 Marc Users Guide: Part I
CHAPTER 2.1
Can Analysis
This problem demonstrates the use oI tables Ior material properties and boundary conditions. Additionally, the
boundary conditions are applied on surIaces created Irom a solid model. The geometry is shown in Figure 2.1-11.
Figure 2.1-11 Creating the Geometric Model
The hollow cylinder is 12 inches high and has a radius oI 4 inches A cylindrical slot is located 6 inches Irom the bottom
and has a radius oI 1 inch. The simulation can be perIormed by exercising the Iollowing steps.
Overview of Steps
Step 1: Create the geometric model
Step 2: Define shell thickness
Step 3: Define Material Properties
Step 4: Apply Boundary Conditions
Step 5: Create Loadcase
Step 6: Job creation
Step 7: Evaluate Results
Step 1: Create the geometric model
The model consists oI a cylindrical shell with a hemispherical dome and a cylindrical slot. One side oI the cylindrical
slot is blended into the larger cylinder. The creation oI the Iinite element model involves three substeps.
A. Form solid model
B. Convert solid Iaces into surIaces
C. Use the automatic Delaunay mesher to create the mesh.
X
Y
Z
Cylinder
r = 4 in
h = 12 in
Blend radius
r = 0.2 in
Hole
r = 1 in
center @ z = 6
Spherical Cap
307
CHAPTER 2.1
New-style Table Input
It should be noted that since the automatic mesh generator is used, the Iaces oI the elements are automatically attached
to the surIaces. Additionally, element edges are attached to the curves.
FILES
SAVE AS
table_bc_shell
OK
RETURN
MESH GENERATION
SOLID TYPE
CYLINDER
RETURN
SOLIDS ADD (create cylinder)
0 0 0
0 0 12
4 4
FILL
SOLID TYPE
SPHERE
RETURN
SOLIDS ADD (create sphere)
0 0 12
4
COORDINATE SYSTEM
SOLIDS
BOOLEANS
UNITE
1 2
# END LIST
PLOT
SOLIDS
SOLID
DRAW
MAIN
CONFIGURATION
VISUALIZATION
308 Marc Users Guide: Part I
CHAPTER 2.1
The result is shown in Figure 2.1-12.
LIGHTING (on)
VIEW1 (on)
RETURN
RETURN
MESH GENERATION
SOLID TYPE
CYLINDER
RETURN
SOLIDS ADD
-6 0 6
6 0 6
1 1
COORDINATE SYSTEM
SOLIDS (create slot by subtracting cylinder from solid)
BOOLEANS
SUBTRACT
1 2 #
MISCELLANEOUS
SPLIT FACES
ALL: EXIST.
BLEND
RADIUS
0.2
CHAMFER
ROLLING (turn on rolling blend)
EDGE
1:5
RETURN
309
CHAPTER 2.1
New-style Table Input
Figure 2.1-12 Solid Representation of Can
CONVERT
FACES TO SURFACES
ALL: EXIST.
PLOT
POINTS (off)
SOLIDS (off)
SURFACES
SETTING
SOLID
RETURN
IDENTIFY (For application of boundary conditions
and for mesh generation, it is important
that the surface orientation is consistent.)
BACKFACES
DRAW (the orientation is shown in Figure 2.1-13)
310 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-13 Orientation of Surfaces before Slot is Flipped
MAIN MENU
PREPROCESSING
MESH GENERATION
CHECK
FLIP SURFACES (make sure all surfaces have
1 2 4 6 # a consistent orientation)
FILL
RETURN
SWEEP
ALL
RETURN
AUTOMESH
PRELIMINARY
CURVE DIVISIONS
FIXED AVG LENGTH
AVG LENGTH
0.6
APPLY CURVE DIVISION
ALL: EXIST.
TOOLSS
MATCH CURVE DIVISIONS
0.1
311
CHAPTER 2.1
New-style Table Input
Figure 2.1-14 Finite Element Mesh Created with Delaunay Mesh Generator
ALL: EXIST.
RETURN
CHOOSE
SURFACE MESHING (create finite element mesh on the surface)
TRIANGLES (DELAUNAY)
SURFACE TRI MESH!
ALL: EXIST. (the finite element mesh is show in Figure 2.1-14)
PLOT
SURFACES (off)
ELEMENTS
SETTING
EDGES
OUTLINE
DRAW (this allows one to check for gaps
in the model, see Figure 2.1-15)
RETURN
RETURN
RETURN
312 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-15 Finite Element Mesh - Elements in Outline Mode to Verify No Gaps in Model
SWEEP
TOLERANCE
0.1
SWEEP
NODES
ALL: EXIST.
PLOT
NODES (off)
CURVES (off)
DRAW
RETURN
RETURN
AUTOMESH
CURVE DIVISIONS
CLEAR CURVE DIVISION
ALL: EXIST.
RETURN
MAIN
313
CHAPTER 2.1
New-style Table Input
Step 2: DeIine shell thickness
The shell has a thickness oI 0.1 applied to all elements.
Step 3: DeIine Material Properties
The shell is made up oI steel. It is anticipated that the stress will exceed the yield stress. The workhardening data is
entered as a table.
GEOMETRIC PROPERTIES
STRUCTURAL ELEMENTS
3-D
SHELL
THICKNESS
0.1
OK
ELEMENTS ADD
ALL EXIST.
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
workhardening
TYPE
eq_plastic_strain
FUNCTION VALUE F
MIN
20000
MAX
30000
ADD
0 20000
0.1 23000
0.3 25000
0.6 26000
1.0 27000
314 Marc Users Guide: Part I
CHAPTER 2.1
MORE
INDEDPENT VARIALBE V1
LABEL
Equivalent Plastic Strain
FUNCTION VALUE F
LABEL
Flow Stress (the flow stress table is shown in Figure 2.1-16)
RETURN
SHOW MODEL (toggle)
RETURN
NEW
STANDARD
STRUCTURAL
YOUNGS MODULUS
1.e7
POISSONS RATIO
0.3
PLASTICITY
YIELD STRESS
1.0
TABLE
workhardening
OK (twice)
ELEMENTS ADD
ALL: EXIST.
MAIN
315
CHAPTER 2.1
New-style Table Input
Figure 2.1-16 Flow Stress Table
Step 4: Apply Boundary Conditions
The Can has three boundary conditions:
1. The Can is constrained by prescribing a Iixed displacements in the z-direction to the bottom surIace and
clamping the curve at the bottom edge.
2. Applying a nonuniIorm load to the surIace representing the circular slot.
3. Apply a uniIorm pressure to the spherical cap.
The nonuniIorm load is applied by deIining a bilinear equation, where the independent variables are the time and the
x-coordinate position. The load on the dome is ramped up as a Iunction oI time.
BOUNDARY CONDITIONS
STRUCTURAL
NAME
no_axial (axial constraint on surface)
FIXED DISPLACEMENT
DISPLACEMENT Z
0
OK
SURFACES ADD
9 # (BOTTOM SURFACE)
RETURN
DRAW BOUNDARY CONDS ON MESH
NEW (clamp condition on curve)
2
1 0
Flow Stress (x10000)
Equivalent Plastic Strain
workhardening
3
1
3
5
4
2
1
316 Marc Users Guide: Part I
CHAPTER 2.1
STRUCTURAL
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Y
ROTATION X
ROTATION Y
OK
CURVES ADD
16 33 43 44 #
NEW
NAME
press_in_hole (pressure in slot)
TABLES
NEW
2 INDEPENDENT VARIABLES
INDEPENDENT VARIABLE V1
TYPE
x0_coordinate
MIN
-4
MAX
4
FUNCTION VALUE F
MAX
4
INDEPENDENT VARIABLE V2
TYPE
time
FORMULA
ENTER
(4-abs (v1)) * v2
MORE
FUNCTION VALUE F
LABEL
time
INDEPENDENT VARIABLE V1
317
CHAPTER 2.1
New-style Table Input
Figure 2.1-17 Table Describing Pressure in Slot
LABEL
x-coordinate
PREVIOUS
SHOW IDS
0
RETURN (this function is shown in Figure 2.1-17)
FACE LOAD
PRESSURE
800
TABLE
table2
OK
SURFACES ADD
4 6 #
NEW
NAME
domeload (uniform pressure on dome)
TABLES
NEW
INDEPENDENT VARIABLES V1
F
4
table2
0 -4
x-coordinate
4
1
318 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-18 All Boundary Conditions Shown
Step 5: Create Loadcase
The three boundary conditions are combined into a single loadcase. The adaptive time stepping procedure is used.
Because the shell is thin and subjected to an external pressure on the dome, the nonpositive deIinite Ilag is activated.
All other parameters are deIault.
NAME
ramp
TYPE
time
ADD
0 0
1 1
RETURN
FACE LOAD
PRESSURE
200
TABLE
ramp
OK
SURFACES (all of the boundary conditions applied
8 5 # on the geometry are shown in Figure 2.1-18)
MAIN
319
CHAPTER 2.1
New-style Table Input
Step 6: Job creation
It is anticipated that the plastic strains may be large in the slot, so the LARGE STRAIN (plastic) option is invoked.
Furthermore, the FOLLOWER FORCE option is activated so the pressure load is always based upon the current
geometry. The equivalent stress and plastic strains are written to the post Iile Ior the outside and middle layer. Using
Mentat, the deIault number oI layers is Iive, so output is obtained Ior layers 1, 3, and 5. The three-node thin shell
element is used in this analysis.
MAIN MENU
ANALYSIS
LOADCASES
ANALYSIS CLASS
STRUCTURAL
NEW
STATIC
PROPERTIES
MULTI-CRITERIA
SOLUTION CONTROL
NON-POSITIVE DEFINITE
OK (twice)
MAIN
MAIN MENU
ANALYSIS
JOBS
NEW
STRUCTURAL
PROPERTIES
lcase1 (add loadcase)
ANALYSIS OPTIONS
LARGE STRAIN
FOLLOWER FORCE (toggle)
OK
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress
LAYERS
OUT & MID (toggle)
320 Marc Users Guide: Part I
CHAPTER 2.1
Step 7: Evaluate Results
The objective oI the simulation is to examine the plastic strains, the deIormations, and the stress in the vicinity oI the
slot. The satisIaction oI the Ilow stress curve are also veriIied. Finally, the eIIectiveness oI the adaptive time stepping
procedure is examined.
AVAILABLE ELEMENT SCALARS
Total Equivalent Plastic Strain
LAYERS
OUT & MID (toggle)
OK
OK
TITLE
Boundary Conditions on Geometric Entities Driven by Table and Equation
ELEMENT TYPES
STRUCTURAL
3-D SHELL/MEMBRANE
138
OK
ALL: EXIST.
RETURN
RUN
ADVANCED JOB SUBMISSION
WRITE INPUT FILE
EDIT INPUT FILE
OK
SUBMIT
MONITOR
OK
MAIN MENU
POSTPROCESsING
RESULTS
OPEN
table_bc_shell_job1.t16
OK
321
CHAPTER 2.1
New-style Table Input
PLOT
ELEMENTS (on)
CURVES (off)
NODES (off)
POINTS (off)
SURFACES (off)
ELEMENTS SETTING
OUTLINE
FILL
DRAW
MAIN
CONFIGURATION
VISUALIZATION
LIGHTING (on)
VIEW 1 (on)
MAIN MENU
POSTPROCESsING
RESULTS
LAST
CONTOUR BANDS
SCALAR
Total Equivalent Plastic Strain Top Layer
OK
SELECT
METHOD
USER BOX
RETURN
ELEMENTS
-10 10
-0.002 10
-100 1000
MAKE VISIBLE
RETURN
rotate model using DYN. MODEL
DEF ONLY
322 Marc Users Guide: Part I
CHAPTER 2.1
The equivalent plastic strain is shown in Figure 2.1-19.
Figure 2.1-19 Equivalent Plastic Strain in Slot
The equivalent stress is shown in Figure 2.1-20.
SCALAR
Equivalent Von Mises Stress Top Layer
OK
SELECT
METHOD
SINGLE
RETURN
ELEMENTS
ALL: EXIST.
MAKE VISIBLE
RETURN
PLOT
ELEMENTS SETTINGS
OUTLINE
DRAW
RETURN
RETURN
-2.323e-003
5.285e-002
1.080e-001
1.632e-001
2.184e-001
2.736e-001
3.287e-001
3.839e-001
4.391e-001
4.943e-001
5.494e-001
lcase1
Total Equivalent Plastic Strain Top Layer
Inc: 61
Time: 1.000e+000
Y
X
Z
323
CHAPTER 2.1
New-style Table Input
Figure 2.1-20 Equivalent Stress
You can observe that large plastic strain occurs. In examining node 569 located in the center oI the slot, the tracking
oI the yield stress is compared with the deIined Ilow stress. This is shown in Figure 2.1-21.
HISTORY PLOT
SET LOCATIONS
n: 569 (or whatever node is at center of slot)
#
ALL INCS
SHOW IDS
5
LIMITS
MAX X
1
MIN Y
20000
MAX Y
30000
ADD CURVES
ALL LOCATIONS
Total Equivalent Plastic Strain Top Layer
Equivalent von Mises Stress Top Layer
4.048e-006
2.583e+003
5.166e+003
7.749e+003
1.033e+004
1.291e+004
1.550e+004
1.808e+004
2.066e+004
2.325e+004
2.583e+004
lcase1
Equivalent Von Mises Stress Top Layer
Inc: 61
Time: 1.000e+000
X
Y
Z
324 Marc Users Guide: Part I
CHAPTER 2.1
Figure 2.1-21 Stress-Strain Behavior at Node 569
You can observe that the behavior Iollows the stress-strain law that was deIined in Figure 2.1-16. At increment 23, this
node unloads elastically, and a Iew increments later, it reloads.
The next step is to examine the application oI the load, this can be done by displaying the history oI the time.
The result is shown in Figure 2.1-22. The user observes that Irom increment 20 to 45, there is only a slight increase in
the time.
CLEAR CURVES
RETURN
ADD CURVES
GLOBAL
INCREMENT
TIME
FIT
3
2
0
20
25 30 35 40 45
50
55
60
lcase1
Total Equivalent Plastic Strain Top Layer Node 569
Equivalent Von Mises Stress Top Layer Node 569 (x10000)
1
1
325
CHAPTER 2.1
New-style Table Input
Figure 2.1-22 Time versus Increment Number
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
can.proc Mentat procedure Iile to run the above example
1
5
10
15
20
25 30 35 40
45
55
60
lcase1
Increment (x10)
Time
0 6.1
0
50
0
1
326 Marc Users Guide: Part I
CHAPTER 2.1
Chapter 2.2: Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
2.2
Thermo-Mechanical Analysis
of Cylinder Head Joint with
Quadratic Contact

Summary 328

Simulation of a Cylinder Head Joint 330

Input Files 358

328 Marc Users Guide: Part I
CHAPTER 2.2
Summary
Title Thermo-mechanical analysis oI a cylinder head
Problem Ieatures Automated mesh splitting Ior bolt modeling
Geometry
Material properties Steel and Gasket Material Irom Chapter 3.17
Analysis type Thermo-mechanical coupled
Boundary conditions Fixed block, Bolted head, Pressure head, Temperature history
Element type Parabolic element types 57, 127, and linear composite brick, 149
FE results Gasket pressure history
cover_lug
cover_plate
cover_roof
gasket_body
gasket_ring
left_bolt
lower_part
right_bolt
X
Y Z
r = 24 mm
Units: N, mm, minute, C
+
2 Bolts 8 mm DIA
19
2021
22
23
24
25
26
28
29
30
0
job1
Increment (x10)
Gasket Pressure (x10)
0 3
18
-0.143
17
15
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
16
4.804
Node 2085
27
28
29
30
0
17
18
19
20
21
16
22 2425
26
27
23
Node 2337
15
13
1
2
3
4
5
14
6 8 9 10
11
12
7
329
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
This chapter introduces a number oI new Ieatures in Marc and Mentat in the area oI engine modeling:
A new tying type called overclosure tying, Ior prestressing structures like bolts and rivets. The new tying
combines and generalizes the tying and servo link that have traditionally been used in Marc Ior this purpose
(see, Ior example, Chapter 3.17 oI this Marc Users Guide) and can be used in combination with the
automatic contact algorithm.
New tools Ior automated mesh splitting and tying generation in Mentat, that allow Iast and easy set up oI
bolt models.
Improvements made to the automatic contact algorithm in the way user-coordinate transIormations are
treated. It is now possible to deIine coordinate transIormation Ior nodes that come in contact.
The possibility to run a job in parallel using a single input Iile without the need Ior a preprocessor to generate
the domains and split the model data into input Iiles Ior each domain.
Improvements made to Mentat in terms oI visualization oI the local directions oI coordinate transIormations
deIined at the nodes and oI the thickness direction oI gaskets and composite solid elements.
For this purpose, a described analysis oI a cylinder head joint in Chapter 3.17 oI this Marc Users Guide is suitably
modiIied. In brieI, the modiIications involve the Iollowing:
change to the shape oI the cylinder head cover
change Irom linear elements to quadratic elements Ior some components oI the cylinder head joint
change Irom mechanical analysis to thermo-mechanically coupled analysis
incorporation oI temperature dependence Ior gasket properties
addition oI thermal loading
incorporation oI the new way oI modeling bolt preload.
Figure 2.2-1 Finite Element Mesh of the Cylinder Head Joint
cover_lug
cover_plate
cover_roof
gasket_body
gasket_ring
left_bolt
lower_part
right_bolt
X
Y Z
r = 24 mm
Units: N, mm, minute, C
+
2 Bolts 8 mm DIA
330 Marc Users Guide: Part I
CHAPTER 2.2
Simulation of a Cylinder Head Joint
The model consists oI the cylinder head cover and a small portion oI the lower part oI the cylinder head (see
Figure 2.2-1). Both the cover and the lower part are made oI steel. A thin gasket layer seals the joint between the cover
and the lower part. The joint is Iastened by two steel bolts.
The assembly is loaded in six stages. In the Iirst two stages, the Iastening oI the joint is simulated by applying a
pretension oI 12 kN to each oI the bolts. AIter the bolts have been loaded, the three-stage thermo-mechanical loading
cycle starts. First, the base oI the assembly is heated to 200C, while an interior pressure oI 1.2 MPa is simultaneously
applied to the cover and the lower part oI the assembly. Next, the base is cooled down to -20C, while retaining the
interior pressure. In the IiIth stage, the pressure is removed and the temperature oI the base is increased again to room
temperature oI 20C. The Iinal loadcase oI the analysis consists oI disassembling the joint by loosening the bolts.
Mesh
The Iinite element mesh used in this chapter is somewhat diIIerent Irom that used in Chapter 3.17 . The changes are
highlighted here:
The shape oI the engine head cover has been cosmetically changed Irom a spherical to a cylindrical shape and
meshed with 20-noded hexahedral elements.
To demonstrate the use oI the new overclosure tyings in combination with the automatic contact algorithm,
the radius oI the lugs has been reduced such that the bolts Iit exactly into the lugs.
The bolts are meshed with 20-noded hexahedral elements.
The lower part oI the assembly has been meshed with 10-noded tetrahedral elements.
To demonstrate the use oI coordinate transIormations Ior prestressing the bolts and to show that
transIormations can be used on nodes that come into contact, the entire model has been rotated about -30
around the global x-axis.
The base model obtained aIter meshing oI the diIIerent components is available in thermogask_mesh.mud. Suitable
element sets like gasket, cover, bolts, lower_part have already been identiIied in this Iile. A procedure Iile that
then reads in this Iile and incorporates the other changes mentioned below is available in thermogask.proc. The user
is also reIerred to the comments in the procedure Iile Ior more details.
Geometric Properties
The gasket used in this example is modeled as a Ilat sheet with a thickness oI one millimeter and consists oI two
regions with diIIerent material properties: the body and the ring (see Figure 2.2-2). For the gasket material, the
behavior in the thickness direction, the transverse shear behavior, and the membrane behavior are Iully uncoupled. The
thickness direction oI the gasket elements must be speciIied by means oI a geometric property oI type 3-D SOLID
COMPOSITE/GASKET. The Iinite element mesh oI the gasket has been created in such a way that Ior all elements in
the gasket, the thickness direction is given by the direction Irom FACE 4 (1-2-3-4) to FACE 5 (5-6-7-8). In this model,
the two sides oI the gasket actually have diIIerent orientations. This is not signiIicant here because a gasket element
has only one layer (integration point) through the thickness. II the model was composed oI composite brick elements
with multiple layers, this would be important.
331
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Mentat has introduced the possibility to visualize the thickness direction oI solid composite and gasket elements via
the SOLID COMP./GASKET PLOT SETTINGS menu. The latter can be accessed both Irom the various GEOMETRIC
PROPERTIES menus and Irom the PLOT-> ELEMENT SETTINGS menu. II switched on, an arrow is drawn that points
Irom the bottom Iace oI the solid composite or gasket element to the top Iace oI the element. Note that visualization
oI the thickness direction is available only iI the elements are drawn in wireIrame mode. The picture on the right-hand
side oI Figure 2.2-2 depicts the thickness directions Ior some oI the gasket elements in the present model.
Figure 2.2-2 The Finite Element Mesh of the Gasket and Thickness Direction of the Gasket Elements
Material Properties
Details Ior temperature independent gasket properties are available in Chapter 3.17. The incorporation oI thermal
dependence in the gasket behavior is emphasized in this section.
As already mentioned in the preceding section, Ior the gasket material, the behavior in the thickness direction, the
transverse shear behavior, and the membrane behavior are Iully uncoupled. The transverse shear and membrane
behavior are linear elastic, characterized by a transverse shear modulus and the in-plane Young`s modulus and
Poisson`s ratio, respectively. In the thickness direction, the behavior in tension is also linear elastic and is governed by
a tensile modulus, deIined as a pressure per unit length. All the moduli used to characterize the in-plane membrane,
transverse shear, and tensile behavior can be varied with temperature using single variate tables.
Thickness Direction Gasket Properties
For elastic-plastic compressive behavior in the thickness direction, the user speciIies the loading path, the yield
pressure above which plastic deIormation develops and up to ten unloading paths. Mandatory pressure-closure
relationships can be directly input Ior speciIying the loading and unloading paths. Additionally, optional
pressure-temperature relationships can be simultaneously input Ior these paths. Pressure varying as a Iunction oI both
closure distance and temperature is input using the Multi-Variate table capability in Mentat. A table varying as a
Iunction oI temperature can be speciIied Ior the yield stress. While speciIying the loading path, unloading paths, and
yield pressure as a Iunction oI temperature, care should be exercised that the yield pressure at a particular temperature
Note: The arrows indicating the thickness direction point Irom the bottom Iace to the top Iace oI the elements.
332 Marc Users Guide: Part I
CHAPTER 2.2
should intersect the loading curve at that temperature, and that the unloading curve at a particular temperature should
intersect the loading curve at that temperature.
The structural element types that are used to model the mechanical aspects oI the gasket material are 149 (3-D solid),
151 (plane strain) and 152 (axisymmetric). The associated heat transIer element types that are used to model thermal
aspects oI the gasket material are 175 (3-D solid), 177 (planar) and 178 (axisymmetric). The heat transIer properties
oI the gasket are speciIied through the mass density, isotropic thermal conductivity and speciIic heat which can also
be varied with temperature.
As mentioned earlier, the gasket used in this example is modeled as a Ilat sheet with a thickness oI one millimeter and
consists oI two regions with diIIerent material properties, the body and the ring (see Figure 2.2-2). For both regions,
the data oI the loading path and one unloading path are created using Multi-Variate Tables in Mentat. The variation oI
the data with temperature is assumed in a simple Iorm. The temperature is considered to vary between -20C and
200C. For the ring portion oI the gasket, the pressure reduces by 10 over this temperature range, while Ior the body
portion oI the gasket, the pressure reduces by 20.
The creation oI a typical multi-variate table Ior the pressure-closure-temperature relationship is described. The table
consists oI two independent variables and one dependent variable. The two independent variables are closure distance
and temperature, respectively; while the dependent variable (reIerred to as Iunction in Mentat) is pressure. The closure
distance varies Irom 0.0 to 0.175, the temperature varies Irom -20C to 200C. It is Iurther assumed that the pressure
values reduce by 10 over this temperature range. The creation oI such a multi-variate table can be handled using one
oI two available procedures.
The Iirst procedure is to create two single-variate tables and then use the MULTIPLY TABLE option in Mentat to create
a third multi-variate table. Note that this procedure is only Ieasible when F(V1,V2) can be written as F(V1,V2)
G(V1)H(V2). The single-variate tables, G(V1) and H(V2), can be created by reading in an existing Iile or by adding
data points or by entering a Iormula.
FILES
OPEN
thermogask_mesh.mud
SAVE AS
new
OK
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
temperature
ADD
333
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
The table gasket_body_loading is a two variate table obtained by multiplying the current table which has
gasket_closure_distance as its independent variable with body_temp which has temperature as its
independent variable.
The second procedure is to create the two-variate table directly. This procedure is more general than the Iirst procedure
since it allows values oI the independent variables V1, V2 and the corresponding Iunction values F(V1,V2) to be
entered directly. This is accessed by clicking on the ADD ALL POINTS button in Mentat and answering a number oI
questions at the command prompt. These questions pertain to the number oI data points Ior V1, number oI data points
Ior V2, the numerical values oI V1, the numerical values oI V2, and Iinally the values oI F(V1,V2).
-20 1.0
200 0.9
FIT
NAME
body_temp
READ
RAW
ch02_body_loading.raw
OK
TYPE
gasket_closure_distance
MULTIPLY TABLE
body_temp
NAME
gasket_body_loading
NEW
2 INDEPENDENT VARIABLES
INDEPENDENT VARIABLE V1
TYPE
gasket_closure_distance
INDEPENDENT VARIABLE V2
TYPE
temperature
ADD ALL POINTS
7
2
0.0 0.027 0.054 0.081 0.108 0.135 0.175
334 Marc Users Guide: Part I
CHAPTER 2.2
Note that iI any errors are made while entering the data during the second procedure, there is no chance to immediately
correct it. The best technique is to continue entering the rest oI the requested inIormation and correct the Iunction
values aIterwards using the EDIT command.
The above commands illustrate how the 2-variate table gasket_body_loading can be created in Mentat. In a similar
manner, other 2-variate tables gasket_body_unloading, gasket_ring_loading, gasket_ring_unloading
are created.
In certain situations, data may only be available over a limited range oI values. In those situations, Ior independent
variable values that are beyond the data provided, the user can instruct the program iI the Iunction values in the table
should be extrapolated or kept constant when creating a table with Mentat, the deIault is that extrapolation will be
perIormed. For example, iI the gasket pressure needs to be obtained via extrapolation Ior temperature values not given
in the table, this would be set up as:
-20.0 200.0
0.0 2.08 8.32 18.72 33.28 52.0 56.0
0.0 1.872 7.488 16.848 29.952 46.8 50.4
FIT
NAME
gasket_body_loading
TABLES
EDIT
gasket_body_loading
MORE
INDEPENDENT VARIABLE V2
EXTRAPOLATE
RETURN
READ
ch02_gasket_body_unloading
TYPE
gasket_closure_distance
NAME
ch02_gasket_body_unloading
READ
ring_loading
TYPE
gasket_closure_distance
NAME
335
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Temperature dependent yield pressure, tensile modulus. and transverse shear modulus Ior the body and ring regions
are speciIied through appropriate single variate tables, body_temp and ring_temp, respectively. For the body, these
moduli are reduced by 10 over the temperature range oI (-20C, 200C), and Ior the ring, they are reduced by 20.
Membrane/Thermal Gasket Properties
For mechanical membrane properties and thermal properties, the GASKET material reIers to an existing isotropic
material. Temperature dependence oI these properties can also be speciIied through appropriate tables Ior the
isotropic material.
In the present example, no temperature dependence oI the membrane/thermal properties is considered. The mechanical
properties are taken to be the same as those in Chapter 3.17. For the thermal properties, the gasket is intended to behave
as an insulator and the properties assumed herein are consistent with those Ior a hard type oI rubber. The units Ior the
thermal quantities are consistent with those Ior the mechanical quantities: N Ior Iorce, mm Ior length, minute Ior time,
and C Ior temperature.
ring_loading
READ
ring_unloading
TYPE
gasket_closure_distance
NAME
ring_unloading
RETURN
ANALYSIS CLASS
COUPLED
NEW
STANDARD
STRUCTURAL
YOUNGS MODULUS
120 (N/mm
2
)
THERMAL EXP. (twice)
ALPHA
5e-5 (1/C)
OK
OK
THERMAL
K
6.5378
336 Marc Users Guide: Part I
CHAPTER 2.2
SPECIFIC HEAT
9.79056E12
MASS DENSITY THERMAL
VALUE
1.15689E-13
OK
NAME
gasket_body_membrane
NEW
STANDARD
STRUCTURAL
YOUNGS MODULUS
100
THERMAL EXP. (twice)
ALPHA
1e-4
OK
OK
THERMAL
K
6.5378
SPECIFIC HEAT
9.79056E12
MASS DENSITY THERMAL
VALUE
1.15689E-13
OK
NAME
gasket_ring_membrane
337
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Gasket Material Specification
AIter deIining appropriate tables Ior the through thickness behavior and deIining an isotropic material Ior the
membrane/thermal behavior, all this inIormation is jointly speciIied Ior the body and ring regions oI the gasket
as Iollows:
NEW GASKET
STRUCTURAL
YIELD PRESSURE
52
TABLE
body_temp
TENSILE MODULUS
72
TABLE
body_temp
INITIAL GAP
1/11
LOADING PATH TABLE
gasket_body_loading
UNLOADING PATHS TABLE 1
gasket_body_unloading
TRANSVERSE SHEAR BEHAVIOR MODULUS
40
TABLE
body_temp
OK
GENERAL
BASE MATERIAL
gasket_body_membrane
OK
ELEMENTS ADD
SET
gasket_body
OK
NAME
gasket_body
NEW GASKET
338 Marc Users Guide: Part I
CHAPTER 2.2
Material Specification for Other Components
The cylinder head cover, the lower part, and the bolts are all made oI steel. No temperature dependence is considered
Ior the mechanical quantities. For thermal properties, the thermal conductivity is varied with temperature. A table
cond_steel is speciIied Ior the steel with the Iollowing values: (-20,830), (0,830), (100,965), (200,1053) where the
Iirst value represents the temperature and the second value represents the thermal conductivity.
STRUCTURAL
YIELD PRESSURE
42
TABLE
body_temp
TENSILE MODULUS
64
TABLE
body_temp
LOADING PATH TABLE
ring_loading
UNLOADING PATHS TABLE
ring_unloading
TRANSVERSE SHEAR BEHAVIOR MODULUS
35
TABLE
body_temp
OK
GENERAL
BASE MATERIAL
gasket_ring_membrane
OK
ELEMENTS ADD
SET
gasket_ring
NAME
gasket_ring
NEW
STANDARD
STRUCTURAL
339
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
YOUNGS MODULUS
2.1e5 (N/mm
2
)
POISSONS RATIO
0.3
THERMAL EXP. (twice)
ALPHA
1.5e-5 (1/C)
OK
OK
THERMAL
K
1
SPECIFIC HEAT
1.65686E+12
MASS DENSITY THERMAL
VALUE
2.17139E-12
OK
ELEMENTS ADD
SET
cover
lower_part
bolts
NAME
steel
MAIN
340 Marc Users Guide: Part I
CHAPTER 2.2
Modeling Tools
Transformations
As mentioned beIore, compared to the model Irom Chapter 3.17, the present model is rotated about -30 around the
global x-axis. To make sure that boundary conditions are applied correctly, a local coordinate system is deIined by
creating a transIormation. The orientation oI the local coordinate system is obtained by rotating the global coordinate
system about -30 around the global x-axis:
The local z-direction oI this coordinate system thus coincides with the axial direction oI the bolts and the local
x-direction with the global x-direction. Several nodes, including all nodes at the base oI the lower part and the bolts
are added to this transIormation, so that boundary conditions applied to these nodes will act in the local x-, y- and
z-directions.
Since Marc 2007, an important restriction has been removed that existed in previous versions on the use oI
transIormations in combination with the automatic contact algorithm. In previous versions, transIormations could not
be deIined Ior nodes which come in contact. The reason was that the contact algorithm imposes certain constraints on
nodes in contact. These contact constraints are set up in a local coordinate system that is likely to be diIIerent Irom the
local coordinate system deIined by the transIormation. TransIormations are now allowed Ior any nodes including the
ones coming in contact.
MODELING TOOLS
TRANSFORMATIONS
NEW
ROTATE
-30 0 0
SELECT
SELECT SETS
bolts
lower_part
OK
SELECT BY
NODES BY
ELEMENTS
ALL: SELECTED
RETURN
RETURN
NODES ADD
ALL:SELECTED
341
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Mentat has introduced the possibility to visualize the local coordinate directions oI the coordinate systems at the nodes
via the TRANSFORMATION PLOT SETTINGS menu. The latter can be accessed both Irom the TRANSFORMATIONS
menu and Irom the PLOT-> NODE SETTINGS menu. II transIormations are drawn, the local coordinate systems are
indicated by three arrows, pointing Irom the node in the direction oI the local x-, y- and z-axis. To distinguish the
directions, the local x-direction is indicated by a red arrow, the local y-direction by a green arrow and the local
z-direction by a blue arrow. Plotting oI each oI these arrows can be controlled individually by the FIRST, SECOND,
and THIRD DIRECTION buttons in the TRANSFORMATION PLOT SETTINGS menu. Note that transIormation are
displayed only iI the nodes are drawn. Figure 2.2-3 depicts the local coordinate systems oI the nodes at the base oI the
lower part oI the model.
Figure 2.2-3 The Local Coordinate Systems of the Nodes at the Base of the
Lower Part Defined by the Transformation
Matching Boundaries and Overclosure Tyings Bolt Modeling
As mentioned earlier, in the Iirst two stages oI the analysis, the Iastening oI the joint is simulated by applying
pretension loads oI 12 kN to each oI the bolts. In the Iirst stage, the leIt bolt (see Figure 2.2-1) is pretensioned while
the right bolt is locked, and in the second stage, the right bolt is loaded while the length oI the leIt bolt is Iixed. During
the subsequent three-stage thermo-mechanical loading cycle, the bolts will be locked and in the Iinal stage oI the
analysis, the joint will be disassembled by loosening the bolts.
The pretension Iorce in the bolt is simulated using the standard TYING option existing in Marc. The basic idea is that
the Iinite element mesh oI the bolt is split across the shaIt oI the bolt in two disjoint parts with congruent meshes at
the split and that corresponding nodes on both sides oI the split are connected to each other and to a special node, called
the control node oI the bolt, by multi-point constraints (see Figure 2.2-4). The latter are used to create an overlap
between the two parts oI the bolt in the axial direction and in this way introduce a tensile stress in the bolt.
In Chapter 3.17, the Iinite element meshes oI the bolts are split up into a top and a bottom part during the mesh
generation process. A combination oI tyings oI type 203 (to prevent relative tangential motion oI the two parts) and
servo links is used to pretension the bolts. The servo links are chosen in such a way that a pretension Iorce can be
342 Marc Users Guide: Part I
CHAPTER 2.2
applied to the bolt simply by applying a POINT LOAD boundary condition to the control node oI the bolt. Alternatively,
the bolt can be tightened, locked or loosened by applying a FIXED DISPLACEMENT boundary condition to the
control node.
Figure 2.2-4 Pre-stressing a Structure by Creating an Overlap Between the Top and the
Bottom Part using Overclosure Tyings
Since the 2007 releases oI Marc and Mentat, this technique has been streamlined. Mentat provides tools Ior
automatically splitting a continuous Iinite element mesh along a list oI edges (in 2-D) or Iaces (in 3-D). The new
boundaries on both sides oI the split, that are created in this process, are stored pair-wise along with the rest oI the
model in the mud or mfd Iile. This so-called matching boundary information can subsequently be used to automatically
generate tyings, servo links, or springs between the corresponding nodes on both sides oI the split. This greatly speeds
up the modeling process, as no special measures have to be taken during the meshing phase. A bolt can simply be
modeled by a continuous Iinite element mesh that is subsequently split up in two parts which are then connected by
multi-point constraints.
In addition, a new tying type that combines and generalizes the tying type 203 and the servo link used in Chapter 3.17.
The new type (69) is called overclosure tying and has one tied node and two retained nodes. Like the servo links,
overclosure tyings should be applied in such a way that the tied node and the Iirst retained node oI the tyings are the
corresponding nodes at the split. The second retained node oI the tyings should be a Iixed external node which is shared
by all tyings oI a bolt. This node is also called the control node oI the tying, since it can be used to control the size oI
the gap or overlap between the parts:
1. The displacement oI the control node in a particular direction is equal to the size oI the gap or overlap in that
direction; and
2. The Iorce on the control node is equal to the sum oI the Iorces on the tied nodes oI all overclosure tyings which
share that control node. It is equal (but with opposite sign) to the sum oI the Iorces on the Iirst retained nodes
oI the overclosure tyings.
control node
bottom part
overclosure tyings
(tied)
(first retained)
top nodes
top part
mesh split
bottom nodes
(second retained)
undeformed
F
1,bot
F
2,bot
F
2,top
F
control
top part
bottom part
F
1,top
u
1,top
u
2,top
u
2,bot
u
1,bot
u
control
overlap u
control
deformed
343
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
The tying is Iunctionally equivalent to n servo links oI the Iorm used in Chapter 3.17, on the n displacement and
rotational degrees oI Ireedom oI the nodes oI the bolt and the control node. The important diIIerences with the
approach oI Chapter 3.17 are:
1. The overclosure tying always acts on the displacement and rotation components in the global coordinate
system, while the tying type 203 and the servo link act on the components in the local coordinate systems oI
the nodes (iI a coordinate transIormation has been deIined); and
2. The control node oI an overclosure tying has the same displacement and rotational degrees oI Ireedom as the
nodes oI the bolt, while the control node in Chapter 3.17 has only one degree oI Ireedom.
The latter implies that suIIicient boundary conditions have to be applied to the control node oI overclosure tyings to
suppress any rigid body modes.
The advantages oI the new tying type over the servo link approach are twoIold. First oI all, only a single tying needs
to be created between corresponding nodes on both sides oI the split instead oI a tying and a servo link. Secondly, the
bolt can be loaded in any direction or combination oI directions by applying FIXED DISPLACEMENT or POINT LOAD
boundary conditions to the control node in the appropriate directions. II the loading is not parallel to one oI the global
coordinate directions (Ior example, because the bolt is not aligned with one oI the global axes, as in the present model),
a coordinate transIormation can be deIined at the control node, such that one oI the local directions coincides with the
loading direction. By contrast, the bolts oI Chapter 3.17 can be loaded in one direction only and coordinate
transIormations are needed on all nodes oI the split to ensure that the pretension is applied in the correct direction.
In nonmechanical passes oI a coupled analysis (e.g., in the heat transIer pass oI a thermo-mechanical analysis), the
overclosure tying reduces to tying type 100 between the tied and the Iirst retained node, thus ensuring continuity oI
the primary Iield variable (e.g., temperature) cross the split.
Overclosure tyings can be used in combination with the automatic contact algorithm; that is, nodes at the surIace oI
the same contact body can be connected by overclosure tyings. Since constraints imposed by contact can potentially
conIlict with the constraint imposed by the tying, the tied nodes oI these tyings cannot come into contact, so the contact
status oI the tied nodes is always 0. However, this will not lead to penetration oI these nodes, as they are Iully tied to
the retained nodes via the tying. For more details about the overclosure tyings, please reIer to Chapter 9 in Marc
Volume A: Theory and User Information.
Assuming that a continuous Iinite element mesh has been created, the general procedure Ior pretensioning a bolt is
as Iollows:
1. Create a new pair oI matching boundaries oI the appropriate dimension (1-D Ior meshes consisting oI beam,
truss, or axisymmetric shell elements; 2-D Ior meshes consisting oI 2-D solid and 3-D shell elements; 3-D Ior
meshes consisting oI 3-D solid elements) and split the Iinite element mesh oI the bolt in two parts using one oI
the automatic mesh splitting methods.
2. Use the matching boundary inIormation to connect each pair oI corresponding nodes on the matching
boundaries to each other and to a common control node, using the MATCHING BOUNDARY NODAL TIES
submenu oI the MATCHING BOUNDARIES menu.
3. Apply POINT LOAD and/or FIXED DISPLACEMENT boundary conditions to the control node oI the
overclosure tyings to apply the pretension Iorce to the bolt or to prescribe the tightening (change) oI the bolt.
344 Marc Users Guide: Part I
CHAPTER 2.2
In the present example, two 3-D matching boundaries are created (one Ior each bolt) and the PLANE method is
employed to automatically split the meshes oI the bolts across the shaIt (please reIer to the Mentat online help oI the
MATCHING BOUNDARIES menu Ior the other available methods). The latter splits the mesh along a list oI Iaces by
disconnecting the elements on one side oI a plane Irom the elements on the other side. The plane is deIined by a normal
vector and a node. In this case, the normal vector to the plane is the axial direction oI the bolts or the local z-direction
oI the previously deIined transIormation. The normal vector be can supplied either by providing its components with
respect to the global coordinate system or by clicking two nodes on the axis oI the bolt using the FROM/TO method.
In this case, the Iormer method is employed, and the dsin and dcos Iunctions (which return, respectively, the sine
and cosine oI an angle speciIied in degrees) are used to speciIy the global y- and z-components oI the vector:
Figure 2.2-5 Bolts Split up into a Top and Bottom Parts with Matching Boundaries Connected
by Overclosure Tyings
The matching boundary inIormation generated by the mesh splitting process is displayed graphically by drawing the
Iaces at the boundaries thicker than usual and using diIIerent colors to distinguish the Iaces on the side oI the positive
normal to the plane, reIerred to as 'side A, Irom those on the opposite side oI the plane, 'side B (see Figure 2.2-5).
By deIault, the Iaces on side A are drawn in magenta and the Iaces on side B in green.
MODELING TOOLS
MATCHING BOUNDARIES
NEW
3-D (3-D SOLID)
AUTOMATIC MESH SPLITTNG METHOD: PLANE
NORMAL
0 dsin(30) dcos(30)
SPLIT MESH
9916
Side A
Side B
345
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
The matching boundary inIormation is used to connect corresponding nodes on both boundaries to each other and to
a common control node by overclosure tyings using the MATCHING BOUNDARY NODAL TIES menu. The ADD NODE
submenu is employed to create the control node and to add it to the previously created transIormation, so that the local
z-direction oI the control node coincides with the axial direction oI the bolt:
The overclosure tyings are created by selecting the new node as the second retained oI the tyings to be generated and
clicking the CREATE TIES button:
This generates Ior each pair oI corresponding nodes on the matching boundaries a separate tying (see Figure 2.2-5).
The way in which these tyings are created can be controlled by the user. The deIault settings (the tied node in each
tying is the node on side B, the Iirst retained node is the corresponding node on side A, and the second node is a Iixed
external node) are such that a Iorce applied to the control node in the direction oI the positive normal to the plane used
to create the split results in an overlap oI the two parts in that direction, and hence, in a tensile stress in the bolt in
that direction.
NODAL TIES CREATE
ADD NODE
ADD
-36.05 40*dsin(30) 40*dcos(30)
TRANSFORMATIONS
NODES ADD
11049
RETURN
RETURN
RETAINED: NODE 2
NODE
11049
CREATE TIES
346 Marc Users Guide: Part I
CHAPTER 2.2
Contact
Figure 2.2-6 Contact Body Definition for Joint Assembly
The automatic contact algorithm is used to describe the contact between the gasket and the metal parts oI the joint and
between the bolts and the cylinder head cover. Moreover, a contact symmetry surIace is used to take symmetry
conditions into account.
The deIinition oI the various contact bodies is shown in Figure 2.2-6. The Iirst contact body consists oI the HEX8
gasket elements. The second contact body consists oI the HEX20 cover elements. The third contact body consists oI
the HEX20 bolt elements. Note that some nodes on the surIace oI this body are connected by overclosure tyings and
recall Irom the preceding section that this is allowed. The Iourth contact body consists oI the TET10 lower part
elements. The last contact body is the symmetry plane. Note that the nodes at the base oI the bolts and the lower part
have a local coordinate system deIined by the transIormation. Some oI these nodes will be in contact with the
symmetry plane or (Ior the nodes oI the bolt) with the lower part. In previous Marc versions, this was not allowed;
since Marc 2007, this restriction has been removed.
The gasket is glued to the cover and lower part and is not allowed to separate. Normal touching contact is used between
the gasket and the bolts and between the cover and bolts. Glued contact is used between the bolts and lower part. Also,
a contact heat transIer coeIIicient oI 10 N/mm/min/C is used Ior any metal-metal contact and a contact heat transIer
coeIIicient oI 0.5 N/mm/min/C is used Ior any gasket-metal contact. A CONTACT TABLE is created to activate
these options.
CONTACT
CONTACT TABLES
NEW
PROPERTIES
1 2
347
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Initial Conditions
The temperature oI the model is initialized to 20C (room-temperature) by means oI a TEMPERATURE
initial condition.
CONTACT TYPE: GLUE
THERMAL PROPERTIES
CONTACT HEAT TRANSFER COEFFICIENT
0.5
1 3
CONTACT TYPE: TOUCHING
CONTACT HEAT TRANSFER COEFFICIENT
0.5
1 4
CONTACT TYPE: GLUE
CONTACT HEAT TRANSFER COEFFICIENT
0.5
1 5
CONTACT TYPE: TOUCHING
2 3
CONTACT TYPE: TOUCHING
CONTACT HEAT TRANSFER COEFFICIENT
10.0
2 5
CONTACT TYPE: TOUCHING
3 4
CONTACT TYPE: GLUE
CONTACT HEAT TRANSFER COEFFICIENT
10.0
3 5
CONTACT TYPE: TOUCHING
4 5
CONTACT TYPE: TOUCHING
INITIAL CONDITIONS
THERMAL
NEW
348 Marc Users Guide: Part I
CHAPTER 2.2
Boundary Conditions
The boundary conditions applied in this model are similar to those applied in Chapter 3.17. The main diIIerence is in
the time variation oI the loading. Since the current analysis is modeled as a transient thermal coupled with a static
mechanical analysis, time is a physical quantity and any time variations Ior the loading need to be physically based.
Time is expressed in minutes in the current analysis.
Recall that the control nodes oI the bolts have the same degrees oI Ireedom (three) as the nodes oI the bolt. The
displacements oI the control nodes in the local z-direction oI these nodes (the axial direction oI the bolts) are equal to
the amount oI overlap between the top and bottom parts oI the bolts in this direction, and hence equal to amount oI
tightening oI the bolts. The displacements in the local x- and y-direction represent the relative tangential motion oI the
two parts. Throughout the analysis, the latter is suppressed by applying FIXED DISPLACEMENT boundary conditions
to the control nodes, oI which the second and third degree oI Ireedom are Iixed to 0 mm.
Two POINT LOAD boundary conditions are used to load the bolts with a pretension oI 12 kN. The time duration Ior
each oI these pretensioning events is taken as 1 minute. Since only halI oI the bolts is taken into account in the model,
halI oI the pretension load is applied to the control nodes oI the bolts. The loads are applied to the third degree oI
Ireedom oI the control nodes. Tables are used to deIine the loading history oI both bolts.
In the Iirst loadcase, when the leIt bolt is preloaded, the right bolt is unlocked and unloaded and can, thereIore, Ireely
shorten or lengthen. As this may introduce a rigid body mode oI the top part oI the bolt, the latter is pushed onto the
cover by applying a small Iorce oI 1 N in the axial direction oI the bolt. This Iorce is removed again in the second
loadcase when the right bolt is preloaded. In that loadcase, the leIt bolt remains locked. Locking oI the leIt bolt in this
loadcase and oI both bolts in the subsequent thermo-mechanical loading cycle is simulated by applying FIXED
DISPLACEMENT boundary conditions to the control nodes, oI which the third degree oI Ireedom is Iixed to 0 mm in
the loading cycle. The loosening oI the bolts in the Iinal loadcase oI the analysis is simulated by gradually releasing
the Iorces on the control nodes in the local z-direction oI these nodes.
In the three-stage thermo-mechanical loading cycle that Iollows the prestressing oI the bolts, the cylinder head joint is
subjected to a combination oI mechanical and thermal loads. The mechanical loading consists oI a pressure oI 1.2 MPa
applied to the interior oI the cylinder head cover and the lower part over a period oI 5 minutes, retained Ior 25 minutes
and then gradually removed over another 5 minutes. The TABLE that deIines the history oI the pressure is oI type time
and is deIined by the points (0,0), (2,0), (7,1), (32,1), (37,0) and (38,0).
TEMPERATURE
CONTINUUM ELEMENTS TEMPERATURE (yellow light comes on)
20
OK
NODES ADD
ALL: EXIST.
NAME
initial_temperature
349
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
The thermal part oI the loading cycle consists oI an increase oI the temperature at the base oI the assembly to 200C
over 5 minutes, a decrease to -20C over 25 minutes and again an increase back to room temperature (20C) over 5
minutes. This is achieved by applying a FIXED TEMPERATURE boundary condition to all nodes at the base oI the
model, setting the TEMPERATURE (TOP) to 1 and employing a TABLE to a table oI type time deIined by the points
(0,20), (2,20), (7,200), (32,20), (37,20) and (38,20).
Finally, to suppress rigid body motions, the displacements in the local z-direction oI all nodes at the bottom oI the
lower part oI the cylinder head assembly are suppressed as well as the displacements in the local x-direction oI the
nodes at the bottom oI the lower part that lie in the local yz-plane. The applied loads are depicted in Figure 2.2-7.
Figure 2.2-7 Boundary Conditions applied to the Cylinder Head Joint
Load Steps and Job Parameters
The job consists oI six loadcases. The user is reIerred to Chapter 3.17 Ior details on the boundary conditions applied
in each loadcase. All the loadcases are run as quasi-static thermo-mechanically coupled analyses using Iive increments
each. For the loadcases in which there are temperature changes, a temperature error in estimate oI 50C is used. This
improves the accuracy oI the temperature dependent material property estimation by allowing the solution to iterate iI
the diIIerence between the calculated temperature and estimated temperature is greater than 50C. The setup Ior a
typical loadcase is shown below:
LOADCASES
COUPLED
NEW
QUASI-STATIC
LOADS
350 Marc Users Guide: Part I
CHAPTER 2.2
The analysis is set up as a coupled analysis in the JOBS menu and all six loadcases are preIormed in sequence. The
quadratic segments in contact bodies cover, lower_part, and bolts can be treated in one oI two ways:
GENUINE is the deIault scheme this requires that the separation checking be based on stresses rather than Iorces.
Furthermore, the separation checking Ior quadratic contact should be based on nodal stresses obtained by extrapolating
Irom integration point stresses rather than those obtained as the ratio oI an eIIective Iorce to eIIective area. Control oI
these buttons are available under ADVANCED CONTACT CONTROL in the JOBS menu in Mentat. It should be noted
that when Mentat detects quadratic elements in contact bodies, the GENUINE scheme and separation checking based
on extrapolated stresses is automatically set. The button sequence shown here Ior these advanced contact options is
mainly Ior instructive purposes.
deactivate:
prestress_left_bolt
prestress_right_bolt
push_right_bolt
OK
CONTACT
CONTACT TABLE
ctable1
CONVERGENCE TESTING
MAX ERROR IN TEMPERATURE ESTIMATE
50
TOTAL LOADCASE TIME
5
STEPPING PROCEDURE
FIXED
PARAMETERS
# STEPS
5
OK
OK
NAME
loading
GENUINE wherein midside nodes are independently checked Ior contact, separation, penetration, etc.;
LINEARIZED wherein midside nodes oI a Iace are tied to the corresponding corner nodes oI that Iace.
351
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Under the ANALYSIS OPTIONS menu, the LARGE DISPLACEMENT option is selected. In addition to Equivalent Von
Mises Stress (Marc post code 17), Gasket Pressure (Marc post code 241), Gasket Closure (Marc post code 242), and
Plastic Gasket Closure (Marc post code 243), you can also choose Temperature (Integration Point) (Marc post code
180).
For the lower part oI the assembly, element type 127 (TET10 element) is used. For the cover and the bolts, element
type 57 (reduced integration HEX20 element). For the gasket, element type 149 is selected.
JOBS
NEW
TYPE
COUPLED
PROPERTIES
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
ADVANCED CONTACT CONTROL
DEFORMABLE-DEFORMABLE METHOD
SINGLE-SIDED
QUADRATIC SEGMENTS
GENUINE
SEPARATION CRITERION
STRESS
DERIVATION
EXTRAPOLATION
ELEMENT TYPES
ANALYSIS TYPE
COUPLED
ANALYSIS DIMENSION
3-D
SOLID
57
cover
127
lower_part
57
352 Marc Users Guide: Part I
CHAPTER 2.2
Save Model and Run Job
Write out the Marc input Iile thermogask_job1.dat and run the job in serial mode, using the SUBMIT 1 button in
the RUN menu:
To run the job in parallel mode using the domain decomposition method, previous Marc versions required that the
model was decomposed into domains using the DOMAIN DECOMPOSITION menu in Mentat. In that case, Mentat
would split the model data into input Iiles Ior each domain. While this is still possible, Marc can also run the job in
parallel using the same single input Iile that was created Ior the serial run. In that case, the domain decomposition is
done internally within Marc. The input Iile is read on one processor, decomposed into domains, and the domain data
is passed to the other processors.
To run the job in single input Iile parallel mode, use the -nprocds option to the a 2010 run_marc script to speciIy
the number oI domains:
path_to_marc2010/tools/run_marc -jid thermogask_job1 -nprocds 2
Please reIer to the Marc and Mentat Release Guide (2010), Ior more inIormation on the single input Iile
parallel option.
bolts
OK
SOLID COMPOSITE/GASKET
149
gasket
OK
FILE
SAVE AS
thermogask.mud
OK
RETURN (twice)
JOBS
RUN
SUBMIT 1
MONITOR
OK
RETURN
MAIN
353
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
View Results
To monitor the temperature distribution on the assembly, make a contour plot oI the temperature, set the range, and
the legend, and monitor the results.
Figure 2.2-8 shows a contour plot oI the temperature distribution at the end oI the third loadcase when the joint has
been Iastened, the temperature at the base oI the assembly has been increased to 200C and the interior pressure has
been applied. Due to the insulating properties oI the gasket, the heat transmitted to the cover through the gasket is quite
small.
RESULTS
OPEN DEFAULT
PLOT
DRAW
switch off NODES
RETURN
SCALAR PLOT SETTINGS
RANGE
MANUAL
SET LIMITS
-20
200
# LEVELS
22
LEGEND
FORMAT: INTEGER
RETURN
RETURN
SCALAR
Temperature
CONTOUR BANDS
MONITOR
354 Marc Users Guide: Part I
CHAPTER 2.2
Figure 2.2-8 Contour Plot of the Nodal Temperatures at the End of the Third Loadcase
Figure 2.2-9 shows a contour plot oI the temperature distribution at the end oI the Iourth loadcase when the joint is still
Iastened and the temperature at the base oI the assembly has been decreased to -20C.
Figure 2.2-9 Contour Plot of the Nodal Temperatures at the End of the Fourth Loadcase
In order to assess the thermo-mechanical eIIects on the gasket response, variations oI the gasket pressures at nodal
points in the gasket body and ring, where the plastic gasket closure is a maximum, are computed and displayed in
Figure 2.2-10.
355
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
As the gasket is heated, the gasket pressure drops in accordance with the compression data that has been provided, and
that upon cooling, the pressure increases back to the previous level. Also, as expected, the pressure drop-oII and
subsequent increase is more pronounced in the gasket body than in the gasket ring.
Figure 2.2-10 Variation of the Gasket Pressure with Increment Number in Body and Ring
Finally, in Figure 2.2-11 and Figure 2.2-12, the time variation oI the bolt Iorces is depicted. Figure 2.2-11 shows the
bolt Iorces in the axial (pretension) direction oI the bolts. In the loadcase in which the bolts are prestressed, these Iorces
RESULTS
SCALAR PLOT SETTINGS
HISTORY PLOT
SET NODES
2085 2337 #
COLLECT DATA
0 30 1
NODES/VARIABLES
ADD VARIABLE
Increment
Gasket Pressure
FIT
RETURN
19
2021
22
23
24
25
26
28
29
30
0
job1
Increment (x10)
Gasket Pressure (x10)
0 3
18
-0.143
17
15
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
16
4.804
Node 2085
27
28
29
30
0
17
18
19
20
21
16
22 2425
26
27
23
Node 2337
15
13
1
2
3
4
5
14
6 8 9 10
11
12
7
356 Marc Users Guide: Part I
CHAPTER 2.2
are given by the external Iorces on the control nodes in the local coordinate system oI these nodes. This is also the case
in the Iinal loadcase when the bolt Iorces are released. In the loadcases where the bolts are locked, the bolt Iorces are
given by the reaction Iorces on the control nodes in the local coordinate system oI the nodes.
Figure 2.2-12 shows the bolt Iorces in the x-direction oI the model. This is basically the total shear Iorce on the
matching boundaries in that direction.In the three-stage thermo-mechanical cycle, both bolts are sheared in outward
directions pointing away Irom the center oI the cover due to the applied internal pressure oI the cover.
RESULTS
SCALAR PLOT SETTINGS
USE NODAL TRANSFORMATIONS
RETURN
HISTORY PLOT
SET NODES
11049 11977 #
COLLECT DATA
0 30 1
NODES/VARIABLES
ADD VARIABLE
Time
Reaction Force Z
ADD VARIABLE
Time
External Force Z
FIT
RETURN
357
CHAPTER 2.2
Thermo-Mechanical Analysis of Cylinder Head Joint with Quadratic Contact
Figure 2.2-11 History Plot of the Bolt Forces in the Axial Direction
Figure 2.2-12 History Plot of the Bolt Force in the X-direction
358 Marc Users Guide: Part I
CHAPTER 2.2
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
thermogask.proc Mentat procedure Iile to run the above example
thermogask_mesh.mud Original geometry read by procedure Iile
ch02_body_loading.raw Gasket material curve read by procedure Iile
ch02_ring_loading.raw Gasket material curve read by procedure Iile
ch02_body_unloading.raw Gasket material curve read by procedure Iile
ch02_ring_unloading.raw Gasket material curve read by procedure Iile
Chapter 2.3: RBE3 (General Rigid Body Link)
2.3
RBE3
(General Rigid Body Link)

Chapter Overview 360

Soft and Rigid Connections 360

Submit Job and Run the Simulation 369

Input Files 372

360 Marc Users Guide: Part I
CHAPTER 2.3
Chapter Overview
This chapter describes the use oI RBE2 and RBE3 in Marc. In this example, RBE2 is used to simulate really rigid
connection while RBE3 is used to simulate soIt connection. Rigid connection means that the displacement oI nodes is
under controlled while soIt connection means that the distribution oI Iorces is under controlled.
Soft and Rigid Connections
A rectangular tube with a stopper is loaded on one end and partially Iixed on the other end as shown in Figure 2.3-1.
The cross section oI the tube is 100 x 50 mm
2
. The thickness is 5 mm. The length is 1000 mm.
Figure 2.3-1 Schematic Model
The size oI the connection pad is about 40x40 mm
2
. The connection between the rod and the tube is assumed to be
soIt and is about 50 mm above it. The supported (pinned) end oI the rod is locate at 500 mm above the tube. The
simulation oI the soIt connection is done using RBE3. In this case, the displacement oI the tube along the connection
pad is Iree while the Iorce distribution is controlled using the simple RBE3 Iormulation. Another possibility is using
rigid RBE2 connection which could result in an overstiII simulation. The end sections oI the tube are assumed to be
rigid. They are simulated using RBE2.
A concentrated load is applied on one end oI the tube and the other one is Iixed on all degrees oI Ireedom except the
rotation about y-axis. Initially there is a gap oI 20 mm between the tube and the cylindrical stopper. The Iinite element
model is shown in Figure 2.3-2.
F
pinned
Rigid, Iixed
x
y
z
RBE3,
RBE2,
Rectangular
tube
RBE2:
1,2,3,4,6
Iixed
soIt connection
rigid connection
Contact
361
CHAPTER 2.3
RBE3 (General Rigid Body Link)
Figure 2.3-2 Finite Element Model
Mesh Generation
The Iinal mesh can be seen in Figure 2.3-2. One extra (reIerence) node is also created to deIine RBE3`s.
MESH GENERATION
PTS ADD
0 0 0
580 0 0
580 50 0
0 50 0
620 0 0
1000 0 0
1000 50 0
620 50 0
DUPLICATE
TRANSLATIONS
0 0 50
POINTS
EXIST.
362 Marc Users Guide: Part I
CHAPTER 2.3
SRFS
ADD
1 2 3 4
2 5 8 3
5 6 7 8
4 3 11 12
3 8 16 11
8 7 15 16
9 10 11 12
10 13 16 11
13 14 15 16
CONVERT
DIVISIONS
16 5
SURFACES TO ELEMENTS
1 4 7 #
DIVISIONS
4 5
SURFACES TO ELEMENTS
2 5 8 #
DIVISIONS
14 5
SURFACES TO ELEMENTS
3 6 9 #
RETURN
SWEEP
SWEEP
NODES
EXIST.
ELEMENTS
EXIST.
REMOVE UNUSED
NODES
RETURN
ELEMS
REM
363
CHAPTER 2.3
RBE3 (General Rigid Body Link)
281 282 258 284 285 286 287 288 #
SWEEP
REMOVE UNUSED
NODES
RETURN
SYMMETRY
SYMMETRY PLANE
NORMAL
0 1 0
SYMMETRY
EXIST.
RETURN
SWEEP
SWEEP
NODES
EXIST.
ELEMENTS
EXIST.
REMOVE UNUSED
NODES
RETURN
NODES
ADD
0 0 25
1000 0 25
600 0 100
0 0 500
ELEMENT CLASS
LINE(2)
RETURN
ELEMS
ADD
1223
1222
RETURN
364 Marc Users Guide: Part I
CHAPTER 2.3
Geometric Properties
The thickness oI the plate is 5 mm. The area oI the rod is 4 mm
2
.
SURFACE TYPE
CYLINDER
RETURN
SURFACES
ADD
600 -75 -70
600 -75 -70
50
50
MAIN
GEOMETRIC PROPERTIES
NEW
3-D
SHELL
THICKNESS
5
OK
ELEMENTS ADD
SELECT ALL QUAD ELEMENTS
NEW
TRUSS
AREA
4
OK
ELEMENTS ADD
SELECT LINE(2) ELEMENTS
MAIN
365
CHAPTER 2.3
RBE3 (General Rigid Body Link)
Material Properties
The material Ior the tube is elastoplastic with 73000 MPa Young`s modulus. The Yield stress is 340 MPa and 400 MPa
at 0 and 0.15 equivalent plastic strain. The Young`s modulus Ior the rod is 210000 MPa. The Yield stress is 550 MPa
and 600 MPa at 0 and 0.15 equivalent plastic strain.
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNGS MODULUS
72E+3
POISSON RATIO
0.3
ELASTIC-PLASTIC
INITIAL YIELD STRESS
1
OK
OK
TABLES
NEW
TYPES:
eq_plastic_strain
DATA POINTS
ADD
0 340 0.15 400
COPY
DATA POINTS
EDIT
1
0 550
2
0.15 600
RETURN
SHOW MODEL
ELEMENTS
ADD
SELECT ALL QUAD ELEMENTS
366 Marc Users Guide: Part I
CHAPTER 2.3
Contact
DeIine a contact between the tube and the rigid cylinder.
ISOTROPIC
ELASTIC-PLASTIC
TABLES
table1
OK (twice)
NEW
ISOTROPIC
YOUNGS MODULUS
210E+3
POISSONS RATIO
0.3
ELASTIC-PLASTIC
INITIAL YIELD STRESS
1
TABLES
table2
OK (twice)
ELEMENTS
ADD
SELECT LINE(2) ELEMENTS
MAIN
CONTACT BODIES
DEFORMABLE
OK
ELEMENTS
ADD
select all quad elements
NEW
RIGID
SURFACES
ADD
367
CHAPTER 2.3
RBE3 (General Rigid Body Link)
Links
One RBE3 and two RBE2`s are deIined. The RBE3 is used to create soIt connection between the rod and the
connection pad on the tube. A RBE2 on one end oI the tube is deIined to allow simple application oI revolute support
while the other RBE2 is created where the applied point load is applied.
19 #
MAIN
LINKS
RBE3S
REFERENCE NODE
NODE
1222 (node at end of truss)
DOF
1 2 3 4 5 6 # (all DOF selected)
CONNECTED NODES
DOF
1 2 3 # (only displacement DOF)
COEFF.
1.0
ADD
903 984 985 986 987 986 221 238 255 37 (nodes on pad)
379 380 381 376 371 983 #
RETURN
RBE2S
RETAINED (REFERENCE)
NODE
1221 (node at center of box beam)
TIED NODES
NODE
ADD
SELECT ALL NODES OF THE TUBE AT X=1000
DOF
1 2 3 (only displacement dof)
NEW
RETAINED (REFERENCE)
368 Marc Users Guide: Part I
CHAPTER 2.3
Boundary Conditions
All degrees oI Ireedom on the edges oI the tube are Iixed except the rotation about the y-axis. The end rod is simply
supported. Concentrated load F
z
oI -22.6 kN is applied at the other end oI tube.
NODE
1152
TIED NODES
NODE
ADD
SELECT ALL NODES OF THE TUBE AT X=0
DOF
1 2 3 4 5 6 (all dof)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
ALL DISPLACEMENT DOFS SELECTED
OK
NODES ADD
1223 #
NEW
FIXED DISPLACEMENT
ALL DOFS SELECTED EXCEPT ROTATION Y
OK
NODES ADD
1220 #
NEW
POINT LOAD
FORCE Z= -22600
OK
NODES ADD
1221 #
MAIN
369
CHAPTER 2.3
RBE3 (General Rigid Body Link)
Loadcases
One loadcase is deIined. The convergence testing is done using residuals and displacements with the relative tolerance
oI 0.01 Ior both oI them An automated load step with deIault setting is used.
Submit Job and Run the Simulation
The line(2) element must be assigned as element type 9 (truss). The simulation is run with LARGE DISPLACEMENT
option. Extra output Ior tying Iorces is request Ior postprocessing.
LOADCASES
MECHANICAL
STATIC
CONVERGENCE TESTING
RESIDUALS AND DISPLACEMENTS
RELATIVE FORCE TOLERANCE
0.01
RELATIVE DISPLACEMENT TOLERANCE
0.01
OK
ADAPTIVE: MULTI CRITERIA
OK
MAIN
JOBS
NEW
MECHANICAL
lcase1
INITIAL CONDITIONS
unselect apply3
OK
ANALYSIS OPTION
select LARGE DISPLACEMENT
select UPDATED LAGRANGE PROCEDURE
OK
OK
JOB RESULTS
SELECTED NODAL QUANTITIES
370 Marc Users Guide: Part I
CHAPTER 2.3
Results
The deIormed conIiguration at the end oI the simulation is shown in Figure 2.3-3(a). As expected, the deIormation oI
the end tube remains rigid Figure 2.3-3(b). The deIormation oI the tube along the connection does not remain rigid. It
can deIorm Ireely as seen in Figure 2.3-3(c). The tying Iorce distribution Iollow the simple RBE3 Iormulation as shown
in Figure 2.3-3(d).
CUSTOM
DISPLACEMENT
TYING FORCE
OK
OK
ELEMENT TYPES
MECHANICAL
3-D TRUSS/BEAM
choose element type 9
select all line(2) elements
RETURN
RETURN
RUN
SUBMIT
371
CHAPTER 2.3
RBE3 (General Rigid Body Link)
Figure 2.3-3 Results of the Analysis
(a) Deformed configuration
(c) Deformed connection pad (d) Tying force distribution
(b) Rigid end form
372 Marc Users Guide: Part I
CHAPTER 2.3
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
rbe3.proc Mentat procedure Iile to run the above example
rbe3.mud Associated model Iile
rbe3.dat Associated Marc input Iile
Chapter 2.4: Arc Welding Process Simulation
2.4
Arc Welding Process
Simulation

Chapter Overview 374

Welding Process Simulation of Cylinder-Plate Joint 374

Input Files 392

374 Marc Users Guide: Part I
CHAPTER 2.4
Chapter Overview
In many manuIacturing processes, smaller components are joined together by a variety oI joining techniques to Iorm
the main structure. Welding is one such commonly used joining technique. An undesirable side-eIIect oI welding is
the generation oI residual stresses and deIormations in the component and the quality oI the weld has a substantial
impact on the Iatigue liIe oI the structure. These resultant deIormations may render the component unsuitable Ior
Iurther use. Also, the residual stresses Iorm the input Ior subsequent manuIacturing or structural processes.
Finite element analysis oI welding processes has been undertaken by a large number oI investigators. With Marc, this
has required the writing oI specialized user subroutines in the past. Such user subroutines are usually speciIic to the
problem under consideration and require sophisticated code to deal with Iiller element treatment, complex weld paths,
etc. With a view to simpliIy the simulation oI welding with Marc, several welding simulation capabilities have been
introduced into Marc.
This chapter demonstrates the various Ieatures available in Marc to simulate the welding process. For this purpose, the
simulation oI the welding oI a cover plate to a cylinder is described in depth. The objective is to demonstrate the
various options available to simulate the weld thermal loading, weld motion, Iiller element treatment and time
stepping. In order to keep the problem small and run within a reasonable time interval, the mesh used is somewhat
coarse, reduced integration elements are used and the time stepping thermal tolerances are rather loose.
Welding Process Simulation of Cylinder-Plate Joint
A solid cylinder with a number oI holes machined through it is joined to a thin cover plate. The joining is achieved
through two short Iillet welds placed at the junction oI the cylinder and two Ilanges oI the cover plate. The objective
oI the welding simulation is to study the temperatures generated during the welding process and investigate the
residual stresses in the component aIter welding. The Iinite element mesh oI the cylinder-plate joint is shown in
Figure 2.4-1.
Only halI the model is considered herein. Though the Iour welds (2 on each halI) are placed sequentially and the whole
model should be considered Ior a Iull description oI the welding eIIects, in order to understand the local stresses and
deIormations eIIects introduced by each weld, only the halI-model is considered here.
The solid cylinder oI radius 100 mm is modeled using hexahedral elements. The cover plate oI radius 125 mm and
Ilanges are modeled using shell elements. Weld Iiller 1 is modeled using shell elements and weld Iiller 2 is modeled
using solid elements. The various components are identiIied in Figure 2.4-1. The thermal loading comprises oI the heat
input Irom the welds which increases the temperatures at the joints to about 1200

C.
375
CHAPTER 2.4
Arc Welding Process Simulation
Figure 2.4-1 Finite Element Mesh of Cylinder-Plate Joint
Procedure File
The analysis has been completely set up using Mentat. The procedure Iile to demonstrate the example is called
weld.proc under mentat2010/examples/marc_ug/s2/c2.4.
To run the procedure Iile and build the model Irom start to Iinish, the Iollowing button sequence can be executed
in Mentat:
II one wishes to understand each and every command in the procedure Iile, the procedure Iile can be sequentially
executed through the Iollowing button sequence:
UTILS
PROCEDURES
EXECUTE
weld.proc
UTILS
PROCEDURES
LOAD
376 Marc Users Guide: Part I
CHAPTER 2.4
Every STEP click executes the next command in the procedure Iile and simultaneously shows the associated menu
and button click. When the model is being dynamically rotated or translated, due to the large number oI
rotations/translations, it is highly advisable to run through those portions quickly by clicking on START/CONT to
execute the commands continuously and clicking on STOP when the model motion is completed.
Mesh Generation
The generation oI the Iinite element mesh is not discussed in detail here. Instead, the reader is reIerred to the procedure
Iile and the comments in that Iile. A Iiner mesh is used Ior the Ilanges, weld Iillers, and Ior the cylinder in the
vicinity oI the welds. It is important to use a Iine mesh in the vicinity oI the weld in order to capture the thermal
gradients accurately.
For the set oI parameters used in the present example, the resulting Iinite element mesh is depicted in Figure 2.4-1. All
dimensions are in mms. There are a total oI 2480 elements and 3314 nodes.
Geometric Properties
The thickness oI the cover plate wall and Ilanges are speciIied as 1 mm and 2 mm, respectively. The CONSTANT
TEMPERATURE option is speciIied Ior the solid elements (both the solid cylinder and the second weld Iiller elements)
as Iollows:
It is recommended in general literature that when Iirst-order Iull integration elements are used Ior the thermal part oI
the welding analysis, second-order elements should be used Ior the corresponding mechanical analysis. This allows
accurate capture oI stresses due to linear thermal strains. The CONSTANT TEMPERATURE option allows the use oI
Iirst-order elements Ior both the thermal and mechanical passes without inducing artiIicial stresses due to linear
thermal strains. Note that Ior the reduced integration elements used herein with just one integration point, the
CONSTANT TEMPERATURE option is not really needed.
As previously mentioned, the Iirst weld Iiller is modeled with shell elements while the second weld Iiller is modeled
with solid elements. The cross-sectional area Ior each weld is 12.5 mm
2
. The equivalent thickness oI the shell weld
Iiller is obtained by:
(Perimeter length oI shell weld Iiller cross-section) x (Equivalent Thickness) 12.5.
This results in a value oI 0.801 mm Ior the shell thickness oI the Iirst weld Iiller.
weld.proc
STEP
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS
3-D
SOLID
CONSTANT TEMPERATURE
377
CHAPTER 2.4
Arc Welding Process Simulation
Note that as the thickness value is provided, the shell elements are plotted in expanded mode. The menu to control this
expanded plotting mode can be accessed as Iollows:
Material Properties
The material database in Mentat is used to deIine the temperature dependent material properties oI the cylinder-cover
structure and the weld Iillers. It is assumed that both the cylinder-cover plate and the Iillers are made oI steel material.
Based on the assumed composition oI the materials, the cylinder, cover-plate and Ilanges are given the properties oI
100Cr6 and the weld Iillers are given the properties oI 41Cr4. The material database is accessed as Iollows:
Note that the units Ior the material properties in this database are:
Length (milli meter), Mass (mega gram), Time (second), and Temperature (centigrade).
It is important to ensure that other provided data like dimensions, temperature boundary conditions, etc. are in
consistent units.
Note also that, when the material database is used, the temperature dependence oI mechanical properties like Young`s
modulus, Poisson`s ratio, CoeIIicient oI thermal expansion and oI thermal properties like SpeciIic heat, Conductivity
is read in through tables. The X-axis oI these tables (Temperature) extends Irom about -100
o
C to 1500
o
C. It is
important to note that iI the temperatures in the problem were expected to exceed these limits, the provided data should
be extended. Also, the provided tabular data can be modiIied/extended; e.g., the thermal conductivity can be increased
signiIicantly Ior high enough temperatures to model stirring eIIect in molten metal. These extensions are not made in
the present study. Also, latent heat oI solidiIication is not considered here. It can be easily incorporated iI desired by
modiIying the speciIic heat or by using the LATENT HEAT option in Marc. It should also be noted that solid-solid phase
transIormation capability are not considered in this example. The T-T-T parameter and TIME-TEMP option are
available in Marc Ior deIining solid-solid phase transIormations. It should, however, be noted that these options are
not supported by the GUI and that the data requirements Ior these options are signiIicant.
Finally, note that the yield stress and its dependence on plastic strain, strain rate, and temperature is not directly
provided in the GUI. This is accessed Irom the AFFLOWMAT directory at run-time. The *.out` Iile produced by
Marc at run-time indicates the name oI the Iile being accessed Ior the Ilow stress data.
PLOT
ELEMENTS
SETTINGS
RELATED PLOT SETTINGS
SHELL
PLOT EXPANDED
REGEN
MATERIAL PROPERTIES
READ
100Cr6
378 Marc Users Guide: Part I
CHAPTER 2.4
Weld Path Setup
Two weld paths are setup here, one Ior each weld source. Prior to setting up weld path 1, two poly-line curves are
deIined at the root and throat oI the Iirst weld Iiller, as shown in Figure 2.4-2.
Figure 2.4-2 Weld Path Definition using Poly-Line Curves
Note that both curves should have the same number oI points and the direction in which the curves are deIined should
be the same. Also, the order in which the points are clicked is important. The Iirst weld path is then deIined in Mentat
as Iollows:
As the weld path is created, the local X-Y-Z axes oI the weld path are shown in Mentat. The Z-axis is along the weld
motion direction, the Y-axis indicates the orientation direction oI the weld arc, and the X-axis indicates the width
MODELING TOOLS
WELD PATHS
PATH INPUT METHOD
CURVES
CURVES ADD
pick primary curve
ORIENTATION INPUT METHOD
CURVES
CURVES ADD
pick auxiliary curve
ANGLE
180
Auxiliary Curve
Primary Curve
379
CHAPTER 2.4
Arc Welding Process Simulation
direction oI the weld. The DRAW WELD PATHS option allows the path to be shown as a yellow line with the associated
local weld directions indicated on the path. The ANGLE value oI 180
o
allows the weld orientation direction to
be reversed.
Prior to setting up weld path 2, an auxiliary node is deIined at the center oI the model at 0,0,-10. This node is used to
deIine the orientation oI the weld arc. Note that the number oI auxiliary nodes can either be 1 (as in this model) or
equal to the number oI primary nodes deIining the weld path. Once again, the order in which the nodes are clicked
is important.
The second weld path is then deIined using nodes as Iollows:
The ANGLE value oI 45
o
allows the weld orientation direction to be rotated about the weld path direction, as shown
in Figure 2.4-3. The ARC INTERPOLATION option is turned on Ior this path.
Figure 2.4-3 Weld Path Definition using Nodes
MODELING TOOLS
WELD PATHS
PATH INPUT METHOD
NODES
NODES ADD
pick Primary line of nodes
ORIENTATION INPUT METHOD
NODES
NODES ADD
pick Auxiliary node
ANGLE
45
Primary line of Nodes
Auxiliary Node
380 Marc Users Guide: Part I
CHAPTER 2.4
Weld Filler Setup
Three optional Ieatures can be deIined Ior any weld Iiller:
Melting Point Temperature iI this is set, the weld Iiller is introduced in the model at this temperature. II
not set, the usual approach is to heat up the Iiller through direct weld Ilux boundary conditions.
Filler Bounding Box iI the deIault is used, weld dimensions set on the WELD FLUX option are used to
deIine the Iiller bounding box in order to identiIy when Iiller elements are active in the model. II the deIault is
not used, the bounding box dimensions in the local X, Y, Z and -Z directions are set here.
Initial Status can be set as either Deactivated (usual option) or Quiet. The deactivated option should be
used when large motions are not expected in the model and large deIormations are not expected in the vicinity
oI the Iiller elements. II these conditions are not satisIied, the quiet option could be used.The quiet option
requires an appropriate property scaling Iactor to be set (deIault is 1e-5). It should be noted that the quiet
option is susceptible to ill-conditioning, and the property scale Iactor may have to be massaged in order to
avoid problems.
Two weld Iillers are set up in the current model.
Weld Filler 1
The Iirst weld Iiller, comprising oI shell elements, is set up without a melting point temperature. Weld Ilux boundary
conditions, described later in the Initial/Boundary Conditions section, is used to heat up the weld Iiller directly. The
Iiller bounding box dimensions are set here. In the X- (width) and Y- (depth) directions, coarser dimensions (10 mm)
are used in order to ensure that the entire cross-section oI the Iiller element set is activated simultaneously. In the Z-
(length) direction, the Iiller bounding box values correspond to the physical Iiller length that participates in the weld
pool. This is set to 5 mm. The initial status is set to deactivated.
Weld Filler 2
The second weld Iiller, comprising oI solid elements, is set up with a melting point temperature oI 1200
o
C. The
temperature ramp time is leIt as 0, which implies that the temperature is introduced instantaneously. DeIault Iiller
MODELING TOOLS
WELD FILLERS
FILLER BOUNDING BOX DEFAULT
X 10
Y 10
+Z 5
-Z 5
INITIAL STATUS
DEACTIVATED
ELEMENTS ADD
Add elements belonging to weld filler 1
381
CHAPTER 2.4
Arc Welding Process Simulation
bounding box values are to be used which implies that the bounding box dimensions are equal to: 1.5 times the weld
width in the X-direction (15 mm), 2 times the weld width in the Y-direction (20 mm), the weld Iorward length in the
Z-direction (2 mm) and the weld rear length in the -Z-direction (8 mm). Note again that in the local X- and Y-
directions, the bounding box dimensions can be loosely set to larger values in order to ensure that the entire solid
cross-section is activated simultaneously; whereas, in the Z-direction, the bounding box dimension is tightly coupled
with the associated weld pool dimensions. The initial status oI the solid weld Iiller is also set to deactivated.
Contact Body Setup
The weld Iillers can be linked to the other components in the model either through homogeneous meshing or
through contact bodies. In the current model, the mesh Ior weld Iiller 1 (shell) is continuous with the Ilange and
cylinder meshes.
Weld Iiller 2 (solid) is deIined as a contact body. The Ilange in the vicinity oI weld Iiller 2 (shell) and the cylinder
(solid) are also deIined as contact bodies. The contact body setup is shown in Figure 2.4-4.
Figure 2.4-4 Contact Body Definition for Weld Filler 2
MODELING TOOLS
WELD FILLERS
MELT POINT TEMP
1200
INITIAL STATUS
DEACTIVATED
ELEMENTS ADD
Add elements belonging to weld filler 2
382 Marc Users Guide: Part I
CHAPTER 2.4
A contact table is then set up between the three contact bodies. Weld Filler 2 is glued to the Ilange and to the solid.
The Ilange is allowed to touch the solid. A heat transIer coeIIicient oI 100 N/mm
2
/sec/
o
C is used between the bodies.
Note that the units used Ior the heat transIer coeIIicient should be consistent with the other dimensions and properties
used in the model.
Initial/Boundary Conditions
All the nodes in the model are set to an initial temperature oI 30
o
C. Due to the presence oI shell elements with linear
temperature through-thickness variation, both the top and bottom temperature values are set to 30
o
C.
The solid cylinder and shell cover plate are Iixed in the X-direction along the centerline. The Iixtures holding the
system are modeled by Iixing the base oI the solid cylinder and shell cover plate in the X-, Y-, and Z-directions.
A Iace Iilm boundary condition is applied to all the exposed Iaces oI the cylinder, Ilanges and cover plate. The sink
temperature is set to 30
o
C and the Iilm coeIIicient is taken as 0.02 N/mm
2
/sec/
o
C. For the shell elements, the Iilm
boundary conditions are applied to both the top and bottom Iaces. The Iace Iilm boundary condition is not applied to
the Ilange and cylinder Iaces that are covered by the weld Iillers.
The weld Iluxes applied to the solid cylinder and the weld Iillers are described in detail here.
Weld Flux Associated with Weld Filler 2
This is a volume weld Ilux that is applied to the elements in the vicinity oI weld Iiller 2. No power is provided since
the heat input is to come Irom the molten Iiller elements. The dimensions are speciIied as width 10 mm, depth 0
mm, Iorward length 2 mm and rear length 8 mm. Note that since the provided Ilux has zero magnitude, the width,
Iorward length, and rear length are only used here to deIine the Iiller box dimensions. The Initial Weld Position is taken
as deIault (internally set to the Iirst point oI the weld path). The velocity is set to 2 mm/sec. Weld path 2 is chosen Ior
the weld path and weld Iiller 2 is used Ior the weld Iiller. The button clicks Ior setting up weld Ilux 2 are as Iollows:
BOUNDARY CONDITIONS
THERMAL
VOLUME WELD FLUX
FLUX
DIMENSIONS
WIDTH
10
DEPTH
0
FORWARD LENGTH
2
REAR LENGTH
8
383
CHAPTER 2.4
Arc Welding Process Simulation
Since the Ilux is 0 and all the heat input in this boundary condition is Irom the molten Iiller, it is not very critical to
identiIy which elements receive the Ilux. Note, however, that it is important to apply this boundary condition to at least
one element so that it gets written out to the input Iile.
Weld Fluxes associated with Weld Filler 1
Two weld Iluxes (weld Ilux 3 and weld Ilux 1) are applied to weld Iiller 1 and to the elements in the vicinity oI this
weld Iiller, respectively. While it is certainly convenient to apply the temperature oI the weld Iiller directly as shown
Ior weld Ilux 2, the objective oI these two boundary conditions is to demonstrate the use oI actual weld Iluxes to heat
up the weld Iiller and the surrounding elements. It is assumed that the total heat input Irom the weld torch can be
divided up into the heat input going to the weld Iiller (weld Ilux 3) and the heat input going to the surrounding material
(weld Ilux 1). The total heat input Irom the weld torch is taken as 1.5e6 Nmm/sec (about 1.4 BTU/hour). 66 oI this
heat (1E6) is assumed to be directly taken by the solid cylinder and 33 (5E5) is assumed to be taken by the weld
Iiller. Furthermore, it is assumed that weld Ilux 1 should have a double ellipsoidal variation over the cylinder while
weld Ilux 3 should be nearly uniIorm over the weld Iiller.
Weld Flux 1: A conventional double ellipsoidal volume weld Ilux (weld Ilux 1) with appropriate dimensions is set up
Ior the solid cylinder as Iollows:
MOTION PARAMETERS
VELOCITY
2
WELD PATH
weldpath2
WELD FILLER
weldfill2
ELEMENTS ADD
Pick a few elements in the vicinity of weld filler 1
BOUNDARY CONDITIONS
THERMAL
VOLUME WELD FLUX
FLUX
MAGNITUDE
POWER
1e6
EFFICIENCY
0.7
DIMENSIONS
WIDTH
384 Marc Users Guide: Part I
CHAPTER 2.4
Weld Flux 3: A disk shaped Iace weld Ilux (weld Ilux 3) is set up Ior weld Iiller 1. The path Iollowed by the weld Ilux
is identical to weldpath1 with the exception that it is oIIset Irom the given path by 3.53 mm in the local negative
Y-direction.
Since the stipulation is that the weld Iiller should receive uniIorm heat, additional modiIications to the standard disc
model in Marc are necessary. The Gaussian expression Ior the heat source is given by the expression below:
In order to heat the weld Iiller uniIormly, it is necessary that the exponential Iunctions in the above expression have a
value oI about 1 Ior representative x and z values oI the weld Iiller. This can only be achieved by assuming a very large
radius (r 30 mm) Ior the Iace weld Ilux. When such a large value is used Ior r, Ior values oI x and z in the range oI
3 to 5 mm (note that this range corresponds to the width oI the actual weld Iiller elements), q(x,y,z) is nearly uniIorm.
This non-physical assumption Ior the weld radius, however, requires two additional parameters to be Ilagged Ior the
Iace weld Ilux.
The Iirst parameter is the scale Iactor s. Note that the integral oI the Iace weld Ilux over the surIace oI the weld Iiller
should still equal Q. So, the scale Iactor s is given by:
4
DEPTH
2
FORWARD LENGTH
2
REAR LENGTH
8
MOTION PARAMETERS
VELOCITY
2
WELD PATH
weldpath1
ELEMENTS ADD
Pick all the solid elements that can potentially
receive the heat input
q x y z , , ( )
3Q
tr
2
--------
3x
2

r
2
------------
\ .
| |
3z
2
( )
r
2
---------------- exp exp
s
3Q
t900
------------
3x
2

900
------------
\ .
| |
exp x
3z
2

900
-----------
\ .
| |
exp
5
5
}
d z d
3.53
3.53
}
Q
385
CHAPTER 2.4
Arc Welding Process Simulation
By assuming the exponential terms to be nearly unity, integrating over the entire cross section and taking into account
both top and bottom Iaces Ior the weld Ilux, s can be given by:
This yields a value oI s 3.019. Due to the approximations involved in the integral, s is set to 3.25 in the current model.
The second parameter is the maximum weld distance. This reIers to the maximum distance beyond which weld Ilux
is not considered. It can be leIt undeIined iI physical values are used Ior the weld dimensions. However, since r 30
mm is not a physical dimension, the maximum distance within which nonzero Ilux values are to be considered needs
to be set. In the current example, the maximum weld distance is set to 5 mm, which implies that Ior integration points
that are located more than 5 mm Irom the weld origin, the weld Ilux is taken as 0. This is particularly important to
restrict the number oI Iiller elements participating in the weld pool in the z direction.
The Iace weld Ilux Ior weld Iiller 1 is then set up as Iollows:
BOUNDARY CONDITIONS
THERMAL
FACE WELD FLUX
FLUX
MAGNITUDE
POWER
5e5
EFFICIENCY
0.7
SCALE FACTOR
3.25
DIMENSIONS
SURFACE RADIUS
30
MAXIMUM DISTANCE
5
MOTION PARAMETERS
VELOCITY
2
WELD PATH
weldpath1
OFFSET-Y
s
900t
3
------------
1
5 7.5 2 ( )
------------------------------
1
10
----- -
1
2
---
x - term z term top and bottom faces
386 Marc Users Guide: Part I
CHAPTER 2.4
Loadcase Definition
Two thermo-mechanically coupled loadcases are used to conduct the welding analysis. Loadcase 1 is used to simulate
the weld at Iiller 1 and loadcase 2 is used to simulate the weld at Iiller 2. Adaptive Stepping (AUTO STEP) is used Ior
loadcase 1 while Iixed stepping (TRANSIENT NON AUTO) is used Ior loadcase 2.
Loadcase 1
The weld Iluxes associated with weld Iiller 1 (weld Ilux 1 and weld Ilux 3) are applied in this loadcase. Weld Ilux 2 is
deselected. The maximum error in temperature estimate is set to 30
o
C. This is an important quantity to speciIy and
ensure that the thermal analysis is conducted with converged temperature dependent material properties. The total
loadcase time is set to 10 seconds. The ADAPTIVE STEPPING - MULTI-CRITERIA stepping procedure is used. All
deIaults are used Ior the time stepping. A temperature user criterion is speciIied with an allowable temperature
increment oI 200
o
C. This supersedes the deIault temperature criterion oI 20
o
C. The appropriate button clicks to set up
the loadcase are as Iollows:
weldpath1
FACES ADD
Pick the top and bottom faces of weld filler 1
LOADCASES
COUPLED
QUASI-STATIC
LOADS
deselect weld_flux2
CONTACT
select ctable1
CONVERGENCE TESTING
MAX ERROR IN TEMPERATURE ESTIMATE
30
TOTAL LOADCASE TIME
10
STEPPING PROCEDURE
ADAPTVE MULTI-CRITERIA
USER-DEFINED CRITERIA
TEMPERATURE INCREMENT PARAMETERS
200
[] PROCEED WHEN NOT SATISFIED
387
CHAPTER 2.4
Arc Welding Process Simulation
Loadcase 2
The weld Iluxes associated with weld Iiller 2 (weld Ilux 2) are applied in this loadcase. Weld Ilux 1 and Weld Ilux 3
are deselected. The maximum error in temperature estimate is set to 30
o
C. This tolerance is specially important to
speciIy Ior Iixed stepping loadcases since no other checks on allowable temperature change are made in the case oI
Iixed stepping. The total loadcase time is set to 10 seconds. The FIXED STEPPING procedure is used with a total oI
50 increments (0.2 seconds per increment). The appropriate button clicks to set up the loadcase are as Iollows:
Job Parameters
A coupled job is set up and the deIined loadcases are selected. The shell contact is simpliIied by only checking on the
top surIace and ignoring the thickness. This is necessary since the model has been built by putting weld Iiller 2 on the
Ilange midsurIace. The bias Iactor is taken as 0.95.
The LARGE STRAIN ADDITIVE (plasticity,3) procedure is chosen. LUMPED MASS AND CAPACITY option is Ilagged.
This is an important option to use Ior welding problems since it reduces thermal oscillations induced by the sudden
thermal shocks in the system. The layer von Mises stress, equivalent plastic strain, and temperatures are requested.
Additional print-out in the *.out` Iile Ior contact and welding are requested as Iollows:
LOADCASES
COUPLED
QUASI-STATIC
LOADS
deselect weld_flux1 and weld_flux3 and
select weld_flux2
CONTACT
select ctable1
CONVERGENCE TESTING
MAX ERROR IN TEMPERATURE ESTIMATE
30
TOTAL LOADCASE TIME
10
STEPPING PROCEDURE
FIXED
PARAMETERS
# STEPS
50
JOBS
COUPLED
388 Marc Users Guide: Part I
CHAPTER 2.4
A restart Iile at the end oI every loadcase can be requested as Iollows:
The large increment Irequency allows Marc to only write out the restart Iile at the end oI every loadcase (assuming
that the loadcase takes Iewer than 500000 increments).
JOB RESULTS
OUTPUT FILE
CONTACT
WELDING
JOBS
COUPLED
JOB PARAMETERS
RESTART
<> WRITE
INCREMENT FREQUENCY
500000
389
CHAPTER 2.4
Arc Welding Process Simulation
Results and Discussion
Figure 2.4-5 Comparison of Theoretical and Calculated Heat Input for Volume Weld Flux
and Face Weld Flux at Filler 1
A good accuracy check is to compare the theoretical and calculated heat inputs Ior the weld Iluxes. Assuming that the
entire heat input acts on the structure, the theoretical heat input Ior the weld Iluxes are given by H q Q. The
calculated heat input is obtained in the *.out` Iile by requesting the additional print-out Ior welding.
The theoretical heat input Ior weld Ilux 1 (volume weld Ilux) is 7e5 N mm/sec. This is based on the assumption that
the entire double ellipsoid is acting on the solid.
Since the heat input is oriented at 45
o
to the surIace, this is not strictly valid in the current case. It is still seen that the
calculated heat input is relatively close to the theoretical value.
The theoretical heat input Ior weld Ilux 3 (Iace weld Ilux) is 3.5e5 Nmm/sec. The calculated heat input is seen to be
reduced at the beginning and at the end. This is because only halI oI the Gaussian distribution is captured by the weld
Iiller elements at the beginning and end. For intermediate stages, it is seen that the calculated heat input has a wavy
pattern. This wavy pattern coincides with the activation oI the Iiller elements. With a Iiner Iiller element mesh, the
waviness would reduce. For the purposes oI the current demonstration, it is deemed that the accuracy oI the calculated
heat input is suIIicient. The weld Ilux parameters and/or the mesh size can be adjusted to make the correspondence
between the calculated and theoretical heat inputs closer. Tables as a Iunction oI time could be employed to improve
the comparison especially at the beginning and end stages.
Calculated Heat Input
Theoretical Heat Input
for Weld Flux 1
Calculated Heat Input for Weld Flux 3
for Weld Flux 1
Theoretical Heat Input
for Weld Flux 3
390 Marc Users Guide: Part I
CHAPTER 2.4
The von Mises Stress (Layer 1) and Temperature proIile (Top) at the end oI loadcase 1 (aIter laying weld Iiller 1) is
shown in Figure 2.4-6 and Figure 2.4-7 respectively. It is seen that the highest temperatures are close to 1200
o
C and
the highest residual stresses are in the solid cylinder elements close to the weld Iiller.
Figure 2.4-6 von Mises Stress Contours at the End of Loadcase 1
Figure 2.4-7 Temperature Profile at the End of Loadcase 1
391
CHAPTER 2.4
Arc Welding Process Simulation
The von Mises Stress (Layer 1) and Temperature proIile at the end oI loadcase 2 (aIter laying oI both weld Iiller 1 and
weld Iiller 2) are shown in Figure 2.4-8 and Figure 2.4-9 respectively. It is seen that the largest temperature oI 1200
o
C
is at the right end oI the Iiller and portions oI the Iiller that have moved out oI the weld pool are signiIicantly cooler.
The residual stresses are signiIicant in the Ilange, shell wall and solid regions.
Figure 2.4-8 von Mises Stress Contours at the End of Loadcase 2
Figure 2.4-9 Temperature Profile at End of Loadcase 2
392 Marc Users Guide: Part I
CHAPTER 2.4
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
weld.proc Mentat procedure Iile to run the above example
weld.mud Associated model Iile
weld.dat Associated Marc input Iile
Chapter 2.5: FEM Simulation of NC Machining and PRE STATE
2.5
FEM Simulation of NC
Machining and PRE STATE

Chapter Overview 394

Example 1: Pocket Cutting 395

Example 2: Thin Frame Cutting 412

Example 3: Imported Initial Stresses 420

Import with PRE STATE Feature 425

Input Files 430

394 Marc Users Guide: Part I
CHAPTER 2.5
Chapter Overview
In manuIacturing industry, NC machining is a material removal process that is widely used to produce a part with the
desired geometry. AIter removal oI the machined material, re-establishment oI equilibrium within the remaining part
oI the structure causes some distortion due to the relieI oI the residual stresses in the removed materials. The
deIormation caused by this process usually depends on the residual stress level and its distribution inside the part. It
also depends on the Iinal geometry oI the part aIter machining. For a part with Iinal geometry that includes thin wall
or large plate structures, the deIormation can be so large that it causes severe distortions oI the shape. The highly
distorted part may no longer be able to serve its designated Iunctionality. Such types oI Iailures result in high scrap
rates and increased manuIacturing costs.
Finite element procedure is a powerIul tool to analyze the potential distortion caused by the deIormation during the
machining process. With the FEM results, it is then possible Ior engineers to predict the potential Iailures and reduce
overall manuIacture costs.
The capability Ior the simulation oI NC machining (material removal) processes was initially introduced in the
MSC.Marc version 2003. The Iollowing basic Ieatures in MSC.Marc version 2003 had been developed in order to
enable Marc to conduct the automatic simulation oI NC machining processes:
1. InterIace between Marc and CAD/NC data that describes the cutter shape and cutter path;
2. Detection oI FE mesh-cutter intersection;
3. Automatic deactivation oI elements that are cut;
4. Visualization oI the machining process during postprocessing oI FEA results.
In the current releases oI Marc since version 2007, a number oI additional enhancements have been made to improve
the accuracy, computational eIIiciency, and user-Iriendliness. In summary, the new enhancements and improvements
are listed as Iollowing:
1. The support oI CYCLE statement has been expanded Irom CYCLE/DRILL only to CYCLE/(DRILL, DEEP,
TAP, BORE, and CBORE).
2. More eIIicient and accurate cutter-mesh intersection detection has been implemented.
3. Loadcase time synchronization is now allowed so that time-dependent contact and user boundary conditions
can be used in conjunction with machining.
4. Cutter visualization is allowed during postprocessing oI simulation results.
5. Local adaptive remeshing Ieature is added Ior NC machining analysis. For NC machining simulation purpose,
this Ieature has been Iurther enhanced so that multi-level element splitting, regular and irregular adaptive
remeshing are possible.
a. Multi-level splitting oI an element: This allows an element to be subdivided into the maximum allowed
subdivision levels within one incremental step.
b. Regular adaptive remeshing: Elements that are partially intersected will be subdivided at each increment.
Note that this can cause mesh size and computational time to increase signiIicantly.
395
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
c. Irregular adaptive remeshing: Elements that are partially intersected are not subdivided during the Iirst
coarse stage oI machining. These elements are subdivided during a second Iine stage oI machining, wherein
all the splitting is conducted in the last increment oI the loadcase. This two-stage remeshing process can save
signiIicantly on the computational time and memory usage.
4. Automated residual stress import: Based on the source oI residual stresses, Marc can accept stress input data
Irom Marc result data Iiles obtained Irom previous numerical analyses or residual stress data saved in text
Iormat data Iile Irom experimental analyses.
Three methods have been made available in Marc to import residual stresses into the model prior to machining:
a. Pre State: This method directly transIers data Irom a previously obtained Marc post Iile into the new Marc
Machining model.
b. Text data file: By this method, Marc reads in the residual stress data stored in a text Iormat data Iile. The
stress data are automatically mapped into the FE model by Marc.
c. Table format: User can deIine the residual stresses as tables deIined in space.
With exception oI the enhancements made available oI the interIace to import residual stresses in the FE model, the
interIace Ior cutting path and the FE mesh is the same as MSC.Marc version 2003. The cutter path data are stored in
either APT source or CL data Iormat. The APT source is the NC data output by CAD soItware CATIA. The CL data
is the cutter location data provided by APT compilers.
This chapter describes the usage oI the enhanced capability in the Marc program Ior simulation oI NC machining
(material removing) processes. Three examples are chosen to demonstrate the utilization oI the major new Ieatures:
Example 1: Pocket Cutting
Example 2: Thin Frame Cutting
Example 3: Imported Initial Stresses
For each example, Mentat procedures show how to enclose this Iunctionality into the machining process simulation
by deIining the Marc models together with the NC data in either APT source or the CL data is described in a
step-by-step manner.
Example 1: Pocket Cutting
Input data
The input data required Ior the simulation oI machining process with Marc including both sides oI CAD interIace that
deIines the NC machining process and Marc which deIines the input Ior Iinite element analysis. They are summarized
as Iollowing:
NC data to deIine the cutter geometry and cutter path Ior the machining process. (.apt or .ccl Iiles). For details
oI the Iormat oI the apt or ccl Iiles, please reIer to Marc Volume A: Theory and User Information Manual and
the reIerences listed there.
396 Marc Users Guide: Part I
CHAPTER 2.5
Marc input data includes the Iile names Ior cutter path deIinition and Iinite element model deIinition Ior the
workpiece. The workpiece can be also imported Irom an IGES data which is written by CAD soItware.
Initial stress data beIore the cutting process started. In the current example, the initial stress is the course
oI distortion aIter cutting. For this particular example, the initial stresses are provided in a 2-D model. We
just converted 2-D data into our 3-D model using corresponding adjustments regarding unit and
dimension deIinitions.
Initial Geometry and Stresses
The FE model is created by preprocessing capability oI Mentat. The purpose oI preprocessing is to generate the model
and input data Ior Marc to analyze the metal cutting process. AIter the enhancements added in the current version, it
allows the user to speciIy the Iile name oI cutter path data.
The geometry oI the part beIore cutting is shown in Figure 2.5-1. The initial part is a block with size oI length by width
by thickness 28x14x4.5 (inches). The residual stresses are predeIined in the model already, and is deIined in Mentat
shown in Figure 2.5-2.
Figure 2.5-1 Initial Part Geometry
Notes:) 1)In the current version, circular motion is required to transIorm into point-to-point motion type when output by CAD
NC soItware. In addition, the TRACUT and COPY statements are necessary to be explicitly interpreted into cutter
motion statements. Major statement CYCLE is supported in combination with DRILL minor statement Ior the
deIinition oI drilling motion type.
2)The Ilipping over oI a part during the course oI machining process is supported by converting the Ilip over oI the part
into the rotation oI cutter axis. MLTAXS statement is used to deIine the rotation oI cutter axis.

width
thicknes
length
thickness
length
width
397
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Figure 2.5-2 Definition of Cutting Processes
The cutting process includes two cut steps:
The Iirst step is to cut 2 inches oII the upper surIace as shown in Figure 2.5-2. The cutting depth oI each
cutting step is deIined by the cutter path data Iile m2q0090s1.ccl.
The second step is to cut two pockets over the lower surIace oI the part aIter the Iirst cut step is done. The
cutter path Ior this step is deIined by the cutter path data Iile m2q0090s2.ccl. The ccl Iiles are created based on
the APT sources generated by the CAD soItware CATIA.
Between the Iirst and second step, the part is supposed to be Ilipped over, so that the cutter axis is unchanged in the
second cut step. However, Ior the convenience oI FE model deIinition and analysis, the Ilipping over oI the part is
equivalently simulated by the rotation oI the cutter. ThereIore, the second cut is conducted by rotating the cutter into
the opposite direction, as shown in Figure 2.5-2.
See below the step-by-step commands to execute the procedure to deIine the boundary conditions and loadcases in
order to conduct the cutting processes sequentially and automatically.
Here, we will create the model by reading a predeIined model Iile. This assumes that the users are Iamiliar with the
model generation. The model Iile to be read in is ex_r01.mud. The cutter path Iiles are deIined and saved in the current
working directory. First, we will read in this mud Iile using the procedure Iile: mc_nfg.proc. As show in Figure 2.5-3,
aIter clicking on the command LOAD, and selecting the Iile name: mc_nfg.proc, then click command START/CONT,
the Iile ex_r01 is read into Mentat. The sequence is recorded as (aIter Mentat is started):
UTILS
PROCEDURES
LOAD
mc_nfg.proc
Surface for the first cut step
Cutter and its axis definition1
Surface where the second
cut step starts
Cutter and its axis definition2
398 Marc Users Guide: Part I
CHAPTER 2.5
Up to now, the initial model Iile ex_r01.mud has been read in. The next step is to work on this model by deIining
boundary conditions and loadcases beIore submitting the job. So, we click the START/CONT button again.
Totally, 28224 brick elements and 32205 nodes are deIined in the model. Figure 2.5-3 shows the model and its
initial stresses.
Figure 2.5-3 Model with Initial Stresses before Machining (a) o
x
, (b) o
y
, (c) o
z
, (d) o
xz
, o
zy
= o
xy
=0
By now, the model is in and all the elements have been applied to initial stress. The initial stress inIormation was
provided by the company. What we did is to convert 2-D data into our 3-D model using the corresponding adjustments
regarding unit and dimension deIinitions. Isotropic material property parameters are used, which are deIined: E
(Young's Modules) 1000, Poisson`s ratio 0.3.
OK
START/CONT
START/CONT
(a) (b)
(c) (d)
399
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Next step oI model deIinition is to deIine boundary conditions and loadcases. This procedure is very long and recorded
in procedure Iile: machining_rcd. By loading this procedure Iile and click on START/CONT, Mentat will complete all
the tasks automatically. For better understanding, the user may use the STEP button to conduct the procedure
step-by-step. In this Marc Users Guide, we only selectively demonstrate the key steps that are needed to generate the
input data Ior the metal cutting analysis.
There are a total oI Iour loadcases deIined in this model. They are:
1. Cut the top part of the workpiece. The cut Iile used here is m2q0090s1.ccl.
2. Release the bottom boundary condition and apply to the upper face. This loadcase is the one to Ilip over
the part by switching the boundary conditions Irom bottom to the newly generated top surIace.
3. Cut the pocket from the lower face. This loadcase is the one used to cut the pocket on the lower side oI the
part. The cut Iile used here is m2q0090s2.ccl.
4. Final release (springback). This loadcase is to Iinally release all the boundary conditions, except those
required to clear the rigid body motion oI the part.
The total sets oI boundary conditions deIined by this procedure are:
Fix_bottom: This set Iixes the x-y-z displacement oI all the nodes at the bottom surIace. It is used in
loadcase 1.
Fix_middle: This set Iixes the x-y-z displacement oI all the nodes at the top surIace oI the part aIter the Iirst
cut. It is used in loadcases 2 and 3.
Fix_xyz: This set Iixes the x-y-z displacement oI node 2266.
Fix_x: This set Iixes the x displacement oI node 9.
Fix_y: This set Iixes the y displacement oI node 32065.
Fix_z: This set Iixes the z displacement oI node 32058. Boundary Condition sets 3 to 6 are used in the
loadcase 4.
The Mentat commands to deIine all the loadcases are shown as below:
Loadcase1 (Cut the top part of the workpiece):
MAIN
LOADCASES
NEW
NAME
cutface1
MECHANICAL
STATIC
LOAD
fixbottom: (to the B.C. for the loadcase)
CONVERGENCE (defining convergence criteria)
400 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-4 Definition of First Cutting Loadcase
RESIDUAL
AUTO SWITCH
Relative Force Tolerance
0.01
OK
CONSTANT TIME STEP
STEPS
10
OK
AUTO TIME STEP CUT BACK (OFF)
OK
DEACTIVATION / NC MACHINING (enter the parameters)
NC MACHINING
FILE
m2q0090s1.ccl
(name cutter path definition)
LAST INCREMENT (select remeshing method)
(take defaults for other parameters)
RETURN
TITLE
cut the top part of the workpiece
401
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Now, as shown in Figure 2.5-4, the Iirst loadcase has been deIined. Next step is to deIine the loadcase to Ilip over the
part aIter the Iirst cut step is completed.
Loadcase2 (Release the bottom boundary condition and apply to the top face):
Now, the second loadcase has been deIined. Next step is to deIine the loadcase to cut the pockets over the other side
oI part. The procedure is recorded as Iollowing:
Loadcase3 (cut the pocket from the lower face part):
MAIN
LOADCASES
NEW
NAME
release_bot
MECHANICAL
STATIC
LOAD
fixbottom (OFF to free B.C. for the loadcase 1)
fixmidface (apply B.C. on middle surface)
CONVERGENCE (defining convergence criteria)
RESIDUAL
AUTO SWITCH
Relative Force Tolerance
0.01
OK
CONSTANT TIME STEP
STEPS
10
OK
AUTO TIME STEP CUT BACK (OFF)
OK
MANUAL (for Inactive Elements)
TITLE
Release the bottom B.C. and apply to the top face
OK
MAIN
LOADCASES
402 Marc Users Guide: Part I
CHAPTER 2.5
When the second cutting step is Iinished, we need to do the analysis oI springback. This process requires Marc to Iree
all the restraints except those that are needed to prevent rigid body motion. So, we deIine the minimum boundary
condition Ior this loadcase (only 6 DOF`s are Iixed Ior the whole model). The procedure is recorded as Iollowing:
NEW
NAME
cut pocket
MECHANICAL
STATIC
LOAD
fixbottom (to apply B.C.)
CONVERGENCE (defining convergence criteria)
RESIDUAL
AUTO SWITCH
Relative Force Tolerance
0.01
OK
CONSTANT TIME STEP
STEPS
10
OK
AUTO TIME STEP CUT BACK (OFF)
OK
DEACTIVATION / NC MACHINING (enter parameters)
NC MACHINING
FILE
m2q0090s2.ccl
(name cutter path definition)
LAST INCREMENT (select remeshing method)
(take defaults for other parameters)
RETURN
TITLE
cut the pocket from the lower face part
OK
403
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Loadcase4 (Final release (springback):
MAIN
LOADCASES
NEW
NAME
final_release_bc
MECHANICAL
STATIC
LOAD
fixmidface (free B.C.)
fix_xyz (fix x, y and z)
fix_x (fix x)
fix_y (fix y)
fix_z (fix z)
CONVERGENCE (defining convergence criteria)
RESIDUAL
AUTO SWITCH
Relative Force Tolerance
0.01
OK
CONSTANT TIME STEP
STEPS
2
OK
AUTO TIME STEP CUT BACK (OFF)
OK
MANUAL (for Inactive Elements)
TITLE
final release (spring back)
OK
404 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-5 Definition of Final Springback Analysis
Local Mesh Adaptivity Definition
AIter the loadcases are deIined, it is necessary to deIine the Marc job. The Iollowing procedure records the job
deIinition with this model.
First, it is necessary to deIine the parameters Ior local adaptive remesh:
MAIN
MESH ADAPTIVITY
LOCAL ADAPTIVITY CRITERIA
MORE
ELEMENT WITHIN CUTTER PATH
MAX # LEVELS
1
OK
ADD
EXIST (choose all existing elements)
405
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Figure 2.5-6 Definition of Parameters for Local Adaptive Remeshing
BeIore job deIinition, we deIine the element type Ior this analysis by:
Then the job deIinition is done with the Iollowing procedure.
MAIN
JOB
ELEMENT TYPES
MECHANICAL
3-D Solid
select 7
OK
EXIST (choose all existing element)
MAIN
JOB
NEW
NAME
metal_cut
MECHANICAL
Select loadcases 1, 2, 3, and 4 sequentially
(Applying loadcases)
406 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-7 Definition of Machining with Mesh Adaptivity
INITIAL LOADS (click OFF all the b.c.)
INITIAL CONDITIONS (check if they are all on)
ANLYSIS OPTIONS (use defaults for this)
MESH ADAPTIVITY (use defaults for this, Figure 2.5-7)
OK
JOB RESULTS (select the results that are interested)
CENTROID (reduce the post file size by clicking this button)
OK
JOB PARAMETER
SOLVER (choose correct solver)
ITERATIVE SPARSE
INCOMPLETE CHOLESKI
OPTIMIZATION
OK
(Iterative solver is used to reduce memory
requirement and total computation time)
OK
OK
407
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
AIter the job is deIined, to run the job and see the results, it is necessary to do Iollowing:
Up to now, the FE analysis oI machining (namely, metal cutting) process has started. By clicking the MONITOR button,
Mentat instantly shows the progress oI the calculation.
Visualization of Results
Enhancements have been made Ior better visualization oI the results oI Metal Cutting analysis. Particularly, the
elements being cut oII Irom the part are not displayed in Mentat, so that only the remaining part oI the FE model is
displayed Ior postprocessing purposes.
To check the results, we Iirst need to check whether the cutter paths have been Iollowed exactly. Secondly, we need to
check the deIormations and stresses leIt in the remaining part. By analyzing the deIormation/displacement results, we
can see the eIIect oI residual stress and machining process to the geometry oI the Iinal part.
As shown in Figure 2.5-11, we see that the cutter path has been processed properly during the FE analysis. The part
displayed strong deIormation aIter springback, as shown in Figure 2.5-11. The maximum displacement oI the part is
about 20 times larger aIter springback (Irom 0.0005688 increases to 0.01055 in.). Figure 2.5-12 shows the deIormation
pattern aIter machining process with the residual stress provided.
For places with corner oI small radius, a Iine mesh is required in order to have better resolution oI the part shape aIter
machining. Figure 2.5-13 shows the Iiner mesh aIter local adaptive remeshing at one oI the corner areas.
In Figure 2.5-8, the model displays the results aIter each loadcases, respectively.
MAIN
JOB
RUN
SUBMIT (1)
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
Total Displacement
OK
FILL
DYN. MODEL
MONITOR
408 Marc Users Guide: Part I
CHAPTER 2.5
1. Cut upper Iace
Figure 2.5-8 Machining of the Upper Surface
2. Flipping over (switch boundary conditions)
Figure 2.5-9 Flip over the Part after First Cutting Process
409
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
3. Cut pockets
Figure 2.5-10 Process of Pocket Cutting
Figure 2.5-11 Geometry after Pocket Cutting
410 Marc Users Guide: Part I
CHAPTER 2.5
4. Final release (springback)
Figure 2.5-12 Final Geometry and Deformation after Springback (with scaling)
Figure 2.5-13 Visualization of Mesh after Adaptive Remeshing
411
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Verification of Material Removal
Material removal during machining causes a redistribution oI residual stresses that change the dimensions oI the Iinal
part. An experimental veriIication oI this machining behavior is included as a 1.5 inch thick stock aluminum beam that
is bent to a prescribed radius. A 2.5 inch slot is cut through 75 oI its thickness. The displacements oI the ten gage
points shown in Figure 2.5-14 are measured beIore and aIter the machining process.
Figure 2.5-14 Machining Simulation and Test of Aluminum Beam:
Experimental Versus Simulation Deflection of the Ten Gage Points
0 2 4 6 8 10
0.00
0.05
0.10
0.15
0.20
Test 2
FEA Prediction
Test 1
Gage Point
Gage Point Displacement (in)
412 Marc Users Guide: Part I
CHAPTER 2.5
Example 2: Thin Frame Cutting
Input Data
The purpose oI this example is to compare the local adaptive remeshing methods Ior NC machining analysis. One
method is the regular adaptive remeshing, which the adaptive remeshing is conducted progressively with the cutter
motion. The other method is the irregular adaptive remeshing which the adaptive remeshing is conducted at the end
oI the cutting loadcase aIter the cutting path is completed.
Figure 2.5-15 Initial Geometry of the Workpiece
The input data available Ior this example is summarized as:
NC data: APT Iormat data is used Ior this example. There are three cutting passes (each pass has its own APT
Iile). They are: p1.apt, a2B.apt, and b3A.apt. The three cutting paths are illustrated in Figure 2.5-15.
Initial Marc model: For this example, it is imported Irom the .mud Iile which includes the Iinite element
model deIinition Ior the workpiece and the Iile names Ior cutter path deIinition. The workpiece is a slab with
dimensions oI 3.9x3.9x20. Totally, 13440 brick elements and 18135 nodes are deIined in the model.
Initial stress data before the cutting process started: For this particular example, the initial stresses are set
by the user with constant value oI s
xx
10000, o
yy
o
zz
o
zx
o
xy
o
yz
0.
As the Marc model has been predeIined, it is assumed that the user is already Iamiliar with model deIinition procedures
Ior material properties and boundary condition, etc. So, we have set a procedure Iile to read in the predeIined model
and continue Irom there to deIine the corresponding loadcases and job parameters to run the machining analysis oI this
cutting example.
p1.apt
a2B.apt
413
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
This is done by reading in the procedure Iile: example2.proc. As shown in Figure 2.5-3, aIter clicking the LOAD button,
and selecting the Iile name example2.proc, then click command START/CONT. The sequence is shown as (aIter Mentat
is started):
Up to now, by clicking on START/CONT button, the procedure Iile is executed. First, the initial model Iile
machining_mp1_cm_r1.mud is read in, initial stress is deIined, then job parameters and loadcases are set. Finally, the
job is submitted and the visualization oI results are started. The Mentat command is:
Initial Stress and Local Adaptive Remeshing Definition
First, the initial stress is deIined by (see Figure 2.5-16).
UTILS
PROCEDURES
LOAD
example2.proc
OK
START/CONT
MAIN
INITIAL CONDITIONS
NEW
MECHANICAL
STRESS
Click on the 1st component of stress
Enter value: 10000
OK
ADD
EXIST (choose all existing elements)
414 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-16 Definition of Initial Stress
Then, it is necessary to deIine the parameters Ior local adaptive remesh. This is done in a similar way as shown in
Figure 2.5-6 as:
Loadcases and Machining Job Definition
There are a total oI six loadcases deIined in this model. They are:
1. The first cutting path defined by APT file: p1.apt. In this loadcase, boundary condition set m_clamp is
applied.
2. Release pos1: To remove the current m_clamp boundary condition set and apply the release boundary
condition set oI m_clamp.
MAIN
MESH ADAPTIVITY
LOCAL ADAPTIVITY CRITERIA
MORE
(ELEMENT WITHIN CUTTER PATH)
MAX # LEVELS
1
OK
ADD
EXIST (choose all existing elements)
415
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
3. The second cutting path defined by APT file: a2B.apt. In this loadcase, boundary condition set m_clamp is
re-applied.
4. Release pos2: To remove the current m_clamp boundary condition set and apply the release boundary
condition set m_clamp.
5. The third cutting path defined by APT file: b2A.apt. Similarly, boundary condition set m_clamp is applied
in this loadcase.
6. Net Release: In this loadcase, both boundary condition sets: m_clamp and release are applied. So, this
workpiece is not totally Ireed Irom the clamps.
In the Mentat procedure Iile, the parameter deIinition oI all the machining (metal cutting) loadcases has been recorded.
However, due to the similarity, only the Iirst one is shown below:
Loadcase1 (Cut path of p1.apt):
MAIN
LOADCASES
EDIT
machine_pos1
OK
DEACTIVATION / NC MACHINING (enter parameters)
NC MACHINING
FILE
p1.apt (name cutter path definition)
LAST INCREMENT (select remeshing method)
(take defaults for other parameters)
RETURN
416 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-17 Definition of the First Case for Cutting Path: p1.apt
Similarly, the loadcases Ior the other two cut paths (a2B.apt and b2A.apt) can be deIined.
FEA Results
This analysis uses the local adaptive remeshing at each cutting loadcase. The adaptive remeshing is only conducted at
the last increment oI each loadcase. Figure 2.5-18 shows the intermediate stage oI the Iirst cut path. In Figure 2.5-18,
the original mesh is still present Ior the Iirst cutting path until the last increment is completed (Figure 2.5-19). AIter
three cutting paths, the workpiece and deIormation are displayed in Figure 2.5-21, respectively.
Figure 2.5-18 Intermediate Stage of the First Cutting Path
417
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
1. The workpiece aIter cutting path, p1.apt:
Figure 2.5-19 Workpiece after First Cutting Path
2. The workpiece aIter cutting path, a2B.apt:
Figure 2.5-20 Workpiece after the Second Cutting Path
418 Marc Users Guide: Part I
CHAPTER 2.5
3. The workpiece aIter cutting path, b2A.apt:
Figure 2.5-21 Workpiece after the Third Cutting Path
For comparison purpose, the regular adaptive remeshing method is also used to conduct the analysis. User can achieve
this by selecting the EACH INCREMENT button when deIining the adaptive remeshing method Ior the machining
loadcases. In this way, the adaptive remeshing is perIormed instantly at each increment iI an element is Iound partially
intersected by the cutter path. As shown in Figure 2.5-22, some elements are already subdivided at the intermediate
stage oI the Iirst cutting path because oI the intersection with the cutter. Comparing to the previous analysis with
adaptive remeshing only at the Iinal increment oI the loadcase, additional new elements and nodes are generated due
to the adaptive remeshing. Correspondingly, more CPU time and larger memory capacity are required.
Table 2.5-1 compares the general inIormation oI the two analysis. It can be seen that the analysis with instant adaptive
remeshing at each increment whenever any element needs to subdivided causes the model to become extremely large.
ThereIore, an increase in computer memory and CPU time are needed to complete the analysis. However, as shown
in Figure 2.5-23, the results are not signiIicantly diIIerent as compared with the results shown in Figure 2.5-21.
For this example, using the irregular method Ior adaptive remeshing can signiIicantly reduce the maximum number oI
elements and nodes generated during the remeshing process with nearly the same accuracy.
Table 2.5-1 Comparison of the Two Adaptive Remeshing Methods for Machining Analysis
Adaptive Remeshing
Method
Total
Number
Increments
Max. #
Nodes
Max. #
Elements
Memor
y (MB)
Total CPU
Time (sec)
Regular (Each increment) 362 105280 91200 1154 34429
Irregular (Last increment) 326 065892 51200 0693 20688
419
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Figure 2.5-22 Intermediate Stage of the First Cutting Path with Instant Adaptive Remeshing
Figure 2.5-23 Workpiece after Final Cutting Path with Instant Adaptive Remeshing
420 Marc Users Guide: Part I
CHAPTER 2.5
Example 3: Imported Initial Stresses
Overview
As mentioned beIore, the existing three possible approaches Ior the user to import the initial stresses are: Pre State,
Text Data File, and Table Input. This section demonstrates usages oI the Text Data File and the Pre-State approach
Ior the import oI the initial stress into the FE Model Ior machining analysis.
Import with Text Data File
Input Data
The text data Iile stores the initial stress data oI the workpiece oI the machining process. These data, generated either
by analytical or experimental methods, are saved in a text Iile in the Iormat as described in Appendix B oI the Marc
Volume A: Theory and User Information Manual.
The initial model is a block as shown in Figure 2.5-24. The red-lined surIace is the tool surIace created in the model
Ior the visualization oI cutter motion. One side oI the block is Iixed during the cutting process (see the arrows).
Figure 2.5-24 Initial Geometry of the Workpiece
421
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Model and Loadcase Definition
The whole procedure oI running this example has been recorded in procedure Iile: example3a.proc. The user can
Iollow that Iile to reproduce the input deck and visualize the results. To run this procedure Iile within Mentat, the user
only needs to Iollow the commands as shown below:
In order to make it easier to understand how to deIine a text data as initial stress and to deIine the tool surIace as the
cutter so that the cutter motion can be visualized, two extra procedures are shown below, respectively:
Use the command procedure below to deIine the initial stress data Iile (Figure 2.5-25):
UTILS
PROCEDURES
LOAD
example3a.proc
OK
START/CONT
MAIN
INITIAL CONDITIONS
NEW
MECHANICAL
STRESS
TEXT FILE
Initialstressn
OK
3-D
Click on each areas to define each components of
the two base vectors required for this analysis
1
0
0
0
1
0
OK
ADD
EXIST (choose all existing elements)
422 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-25 Definition of Text Data File for Initial Stress
Below is the Mentat command procedure to deIine the tool surIace to represent the cutter in the model (Figure 2.5-26).
Note that this tool surIace must be located at position (0,0,0) beIore the Iirst cutting process starts.
MAIN
LOADCASES
NEW
NAME
lcase1
MECHANICAL
STATIC
. . .
OK
DEACTIVATION / NC MACHINING (enter parameters)
NC MACHINING
FILE
extru-test.apt
BODY (select contact surface as cutter)
cbody2 (see Figure 2.5-26)
RETURN
423
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Figure 2.5-26 Definition of the Tool Surface to represent the Cutter
Job Definition
In this example, the contact tool surIace has been deIined as the cutter surIace, so it is necessary to prevent this tool
surIace Irom contact detection. ThereIore, as contact table: ctable1 has been created. As shown in Figure 2.5-27, it can
be seen that the contact surIace: cbody2, is deactivated Ior contact detection. In addition, the local adaptive remeshing
is also adopted with maximum oI one level splitting Ior each element.
Figure 2.5-27 Contact Table to prevent the Cutter Surface from Real Contact
424 Marc Users Guide: Part I
CHAPTER 2.5
Visualization of Results
Figure 2.5-28 shows the initial stress at increment zero. The stress values displayed in the FE model are those
transIerred Irom the text data Iile: initialstressn. This puts the cutter position at an intermediate stage oI the cutting
process (Figure 2.5-29).
Figure 2.5-28 Initial Stress at Increment Zero
Figure 2.5-29 Cutter Position in the Machining Process
425
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Import with PRE STATE Feature
PRE STATE is a Ieature that enables users to transIer result data Irom a previous Marc analysis to a new Marc analysis
as an initial state. This Ieature also allows users to expand a 2-D model oI a plane strain or axisymmetric application
to a 3-D model, transIerring the history data automatically Irom 2-D to 3-D as the initial conditions. In this sense,
AXITO3D (Irom axisymmetric 2-D to 3-D) Ieature supported in 2003 release is now a special case oI PRE STATE. This
Ieature also allows users to select contact bodies that are needed in the new analysis.
The Iollowing example shows its application with NC machining simulation. To use this Ieature, users will simulate
the process leading to the NC machining. AIter the pre-state values are obtained, the deIormed workpiece needs to be
merged into the new model in order to start the new analysis. The examples described here shows how to use this
Ieature. The two procedures are described, respectively, Ior the pre-analysis and Iinal machining analysis.
It is necessary to notice that the pre-state Ieature is generally available in Marc, not limited to machining analysis. For
more inIormation about this Ieature, reIer to Marc Volume A: Theory and User Information and Marc Volume C:
Program Input.
Input Data
The initial model is a block (Figure 2.5-24), which is the same one as used in the example below (Figure 2.5-30). The
only diIIerence is that there is no pre-deIined initial stress. The boundary conditions are applied in order to pre-deIorm
the body beIore the machining process started. The boundary condition set compress is to compress the workpiece
Irom the right side. The boundary condition set tension is to pull some nodes on the Iront end oI the workpiece. The
boundary condition set appy1 is to Iix the all nodes on the leIt side.
Figure 2.5-30 Workpiece and Boundary Conditions
426 Marc Users Guide: Part I
CHAPTER 2.5
The Iirst step is to run the analysis using 3dcut1.mfd. This provides the initial conditions and the deIormed mesh Ior
the NC machining simulation. When the analysis is completed, read in the result Iile. Assuming the NC machining
simulation starts at the end oI the previous analysis, users need to scan to the last increment. Apply REZONE MESH
to extract the mesh conIiguration at the end oI the analysis. This mesh will be used in the NC machining simulation.
The Iollowing steps are used to extract the deIormed mesh in Mentat:
Step 1: read in 3dcut1_job1.t16
Step 2: scan to the last increment
Step 3: apply REZONE MESH
Step 4: save the mesh in a model Iile: test_1.mfd.
In the new model Iile, 3dcut2.mfd, users can replace the mesh using MERGE with test_1.mfd.
Pre-deformed FE Mesh
There are two methods to deIine the mesh Ior the FEA model aIter the pre-deIormation:
1. To import the post data into Mentat and take the last deIormed workpiece as the initial geometry oI the new FE
model. The advantage oI this method is that the user can take into account the pre-deIormed geometry oI the
workpiece during the model design stage.
2. To use the original (undeIormed) body and additionally select total displacement as one oI the pre-state
variables that needs import into the new model during the FE analysis by Marc. This method is simple and easy
to use, providing there is no global remeshing or any mesh changes in the previous analysis.
For the demonstration purpose, the Iirst method is used in this example. The whole procedure run oI this example has
been recorded in procedure Iile: example3b.proc. The user can Iollow that Iile to reproduce the input deck and
visualize the results. To run this procedure Iile within Mentat, Iollow the commands as shown below:
Within procedure Iile: example3b.proc, the deIormed mesh is automatically merged into the new model Ior machining
analysis. The command procedure is recorded as below:
UTILS
PROCEDURES
LOAD
example3b.proc
OK
START/CONT
MAIN
RESULTS
OPEN
3dcut1_job1.t16
OK
427
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
The new model is created at the end oI the above procedure. The user can either generate a new model Irom test_1.mfd
or merge it into the model Iile 3dcut2.mfd by replacing the FE mesh in 3dcut2.mfd with test_1.mfd.
Import of the Pre-State Results
Using the Iollowing procedure below completes the deIinition oI the pre-state results (Figure 2.5-31):
DEF ONLY
LAST
TOOLS
REZONE MESH
MAIN
FILES
SAVE AS
test.1.mfd
OK
MAIN
MAIN
INITIAL CONDITIONS
NEW
MECHANICAL
PREVIOUS ANALYSIS STATE
STRESS
STRAIN
PLASTIC STRAIN
TOTAL EQUIVALENT PLASTIC STRAIN
POST FILE
3dcut1_job1.t16
OK
INCREMENT
LAST
SELECT BODY
cbody1 (defaults)
OK
OK
428 Marc Users Guide: Part I
CHAPTER 2.5
Figure 2.5-31 Definition of the Pre-State Option for Initial Conditions
Job and Loadcase Definition
Now, we can see that the initial condition is automatically activated when the job is created, as shown in Figure 2.5-32:
Figure 2.5-32 Job Definition with PRE-STATE Initial Conditions
429
CHAPTER 2.5
FEM Simulation of NC Machining and PRE STATE
Visualization of Results
In conIirming that the initial stress set by the Pre-State Ieature had successIully transIerred the initial data into the new
model, you can compare the stress data at the last increment in the previous analysis oI the stress data at increment
zero in the current analysis (Figure 2.5-33). The Iinal stress and displacement aIter machining are shown in
Figure 2.5-34.
Figure 2.5-33 Stress Data before (a) and after (b) Transferred by Pre-State Feature
Figure 2.5-34 The Equivalent Stress (a) and Displacement (b) after Machining
(a) (b)
(a) (b)
430 Marc Users Guide: Part I
CHAPTER 2.5
Input Files
To run the examples, to type in the commands or execute the Iollowing procedure Iiles:
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Example1: MENTATDIR/mentat -pr mcnIg.proc
Example2: MENTATDIR/mentat -pr example2.proc
Example3: MENTATDIR/mentat -pr example3a.proc
MENTATDIR/mentat -pr example3b.proc
File Description
3dcut1.mfd deIormed mesh Ior the NC machining simulation
3dcut2.apt cutter path data
3dcut2.mfd new model Iile
a2B.apt cutter path data
b3A.apt cutter path data
ex_prestate.mfd model Iile to deIine pre-stress
ex_r01.mud predeIined model Iile
example2.proc procedure Iile to run example 2
example3a.mfd predeIined model Iile called by example3a.proc
example3a.proc procedure Iile to run example 3a
example3b.proc procedure Iile to run example 3b
extru-test.apt cutter path data
initialstressn text Iile oI initial stresses
m2q0090s1.ccl cutter path data Iile
m2q0090s2.ccl cutter path data Iile
machining.proc post processes results Irom machiningmp1cmr1.mud
machining_mp1_cm_r1.mud model Iile used in example 2
machining_rcd procedure Iile that applies loads and BCs
mc_nfg.proc procedure Iile to run example 1
p1.apt cutter path data
prestate.proc procedure Iile to load prestress & visualize
prestate_table.proc procedure Iile to postprocess 3dcut1.mId
Chapter 2.6: Parallelized Local Adaptive Meshing
2.6
Parallelized Local Adaptive
Meshing

Chapter Overview 432

Simulation 432

Input Files 435

Animation 435
432 Marc Users Guide: Part I
CHAPTER 2.6
Chapter Overview
The sample session described in this chapter demonstrates the process oI bending a tube around a mandrel. The
simulation will use local adaptive meshing with parallel processing using a single input Iile. Local adaptive meshing
will add elements and thus improve the accuracy oI the simulation. In prior versions, parallel processing was not
available Ior local adaptive meshing. Furthermore, parallel processing also needed each domain to be written to a
separate input Iile. These limitations have been removed as demonstrated herein.
The goal oI the analysis is to demonstrate:
Local adaptive meshing with parallel processing
The use oI a single input Iile
Simulation
A metal tube will be bent ninety degrees around a mandrel as shown in Figure 2.6-1. The local adaptivity criterion is
based upon relative equivalent plastic strain with a threshold value oI 0.75 and two levels oI subdivision. Figure 2.6-2
shows a close up oI the total equivalent plastic strain contours in the Iinal position.
Figure 2.6-1 Metal Tube Bent Around Mandrel
433
CHAPTER 2.6
Parallelized Local Adaptive Meshing
Figure 2.6-2 Metal Tube in Final Position (Plastic Strain Contours)
The simulation uses symmetry and only halI oI the tube is modeled. Poisson`s ratio is 0.3; Young`s modulus is 200,000
Mpa with initial yield oI 200 Mpa with work hardening. To run the simulation in parallel using the single input Iile
mode, simply submit the single input Iile, say tubebend_job1.dat with the Iollowing procedure:
Procedure: run single input file in parallel using 4 processors
../marc2010/tools/run_marc -j tubebend_job1 -v n -b n -nps 4
where the option -nps 4 indicates the number oI processors, 4, in the single input mode. Running the job in parallel
will produce n+1 post Iiles that can be read into Mentat individually or consolidated by choosing the root post Iile.
Figure 2.6-3 identiIies the 4 domains that were automatically chosen. In this case, the number oI elements in each
domain vary because oI the adaptive meshing during the analysis. Originally each domain had only 576 elements,
however, at the end oI the analysis there are 576, 1563, 3383 and 688 elements in domains 1, 2, 3, and 4, respectively
as depicted in Table 2.6-1. For this implementation, the domains are Iixed and elements are not re-balanced among
the domains.
Figure 2.6-3 Automatically Generated Domains
434 Marc Users Guide: Part I
CHAPTER 2.6
Table 2.6-1 compares Iour diIIerent runs oI the same simulation. The baseline is a case with no adaptive meshing using
one CPU that takes 3287 seconds to run 100 increments with 2304 elements. The maximum total equivalent plastic
strain at the last increment is 0.4846.
Adding parallel processing with Iour processors shows a speedup Iactor oI 1.8 with no change in the maximum total
equivalent plastic strain at the last increment. Adding more elements using local adaptive meshing increases the
maximum total equivalent plastic strain at the last increment by about 25. Clearly the more elements used in the
simulation will capture the solution better. Furthermore, since local adaptive meshing is now available with parallel
processing, this more accurate solution can be obtained quicker. In this example the speedup Ior local adaptive
meshing was 1.6 but the total number oI elements generated diIIered running parallel. This is because neighboring
elements in diIIerent domains that require subdivision with local adaptive meshing are not allowed in the parallel
version at this time. ThereIore the parallel version tends to add Iewer elements than the single model simulation. In
this release, new domains are not created aIter local adaptive reIinement occurs.
Table 2.6-1 Comparison of Results for Parallel and Adaptive Meshing
Compare Parallel and
Local Adaptive Meshing
Local Adaptive Meshing No Adaptive Meshing
4 CPU 1 CPU 4 CPU 1 CPU
Single File Single File
Mwords 45.6 44.5 11.9 13.2
Seconds 3910 6238 1796 3287
Increments 100 100 100 100
# Recycles 1173 1166 1374 1374
Max Plastic Strain 0.6198 0.6499 0.4845 0.4846
# Elements 6210 7616 2304 2304
Domain 1 576 NA 576 NA
Domain 2 1563 NA 576 NA
Domain 3 3383 NA 576 NA
Domain 4 688 NA 576 NA
Speedup 1.6 NA 1.8 NA
Change Max Plastic Strain 21.8 25.4 0.0 Baseline
435
CHAPTER 2.6
Parallelized Local Adaptive Meshing
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Animation
Click on the Iigure below to play the animation (ESC to stop).
File Description
tubebend.proc Mentat procedure Iile to run the above example
tubebend.mud Associated model Iile
436 Marc Users Guide: Part I
CHAPTER 2.6
Chapter 2.7: New Magnetostatic Elements
2.7
New Magnetostatic Elements

Chapter Overview 438

Magnetostatic Field Around a Coil 438

Input Files 446

438 Marc Users Guide: Part I
CHAPTER 2.7
Chapter Overview
This chapter describes the use oI three new magnetostatic elements in Marc. These elements are 4-node and 10-node
tetrahedral elements, and a 2-node line element. With these new tetrahedral elements, it is possible to use automatic
meshers which will Iacilitate meshing oI complex structures. The purpose oI the line element is to deIine an external
loading; the current in a wire. This element does not have material or geometric properties. The line element can be
either placed on element edges oI the solid elements or embedded in these solid elements. The direction oI the current
is in the direction oI the line elements, Iollowing the connectivity.
Magnetostatic Field Around a Coil
This example demonstrates the use oI the 10-node magnetostatic tetrahedral element with the use oI the magnetostatic
line element in Marc. The Iunction oI the latter is to simpliIy deIining a current as an external loading. A one wire coil
in air is analyzed. The results will be compared with an analytical solution using the Biot-Savart law. A schematic view
oI the model is shown in Figure 2.7-1.
Figure 2.7-1 Schematic View of the Coil with Surrounding Air
Mesh Generation
The mesh is generated previously, and can be seen in Figure 2.7-2. The mesh is reIined around the location where the
coil will be to better capture the gradient oI the magnetic Iield near the coil. Due to symmetry, only a quarter oI a
cylinder is modeled. A curve with a radius oI 0.3 m is added in the center oI the densely meshed area. This curve is
then converted to line elements. The number oI line elements should match the density oI the solid elements, so that
the size oI the line elements is at least the same as the average edge length oI the solid elements.
FILE
NEW
RESET PROGRAM
OPEN
mesh_mag.mfd
OK
3.0 m
radius 0.3 m
radius 2.0 m
air
wire
face A
face B
439
CHAPTER 2.7
New Magnetostatic Elements
RETURN
MESH GENERATION
CURVE TYPE
CENTER/RADIUS/ANGLE/ANGLE
RETURN
CRVS ADD
0 0 0
0.3
0
90
MOVE
ROTATION ANGLES (DEGREES)
0 -90 0
CURVES
1 #
RETURN
CONVERT
DIVISIONS
24 1
CURVES TO ELEMENTS
1 #
SELECT
SELECT BY
CLASS
line(2)
OK
RETURN
RETURN
RETURN
440 Marc Users Guide: Part I
CHAPTER 2.7
Figure 2.7-2 Finite Element Mesh
Material Properties
The permeability oI the air surrounding the coil is 1.2566 x 10
-6
H/m. The line elements which Iorm the coil do not
need material properties since they are only used to deIine the loading.
Inserts
To transIer the current Irom the line elements to the solid elements, the INSERT option is used. The line elements are
inserted in the solid elements where the solid elements are the host elements, and the line elements the embedded
MATERIAL PROPERTIES
NAME
air
MANETOSTATIC
PERMEABILITY
1.2566e-6
OK
ELEMENTS ADD
ALL UNSELECT
RETURN
A
B
441
CHAPTER 2.7
New Magnetostatic Elements
entities. Marc automatically ties the degrees oI Ireedom oI the nodes to be inserted to the corresponding degrees oI
Ireedom oI the nodes oI the host elements.
Boundary Conditions
Symmetry conditions are applied on the two rectangular Iaces oI the quarter section (Iace A and B in Figure 2.7-1) in
such a way, that the potential is Iorced to be perpendicular to the surIace oI the rectangular Iaces. A current oI -0.5 A
is prescribed to the line elements.
LINKS
INSERTS
HOST ELEMENTS ADD
ALL UNSELECT
EMBEDDED ENTITIES ADD
ALL SELECT
RETURN
RETURN
BOUNDARY CONDITIONS
MAGNETOSTATIC
NAME
fix_xy
FIXED POTENTIAL (3-D)
POTENTIAL X
POTENTIAL Y
OK
SELECT
CLEAR SELECT
METHOD
BOX
RETURN
NODES
-10 10
-10 10
-10 0.001
RETURN
NODES ADD
ALL SELECT
442 Marc Users Guide: Part I
CHAPTER 2.7
NEW
NAME
fix_xz
FIXED POTENTIAL (3-D)
POTENTIAL X
POTENTIAL Z
OK
SELECT
CLEAR SELECT
NODES
-10 10
-10 0.001
-10 10
RETURN
NODES ADD
ALL SELECT
NEW
NAME
load
WIRE CURRENT (define current in wire)
CURRENT
-0.5
OK
SELECT
CLEAR SELECT
SELECT SET
insert_embed_elements
OK
RETURN
ELEMENTS ADD
ALL SELECT
RETURN
RETURN
443
CHAPTER 2.7
New Magnetostatic Elements
Loadcases and Job Parameters
A steady state analysis is perIormed. Figure 2.7-3 shows the element type menu, where in 3-D SOLID element type 182
is selected Ior the 10 node tetrahedral elements, and in 3-D WIRE element type 183 is selected Ior the line elements.
Figure 2.7-3 Element Types Menu for Magnetostatics
LOADCASES
MAGNETOSTATIC
STEADY STATE
OK
RETURN (twice)
JOBS
ELEMENT TYPES
MAGNETOSTATIC
3-D WIRE
183
OK
ALL SELECT
3-D SOLID
182
OK
ALL UNSELECT
RETURN (twice)
MORE
MAGNETOSTATIC
lcase1
INITIAL LOADS
fix_xy
fix_xz
444 Marc Users Guide: Part I
CHAPTER 2.7
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened.
load
OK
JOB RESULTS
1st Comp of Magnetic Induction
2nd Comp of Magnetic Induction
3rd Comp of Magnetic Induction
1st Comp of Magnetic Field Intensity
2nd Comp of Magnetic Field Intensity
3rd Comp of Magnetic Field Intensity
OK (twice)
FILE
SAVE AS
coil.mud
OK
RETURN
RUN
SUBMIT(1)
MAIN
RESULTS
OPEN DEFAULT
NEXT
PATH PLOT
NODE PATH
9266 9268 #
VARIABLES
ADD CURVE
Arc Length
1st Component of Magnetic Induction
FIT
445
CHAPTER 2.7
New Magnetostatic Elements
Figure 2.7-4 shows the contour plot oI the 1st component oI the magnetic induction. A subsection oI the elements just
below the coil including the line elements oI the coil is plotted here. The magnetic induction in the plane oI the coil
should be perpendicular to this plane, and changing sign going Irom the inside to the outside oI the coil.
Figure 2.7-4 Contour Plot of the 1st Component of the Magnetic Induction
An analytical solution Ior the magnetic Iield oI this example can be obtained using the Biot-Savart law. The magnetic
induction along the line going through the center axis oI the coil is given by,
with,
The axis oI the coil is shown in Figure 2.7-2, indicated by the arrows. Figure 2.7-5 shows a path plot oI the magnetic
induction along the path going Irom A to B (see Figure 2.7-2). The analytical solution is also shown in this Iigure. The
result corresponds very well with the analytical solution.
magnetic induction along the axis oI the coil
magnetic permeability
radius oI the coil
position on the axis through the coil
current
B
axis
1
2
---
r
2
I
r
2
l
2
( )
3 2
-----------------------------
B
axis

r
l
I
446 Marc Users Guide: Part I
CHAPTER 2.7
Figure 2.7-5 Magnetic Induction along the Axis of the Coil Compared with the Analytical Solution
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
coil.proc Mentat procedure Iile to run the above example
mesh_mag.mfd Associated model Iile
0.00E00
2.00E-07
4.00E-07
6.00E-07
8.00E-07
1.00E-06
1.20E-06
-1.50 -1.00 -0.50 0.00 0.50 1.00 1.50
Path through center oI coil
M
a
g
n
e
t
i
c

i
n
d
u
c
t
i
o
n
coil
analytical
Chapter 2.8: Coupled Electrostatic Structural Analysis of a Capacitor
2.8
Coupled Electrostatic
Structural Analysis of
a Capacitor

Chapter Overview 448

Capacitor Loaded with Charge 448

Input Files 463

448 Marc Users Guide: Part I
CHAPTER 2.8
Chapter Overview
This chapter describes the use oI coupled electrostatic structural analysis in Marc. In this analysis type, the Coulomb
Iorce, the Iorce between charged bodies, links the electrostatic part to the structural part, and the deIormation
inIluences the electrostatic Iield. This is a weak coupling where, in the Iirst pass, the electrostatic Iield is computed
and the corresponding Coulomb Iorces are calculated. In the next pass, the structural response is evaluated, such that
the Coulomb Iorces are treated as additional external Iorces. In a subsequent increment, the deIormed state is used in
the calculation oI the electrostatic Iield. Since the electrostatic solution is a steady state solution, a time dependent
problem will be solved as quasi-static during the electrostatic phase oI the solution.
Capacitor Loaded with Charge
Two parallel plates Iorm a capacitor, that contains a charge (see Figure 2.8-1). One plate is Iixed and electrically
grounded, while the other plate is attached to a spring. Boundary conditions are chosen so that the plate connected to
the spring can only move perpendicular to the Iixed plate. Then, when this plate is loaded with charge, it will move
towards the Iixed plate. The charge is chosen as the applied load instead oI potential since with increasing charge at a
certain moment the potential will decrease as shown later in the results.
Since the two plates are circular, an axisymmetric analysis will be perIormed. Air, both between the two plates and
outside the plates, is taken into account to get a good representation oI the electrostatic Iield. The air is only active in
the electrostatic pass. The position oI the nodes oI the air region 'contacting the plates is updated based on the
displacements oI the plates. In order to avoid getting badly shaped elements due to the motion oI the Iree plate, a region
oI air surrounding the plates is periodically remeshed.
Figure 2.8-1 Schematic Representation of the Capacitor
Mesh Generation
A previously deIined mesh is read in as an mfd Iile. A curve is added along the x-axis, which is needed as a boundary
Ior remeshing. Then the mesh is split around the plates, so that only the inner part oI the air surrounding the plates can
be remeshed. This way the mesh on the outside stays coarse, while the mesh directly surrounding the plates will be
suIIiciently reIined to capture the gradient in the electric Iield. Splitting the mesh is done using MATCHING
BOUNDARIES in MODELING TOOLS. This tool splits a mesh and creates matching boundaries. The latter inIormation
is not needed in this analysis. The radius oI the two plates is 10 mm, the thickness 1 mm, the initial gap is 0.4 mm, and
the radius oI the air modeled around the plates is 50 mm.
Initial gap
fixed plate
movable plate
loaded with charge
V
k
449
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Material Properties
The plates are made oI copper with a Young`s modulus oI 124 GPa and a Poisson`s ratio oI 0.3. The permittivity is
0.001 F/m. The air surrounding the plates will only be active in the electrostatic pass, so no mechanical properties are
needed. The permittivity is 8.854 pF/m.
FILE
NEW
RESET PROGRAM
OPEN
capmesh.mfd
OK
RETURN
MESH GENERATION
CRVS ADD
-0.05 0 0
0.05 0 0
RESET VIEW
RETURN
MODELING TOOLS
MATCHING BOUNDARIES
NEW
2-D (SOLID, 3-D SHELL)
SELECT
METHOD
BOX
RETURN
ELEMENTS
-0.00407 0.00407
-1 0.0136
-1 1
RETURN
SPLIT MESH
ALL SELECT
SELECT
CLEAR SELECT
RETURN (three times)
450 Marc Users Guide: Part I
CHAPTER 2.8
MATERIAL PROPERTIES
NAME (define structural and electrical
conductor properties of conductor)
ISOTROPIC
YOUNGS MODULUS
124e9
POISSONS RATIO
0.3
O
K
ELECTROSTATIC
PERMITTIVITY
0.001
OK
SELECT
METHOD
FLOOD
RETURN
ELEMENTS
3 103 #
RETURN
ELEMENTS ADD
ALL SELECT
NEW (define electrical properties of air)
NAME
air
ELECTROSTATIC
PERMITTIVITY
8.854e-12
OK
SELECT
CLEAR SELECT
ELEMENTS
731 734 #
RETURN
ELEMENTS ADD
451
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Contact
The Coulomb Iorce is calculated at the interIace oI contact bodies. It is important that contact bodies are connected in
the right direction. In general, an insulator is touching a conductor; so Ior this example, air should be touching the
plates. When a body is only active in the electrostatic pass (the air), it must be a so-called ZERO STIFFNESS body
(see Figure 2.8-2). Then, in the CONTACT TABLE section, such a body cannot be touched by a DEFORMABLE body,
thus Iacilitating the required connection
.
Figure 2.8-2 Contact Bodies Menu
ALL SELECT
SELECT
CLEAR SELECT
RETURN (twice)
CONTACT
CONTACT BODIES
NAME
plate_a
DEFORMABLE
OK
SELECT
ELEMENTS
103 #
RETURN
ELEMENTS ADD
ALL SELECT
NEW
452 Marc Users Guide: Part I
CHAPTER 2.8
NAME
plate_b
DEFORMABLE
OK
SELECT
CLEAR SELECT
ELEMENTS
3 #
RETURN
ELEMENTS ADD
ALL SELECT
NEW
NAME
air_inner
ZERO STIFFNESS
OK
SELECT
CLEAR SELECT
ELEMENTS
731 #
RETURN
ELEMENTS ADD
ALL SELECT
NEW
NAME
air_outer
ZERO STIFFNESS
OK
SELECT
CLEAR SELECT
ELEMENTS
734 #
RETURN
ELEMENTS ADD
ALL SELECT
NEW
453
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Boundary Conditions
A spring with a spring constant oI 50 N/m is attached to the moving plate to balance the Coulomb Iorce. The moving
plate (leIt, or top plate in Figure 2.8-1), is loaded with an in-time linear increasing charge, which reachs 20C aIter one
second. This should result in a continuous increase oI the Coulomb Iorce, so that the plates move toward each other.
II a linear increasing potential was applied to the moving plate, the system would, at a certain point, become unstable,
and the plates would collapse. The right plate is Iixed, and the potential is set to 0 volts.
NAME
symmetry
SYMMETRY
OK
CURVES ADD
1 #
RETURN
CONTACT TABLES
NEW
PROPERTIES
1 2
CONTACT TYPE: GLUE
OK
1 3
CONTACT TYPE: GLUE
OK
1 4
CONTACT TYPE: GLUE
OK
1 5
CONTACT TYPE: GLUE
OK
OK
RETURN (twice)
LINKS
SPRINGS/DASHPOTS
NEW
454 Marc Users Guide: Part I
CHAPTER 2.8
PROPERTIES
STIFFNESS SET
50
OK
BEGIN NODE
1
DOF
1
END NODE
103
DOF
1
RETURN (twice)
BOUNDARY CONDITIONS
NAME
fix
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
OK
NODES ADD
1 4 #
RETURN
NEW
NAME
pot_0
ELECTROSTATIC
FIXED POTENTIAL
POTENTIAL (TOP)
0
NODES ADD
4
OK
RETURN
NEW
455
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
NAME
fix_y
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT Y
0
OK
NODES ADD
1 4 104 #
RETURN
NEW
NAME
load
ELECTROSTATIC
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 0
1 1
RETURN
POINT CHARGE
CHARGE(TOP)
2e-8
TABLE
table1
NODES ADD
104
OK
RETURN (twice)
456 Marc Users Guide: Part I
CHAPTER 2.8
Mesh Adaptivity
The mesh surrounding the two plates is remeshed every Iive increments to accommodate the deIormation oI the air
when one plate moves towards the other plate. A triangular mesh is created using the Delaunay method, where the
target element edge length is 0.4 mm.
Loadcases and Job Parameters
A quasi-static analysis is perIormed. MULTI-CRITERIA load stepping method (AUTO STEP in Marc Volume C:
Program Input) is used to control the time step. The time step control is based on a maximum displacement per
increment oI 3 m in the x-direction. The axisymmetric mechanical element 10 is selected Ior the elements oI the
plates, and the axisymmetric electrostatic element 38 is selected Ior the elements oI the air. In perIorming coupled
electrostatic-structural analysis, two procedures are available Ior calculation electrical Iorces. The Iirst method
(deIault) is based upon the nodal charges and is applicable iI the bodies are close to one another. The second method
is based upon the electrical Iield and is more accurate when the bodies are Iurther apart. The deIault procedure is used
here, but in the menu used to select the procedure is shown in Figure 2.8-3.
MESH ADAPTIVITY
GLOBAL REMESHING CRITERIA
DELAUNAY TRIA
INCREMENT
FREQUENCY
5
ELEMENT EDGE LENGTH SET
0.0004
OK
REMESH BODY
air_inner
RETURN (twice)
457
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Figure 2.8-3 Electrostatic-Structural Analysis Options Menu
LOADCASES
ELECTROSTATIC-STRUCTURAL
NAME
capacitor
QUASI-STATIC
CONTACT
CONTACT TABLE
ctable1
OK
GLOBAL REMESHING
adapg1
OK
MULTI-CRITERIA
PARAMETERS
USER DEFINED
DISPLACEMENT INCREMENT
458 Marc Users Guide: Part I
CHAPTER 2.8
PARAMETERS
DISPLACEMENT INC ALLOWED
1
3e-6
OK
OK
OK
OK
RETURN
RETURN
JOBS
NAME
capacitor
ELEMENT TYPES
SELECT
CLEAR SELECT
SELECT CONTACT BODY ENTITIES
plate_a
plate_b
OK
RETURN
ELECTROSTATIC-STRUCTURAL
ELECTROSTATIC-STRUCTURAL ELEMENT TYPES:
AXISYM SOLID
10
OK
ALL SELECT
ELECTROSTATIC ELEMENT TYPES:
AXISYM SOLID
38
OK
ALL UNSELECT
RETURN (twice)
ELECTROSTATIC-STRUCTURAL
capacitor
459
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
INITIAL LOADS
fix
pot_0
fix_y
load
OK
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK
OK
ANALYSIS OPTIONS
LARGE DISPLACEMENT
OK
JOB RESULTS
1st Comp of Electric Field Intensity
2nd Comp of Electric Field Intensity
1st Comp of Electric Displacement
2nd Comp of Electric Displacement
SELECTED NODAL QUANTITIES:
CUSTOM
Electric Potential
External Charge
Reaction Charge
Displacement
Reaction Force
Coulomb Force
Contact Status
OK (twice)
460 Marc Users Guide: Part I
CHAPTER 2.8
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened.
Figure 2.8-4 shows the contour plot oI the x-component oI the electric Iield intensity. You can observe that this Iield
is constant between the two plates, except at the top (outer radius) oI the two plates.
FILE
SAVE AS
capacitor.mud
OK
RETURN
RUN
SUBMIT(1)
MAIN
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODES
104
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Electric Potential
Displacement X
FIT
RETURN
SHOW IDS
O
YMAX
0.0004
461
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Figure 2.8-4 Contour Plot of the X-Component of the Electric Field Intensity
The electrical potential in the vicinity oI the capacitor is shown in Figure 2.8-5.
Figure 2.8-5 Electrical Potential
The Iollowing equation Ior the potential as a Iunction oI the gap opening, can be derived Ior the ideal situation where
the electric Iield intensity is constant between the plates:
V
2k
c
0
A
--------- g
2
g
0
g ( )
462 Marc Users Guide: Part I
CHAPTER 2.8
With the potential, , the spring constant, , the permittivity, , the area oI the plate, , the gap opening, , and the
initial gap opening, . Note that is the gap closing displacement computed by Marc. In Figure 2.8-6, the result
is compared with the analytical solution where the gap closing displacement is plotted as a Iunction oI the electrostatic
potential.
Figure 2.8-6 Potential as a Function of Gap Opening
Also, note that a maximum oI the potential is reached when , or when the current gap opening is oI the
initial gap. II the loading was prescribed with an increasing potential, the plates would become unstable at this point
and collapse. Figure 2.8-7 gives a close up look oI the remeshed area oI the air at the top oI the plates during diIIerent
stages oI the analysis.
V k c
0
A g
g
0
g
0
g
0 100 200 300 400 500 600
0.00000
0.00005
0.00010
0.00015
0.00020
0.00025
0.00030
0.00035
0.00040
Analytical
Marc
Electrostatic Potential [V]
A
i
r

G
a
p

C
l
o
s
i
n
g


D
i
s
p
l
a
c
e
m
e
n
t

(
g
o
-
g
)

[
m
]
g
0
g
1
3
--- g
0

2
3
---
463
CHAPTER 2.8
Coupled Electrostatic Structural Analysis of a Capacitor
Figure 2.8-7 Result of Remeshing at Different Steps (increment 20, 50, 100, and 141) in the Analysis
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Note: The Iigures are zoomed in at the top oI the plates.
File Description
capacitor.proc Mentat procedure Iile to run the above example
cap_mesh.mfd Associated model Iile
464 Marc Users Guide: Part I
CHAPTER 2.8
Chapter 2.9: 3D Contact and Friction Analysis using Quadratic Elements
2.9
3-D Contact and Friction
Analysis using
Quadratic Elements

Chapter Overview 466

Sliding Mechanism 467

Input Files 483

466 Marc Users Guide: Part I
CHAPTER 2.9
Chapter Overview
Various new options have been added to Iurther enhance the capabilities to analyze contact problems. In this chapter,
attention will be mainly paid to the Iollowing items:
new Iriction model;
new style table input to deIine a velocity dependent Iriction coeIIicient;
automatic optimization oI contact constraint equations;
postprocessing contact stresses.
The new Iriction model is intended to solve the limitations oI the two basic Iriction models available in previous
versions oI Marc. The model uses a bilinear approximation oI the theoretical stick-slip step Iunction as shown in
Figure 2.9-1. The slip threshold parameter is, by deIault, determined by the program based on the average element
edge length oI the elements deIining the deIormable contact bodies. A second parameter oI the model is the Iriction
Iorce convergence ratio, which is used to compare the length oI the Iriction Iorce vector oI the current iteration with
the previous iteration. The deIault value oI this parameter is 0.05. Both deIault values oI the parameters have been
designed to produce accurate results in a wide range oI applications. II needed, they can be modiIied by the user.
Figure 2.9-1 Bilinear Friction Model
In order to get more Ilexibility in deIining e.g., boundary conditions, material properties, etc. as a Iunction oI various
independent variables, like position, time, equivalent plastic strain, etc., a new style table input has been introduced
since Marc 2007. In this chapter, use is made oI tables to deIine boundary conditions as a Iunction oI time, but also a
Iriction coeIIicient as a Iunction oI the relative sliding velocity between contact bodies. Notice that the latter would
not have been possible in earlier versions oI Marc without user subroutines.
In Marc, deIormable contact problems are traditionally solved using multipoint constraint equations. For certain
problems, the accuracy oI the solution strongly depends on the order in which the constraint equations have been
deIined. New logic has been added to automatically optimize the constraint equations. The procedure is based on Iirst
deIining all possible constraint equations using true double-sided contact and then, taking into account the average
stiIIness oI the contact bodies involved and the size oI the element segments in the areas oI contact, reducing this to a
set oI optimal constraint equations.
In MSC.Marc 2003, true quadratic contact was introduced. In the area oI postprocessing, the contact stresses were only
available on contacting nodes and not on nodes oI contacted bodies. This limitation has been removed since Marc 2007
: friction force
: friction coefficient
: normal normal
: relative tangential displacement
: slip threshold
f
t

f
n
Au
t
o
I
t
o
I
n
Au
t
467
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
by introducing a procedure which yields a local coordinate system in nodes oI contacted bodies, based on which the
nodal stress tensor can be transIormed to get the contact normal and/or Iriction stress.
The units used herein are Force |N|, Length |mm|, and Time |sec|.
Sliding Mechanism
A sliding mechanism, as shown in Figure 2.9-2, is analyzed. A square block with Ilattened edges can slide in a
U-shaped section which, at its ends, is mounted on two support blocks. The square block has a circular hole in which
a rigid cylinder is inserted. The block is loaded via the cylinder by a vertical Iorce in the global y-direction
and prescribed displacements in global x- and z-direction oI and , in which
denotes the time. The material behavior oI the square block is described using a Neo-Hookean material model deIined
through the MOONEY property menu with , while the material behavior oI the U-shaped section and the
supports is isotropic and linear. Young`s modulus and Poisson`s ratio oI the section are and , and
oI the supports and . Frictional contact between the block and the section is assumed based
on Coulomb`s Iriction law with a velocity dependent Iriction coeIIicient , in which is the relative
sliding velocity. For all components, except the cylinder, 10-node tetrahedral elements with Iull integration (Marc
element type 127) will be used.
Figure 2.9-2 Solid Model of the Sliding Mechanism
Model Generation
First, the Mentat database is cleaned, the view point is set and a colormap with a white background is selected. Then
the Iinite element model is set up by subsequently merging the various components oI the structure, which have been
F
y
300
u
x
25 2tt ( ) sin u
z
0.05 tt ( ) sin ( )
2
t
C
01
100
E 5.0
4
10 v 0.3
E 2.2
5
10 v 0.28
0.03 0.07e
0.01v
v
468 Marc Users Guide: Part I
CHAPTER 2.9
stored in individual Iiles, called support.mfd, section.mfd, block.mfd, and cylinder.mfd. The Iirst three Iiles contain a
solid model oI the component as well as a Iinite element mesh obtained by automatic mesh generation. The last Iile
contains a solid model and the surIaces obtained by conversion oI the solid Iaces into surIaces. AIter reading the
models, element and node sets are generated, which makes it easy later on to assign material properties, deIine contact
bodies, and assign boundary conditions. Finally, an extra node above the block is added, which will be used as the
control node Ior the rigid cylinder to apply the Iorce and prescribed displacements. The Iinite element model is shown
in Figure 2.9-3.
Figure 2.9-3 Finite Element Model
FILES
NEW
OK
RESET PROGRAM
RESET VIEW
VIEW
SHOW VIEW 1
RY+
RY+
RY+
RY+
RX+
MAIN
469
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
VISUALIZATION
COLORS
COLORMAP 2
MAIN
FILES
MERGE
support.mfd
OK
MERGE
section.mfd
OK
MERGE
block.mfd
OK
MERGE
cylinder.mfd
OK
FILL
PLOT
POINTS (off)
CURVES (off)
SOLIDS (off)
ELEMENTS SOLID
SURFACES SOLID
MAIN
MESH GENERATION
SELECT
METHOD
FLOOD
RETURN
ELEMENTS
8 20 (click a node of each of the support blocks)
ELEMENTS STORE
support
OK
ALL: SELECTED
470 Marc Users Guide: Part I
CHAPTER 2.9
ELEMENTS CLR
ELEMENTS
223 (click a node of the section)
ELEMENTS STORE
section
OK
ALL: SELECTED
ELEMENTS CLR
ELEMENTS
1458 (click a node of the block)
ELEMENTS STORE
block
OK
ALL: SELECTED
ELEMENTS CLR
METHOD
SINGLE
RETURN
NODES
5 8 11 12 13 14 64 65 66 74 84 85 86 87 89
17 20 23 24 25 26 94 95 96 104 114 115 116
117 119 (nodes at the bottom of the support)
END LIST (#)
NODES STORE
support_bottom
OK
ALL:SELECTED
NODES CLR
RETURN
NODES ADD
25 60 25
MAIN
471
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
Material Properties
The deIinition oI the material properties is straightIorward. One Mooney and two isotropic materials are deIined and
assigned to the corresponding element sets.
MATERIAL PROPERTIES
NEW
NAME
Support_material
ISOTROPIC
YOUNGS MODULUS
2.2e5
POISSONS RATIO
0.28
OK
ELEMENTS ADD
SET
support
OK
NEW
NAME
Section_material
ISOTROPIC
YOUNGS MODULUS
5e5
POISSONS RATIO
0.3
OK
ELEMENTS ADD
SET
section
OK
NEW
NAME
Block_material
MORE
MOONEY
472 Marc Users Guide: Part I
CHAPTER 2.9
Contact
Three deIormable contact bodies and one rigid contact body are deIined in the Iollowing order: Iirst the support, next
the U-shaped section, then the block, and Iinally the rigid cylinder. The bodies are called Support, Section, Block, and
Cylinder, respectively. Contact body Cylinder will be a load-controlled rigid body with the previously deIined Iree
node as the control node. A contact table is deIined to enter the diIIerent contact conditions between the bodies. Glued
contact is used between the bodies Block and Cylinder and the bodies Section and Support. Frictional contact is used
between the bodies Block and Section. The velocity dependent Iriction coeIIicient is deIined using a table oI type
velocity, as shown in Figure 2.9-4.
Figure 2.9-4 Table Defining Velocity Dependent Friction Coefficient
In order to illustrate the eIIect oI the new contact constraint optimization procedure, a user-deIined detection order Ior
one set oI contact bodies (Block and Section) is used together with the global optimization procedure Ior the other set
C10
100
OK
ELEMENTS ADD
SET
block
OK
MAIN
473
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
(Section and Support). In such cases, a nondeIault order deIined via a contact table takes precedence over the
global procedure.
CONTACT
CONTACT BODIES
NEW
NAME
Support
DEFORMABLE
OK
ELEMENTS ADD
SET
support
OK
NEW
NAME
Section
DEFORMABLE
OK
ELEMENTS ADD
SET
section
OK
NEW
NAME
Block
DEFORMABLE
OK
ELEMENTS ADD
SET
block
OK
NEW
NAME
Cylinder
RIGID
LOAD
474 Marc Users Guide: Part I
CHAPTER 2.9
OK
CONTROL NODE
5374
SURFACES ADD
54 55 56 57
END LIST (#)
RETURN
CONTACT TABLES
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
friction_coef
TYPE
velocity
OK
FORMULA
ENTER
0.03+0.07*exp(-0.01*v1)
MAX (INDEPENDENT VARIABLE V1)
300
STEPS (INDEPENDENT VARIABLE V1)
100
REEVALUATE
FIT
RETURN
NEW
PROPERTIES
1 2 (click entry 1-2)
CONTACT TYPE: GLUE
PROJECT STRESS-FREE (on)
2 3 (click entry 2-3)
CONTACT TYPE: TOUCHING
FRICTION COEFFICIENT
1
TABLE
475
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
Boundary Conditions
The Iollowing boundary conditions have to be deIined: Iixing the bottom oI the support blocks, prescribing the motion
oI the cylinder in global x- and z-direction, and applying the Iorce in global y-direction on the cylinder. Since contact
body Cylinder is a load-controlled rigid body, the prescribed motion and Iorce is assigned to its control node.
friction_coef
OK
3 4 (click entry 1-4)
CONTACT TYPE: GLUE
PROJECT STRESS-FREE (on)
OK (twice)
MAIN
BOUNDARY CONDITIONS
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
motion-x
TYPE
time
OK
FORMULA
ENTER
25*sin(2*pi*v1)
STEPS (INDEPENDENT VARIABLE V1)
100
REEVALUATE
FIT
NEW
1 INDEPENDENT VARIABLE
NAME
motion-z
TYPE
time
OK
476 Marc Users Guide: Part I
CHAPTER 2.9
FORMULA
ENTER
0.05*sin(pi*v1)^2
STEPS (INDEPENDENT VARIABLE V1)
100
REEVALUATE
FIT
RETURN
NEW
NAME
fix-support
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (0)
DISPLACEMENT Y (0)
DISPLACEMENT Z (0)
OK
NODES ADD
SET
support_bottom
OK
RETURN
NEW
NAME
motion
FIXED DISPLACEMENT
DISPLACEMENT X
1
TABLE
motion-x
OK
DISPLACEMENT Z
1
TABLE
motion-z
OK (twice)
477
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
Loadcases
A mechanical static loadcase is deIined, in which the previously deIined contact table and boundary conditions are
selected (note that the boundary conditions are automatically selected iI they have been deIined beIore deIining the
current loadcase). The total loadcase time is 1 (which is also the deIault loadcase time), so that the block gets one
complete cyclic motion in the global x-direction. A Iixed stepping procedure is chosen with 100 steps and the deIault
control settings Ior the Newton-Raphson iteration process are used.
NODES ADD
5374
END LIST (#)
RETURN
NEW
NAME
force-y
MECHANICAL
POINT LOAD
FORCE Y
-300
OK
NODES ADD
5374
END LIST (#)
RETURN
MAIN
LOADCASES
NEW
MECHANICAL
STATIC
CONTACT
CONTACT TABLE
ctable1
OK (twice)
# STEPS
100
OK
478 Marc Users Guide: Part I
CHAPTER 2.9
Jobs
A mechanical job is deIined in which the previously deIined loadcase is selected. The available contact table is also
used Ior initial contact. The Iriction type is switched to the bilinear Coulomb model with deIault parameters (see
Figure 2.9-5). The newly introduced procedure to optimize the contact constraint equations is activated, while the other
contact parameters are leIt deIault. The updated Lagrange procedure Ior rubber is selected, which allows the use oI
regular displacement-based elements instead oI Herrmann elements with additional pressure degrees oI Ireedom. As
post Iile variables, the Cauchy stress tensor is selected as an element tensor, while the displacements, external Iorces,
reaction Iorces, contact normal stress, contact normal Iorce, contact Iriction stress, contact Iriction Iorce, and contact
status are selected as nodal quantities. The element type Ior all Iinite elements is set to 127, the 10-node tetrahedral
element with Iull integration. BeIore submitting the job, the new style table input is activated. This causes the Marc
data Iile to be written in a new Iormat which, e.g., allows all tables to be used directly by Marc in equation Iormat.
Figure 2.9-5 Contact Control: Friction Model and Parameters
TITLE
Sliding Mechanism
OK
MAIN
JOBS
MECHANICAL
lcase1
CONTACT CONTROL
FRICTION TYPE: COULOMB BILINEAR (DISPLACEMENT)
479
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
(pull-down menu)
INITIAL CONTACT
ctable1
OK
ADVANCED CONTACT CONTROL
OPTIMIZE CONTACT CONSTRAINT EQUATIONS
OK(twice)
ANALYSIS OPTIONS
RUBBER ELASTICITY PROCEDURE:
LARGE STRAIN-UPDATED LAGRANGE (roller button)
OK
JOB RESULTS
Cauchy Stress (on)
CUSTOM
Displacement (on)
External Force (on)
Reaction Force (on)
Contact Normal Stress (on)
Contact Normal Force (on)
Contact Friction Stress (on)
Contact Friction Force (on)
Contact Status (on)
OK
OK
ELEMENT TYPES
MECHANICAL
3-D SOLID
127
OK
ALL: EXISTING
RETURN
RETURN
TITLE
Sliding Mechanism
OK
480 Marc Users Guide: Part I
CHAPTER 2.9
Results
In Figure 2.9-6, the initial contact status oI the nodes oI contact body Section is given. Clearly, the nodes oI body
Section are contacting body Support, which is a result oI the procedure to optimize the contact constraint equations.
Figure 2.9-6 Initial Contact Status of the Nodes of Contact Body Section
Figure 2.9-7 shows the contact normal stress on the deIormable bodies Ior increment one. Both the contacting nodes
and the nodes corresponding to contacted segments can be seen to have nonzero values. The distribution is not exactly
symmetric, since the block already has some displacement in the global x-direction.
RUN
NEW-STYLE TABLES (on)
SUBMIT 1
MONITOR
OK
MAIN
481
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
Figure 2.9-7 Contact Normal Stress for Increment 1
Finally, Figure 2.9-8 contains a history plot oI the x-component oI the total Iorce on the cylinder. The nonlinear
response is partly due to the prescribed motion oI the cylinder in the z-direction, but mostly due to the velocity
dependent Iriction coeIIicient, which causes more Iriction at lower sliding velocities. Notice that due to the motion in
the z-direction, the magnitude oI the x-component oI the total Iorce can be larger than the maximum Iriction coeIIicient
times the applied load in the y-direction.
Figure 2.9-8 Total X-force on Contact Body Cylinder as a Function of Time
482 Marc Users Guide: Part I
CHAPTER 2.9
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Contact Status
OK
CONTOUR BANDS
SELECT
CONTACT BODY ENTITIES
Section
OK
MAKE VISIBLE
RETURN
NEXT
SELECT
CONTACT BODY ENTITIES
Support
Section
Block
Cylinder
OK
MAKE VISIBLE
RETURN
SCALAR
Contact Normal Stress
OK
MONITOR
HISTORY PLOT
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD GLOBAL CRV
Time
Force X Cylinder
Fit
483
CHAPTER 2.9
3-D Contact and Friction Analysis using Quadratic Elements
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
friction.proc Mentat procedure Iile to run the above example
block.mfd Associated model Iile
cylinder.mfd Associated model Iile
section.mfd Associated model Iile
support.mfd Associated model Iile
484 Marc Users Guide: Part I
CHAPTER 2.9
Chapter 2.10: Pin to Seal Contact with Various Friction Models
2.10
Pin to Seal Contact with
Various Friction Models

Chapter Overview 486

Problem Description 486

Friction Modeling 487

Results 489

Input Files 490

486 Marc Users Guide: Part I
CHAPTER 2.10
Chapter Overview
The sample session described in this chapter demonstrates various Iriction models oI a rigid pin being inserted into
and extracted Irom a rubber seal. The simulation will use all oI the available Iriction models to discuss their beneIits.
In any simulation with Iriction, it is always best to start with the no Iriction case Iirst whenever possible. This allows
Ior an understanding oI how Iriction impacts the simulation which is generally not intuitive.
The goal oI the chapter is to demonstrate:
The basic insertion/extraction process with and without Iriction
Demonstrate the beneIits oI the new Bilinear Iriction model by comparing to the Arc Tangent and Stick Slip
Coulomb Iriction models.
Problem Description
The model is shown in Figure 2.10-1 where the axisymmetric rubber seal is modeled with a Neo-Hookean material
with C
10
50N/cm
2
.
Figure 2.10-1 Rigid Pin Inserted into and Extracted from Rubber Seal
The pin is inserted into and extracted Irom the seal Ior Iive cases: no Iriction, bilinear, arc tangent (two diIIerent sliding
velocities), and the stick-slip Coulomb Iriction models. The coeIIicient oI Iriction between the pin and seal is 0.230;
whereas, the coeIIicient oI Iriction between the seal to seal contact is 0.500. The seal-to-seal contact is created as the
rubber Iingers bend and touch the surrounding rubber material.
x
r
r = 0 cm
r = 0.635 cm
r = 0.910 cm
1.43 cm
Rigid Pin
Rubber
Seal
=
0.230 Pin/Seal
0.500 Seal/Seal
{
487
CHAPTER 2.10
Pin to Seal Contact with Various Friction Models
Friction Modeling
The preexisting models are shown in Table 2.10-1 with the various Coulomb Iriction types used. By deIault, contact
deIaults to the Irictionless case, and as mentioned beIore, this is the Iirst place to start iI physically possible.
Procedure: to run above models:
A similar procedure is used Ior the remaining Iiles. While running the other Iiles, check the Iriction type selection Ior
the various models as shown in Figure 2.10-2.
Table 2.10-1 Preexisting Models and Coulomb Friction Type Used
Mentat Model File Coulomb Friction Type Used
sealinsert_nf.mud No Friction Case
sealinsert_arctanv1.mud Arc Tangent with deIault sliding velocity
sealinsert_arctanv2.mud Arc Tangent with correct sliding velocity
sealinsert_bilinear.mud Bilinear with deIault settings
sealinsert_stickslip.mud Stick Slip with deIault settings
FILE
OPEN
sealinsert_nf.mud (open model file)
OK
MAIN
JOBS
MECHANICAL
CONTACT CONTROL
FRICTION TYPE
none
OK (twice)
RUN
SUBMIT
488 Marc Users Guide: Part I
CHAPTER 2.10
Figure 2.10-2 Control Parameters for Coulomb Friction Types
When selecting a Iriction type, it is easy to Iorget to set the various parameters unique to each type. For the Arc Tangent
type, the sliding velocity is deIaulted to unity which in most cases is not correct. The bilinear and stick-slip model
parameters deIault to usable values Ior most oI the cases. Hence, when using the Arc Tangent type, you must pay
particular attention to the value oI the sliding velocity. In general the sliding velocity is about 1 to 10 oI the
characteristic sliding velocity as based upon the physics. Even a static contact problem uses time to control the position
oI the rigid bodies and the sliding velocity must be selected correctly. In our suite oI models, there are two diIIerent
Arc Tangent types with the deIault and correct value oI the sliding velocity.
As mentioned, the new Bilinear and existing Stick-Slip Iriction model parameters do not need to be changed, and as
such, are easier to use correctly. Also, intuitively, one expects that the simulation run times increase Irom no Iriction,
arc tangent, bilinear, and stick-slip Iriction types which is shown in this demonstration.
Arc Tangent Control Parameters
Bilinear Control Parameters Stick Slip Control Parameters
489
CHAPTER 2.10
Pin to Seal Contact with Various Friction Models
Results
Figure 2.10-3 plots the insertion and extraction Iorce history Ior the Iive models.
Figure 2.10-3 lnsertion and Extraction Force History All Models (click right figures to play animation ESC
to stop)
The Irictionless case, as expected, has the lowest insertion and extraction Iorce whose peak value is around 9 N.
Adding Iriction dramatically increases the peak insertion Iorce to about 100N, and the extraction Iorce minimum peak
is about -145 N. All three Iriction types produce nearly the same Iorce history as long as the sliding velocity oI the Arc
Tangent type is properly set. The proper sliding velocity Ior this case is determined by the velocity oI the pin during
the insertion which gives a value oI 0.020 cm/sec. Note what happens when the deIault sliding velocity is used, the
eIIectiveness oI the Iriction is dramatically diminished, and is incorrect.
Comparing run times that are shown in Table 2.10-2 Ior the various cases helps understand the beneIits. OI course, the
Irictionless case requires the least amount oI run time, Iollowed by the Arc Tangent, then Bilinear, and the Stick-Slip
model last. As designed, the Bilinear takes a bit more time but has the beneIit oI realistic deIault parameters that do
not underestimate the Iriction Iorces like the Arc Tangent Iriction type.
0 2
-150

100
0
Frictionless Arc Tangent (V=0.02 cm/sec) Arc Tangent (V=1.0 cm/sec)
Bi Linear Stick-Slip
Time
I
n
s
e
r
t
i
o
n
/
E
x
t
r
a
c
t
i
o
n

F
o
r
c
e

X
Insert
Extract
X
Y
Z
Insertion/Extraction Force X
490 Marc Users Guide: Part I
CHAPTER 2.10
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Table 2.10-2 Run Times of Coulomb Friction Type Used
Coulomb Friction Type Used
Normalized
Run Times
No Friction Case 1
Arc Tangent with deIault sliding velocity .90
Arc Tangent with correct sliding velocity 1.45
Bilinear with deIault settings 1.15
Stick Slip with deIault settings 3.20
File Description
sealinsert.proc Mentat procedure Iile to run the above example
sealinsert_arctanv1.mud Associated model Iile
sealinsert_arctanv2.mud Associated model Iile
sealinsert_bilinear.mud Associated model Iile
sealinsert_nf.mud Associated model Iile
sealinsert_stickslip.mud Associated model Iile
Chapter 2.11: Analysis of a Manhole Structural Zooming
2.11
Analysis of a Manhole with
Structural Zooming

Chapter Overview 492

Background Information 492

Global Analysis 492

Local Model and Analysis 493

Conclusion 498

Input Files 498

492 Marc Users Guide: Part I
CHAPTER 2.11
Chapter Overview
This chapter demonstrates the Marc structural zooming capability.
The chapter starts with a brieI description oI the background inIormation. The model Ior analysis involves one cylinder
joined to another cylinder oI a larger radius. A local model with a Iiner Iinite element mesh, Iocusing on the joint oI
two cylinders and its vicinity, is then generated. Based on the global results, an analysis oI the local model is perIormed
to achieve a reIined evaluation oI the stress concentration around the cylinder joint.
Background Information
The problem used to demonstrate structural zooming capability in this chapter is the same as the one described in
Chapter 3.2 in this manual. Also, reIer to this chapter Ior detailed description on model geometry, materials, boundary
conditions/loads, and mesh generation.
In Chapter 3.2, the problem is considered linear. The total value oI loads are applied at increment 0. In order to
demonstrate the use oI structural zooming in a nonlinear analysis, the problem is slightly modiIied to have the loads
applied in 10 equal increments. The large strain nonlinear behavior is modeled using PLASTICITY,3 parameter.
Furthermore, shell thickness in the global model has to be written into post Iile Ior a structural zooming analysis
involving shell elements in the global model.
Global Analysis
The global model manhole.mud is generated in Chapter 3.2 oI this manual. The modiIications regarding nonlinear
analysis and shell thickness, mentioned in Background Information, are taken into account in manhole.mud.
The steps in this section includes:
Open the established model manhole.mud
Run global model
View stress distribution
FILES
OPEN
manhole.mud
OK
FILL
MAIN
JOBS
RUN
SUBMIT (1)
493
CHAPTER 2.11
Analysis of a Manhole with Structural Zooming
Figure 2.11-1 Distribution of Equivalent Stress, Obtained from Global Analysis
Local Model and Analysis
This section will include the Iollowing three steps:
Step 1: Build a local model with a refined mesh
Step 2: Modify boundary conditions and apply GLOBAL-LOCAL boundary conditions
Step 3: Save model, run model, and view results
MONITOR
OPEN POST FILE (RESULTS)
DEF & ORI
CONTOUR BAND
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
494 Marc Users Guide: Part I
CHAPTER 2.11
Step 1: Build a local model with a reIined mesh
To build a local model with a reIined mesh, the elements out oI the considered local area must be deleted Iirst.
Figure 2.11-2 Delete Elements NOT in Considered Local Area
The mesh is then reIined using the Mentat SUBDIVIDE option. One element becomes Iour by deIault because the
subdivision in each direction is 2. AIter cleaning up the model by removing unused node and by sweeping all elements
and nodes, the local mesh is established.
CLOSE
MAIN
MESH GENERATION
ELEMS: REM
449 450 451 452 453 454 455 456 457 458 459 460 461 462 463
464 465 466 467 468 201 202 203 204 205 221 222 223 224 225
241 242 243 244 245 261 262 263 264 265 281 282 283 284 285
301 302 303 304 305 321 322 323 324 325 341 342 343 344 345
346 347 348 349 350 351 352 353 354 355 356 357 358 359 360
366 367 368 369 370 371 372 373 374 375 376 377 378 379 380
386 387 388 389 390 391 392 393 394 395 361 362 363 364 365
381 382 383 384 385 469 470 471 472 396 397 398 399 400 220
240 260 280 300 320 340
END LIST (#)
495
CHAPTER 2.11
Analysis of a Manhole with Structural Zooming
Step 2: ModiIy boundary conditions and apply GLOBAL-LOCAL boundary conditions
All boundary conditions existing in the global model are still available Ior the local model. However, due to the mesh
reIinement, new nodes are added. The relevant boundary conditions Ior these newly added nodes must be speciIied.
To establish a link between the global model and the local model, a list oI connecting nodes must be deIined. The
kinematic boundary conditions oI these nodes are automatically calculated by Marc program, based on the results
obtained Irom global analysis. We reIer to the deIinition oI the connecting nodes as the speciIication oI
GLOBAL-LOCAL boundary conditions.
SUBDIVIDE
ELEMENTS
ALL: EXIST
RETURN
SWEEP
REMOVE UNUSED: NODES
ALL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NODES: ADD
598 601 604 610 613 619 622 628 631 634 637 640 646 649
655 658 664 667 1318 1319 1327 1328 1342 1343 1351 1352
END LIST (#)
RETURN
NEW
GENERAL
GLOBAL-LOCAL
CONNECT NODES TO GLOBAL MODEL
POST FILE
manhole_job1.t16
OK
NODES: ADD
598 599 600 671 672 743 744 815 816 941 942 1067 1068
1193 1194 667 668 669 740 1197 1200 1206 1209 1215 1218
1224 1227 1233 1236 1242 1245 1251 1254 1260 1263 741 812
496 Marc Users Guide: Part I
CHAPTER 2.11
Figure 2.11-3 Define GLOBAL-LOCAL Boundary Conditions
Figure 2.11-4 Connecting Nodes with GLOBAL-LOCAL Boundary Conditions
GLOBAL-LOCAL boundary conditions must be activated under the JOBS.
813 938 939 1064 1065 1190 1191 1316 1317 1269 1272 1278
1281 1287 1290 1296 1299 1305 1308 1314 1345 1348 1351
1381 1384 1417 1420 1528 1438 1441 1456 1459 1474 1477
1492 1495 1510 1513 1405 1402 1366 1369 1327 1330 1333
END LIST (#)
MAIN
JOBS
MECHANICAL
GLOBAL-LOCAL
GLOBAL-LOCAL BOUNDARY CONDITIONS
497
CHAPTER 2.11
Analysis of a Manhole with Structural Zooming
Figure 2.11-5 Activate GLOBAL-LOCAL Boundary Conditions
Step 3: Save model, run model, and view results
The local model has been Iully established so Iar. To avoid over-writing the global model and the global results, the
local model must be saved with a diIIerent Iile name. Use the Iollowing button to save model, run local analysis, and
to view results.
apply6
OK (three times)
FILES
SAVE AS
manhole_shell.mud
OK
MAIN
JOBS
RUN
SUBMIT (1)
MONITOR
OPEN POST FILE (RESULTS)
DEF & ORI
CONTOUR BAND
498 Marc Users Guide: Part I
CHAPTER 2.11
Figure 2.11-6 Distribution of Equivalent Stress, Obtained from Local Analysis
Conclusion
The maximum equivalent stress obtained Irom the local analysis is 4.48e4, which is about 10 higher than the stress
Irom global analysis. In comparison oI Figure 2.11-1 and Figure 2.11-6, a sharper stress concentration is observed in
the local analysis, representing a better evaluation oI stress gradient.
Using the structural-zooming technique, it is also possible to model the local intersection oI the cylinders with brick
elements that use the global results Irom a global shell model.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
File Description
manhole_shell.proc Mentat procedure Iile to run the above example
manhole.mud Associated model Iile
Chapter 2.12: Radiation Analysis
2.12
Radiation Analysis

Chapter Overview 500

Background Information 500

Detailed Session Description 502

Input Files 527

500 Marc Users Guide: Part I
CHAPTER 2.12
Chapter Overview
This chapter demonstrates how to perIorm a heat transIer simulation that incorporates both conduction and radiation.
An analysis oI a pressure vessel will be perIormed using both axisymmetric and three-dimensional techniques.This
chapter will also demonstrate:
The pixel based semi-hemi-cube method Ior calculating viewIactors was introduced in the Marc 2007 release.
Application oI boundary conditions on geometric entities.
Usage oI tables to deIine temperature dependent boundary conditions.
While the latter two have been available in Mentat Ior many years, the way this data is communicated to the analysis
program has changed since the Marc 2007 release.
Background Information
Description
This session demonstrates that analysis oI a large steel vessel that is subjected to a large heat Ilux. The heat is
transmitted both by conduction through the steel and by radiation. Radiation is an inherently nonlinear phenomena.
Additionally, the material properties oI the steel are dependent on the temperature. The axisymmetric analysis will be
perIormed both with and without the radiation included to show the signiIicance. Finally, a three-dimensional analysis
will also be perIormed using symmetry conditions.
The vessel with boundary conditions is shown below (Figure 2.12-1). It consists oI a cylindrical section oI length oI
30 m and outer radius oI 3 m. Each end is closed with a spherical cap. The thickness is 0.3 m. The bottom oI the vessel
is subjected to a constant Ilux.
Figure 2.12-1 Steel Vessel
Idealization
The model has rotational symmetry and is Iirst modeled using axisymmetric elements. The 3-D model is perIormed
to demonstrate the use oI symmetry surIaces and to show some novel modeling techniques.
Full Disclosure
The pressure vessel is modeled with both Iour-node axisymmetric quadrilateral and eight-node brick elements. Both
the geometric and the Iinite element will be constructed. The temperature dependent thermal conductivity and
speciIied heat are shown in Figure 2.12-2. The density is 7800 kg/m
3
. The emissivity is 0.75 on the interior surIace
and 0.2 on the exterior surIace.
501
CHAPTER 2.12
Radiation Analysis
Figure 2.12-2 Thermal Properties at Elevated Temperatures
The internal applied Ilux is 1000 W/m
2
.
Overview of Steps
Step 1: Create Axisymmetric Geometry, Finite Element Mesh, and Associate the Two Together
Step 2: Apply Material Properties
Step 3: Define Geometry of Radiating Cavity
Step 4: Define Initial Conditions and Boundary Conditions
Step 5: Define Emissivity on Cavity Surface
Step 6: Define the Loadcases
Step 7: Define the Jobs and Submit
Step 8: Review the Results
Step 9: Convert Axisymmetric Geometry and Mesh to 3-D
Step 10: Convert the Remainder of the Model, including Adding Symmetry Surfaces
Step 11: Create Planes to be Used for Symmetry Surfaces
Step 12: Loadcase Creation and Job Creation
Step 13: Review Results
502 Marc Users Guide: Part I
CHAPTER 2.12
Detailed Session Description
Step 1: Create Axisymmetric Geometry, Finite Element Mesh, and Associate the Two Together
The geometric model is created Iirst by creating a grid and generating a series oI straight lines and circular arcs. The
circular arcs are created using the center, radius, beginning and ending angle technique.
The result is shown in Figure 2.12-3.
MAIN
MESH GENERATION
SET
U DOMAIN
0 36
U SPACING
1
V DOMAIN
0 3
V SPACING
1
GRID
RETURN
FILL
CURVE TYPE CENTER/RADIUS/ANGLE/ANGLE
ADD CURVE
33.0 0.0 0.0 3 0 90
33.0 0.0 0.0 2.7 0 90
3.0 0.0 0.0 3.0 90 180
3.0 0.0 0.0 2.7 90 180
CURVE TYPE LINE
ADD CURVE
3 7
6 10
9 12
4 1
CHECK
FLIP CURVES
2 4 6 #
503
CHAPTER 2.12
Radiation Analysis
Figure 2.12-3 Geometric Representation of Vessel, Composed of Curves
The Iinite element mesh is then created by using a combination oI local coordinate systems, creating three elements
and subdividing them, then attaching the edges to the curves.
SET
ORIGIN
3 0 0
CYLINDRICAL
U DIVISION
0 4
RETURN
ADD ELEMENT
node (2.7, 90, 0)
node (3.0, 90, 0)
node (3.0, 180, 0)
node (2.7, 180, 0)
SUBDIVIDE
2 9 1
ELEMENT 1 #
RETURN
SET
ORIGIN
33 0 0
504 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-4 Finite Element Model
RETURN
ADD ELEMENT
node (27, 0, 0)
node (30, 0, 0)
node (30, 90, 0)
node (27, 90, 0)
SUBDIVIDE (is shown in Figure 2.12-4)
ELEMENT 20 #
RETURN
ADD ELEMENT
38 37 2 1
ATTACH
EDGES - CURVE
4 (inner left curve)
2:3 3:3 ... 10:3 # (inner left edges)
3 (outer left curve)
11:3 12:1 ... 19:1 # (outer left edges)
7 (left small flat curve)
19:2 10:2 # (left small flat edges)
2 (inner right curve)
505
CHAPTER 2.12
Radiation Analysis
The result is shown is Figure 2.12-5.
Figure 2.12-5 Finite element Model with Edges of Elements Attached to Curves
21:3 22:3 ... 29:3 # (inner right edges)
1 (outer right curve)
30:1 31:1 ... 38.1 # (outer right edges)
8 (right small flat curve)
21:0 30:0 # (right small flat edges)
6 (inner large flat curve)
39:3 # (inner large flat edges)
5 (outer large flat curve)
39:1 # (outer large flat edges)
RETURN
SUBDIVIDE
DIVISIONS
2 3 1
ELEMENTS
39 #
RETURN
506 Marc Users Guide: Part I
CHAPTER 2.12
Step 2: Apply Material Properties
The thermal material properties are deIined by Iirst entering the temperature dependent tables, then associating them
with a material and Iinally associating this with all oI the elements. Commands associated with labeling oI the tables
are omitted here Ior brevity (they are included in the procedure Iile).
The temperature dependent data provided in Figure 2.12-2 is given with respect to degrees Celsius, this will be shiIted
here to degrees Kelvin.
Note: By deIault in Mentat, iI the independent variable is outside oI the range, entered the table that will be extrapolated. To
change this, select the MORE button and turn oII extrapolation.
The analysis program obtains the values oI the thermal conductivity and speciIic heat at each integration point by
evaluating the table and multiplying it by the reIerence value which is one. The surIace emissivity will be deIined in
a separate stage.
MATERIAL PROPERTIES
TABLE
NEW
INDEPENDENT VARIABLE
TYPE
TEMPERATURE
ADD
273 52
773 38
1273 28
FIT
NAME
thermal_conductivity
RETURN
TABLE
NEW
1 INDEPENDENT VARIABLE
TYPE
TEMPERATURE
ADD
273 .43
873 .70
1523 .73
507
CHAPTER 2.12
Radiation Analysis
The temperature dependent properties are shown in Figure 2.12-6 and Figure 2.12-7
Figure 2.12-6 Definition of Temperature Dependent Conductivity with Table
Figure 2.12-7 Definition of Temperature Dependent Specific Heat with Table
FIT
NAME
specific_heat
RETURN
508 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-8 Association of Tables with Material Properties and Elements
Step 3: DeIine Geometry oI Radiating Cavity
In this analysis, the radiating cavity is the closed region oI the vessel. It is deIined by entering the three curves that
were constructed earlier. As it is an axisymmetric structure, there is no reason to enter a symmetry surIace/curve at
r 0. As the cavity is closed, there is no need to deIine the environment temperature that internal heat can escape
through a control node.
HEAT TRANSFER
CONDUCTIVITY TABLE
thermal_conductivity
SPECIFIC HEAT
1.0
SPECIFIC HEAT TABLE
specific_heat
MASS DENSITY
7800
OK
ADD
# ALL:EXIST
MAIN
MODELING TOOLS
CAVITIES
509
CHAPTER 2.12
Radiation Analysis
In this model, the cavities are deIined by curves. This Iacilitates the use oI radiation with either local or global adaptive
meshing. Its continuum elements are used; the orientation oI the curves is not important; this is not the case when shell
elements are used. The menu is shown in Figure 2.12-9.
Figure 2.12-9 Definition of Cavities
Step 4: DeIine Initial Conditions and Boundary Conditions
The initial conditions oI 293K are entered Ior all nodes as shown in Figure 2.12-10. Note that in radiation analysis,
the Ilux associated with radiation is calculated in absolute units.
The user should either deIine temperatures in absolute units, or speciIy the oIIset temperature between user units
and absolute.
Three boundary conditions are deIined, though they are not all used Ior each analysis. This includes:
1. A Ilux (as shown in Figure 2.12-11) oI 1000 W/m
2
is applied on the inside surIace oI one oI the hemispherical
caps. This represents the heating device that is present in a reacting vessel.
2. Internal radiation is applied as shown in Figure 2.12-12. This is based upon the cavity that is as deIined in the
previous step.
3. Radiation to the environment occurs on Irom the external surIaces as shown in Figure 2.12-13. As none oI the
external Iaces can see each other, a viewIactor calculation is not required. The emissivity on the external
surIaces is 0.2, and the environment temperature is 293K.
DIMENSION:2-D
ADD CURVES
2 4 6
ENDLIST
MAIN
510 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-10 Definition of Initial Temperature
INITIAL CONDITIONS
THERMAL
TEMPERATURE
TEMPERATURE
293
OK
NODE:ADD
ALL:EXIST
MAIN
BOUNDARY CONDITIONS
THERMAL (define internal heating)
NAME
heating
EDGE FLUX
FLUX
1000
OK
CURVES:ADD
4
NEW
511
CHAPTER 2.12
Radiation Analysis
NAME
internal rad (define internal radiation)
CAVITY RADIATION
RADIATION
CLOSED
CALCULATE
WRITE TO POST FILE
OK
CAVITIES:ADD
cavity1
OK
NEW
NAME
external rad (define external radiation)
EDGE FILE
SINK TEMPERATURE
293
EMISSIVITY
0.2
OK
CURVES:ADD
3 5 1
RETURN
ID BOUNDARY CONDITIONS
512 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-11 Thermal Flux on Surface
Figure 2.12-12 Radiation Cavity Boundary Condition
513
CHAPTER 2.12
Radiation Analysis
Figure 2.12-13 External Radiation to the Environment
Step 5: DeIine Emissivity on Cavity SurIace
The emissivity associated with a radiating cavity may either be applied as material data applied to the elements or as
a surIace property applied to an edge or Iace. The latter (surIace property) which was introduced in the 2007 release
is the preIerred method. It permits diIIerent emissivities to be applied to the same model or element to reIlect surIace
coatings, polish and/or wear. In this problem, the emissivity on the internal region is constant, but substantially higher
than the emissivity on the outside surIace. This reIlects the degradation oI the surIace due to the chemical and thermal
reactions in the tank.
MAIN
MATERIAL PROPERTIES
SURFACE PROPERTIES
RADIATION PROPERTIES
COEFFICIENT
0.75
OK
CURVES:ADD
4 6 2
MAIN
514 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-14 Surface Emissivity for Radiating Cavity
Step 6: DeIine the Loadcases
In this problem, three loadcases are deIined and they are activated in three jobs. In all three cases, a period oI 300
seconds are analyzed using the adaptive time stepping procedure. The initial time step is 0.1 sec. The Iirst loadcase
has only the heater boundary condition; the second loadcase the heater and the internal radiation, and the third
loadcase has all three boundary conditions.
This procedure allows the user to associate multiple analyses with the same model Iile. Because oI the highly nonlinear
nature oI radiation, a tight tolerance oI 10 is placed on the maximum error in the temperature estimate. This insures
an accurate analysis.
LOADCASES (create 1st loadcase)
HEAT TRANSFER
TRANSIENT
LOADS
internal rad (deactivate)
external rad (deactivate)
OK
CONVERGENCE TESTING
MAX ERROR IN TEMPERATURE ESTIMATE
10
OK
ADAPTIVE:TEMERATURE
PARAMETERS
INITIAL TIME STEP
515
CHAPTER 2.12
Radiation Analysis
Figure 2.12-15 Loadcase 3 with all Three Boundary Conditions Activated
0.1
MAX # INCREMENTS
500
OK
TOTAL LOADCASE TIME
300
OK
COPY (create 2nd loadcase using 1st as a base)
TRANSIENT
LOADS
internal rad (activate boundary condition)
O
K
COPY (create3rd loadcase using 2nd as a base)
LOADS
external rad (activate boundary condition)
OK
MAIN
516 Marc Users Guide: Part I
CHAPTER 2.12
Step 7: DeIine the Jobs and Submit
In this step, parameters associated with the heat transIer analysis are set. Many deIault values are used, but they are
reviewed here to indicate other possibilities. Three jobs are created, each with a single loadcase and subsequently
submitted Ior analysis. The results are compared in a later step.
JOBS (create the 1st job)
RENUMBER ALL
TITLE
AXISYMMETRIC HEATING OF VESSEL
HEAT TRANSFER
LOADCASE SELECT LCASE1
ANALYSIS DIMENSION: AXISYMMETRIC
ANALYSIS OPTIONS
LUMPED CAPACITANCE
RADIATION Note: DeIaults are considered adequate Ior analysis
OK
JOB PARAMETERS
UNITS AND CONSTANTS
TEMPERATURE IN KELVIN
OK (thrice)
COPY (create the 2nd job)
HEAT TRANSFER
CLEAR
LOADCASE SELECT LCASE2
OK
RUN
SUBMIT
OK
COPY (create the 3rd job)
HEAT TRANSFER
CLEAR
LOADCASE SELECT LCASE3
OK
RUN
SUBMIT
517
CHAPTER 2.12
Radiation Analysis
There are several considerations when perIorming a radiation analysis.
First, the user should be careIul in choosing the units and indicate the units to the program. Here all units are in Kelvin,
so under the JOB-> HEAT TRANSFER-> JOB PARAMETERS-> UNITS AND CONSTANTS menu, this needs to be
deIined. Associated with radiation analysis is the SteIan Boltzmann constant. It must be given in consistent units. II
Irequency dependent emissivity is deIined, then it is also necessary to deIine Planck`s 2nd constant and the speed oI
radiation (light) in a vacuum.
The speed must be given in a unit that is consistent with the wavelength unit used to deIine the emissivity.
Second, the viewIactor calculation is approximate; the accuracy is dependent upon user entered parameters. When the
Monte Carlo procedure is used (see BOUNDARY CONDITIONS-> THERMAL-> COMPUTE RADIATION
VIEWFACTOR), the accuracy is controlled by the number oI rays randomly emitted.
In the Pixel Based Semi-Hemi-cube method used in this example, the accuracy is based upon the number oI pixels
used. This is controlled on via the JOBS-> HEAT TRANSFER-> ANALYSIS OPTION-> RADIATION menu (shown in
Figure 2.12-17).
Here, 500 is entered (deIault), which is the number oI pixels between (0-1). The actual calculation goes Irom (-1 to 1)
in both a local x- and y-direction, so the actual number oI pixels is (2x500)
2
1 million.
For axisymmetric models, the viewIactors are actually calculated in 3-D and then reassociated with the 2-D edge. The
AXISYMMETRIC # DIVISIONS button controls the accuracy in this calculation.
Finally, the viewIactors may be neglected by the analysis program or treated explicitly. II the viewIactor is below the
USE VIEWFACTOR control, it is ignored. II the viewIactor is greater than this value but less than the TREAT
VIEWFACTOR IMPLICITLY button, the radiation Ilux associated with this viewIactor is treated explicitly. This may
result in more iterations, but reduces the size oI the stiIIness matrix.
518 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-16 Heat Transfer Units and Constants Menu
Figure 2.12-17 Radiation Parameter Settings
519
CHAPTER 2.12
Radiation Analysis
Step 8: Review the Results
A contour plot and the time history plot oI selective nodes are examined Ior the three runs. The transient behavior oI
the three jobs is compared.
It is assumed that the post Iile associated with the job to be examined has already been opened using the RESULTS->
OPEN command or the OPEN DEFAULT COMMAND.
Observing the results.
Figure 2.12-18 Location of Nodes being Tracked
FILL VIEW
LAST
CONTOUR BANDS
HISTORY PLOT
SET NODES
1 4 7 40 80
COLLECT DATA
NODES/VARIABLES
ADD VARIABLE
Time
Temperature
FIT
80
4
7
1
40
520 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-19 Transient Response for Job1 Heating Only
Figure 2.12-20 Transient Response for Job2 Heating and Internal Radiation
521
CHAPTER 2.12
Radiation Analysis
Figure 2.12-21 Transient Response for Job3 Heating, Internal and External Radiation
In observing that radiation is not included in job1, the leIt side oI the vessel gets the hottest. When internal radiation
is included (job2), some oI the heat radiates to the opposite side, and hence, the maximum temperature is lower. When
both internal and external radiation is included, the vessel temperature is the lowest as expected.
When examining the output oI job2 or job3 aIter the message start oI increment 1, you can see the Iollowing
inIormation regarding the calculation oI the viewIactors.
s t a r t o f i n c r e m e n t 1
calculating viewfactor for cavity 1
allocated 8688 words of memory due to radiation viewfactors
view factors read in from .vfs file
cavity number : 1
number of faces : 48
number of pixels used : 500
number of factors : 2304
minimum viewfactor : 0.0000164
maximum viewfactor : 0.1723900
maximum connectivity in stiffness matrix is 32 at node 75
maximum half-bandwidth is 146 between nodes 2 and 147
The user observes that the number oI radiating Iaces is 48 which is equal to the number oI elements on the inside; this
indicates that applying the cavity onto the geometry was successIul. Then you can observe that there are 2304
calculated viewIactors, as this is an axisymmetric problem, the maximum possible is 48x48 2304. Hence, all possible
522 Marc Users Guide: Part I
CHAPTER 2.12
viewIactors have been Iound. Then one observes that the minimum viewIactor is 0.0000164 and the maximum is
0.17239, or the minimum is 0.009 oI the maximum. Based upon the deIault thresholds, some oI the viewIactors are
treated explicitly and some are neglected. Even so, the inclusion oI the radiation viewIactors signiIicantly increases
the size oI the stiIIness matrix, as the number oI proIile entries increases Irom 581 in job1 to 1233 in job2 and job3.
Step 9: Convert Axisymmetric Geometry and Mesh to 3-D
In this step, the axisymmetric model is expanded into a quarter section oI a 3-D model, and the boundary conditions
aer updated Irom 2-D to 3-D. The combined expand option is used to expand the mesh, geometry, and simultaneously
attach the 3-D Iinite element mesh to the surIaces. When using the combined expand, it expands all entities activated.
Here, nodes and points are turned oII so unnecessary lines and 2-node elements are not created.
The nodal initial conditions are automatically expanded. The distributed boundary condition are copied over, but need
to be updated.
Step 10: Convert the Remainder oI the Model, including Adding Symmetry SurIaces
MESH GENERATION
EXPAND
ROTATION ANGLE
10 0 0
REPETITIONS
9
COMBINED
NODES (deactivate nodes)
POINTS (deactivate points)
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
BOUNDARY CONDITIONS
THERMAL
FACE FLUX (change boundary condition heating
from edge flux to face flux)
OK
SURFACE ADD
28 29 30 31 32 33 34 35 36
523
CHAPTER 2.12
Radiation Analysis
Step 11: Create Planes to be Used Ior Symmetry SurIaces
With the steps below, create two planes that will be used as symmetry surIaces Ior the viewIactor calculation. This is
shown in Figure 2.12-22.
Figure 2.12-22 3-D Model with Symmetry Surfaces
NEXT
CAVITY RADIATION (review radiating cavity boundary condition)
OK
NEXT
FACE FILM (change boundary condition external rad
from edge film to edge flux)
OK
SURFACE ADD
1 2 3 4 5 6 7 8 9
19 20 21 22 23 24 25 26 27
37 38 39 40 41 42 43 44 45
MESH GENERATION
SET
U DOMAIN
-4 40
U SPACING
5
V DOMAIN
524 Marc Users Guide: Part I
CHAPTER 2.12
-5 5
V SPACING
GRID
RETURN
SRF ADD
-5 -5 0
40 -5 5
40 5 0
-5 5 0
SET
FIX V
W DOMAIN
-5 5
W SPACING
1
SRF ADD
-5 -5 -5
40 0 -5
40 0 5
-5 0 5
GRID (turn off grid)
MAIN
MODELING TOOL
CAVITIES
SYMMETRY PLANE
SURFACE 1
56
SURFACE 2
55
MAIN
525
CHAPTER 2.12
Radiation Analysis
Step 12: Loadcase Creation and Job Creation
As this model is based upon the previously created axisymmetric model, the loadcases (1,2,3) and Jobs (1,2,3) are
already created. The Iirst loadcase is reviewed, and then the 3rd job (which has the internal heating, internal radiation,
and external radiation) is submitted.
Step 13: Review Results
The post Iile is opened and the last increment is examined as shown in Figure 2.12-23. As expected, an axisymmetric
distribution oI the temperatures is obtained. A time history plot is made oI the node (4) at center oI the hemisphere,
see Figure 2.12-24. It is almost identical to the behavior shown is Figure 2.12-21.
LOADCASE
HEAT TRANSFER
TRANSIENT
LOADS
MAIN
JOBS
NEXT
NEXT
TITLE
3-d radiation analysis with symmetry
OK
SAVE
prob_2_8_3d
HEAT TRANSFER
3-D
OK
RUN
ADVANCED JOB SUBMISSION
WRITE INPUT FILE
EDIT INPUT FILE
OK
SUBMIT
526 Marc Users Guide: Part I
CHAPTER 2.12
Figure 2.12-23 Contour Plot of Temperatures
527
CHAPTER 2.12
Radiation Analysis
Figure 2.12-24 Transient Response for 3-D Heating, Internal and External Radiation
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
3drad.proc Mentat procedure Iile to run the above example
3drad_post.proc Mentat procedure Iile to run the above example
axirad_solid.proc Mentat procedure Iile to run the above example
axirad_solid_post.proc Mentat procedure Iile to run the above example
prob2-8axi.mud Associated model Iile
prob2_8_3d.mud Associated model Iile
528 Marc Users Guide: Part I
CHAPTER 2.12
Chapter 2.13: Application of BC on Geometry with Remeshing
2.13
Application of BC on Geometry
with Remeshing

Geometry and Finite Element Mesh 530

Detailed Session Description 531

Input Files 544

530 Marc Users Guide: Part I
CHAPTER 2.13
Geometry and Finite Element Mesh
This chapter provides a simple example oI the large deIormation, large strain oI a rubber component subjected to
distributed loads such that remeshing is required. The boundary conditions are applied to geometric entities (points
and curves).
The Iinite element nodes and edges are attached to these geometries and aIter automatic remeshing occurs, the new
Iinite element mesh is reassociated to the geometry to insure that the boundary conditions are correctly applied. The
model is shown in Figure 2.13-1.
Figure 2.13-1 Geometry (All units are in cm)
Overview of Steps
Step 1: Create Geometry and Finite Element Mesh
Step 2: Defining the Material Properties
Step 3: Add All Elements to Contact Body
Step 4: Apply Boundary Conditions
Step 5: Define Criteria for Global Adaptive Remeshing
Step 6: Define Loadcase
Step 7: Create Job and Submit
Step 8: Review Results
r = 17.2
r = 3.0 (-10,0)
(10,0)
(0,-2)
(3,-10) (-3,-10)
531
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Detailed Session Description
Step 1: Create Geometry and Finite Element Mesh
The geometry oI the model exists in IGES Iormat in a Iile called sector.igs. This Iile is read in and meshed.
FILES
IMPORT
IGES
sector.igs
OK
MAIN
FILL
PLOT
CURVE SETTING
HIGH
RETURN
POINTS (turn off)
REDRAW
MAIN
MESH GENERATION
SWEEP (remove duplicate points)
ALL
RETURN
AUTO MESH
CURVE DIVISION
FIXED AVG LENGTH
0.8
APPLY CURVE DIVISION (the seed points are shown
ALL EXIST in Figure 2.13-2)
RETURN
532 Marc Users Guide: Part I
CHAPTER 2.13
Figure 2.13-2 Applying Seed Points to Curves before Meshing
The resulting mesh is shown is Figure 2.13-3.
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT)
QUAD MESH
ENTER CURVES LISTING
ALL EXISTING
MAIN
533
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Figure 2.13-3 Finite Element Mesh
Edges that are in red indicate that they are attached to curves. Nodes that are shown as circles are attached to the points.
Step 2: DeIining the Material Properties
The rubber piece is modeled using the Mooney-Rivlin model, with the properties:
c
10
20.3 N/cm
2
c
01
5.8 N/cm
2
The material properties are constant.
MATERIAL PROPERTIES
MORE
MOONEY
c
10
= 20.3
c
01
= 5.8
OK
ELMENTS ADD
ALL EXISTING
MAIN
534 Marc Users Guide: Part I
CHAPTER 2.13
Step 3: Add All Elements to Contact Body
The global adaptive remeshing procedure is based upon contact bodies. While in this simulation, the load is not large
enough to cause the hole to close upon itselI, all oI the elements are put into a single contact body.
Step 4: Apply Boundary Conditions
The problem has three boundary conditions: the base is Iully constrained, the top arc has a pressure applied, and halI
oI the circle has a load applied to it. In all cases, the boundary condition is applied to a curve. Because the boundary
conditions are applied to a curve as opposed to Iinite element edges, aIter remeshing occurs, the boundary conditions
are automatically applied correctly. The pressure loads are linearly ramped up over a loadcase oI one second to their
reIerence value oI 12N/cm
2
by using a table.
CONTACT
CONTACT BODIES
NEW
DEFORMABLE
OK
ELEMENTS ADD
ALL EXIST
MAIN
APPLY
NAME (apply pressure on top arc)
pressure_on_top
MECHANICAL
TABLE (create ramp function)
NEW
1 independent variable
TYPE
time
NAME
ramp
ADD
0 0
1 1
MORE
LABEL
time
535
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Figure 2.13-4 Ramp Function used to Apply Pressure
LABEL
Scale Factor
SHOW MODEL
RETURN (the ramp function is shown is Figure 2.13-4)
EDGE LOAD
pressure 12
OK
CURVES ADD
4 #
NEW (apply pressure to half of hole)
NAME
pressure_in_hole
EDGE LOADS
PRESSURE
12
TABLE
ramp
OK (twice)
536 Marc Users Guide: Part I
CHAPTER 2.13
Figure 2.13-5 Boundary Conditions on Geometric Entities
CURVES ADD
5 #
NEW (fully constrain base)
NAME
fixed_bottom
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Y
OK
CURVES ADD
2 #
RETURN
ID BOUNDARY CONDITIONS (see Figure 2.13-5)
DRAW BOUNDARY CONDITIONS ON MESH (see Figure 2.13-6)
MAIN
pressure_on_top
pressure_in_hole
fixed_bottom
537
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Figure 2.13-6 Boundary Conditions on Finite Element Entities
Step 5: DeIine Criteria Ior Global Adaptive Remeshing
In this simulation, because oI the large deIormation and in particular shear, the Iinite element mesh may become highly
distorted. To insure an accurate analysis, the adaptive meshing procedure is invoked. The user needs to indicate when
or why remeshing should occur, parameters controlling the new mesh, and the region to which this will be applied. In
this simulation, remeshing may be due to element distortion, change in strain, or increment Irequency. While the initial
mesh used a target seed distance oI 1.0, here the new mesh is based upon a target distance oI 0.8. Both the initial mesh
and all remeshing uses the advancing Iront quadrilateral automatic mesher. The Iinite element edge is automatically
reattached to the curves.
MESH ADAPTIVITY
GLOBAL REMESHING ADAPTIVITY
ADVANCING FRONT QUAD
ADVANCED
STRAIN CHANGE
ELEMENT DISTORTION
OK
INCREMENT
FREQUENCY
8
ELEMENT EDGE LENGTH
0.8
OK
REMESH BODY
pressure_on_top
pressure_in_hole
fixed_bottom
538 Marc Users Guide: Part I
CHAPTER 2.13
The global adaptive meshing menu is shown is Figure 2.13-7.
Figure 2.13-7 Global Adaptive Meshing Menu
Step 6: DeIine Loadcase
A single loadcase oI duration one second is analyzed using a Iixed time step procedure. Because oI the nonlinearities
involved and potential buckling oI the rubber part, a tight convergence criteria is requested.
cbody 1
OK
ID GLOBAL REMESHING CRITERIA
LOADCASES
MECHANICAL
STATIC
GLOBAL REMESHING
adapg (activate global remeshing)
OK
SOLUTION CONTROL
MAX # RECYCLES
30
Contribution of Initial Stress to Stiffness:Deviatoric Stress
OK
539
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Step 7: Create Job and Submit
The output to be placed on the post Iile is selected and upper bounds are speciIied. The large strain -updated Lagrange
procedure is used Ior this rubber analysis. Follower Force is activated, but using the deIormation at the beginning oI
the increment. This is not as accurate, but may lead to less iterations. Note that the deIault element type 11, a
conventional Iour-node element, is used in this analysis. The table driven input is activated; this insures that both the
geometric and Iinite element data is written to the input Iile, and that the boundary conditions and material data use
the new input Iormat.
CONVERGENCE TESTING
RESIDUALS AND DISPLACEMENTS
RELATIVE FORCE TOLERANCE
0.01
RELATIVE DISPLACEMENT TOLERANCE
0.01
OK
CONSTANT TIME STEP
# STEPS
20
OK
JOBS
TITLE
Adaptive Meshing of Rubber part with Pressure on Curves
MECHANICAL
lcase1
MESH ADAPTIVITY
MAX # ELEMENTS
1000
MAX # NODES
1000
OK
ANALYSIS OPTIONS
FOLLOW FORCE (Begin Inc)
LARGE STRAIN-UPDATED LANGRANGE
OK
540 Marc Users Guide: Part I
CHAPTER 2.13
Step 8: Review Results
The objective oI this problem is to see that the boundary conditions are correctly applied aIter remeshing occurs in the
model. The post Iile is opened and a SCAN is perIormed, Irom this, the user observes that initially there are 217
elements in the model. This is increased to 341 in increment 9 and Iinally to 345 in increment 17. The post Iile is
positioned to the last increment, and then the deIormed mesh is examined (Figure 2.13-8).
Figure 2.13-8 Deformed Part
From the red outline, it is clear that the edges are attached to curves as desired. Points, curves, and surIaces are placed
on the post Iile in their original conIiguration. Hence, by comparing the Iinal mesh to the original curves, we can
OUTPUT RESULTS
Equivalent Von Mises Stress
Total Strain Energy
OK
RETURN
RUN
NEW-STYLE TABLE (activate table driven input)
ADVANCED JOB SUBMISSION
WRITE INPUT FILE
EDIT INPUT FILE
OK
SUBMIT
MONITOR
lcase1
Inc: 14
Time: 7.000e-01
X
Y
Z
541
CHAPTER 2.13
Application of BC on Geometry with Remeshing
observe the total deIormation. The externally applied Iorces are shown in Figure 2.13-9; you can see that all oI the
edges attached to the top arc contribute to the Iorce, as well as, the edges on the right side oI the hole.
Figure 2.13-9 Externally Applied Force on Remeshed Curves
Finally, Figure 2.13-10 and Figure 2.13-11 show the elastic strain energy and the equivalent von Mises stress,
respectively. Note that a large portion oI the energy is at the base, where the nearly singular stress Iield exists and the
material Iolds over.
RESULTS
OPEN
load_geom_adapt_job1.t16
OK
PLOT
ELEMENT SETTING
FACES (off)
RETURN
POINTS (off)
NODES (off)
CURVES SETTING
HIGH
RETURN
RETURN
SCAN
20 (skip to increment 20)
0.000e+00
7.638e-01
1.528e+00
2.291e+00
3.055e+00
3.819e+00
4.583e+00
5.347e+00
6.110e+00
6.874e+00
7.638e+00
lcase1
External Force
Inc: 14
Time: 7.000e-01
X
Y
Z
542 Marc Users Guide: Part I
CHAPTER 2.13
DEF ONLY
PLOT
CURVES (off)
ELEMENT SETTING
OUTLINE
RETURN
MORE
VECTOR PLOT (on)
VECTORS
External Force
OK
CONTOUR BANDS
SCALAR
Total Strain Energy Density
SCALAR SETTING
EXTRAPOLATION
TRANSLATE
RETURN
SET LIMITS
0 16
# LEVELS
8
MANUAL
RETURN
SCALAR
Equivalent Von Mises Stress
SCALAR SETTING
SET LIMITS
0 80
543
CHAPTER 2.13
Application of BC on Geometry with Remeshing
Figure 2.13-10 Strain Energy Density
Figure 2.13-11 Equivalent Stress
544 Marc Users Guide: Part I
CHAPTER 2.13
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
adapt.proc Mentat procedure Iile to run the above example
adapt_post.proc Mentat procedure Iile to run the above example
sector.igs IGES input Iile Ior geometry
Chapter 2.14: Glass Forming of a Bottle with Global Remeshing
2.14
Glass Forming of a Bottle with
Global Remeshing

Chapter Overview 546

Detailed Session Description 549

Conclusion 561

Input Files 563

546 Marc Users Guide: Part I
CHAPTER 2.14
Chapter Overview
This example demonstrates a glass Iorming simulation oI a bottle (Figure 2.14-1). The bottle is blow-Iormed. The
purpose oI the simulation is to assist in the design Iorming process, mold shape, and the glass gob to ensure a
successIul end product.
The capability oI global remeshing together with pressure loading and Iixed displacement boundary conditions is
presented. Thermal and mechanical coupled analysis is required. The glass material is modeled by a user subroutine,
although, Marc`s Narayanaswamy model Ior glass could also be used. The bottle thickness, stress, and temperature
distribution can be predicted in the simulation.
Figure 2.14-1 Bottle Glass Forming
Idealization
A glass gob is shown in Figure 2.14-2. A rigid axisymmetric mold is assumed in the analysis. Initial temperature oI
the glass is at 1000C. The mold temperature and the environment sink temperature are both at 20C. A pressure loading
is applied to the glass inner surIace to model the blow Iorming process. A rigid-viscoplastic material model are
adopted Ior the analysis.
547
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Figure 2.14-2 Initial Model Setup
Element type 10 oI 4-node quadrilateral is adopted Ior the glass gob with 265 elements in the initial mesh. For thermal-
mechanical coupled analysis, the element type 40 is used by deIault Ior to be Newtonian Iluid with a viscosity that is
temperature dependent. This can be modeled as a rigid-visco-plastic material in Marc and through a URPFLO user
subroutine. The Ilow stress Iunction can be described as Iollows |ReIerence 1|:
where T is the temperature in C. The viscosity unit is in Poises. A Poises 0.1 Newton.second/m2. ThereIore, the
stress shown above is converted to SI (mm) unit with
To avoid problems with artiIicially high strain rates, an upper bound to the Ilow stress is also provided.
The thermal properties are listed in the Iollowing:
The contact bodies are shown in Figure 2.14-2. The pressure loading applied on the surIace has the magnitude oI
0.0016N/mm
2
Irom 0 to 0.016 seconds. The bottle is Iormed in 0.016 seconds and Iollowed by 1 second oI cooling
time. A Iixed displacement in X is applied to the top oI the glass gob.
Conductivity 40 N/sec/C
SpeciIic Heat
0.5 mm
2
/sec
2
/C
Mass Density
1.0 Mg/mm
3
o
y
3c

10
2.58
4332
T 25
----------------

o
y
3c

10
2.58
4332
T 25
----------------
10
7

548 Marc Users Guide: Part I
CHAPTER 2.14
No Iriction is assumed. The convection coeIIicient between the workpiece and the mold is 40 (N/sec/C/mm) and the
convection coeIIicient to the environment is 0.04 (N/sec/C/mm).
Analysis with Remeshing
Because oI large deIormation, a global remeshing is activated whenever there is an element distortion. AIter
remeshing, the boundary conditions applied to the elements will be transIerred to the new mesh as well as those
history data.
The Iollowing controls are utilized in the global remeshing:
Advancing Iront quad mesher to generate the mesh
Number oI Elements: 500
Curvature control division: 36
The target number oI elements is used to generate the new mesh oI about the same number oI elements. The remeshing
is activated when any one oI the Iollowing criteria is met:
Every 5 increments
Element distortion
The new style table input Iormat is required Ior the global remeshing to work with the boundary conditions. In the new
input Iormat, boundary conditions are deIined in sets and applied later to diIIerent loadcases with the set names. In this
example, set inIormation is utilized in the remeshing to replace boundary conditions with the new mesh.
Overview of Steps
Step 1: Read in Predefined Mold Geometry and the Mesh for the Glass Gob
Step 2: Define Material Properties
Step 3: Define Contact Bodies
Step 4: Define Initial Conditions
Step 5: Assign Boundary Conditions
Step 6: Define Global Remeshing Controls
Step 7: Define Loadcase
Step 8: Define Analysis Controls and Run Job
Step 9: View Simulation Results
549
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Detailed Session Description
Step 1: Read in PredeIined Mold Geometry and the Mesh Ior the Glass Gob
In order to save time, the mold geometry and the mesh are read in Irom a Mentat database Iile:
glass_bottle_geometry.mfd and save it as mytest.mfd. Users need to copy this Iile to the working directory. The
Iollowing steps read in the geometry and mesh:
Step 2: DeIine Material Properties
The material type is deIined as rigid plastic and a URPFLO user subroutine is used. This subroutine is created in a
FORTRAN Iile called glass_bottle_material.f. Users need to copy this Iile into the working directory.
The user subroutine Iile is provided in JOBS menu.
Heat TransIer material properties are deIined under theHEAT TRANSFER menu.
FILES
OPEN
select: glass_bottle_geometry.mfd
OK
SAVE AS
mytest
MAIN
MATERIAL PROPERTIES
ISOTROPIC
RIGID-PLASTIC
METHOD
urpflo
INITIAL YIELD STRESS
1
OK (twice)
HEAT TRANSFER
CONDUCTIVITY
40
SPECIFIC HEAT
0.5
MASS DENSITY
1
550 Marc Users Guide: Part I
CHAPTER 2.14
Step 3: DeIine Contact Bodies
The gob is deIined as a deIormable body and the mold as a rigid body.
DeIine and assign the contact properties to all the elements:
OK
elements: ADD (assign the material properties to all elements)
all: EXISTING
MAIN
CONTACT
CONTACT BODIES
NEW
NAME
glass
Contact body type
DEFORMABLE
MECHANICAL PROPERTIES
select THERMAL PROPERTIES
heat transfer to env
HEAT TRANS. COE
0.04
SINK TEMPERATURE
20
heat transfer due to contact
CONTACT HEAT TRANSFER COEF.
40
OK
elements: ADD
all: EXISTING
NEW
NAME:
mold
RIGID
MECHANICAL PROPERTIES
THERMAL PROPERTIES (select thermal properties)
551
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Use ID CONTACT to show all deIined contact bodies and the orientation oI the rigid and symmetric bodies
(Figure 2.14-3).
Figure 2.14-3 Defined Contact Bodies
TEMPERATURE
20
heat transfer due to contact:
CONTACT HEAT TRANSFER COE.
40
OK
Define and assign the contact properties to the mold.
curves: ADD
select the curve representing the mold <MR>
NEW
NAME
sym
SYMMETRY
OK
curves:ADD
select the symmetry curve <MR>
552 Marc Users Guide: Part I
CHAPTER 2.14
Step 4: DeIine Initial Conditions
The initial temperature oI the glass gob needs to be deIined.
Step 5: Assign Boundary Conditions
We need to deIine pressure the glass blowing and a Iixed boundary condition to Iix the top oI the bottle.
A table Iunction is deIined Ior the pressure.
MAIN
INITIAL CONDITIONS
THEMAL
TEMPERATURE
TEMPERATURE (top)
1000
OK
nodes: ADD
all: EXISTING
MAIN
BOUNDARY CONDITIONS
NEW
NAME
pressure
MECHANICAL
EDGE LOAD
PRESSURE
1
OK
TABLES
NEW
select 1 INDEPENDENT VARIABLE
NAME
pressure
type
TIME
553
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Now Ior the Iixed displacement condition:
Use ID BOUNDARY CONDS to show deIined boundary conditions (Figure 2.14-4).
ADD
0 0
0.016 0.0016
5 0.0016
FIT
RETURN
EDGE LOAD
TABLE
pressure (select table)
OK
edges: ADD
select all internal element edges <MR>
NEW
NAME
fixed
FIXED DISPLACEMENT
DISPLACMENT X (select fixed in X direction)
OK
nodes: ADD
select nodes on the top of the gob <MR>
554 Marc Users Guide: Part I
CHAPTER 2.14
Figure 2.14-4 Defined Boundary Conditions
Step 6: DeIine Global Remeshing Controls
MAIN
MESH ADAPTIVITY
GLOBAL REMESHING CRITERIA
ADVANCING FRONT QUAD
INCREMENT FEQUENCE
5
ADVANCED
ELEMENT DISTORTION
O
K
# elements SET
500 (enter target number of elements)
OK
REMESH BODY
glass (select contact body for remeshing)
555
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Step 7: DeIine Loadcase
DeIine two loadcases here; the Iirst one Ior the blowing and the second one Ior the cooling. For rigid-plastic model,
using only the tensile contribution oI initial stress to the stiIIness matrix is a better control to avoid divergence.
MAIN
LOADCASES
COUPLED
QUASI-STATIC (first loadcase)
GLOBAL REMESHING
adapg1 (select remeshing)
OK
TOTAL LOADCASE TIME
0.016
MULTI-CRITIA
INITIAL FRACTION
0.1
DESIRED # REC. SET: 10
DEFAULT CRIT.
MAX TEMP.:
100
OK
OK
SOLUTION CONTROL
MAX # RECYCLES
20
contribution to stiffness:
TENSILE STRESS
OK
CONVERGENCE TESTING
RESIDUALS OR DISPLACE.
RELATIVE DISP. TOL
0.01
OK (twice)
NEW (second loadcase)
QUASI-STATIC
TOTALL LOADCASE TIME
556 Marc Users Guide: Part I
CHAPTER 2.14
Step 8: DeIine Analysis Controls and Run Job
The two loadcases are activated in the coupled thermal-mechanical analysis. the FOLLOWER FORCE option is
activated because oI the large deIormation in the Iorming oI the bottle. The equivalent plastic strains are written to the
post Iile Ior later display. The Iile which contains the URPFOLO user subroutine is identiIied. When the job is
submitted, this routine will be compiled and linked to standard Marc.
1.0
MULTI-CRITIA
INITIAL FRACTION.:
0.1
DESIRED # REC. SET
10
OK
DEFAULT CRIT.
MAX TEMP.
100
OK
OK
SOLUTION CONTROL
MAX # RECYCLES:
20
contribution to stiffness:
TENSILE STRESS
OK
CONVERGENCE TESTING
RESIDUALS OR DISPLACE
RELATIVE DISP. TOL
0.01
OK (twice)
MAIN
JOBS
COUPLED
available: lcase1 and lcase2 (select loadcases)
INITIAL CONDITIONS
select all conditions
AXISYMMETRIC
557
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Figure 2.14-5 Defined Analysis Options
Select extra result output:
ANALYSIS OPTIONS
LARGE DISPLACEMENT (on)
FOLLOWER FORCE (on)
plasticity procedure
LARGE STRAIN ADDITIVE
OK
JOB RESULTS
select: TOTAL EQU. PLASTIC STRAIN
OK (twice)
SAVE (save the model)
RUN
NEW-STYLE TABLE (on)
USER SUBROUTINE FILE
glass_bottle_material.f (select the filename)
SUBMIT (1)
MONITOR
558 Marc Users Guide: Part I
CHAPTER 2.14
Figure 2.14-6 The RUN JOB Screen
Step 9: View Simulation Results
To view vector plot oI the external Iorces applied to the glass:
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
deformed shape SETTINGS
edges
OUTLINE (display outline only)
SCAN
20 (select increment 20)
OK
MORE
vector
EXTERNAL FORCE (select)
vector plot:
ON
559
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Figure 2.14-7 External Force Vector Plot
To view temperature contour at the end oI Iorming beIore cooling:
MAIN
RESULTS
SCALAR
TEMPERATURE (select)
scalar plot:
CONTOUR BAND
SCAN
35 (select the step at time=0.016)
560 Marc Users Guide: Part I
CHAPTER 2.14
Figure 2.14-8 Temperature after Forming
Similarly, the temperature at the end oI cooling can be viewed in Figure 2.14-9.
Figure 2.14-9 Temperature after Cooling
Also, by selecting the total equivalent plastic strain, we can see the plastic deIormation in Figure 2.14-10.
561
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Figure 2.14-10 Plastic Strain Contour
Conclusion
The external Iorce in Figure 2.14-7 shows that the pressure loading is applied correctly aIter remeshing. The
temperature contours in Figure 2.14-8 and Figure 2.14-9 show temperature changes aIter Iorming and cooling stages.
The bottle wall thickness can also be viewed in these Iigures.
The simulation can be utilized Ior gob shape and process design so that an optimal bottle thickness can be Iormed.
For example, by blowing the glass 10 times slower, the thickness oI the bottle will vary dramatically as cooling eIIect
on the wall that touches the mold Iirst makes material harder to Ilow. This comparison can be seen in Figure 2.14-11.
II the mold temperature is at 500C, this also aIIects the wall thickness. The upper part oI bottle wall is easier to Ilow
and becomes much thinner than the mold that is at 20C. Comparison can be seen in Figure 2.14-12.
The total Iorce required to Iorm the bottle can be seen in Figure 2.14-13.
562 Marc Users Guide: Part I
CHAPTER 2.14
Figure 2.14-11 Different Thickness by Blowing 10 Times Slower
Figure 2.14-12 Different Thickness with Mold at 500C
563
CHAPTER 2.14
Glass Forming of a Bottle with Global Remeshing
Figure 2.14-13 Blowing Force History
References
1. J.M.A.Cesar de Sa, 'Numerical modeling oI glass Iorming processes, Eng.Comput., 1986, Vol.3, December.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
glass_forming.proc Mentat procedure Iile to run the above example
glass_bottle_geometry.mfd Associated model Iile
glass_bottle_material.f User subroutine
564 Marc Users Guide: Part I
CHAPTER 2.14
Chapter 2.15: Marc Adams MNF Interface
2.15
Marc Adams MNF Interface

Chapter Overview 566

Generation of an MNF for HDD HSA Suspension Arm 566

Local Model and Analysis 568

Input Files 572

566 Marc Users Guide: Part I
CHAPTER 2.15
Chapter Overview
This chapter demonstrates the generation oI a Modal Neutral File (MNF) Ior the Marc-Adams interIace. Adams/Flex
allows Ilexible components to be included into Adams models through MNFs that represent the Ilexible components.
Marc- is capable oI generating MNFs that can be integrated into the Adams model. Generating an MNF Irom Marc-
is based on perIorming the most general method oI component mode synthesis techniques, namely the Craig-Bampton
method. Craig-Bampton analysis oI a Hard Disk Drive (HDD) Head Stack Assembly (HSA) suspension arm is used
as an example in this chapter. The HSA arm MNF is later uploaded into an Adams HDD model.
Generation of an MNF for HDD HSA Suspension Arm
Problem Description
A hard disk drive is a complex electromechanical device that employs many technologies. It mainly consists oI a
printed circuit board to communicate with the computer motherboard, a stack oI disks (the storage media), a spindle
motor to rotate the disks, a stack oI recording heads, a suspension arm to carry the head stack, an actuator to move the
head stack assembly to the target data tracks, all contained within a sealed enclosure as shown in Figure 2.15-1.
Figure 2.15-1 HDD Main Components
HDD usually contain more than one disk in a stacked assembly. Data is written onto each disk surIace (top and bottom)
by a separate recording head, so an HDD with Iive disks will normally have ten separate recording heads. Recording
heads are miniature electromagnets that are bonded to a metal suspension gimbal, which is a small arm that holds the
head in position above or beneath a disk. Sets oI gimbals stacked together Ior installation in a disk drive are called a
head-stack assembly (HSA).
When the disks spin up to operating speed, air Ilows at high speeds causing the recording heads to Ily over the surIace
oI the disks. Heads are said to be riding on an 'air bearing. The separation between the recording heads and the disks
during operation, known as the Ilying height, is one oI the important design parameters that controls the perIormance
and durability oI an HDD. The Ilying height in present-day HDD is in the range oI 0.2 - 0.8 micro inches (or 5 - 20
nanometers) and is getting lower as technology progresses. In order to increase the recording density, it is necessary
to decrease the Ilying height so that the signal to noise ratio obtained Irom the read element is within an acceptable
range. Thus, zero spacing is preIerred. However, zero spacing or contact recording would lead to higher Iriction and

Spindle Motor
Disc Stack
Actuator
HSA Suspension
PCB
Arm
567
CHAPTER 2.15
Marc Adams MNF Interface
wear at the head-disk interIace, hence degrading the perIormance oI the HDD. Ideally, the designed Ilying height
should be maintained during operation. In reality, partial contact between the head and disk may occur. Also, vibration
and shock become more oI a concern. Among the controlling Iactors over the interactions between the head and the
disk during Ilying are the suspension arm and disk geometries, materials, and tolerances used in the industry. Due to
reasons outlined above, it is important to study the vibration characteristics oI the HSA suspension arm. In this
example, we are interested in the dynamics oI the HSA suspension arm shown in Figure 2.15-2. This HSA arm
supports ten recording heads.
Figure 2.15-2 HSA Suspension Arm Geometry
HSA Suspension Arm Model
The Iinite element model Ior HSA suspension arm is shown in Figure 2.15-3. It consists oI 8534 brick and shell
elements (element types 7 and 75). The model Iile hdd_hsa_arm.mfd contains the geometry and Iinite element
model. In the Iollowing, we complete the model by adding the necessary boundary conditions and loadcases to
perIorm the Craig-Bampton analysis and generate the MNF. The Craig-Bampton analysis consists oI computing two
sets oI mode shapes: the constraint modes and the Iixed-boundary normal modes.
Figure 2.15-3 HSA Suspension Arm Model
568 Marc Users Guide: Part I
CHAPTER 2.15
Local Model and Analysis
To open the model:
AIter opening the model and examining it, the Iirst step in perIorming the Craig-Bampton analysis is to deIine the
boundary or attachment degrees oI Ireedom that will connect the arm to the rest oI the Adams HDD model. The
boundary degrees oI Ireedom are used to compute the constraint modes as the static shapes obtained by giving each
boundary degree oI Ireedom a unit displacement while holding all other boundary degrees oI Ireedom Iixed.
The axis oI the actuator that drives the arm is at the centerline oI the cylindrical hole oI the arm. To ease the attachment
oI the arm when the MNF is uploaded into the Adams HDD model, an extra node is deIined in the Iinite element model
at the center oI the cylindrical hole and an RBE2 is used to couple all the nodes within the cylindrical surIace to the
node at the center. The six degrees oI Ireedom oI this node are used as boundary degrees oI Ireedom.
The air bearing and contact between the heads and the disks can be represented in the Adams model by springs acting
in the z-direction normal to the planes oI the arm leaIs. Thus, the z degree oI Ireedom oI ten nodes at the location oI
the ten heads on the arm are also used as boundary degrees oI Ireedom.
In Mentat, the boundary degrees oI Ireedom can be deIined in the CRAIG-BAMPTON NODES menu, shown in
Figure 2.15-4, under the MECHANICAL BOUNDARY CONDITIONS menu. AIter deIining the boundary degrees oI
Ireedom, the next step is to create an Adams CRAIG-BAMPTON loadcase and make sure that the CRAIG-BAMPTON
NODES boundary conditions are selected in this loadcase. In Mentat, the Adams CRAIG-BAMPTON loadcase,
shown in Figure 2.15-5, is located under the MECHANICAL LOADCASES menu. The number oI Iixed-boundary
normal modes requested is speciIied in this loadcase menu. In this example, 20 modes should be enough to prevent
mode truncation.
Figure 2.15-4 CRAIG-BAMPTON NODES Menu
FILES
OPEN
hdd_hsa_arm.mfd
OK
569
CHAPTER 2.15
Marc Adams MNF Interface
Figure 2.15-5 Adams CRAIG-BAMPTON Loadcase Menu
The next step is to select the Adams CRAIG-BAMPTON loadcase in the JOBS menu. We should also choose the
LUMPED MASS option Irom the ANALYSIS OPTIONS. From the Adams JOB RESULTS menu, shown in
Figure 2.15-6, we can pick whether or not stress and/or strain modes should be computed. In the same menu, we should
indicate the units used to create the model. In this example, slug, pound-Iorce, inch, and second are used.
Figure 2.15-6 Adams JOB RESULTS Menu
Follow the steps described below to complete the model deIinition:
MAIN
BOUNDARY CONDITIONS
NEW
570 Marc Users Guide: Part I
CHAPTER 2.15
MECHANICAL
MORE
CRAIG-BAMPTON NODES
DISPLACEMNT X
DISPLACEMNT Y
DISPLACEMNT Z
ROTATION X
ROTATION Y
ROTATION Z
OK
NODES
ADD
13771 #
NEW
CRAIG-BAMPTON NODES
DISPLACEMNT Z
OK
NODES
ADD
13012 13055 13120 13121 13212
13213 13305 13306 13397 13398 #
MAIN
LOADCASES
MECHANICAL
Adams CRAIG-BAMPTON
# MODES
20
OK
MAIN
JOBS
MECHANICAL
lcase1
ANALYSIS OPTIONS
LUMPED MASS
OK
JOB RESULTS
571
CHAPTER 2.15
Marc Adams MNF Interface
SuccessIul job completion and generation oI the MNF is indicated by Exit Number 3018. The post Iile could be opened
in Mentat to check the Craig-Bampton mode shapes. The MNF hdd_hsa_arm_mnf_job1.mnf is created in the job
directory. The MNF can now be uploaded into Adams HDD models to represent the Ilexible HSA suspension arm.
Figure 2.15-7 and Figure 2.15-8 show the results oI an Adams simulation in which the generated MNF is used. Flying
height design and parametric studies can be perIormed in the Adams simulation.
Adams
STRESS
STRAIN
SLUG
POUND
INCH
OK (three times)
FILES
SAVE
AS
hdd_hsa_arm_mnf.mfd
OK
To run the job:
MAIN
JOBS
RUN
RESET
SUBMIT (1)
MONITOR
OK
572 Marc Users Guide: Part I
CHAPTER 2.15
Figure 2.15-7 Adams Results for the Vertical Displacement for the Arm's Upper-leaf due to an Input
Figure 2.15-8 Adams Results for the Stress Distribution in the HSA Arm
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
hdd_hsa_arm.proc Mentat procedure Iile to run the above example
hdd_hsa_arm.mud Associated model Iile

Chapter 2.16: Analysis of Stiffened Place Using Beam and Shell Offsets
2.16
Analysis of Stiffened Plate
Using Beam and Shell Offsets

Chapter Overview 574

Analysis of Beam Reinforced Shell Structure using Offsets 574

Input Files 582

574 Marc Users Guide: Part I
CHAPTER 2.16
Chapter Overview
In many Iinite element analyses using beams and shells, it is common to model the beams and shells at a geometric
location that is diIIerent Irom the actual physical location. Such cases are common when shells or beams oI varying
thicknesses are adjacent to each other and the top/bottom shell surIaces or beam Ilanges are to be aligned with each
other. It is convenient to model all the shell nodes at the midsurIace oI one oI the shells or the beam nodes at the neutral
axis oI one oI the beams. The alignment oI the top shell surIaces or beam Ilanges is then achieved by providing a
suitable shell or beam oIIset to the elements. Another common instance is when beams are used as stiIIeners Ior shells.
It is most convenient to model the beam elements at the mid-surIace oI the shell and sharing the shell nodal
connectivity. The Iact that the beam is actually oIIset by halI the plate thickness and halI the height oI the beam section
is achieved by providing a suitable beam oIIset.
There are two methods by which beam/shell oIIsets can be modeled in Marc:
The Iirst method is to place the beams and shells at the actual oIIset position and then tie the nodes oI these
elements back to the original position through manually deIined RBE2 links. While this method is quite
accurate, it is quite cumbersome Ior large models. Furthermore, iI the oIIset elements have to contact other
bodies, it is not possible since all degrees oI Ireedom oI the oIIset element nodes are already tied through the
RBE2 links.
The second method is to use the in-built beam/shell oIIset capability in Marc. This chapter demonstrates the
various Ieatures available in Marc to analyze beam/shell structures with in-built beam/shell oIIsets. The RBE2
approach is only used to compare the accuracy oI the solution obtained using in-built beam/shell oIIsets and
the emphasis in this chapter is placed on describing the setup and solution using the actual in-built beam/shell
oIIset capabilities oI Marc. The dat Iiles Ior the beam-shell oIIset approach (e3x43a.dat) and Ior the RBE2
approach (e3x43b.dat) are also in the demo directory.
Analysis of Beam Reinforced Shell Structure using Offsets
An overhanging Ilat shell that is reinIorced by beams is subjected to a top Iace load. The plate has a variable thickness
along the length and the top surIaces oI the thick and thin shells are aligned at the same level. The top portion oI the
reinIorcement beam cross-sections are welded to the bottom surIace oI the thicker plate. In the geometric model, all
the elements are modeled at the midsurIace oI the thicker shell. Suitable beam/shell oIIsets need to be provided to
account Ior the diIIerence between the geometric model and the physical model.
The Iinite element mesh oI the beam-plate structure is shown in Figure 2.16-1 and Figure 2.16-2. The physical model
with the beams and shells at their actual oIIset locations is displayed in Figure 2.16-1. This model can be used with
RBE2 links set up between the oIIset beams and the shell. The geometric model where the beams are at the shell
midsurIace and in-built beam/shell oIIsets are used, see Figure 2.16-2. The beams and shells with a solid cross-section
in the Iigures clearly indicate where they are modeled in the two cases.
The plate is oI length 6000 mm and width 4000 mm. The plate has a variable thickness along the length (70 mm over
the Iirst 4000 mm and 35 mm over the remaining 2000 mm). The top surIaces oI the thick and thin shells are aligned
at the same level. One reinIorcement beam (beam 1 in Figure 2.16-1) with a cross-sectional radius oI 100 mm and
thickness oI 25 mm is placed along the plate width at the point where the plate thickness transition occurs. Two other
575
CHAPTER 2.16
Analysis of Stiffened Plate Using Beam and Shell Offsets
reinIorcement beams (beam 2 and beam 3 in Figure 2.16-1), each with a cross-sectional radius oI 125 mm and
thickness oI 40 mm, are placed along the length on either side oI the plate. The top portion oI the beam cross-sections
are welded to the bottom surIace oI the plate.
Figure 2.16-1 Finite Element Mesh showing Physical Beam-shell Model
Figure 2.16-2 Finite Element Mesh showing Geometric Beam-shell Model
Note: The beam-shell oIIsets are modeled here using RBE2 links
Note: The beam-shell oIIsets are modeled using the in-built oIIset Ieatures oI Marc.
576 Marc Users Guide: Part I
CHAPTER 2.16
Procedure File
The analysis has been completely set up Irom Mentat. The procedure Iile to demonstrate the example is called
bmshloffset.proc under mentat2010/examples/marc_ug/s2/c2.16.
To run the procedure Iile and build the model Irom start to Iinish, the Iollowing button sequence can be executed
in Mentat:
Mesh Generation
The generation oI the Iinite element mesh is not discussed in detail here. Instead, reIer to the procedure Iile and the
comments in that Iile. Element type 75 is used Ior the shells and element type 14 is used Ior the beams. All shell and
beam elements are modeled at the midsurIace oI the thicker shell. All dimensions are in milli meters. There is a total
oI 180 elements and 176 nodes.
Geometric Properties
The shell thickness is speciIied as 70 mm over the Iirst 4000 mm oI the length and as 35 mm over the next 2000 mm
oI the length. For the thinner shell, an oIIset oI 17.5 mm (oIIset (t1 - t2)/2 where t1 is the thickness oI the thicker
shell and t2 is the thickness oI the thinner shell) is speciIied in order to allow the top shell surIace to be aligned with
that oI the thicker shell.
For beam 1, the beam radius is speciIied as 100 mm and the thickness as 25 mm. An oIIset oI -135 mm (oIIset (t1/2
r1) where t1 is the thickness oI the thicker shell and r1 is the radius oI beam 1) is speciIied in the global Z-direction
in order to allow the apex oI the beam cross-section to be aligned with the bottom surIace oI the shell.
UTILS
PROCEDURES
EXECUTE
bmshloffset.proc
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS
3-D
SHELL
THICKNESS
35
[] USE OFFSETS
OFFSET
17.5
577
CHAPTER 2.16
Analysis of Stiffened Plate Using Beam and Shell Offsets
For beam 2, the beam radius is speciIied as 125 mm and the thickness as 40 mm. The LOCAL(SHELL) option is used
to speciIy the oIIset. In this case, only the oIIset magnitude is speciIied and the oIIset vector is along the normal to the
associated shell element at the node. Since the shell normal is in the global Z-direction, an oIIset magnitude oI -160
mm (oIIset (t
1
/2 r
2
) where t
1
is the thickness oI the thicker shell and r
2
is the radius oI beam 2) is speciIied in order
to allow the apex oI the beam cross-section to be aligned with the bottom surIace oI the shell.
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS
3-D
GENERAL BEAM
THICKNESS
25
RADIUS
100
VECTOR DEFINING LOCAL X AXIS
X
1
BEAM-SHELL OFFSETS >
[] USE OFFSETS
OFFSET VECTOR AT NODE 1
V GLOBAL
Z
-135
COPY 1 TO 2
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS
3-D
GENERAL BEAM
THICKNESS
40
RADIUS
125
VECTOR DEFINING LOCAL X AXIS
Y
1
BEAM-SHELL OFFSETS >
578 Marc Users Guide: Part I
CHAPTER 2.16
For beam 3, the beam radius is speciIied as 125 mm and the thickness as 40 mm. The LOCAL(BEAM) option is used
to speciIy the oIIset. In this case, the oIIset vector is along the local beam coordinate system. The local Z axis is along
the beam (Irom node 1 to node 2), the local X axis is deIined by the user on the same menu and the local Y axis is
deIined by the cross-product oI Z and X. Since the local Z axis oI beam 3 is (-1,0,0) and the local X axis is deIined as
(0,-1,0), the local Y axis oI the beam comes out as (0,0,1). An oIIset oI -160 mm (oIIset t
1
/2 r
3
) where t
1
is the
thickness oI the thicker shell and r3 is the radius oI beam 3) is speciIied along the local Y axis in order to allow the
apex oI the beam cross-section to be aligned with the bottom surIace oI the shell.
[] USE OFFSETS
OFFSET VECTOR AT NODE 1
V LOCAL(SHELL)
X
-160
COPY 1 TO 2
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS
3-D
GENERAL BEAM
THICKNESS
40
RADIUS
125
VECTOR DEFINING LOCAL X AXIS
Y
-1
BEAM-SHELL OFFSETS >
[] USE OFFSETS
OFFSET VECTOR AT NODE 1
V LOCAL(BEAM)
Y
-160
COPY 1 TO 2
579
CHAPTER 2.16
Analysis of Stiffened Plate Using Beam and Shell Offsets
Note that the data entered Ior local X axis on the menu has a close bearing on the local beam coordinate system and,
in turn, on the oIIset vector components speciIied using the LOCAL(BEAM) option. A visual check Ior the correctness
oI the local beam coordinate system can be obtained as Iollows:
Figure 2.16-3 Orientation of Beams
Material Properties
Isotropic, elastic-perIectly plastic material properties are deIined Ior all elements in the model. Young`s modulus is
deIined as 2.1e4 N/mm
2
, Poisson`s ratio as 0.3 and initial yield stress as 40 N/mm
2
.
Boundary Conditions
The nodes at the leIt edge oI the thicker shell are Iixed in all translations and rotations. The top surIace oI the shell is
subjected to a Iace load oI 7.5e-3 N/mm
2
.
GEOMETRIC PROPERTIES
PLOT SETTINGS
BEAM >
BEAM ORIENTATION
DRAW X-Y AXES
580 Marc Users Guide: Part I
CHAPTER 2.16
Loadcase Definition
A mechanical loadcase is deIined to conduct the analysis. Adaptive Stepping MULTI-CRITERIA (Auto Step) is used
Ior the time stepping. Convergence testing is done on both residuals and displacements with tolerance oI 0.01.
Job Parameters
The large strain additive (PLASTICITY,3) procedure is chosen. Select the FOLLOWER FORCE parameter in order to
allow the pressure load Iollow the geometry. The layer von Mises stress and equivalent plastic strain are requested.
Use 5 layers Ior the shell element. Note that Ior the beam element, 16 layers are used by deIault Ior the circular
cross-section. Layer results are requested at layers 1,3,5 (outer,midbottom layers oI shell), 1,9 (layers at beam neutral
axis), and 5,13 (layers at extreme Iibers oI beam).
Results and Discussion
The variation oI the Z component oI the displacement with time is plotted in Figure 2.16-4. The results obtained Irom
the in-built oIIset Iormulation at the center oI the Iree edge oI the thinner shell are compared with the corresponding
RBE2 solution at the same location. The results are nearly identical to each other. It should be noted that Ior the oIIset
solution, only the displacements at the original user-speciIied location are available on the post Iile.
BOUNDARY CONDITIONS
MECHANICAL
FACE LOAD
PRESSURE
7.5e-3
FACES ADD
ALL:
TOP
581
CHAPTER 2.16
Analysis of Stiffened Plate Using Beam and Shell Offsets
Figure 2.16-4 Displacement Z Variation with Time at Center of Free Edge of Shell
The layer 1 equivalent von Mises stress contours obtained Ior the oIIset solution are plotted in Figure 2.16-5. It should
be noted that while calculating elemental quantities like strains, stresses, and associated nodal quantities like reaction
Iorces, elements, and nodes are taken in the actual physical location by applying appropriate oIIset values. It should
also be noted that the contour bands shown in the Iigure are based on the translated values at the element integration
points and with nodal averaging turned oII. This avoids smearing oI the quantities between shells and beams at
common nodal locations. Results obtained Irom the RBE2 solution are identical and are not shown here.
582 Marc Users Guide: Part I
CHAPTER 2.16
Figure 2.16-5 Deformed Configuration and Equivalent Stress Contours for
In-Built Beam/Shell Offset Model
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
bmshloffset.proc Mentat procedure Iile to run the above example
Chapter 2.17: 3-D Tetrahedral Remeshing with Boundary Conditions
2.17
3-D Tetrahedral Remeshing
with Boundary Conditions

Chapter Overview 584

Simulation Examples 584

Rubber Seal Insertion 595

Input Files 604

584 Marc Users Guide: Part I
CHAPTER 2.17
Chapter Overview
This chapter demonstrates the new capability Ior a user to assign boundary conditions to a remeshing body in 3-D.
Remeshing with boundary condition is available in 2-D aIter Marc 2007r1 release. The same Ieature is now extended
to 3-D tetrahedral remeshing.
The boundary conditions can be applied to nodes or element Iaces. They can also be applied to geometry entities such
as points, surIaces or curves (limited to curves with prescribed nodal displacements and temperatures) providing the
geometry entities are attached to the mesh. The Iollowing boundary conditions are tested in the development:
point loads
nodal displacement/temperature/Ilux
Iace distributed load/Ilux
multiple remeshing bodies and boundary conditions
boundary conditions in thermal-mechanical coupled analysis
curves with Iixed displacements
The Iollowing are the limitations to this Ieature:
maximum two boundary conditions can be assigned to the same element Iace
maximum 99 surIaces can be used in geometry attachment
boundary conditions can only be assigned to the boundary, not to the interior oI the remeshing body
element types are restricted to 157 and 134 tetrahedral elements
new table style input Iormat is required
Simulation Examples
By allowing boundary conditions to be used in a remeshing body, it creates many possibilities in simulations. The
Iollowing applications demonstrate some oI these possibilities.
Pressure on a Rubber Cylinder
This example shows a pressure applied to a circular area on the top oI a rubber cylinder. As the pressure increases, the
rubber deIorms to such an extent that remeshing is necessary. As the result, the pressure boundary condition is
transIerred to the new mesh and the simulation continues.
585
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Figure 2.17-1 Pressure Boundary Condition
Figure 2.17-2 Effective Stress Display on a Cutting Plane
586 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-3 Pressure showing here as an External Force Vector after Remeshing
Metal Compression with Prescribed Displacements
Metal compression is normally simulated with rigid die and punch. In this example, simple prescribed nodal
displacements and temperatures are used to demonstrate the capability oI remeshing with boundary conditions in a
thermal-mechanical coupled analysis. This example shows multiple boundary conditions assigned to the same
element Iaces.
Figure 2.17-4 Prescribed Nodal Displacement and Temperature at 20
587
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Figure 2.17-5 Temperature Distribution at the End of Simulation
Rubber Ring Seal with Pressure Testing after Compression
A section oI rubber ring is compressed and then pressured on one side to test possible leakage. This example shows
applications oI geometry attachment with pressure boundary conditions and suppression oI pressure on element Iaces
that are in contact. The geometry attachment is applied to an area that will have pressure aIter compression and is
transIerred to new mesh aIter remeshing. When pressure is applied to this geometry, only element Iaces that are not in
contact are subjected to this pressure. In the Iollowing Iigures, the geometry attachment is shown in red color.
Figure 2.17-6 Pressure is Pre-Applied to the Attached Surface
588 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-7 Geometry Attachment is Transferred to the New Mesh correctly
Figure 2.17-8 Pressure is Applied Automatically to the Element Faces that are not in Contact
589
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Tube Hydro-forming
Hot hydro-Iorming with thick tubes requires remeshing. The tube is subjected to an internal pressure on the inner
surIace, a Iixed nodal displacement on both ends, and a symmetry boundary condition on the symmetry surIaces.
Thermal-mechanical coupled analysis is assumed with initial temperature at 1000C and 500C in the rigid die.
Figure 2.17-9 Boundary Conditions
Figure 2.17-10 Total Equivalent Plastic Strain at an Intermediate Stage
590 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-11 Temperature Distribution at the Final Stage
Rubber Seal Insertion
In using boundary conditions, we can avoid having to use rigid contact surIaces to apply symmetry conditions on a
remeshing body. This well-shown rubber seal example is simulated now without symmetry surIaces and a rigid surIace
to push the rubber seal.
Geometry attachment with prescribed displacement is used to push the rubber seal. This example is demonstrated in
detail later in the Marc User's Guide.
Figure 2.17-12 Boundary Conditions
591
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Figure 2.17-13 Final Deformation showing Geometry Attachment in Red
Rubber Seal and Steel Interaction
This example demonstrates multiple deIormable bodies in contact and the remeshing with pressure boundary
conditions. The rubber between a steel plate and a steel tube is under pressure and pushed against the steel tube. This
causes large deIormation in the rubber, so the remeshing is required.
Figure 2.17-14 Boundary Conditions
592 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-15 End of Deformation
Figure 2.17-16 Pressure showing as External Force Vectors after Remeshing
593
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Glass Forming
Glass Iorming is another type oI application. With the internal pressure, this example is simulating a blow Iorming
process oI a glass container. Thermal-mechanical coupled analysis with rigid-plastic material model is assumed.
Figure 2.17-17 Boundary Conditions
Figure 2.17-18 Temperature Distribution at an Intermediate Stage
594 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-19 Final Deformation
Rubber Bars with Prescribed Displacement on Curves
This example demonstrates curve attachments to a 3-D rubber bar. A prescribed displacement boundary condition is
applied to the curves. These curves are attached to some element edges. Shown in the Iollowing pictures are curve
attachments beIore and aIter the remeshing in the red color.
Figure 2.17-20 Boundary Condition applied to Curves with Attachment
595
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Figure 2.17-21 Curve Attachments after Remeshing and Deformation
Rubber Seal Insertion
A rubber seal with a rectangular cross section (1.8x1.2 cm
2
) is compressed laterally by a prescribed displacement
boundary condition. Because oI the symmetry, only a halI oI the seal is considered. With a thickness oI 0.2 cm, the
model is setup as a 3-D problem. Assuming this is a long rubber seal in the thickness direction, additional two
symmetry surIaces are used. The three symmetric surIaces are constrained with boundary conditions applying to the
nodes and the moving surIace is simulated by an attached surIace with a prescribed displacement.
The rubber seal is modeled using Mooney constitutive model. The material parameters are given as C18N/cm
2
and
C22N/cm
2
. The bulk modulus is 10000N/cm
2
.
The analysis starts with a hexahedral mesh. AIter immediate remeshing, the hexahedral element is converted into
tetrahedral elements. In the rest oI the analysis, the remeshing/rezoning is done based on the strain change check to
prevent severe element distortion. An adaptive meshing based on the surIace curvature is used to generate smaller
elements near the curved areas. It allows the analysis to capture the geometry changes correctly in those areas without
creating excessive number oI the elements to slow down the analysis. Element type 157 is used in the analysis within
the updated Lagrangian Iramework.
Model Generation
We will start with a pre-deIined model Iile and concentrate on the applications oI the new Ieatures. A model Iile
initial_setup.mId is read as Iollows:
FILE
OPEN
Open file
initial_setup.mfd
OK
596 Marc Users Guide: Part I
CHAPTER 2.17
In the model Iile, most oI the basic inIormation is already provided. We will concentrate on the Iollowing:
1. Attach a surIace to element Iace
2. Boundary conditions
3. Global remeshing criteria
4. Loadcase
5. Job submission
6. Results
First oI all, reset some plot controls so that we have a better view oI the model (Figure 2.17-22).
Figure 2.17-22 Initial Model Setup
PLOT
NODES (click to unselect)
POINTS (click to unselect)
ELEMENTS
SOLID (click to select)
DRAW
DYN.MODEL (set dynamic modeling on and rotate model
to a better position)
597
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Attach a Surface to Element Faces
In order to demonstrate boundary conditions assigned to a geometry surIace, we need to attach this surIace to some
element Iaces. Here is how:
You can see the attached element Iaces change the color to dark blue. II you preIer, you can change this color to red
(Figure 2.17-23):
Figure 2.17-23 Surface Attachment
MAIN
MESH GENERATION
ATTACH
FACE
SURFACE
Select surface on top (you have to deselect DYN MODEL first)
Select all element faces on the top
END LIST (#)
MAIN
VISUALIZATION
COLORS
ATTACHED FACES (change it to red)
OK
598 Marc Users Guide: Part I
CHAPTER 2.17
Boundary Conditions
The symmetry boundary conditions are set up by applying the proper constraints to the nodes on the surIaces. These
boundary conditions are already done in the initial model. Now, we are going to add a prescribed nodal displacement
condition to the attached surIace that pushes the rubber seal:
DeIine a time table Ior the prescribed displacement:
DeIine prescribed displacement:
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
NAME
pres_y (enter a new name)
TABLE
NEW
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 0 (enter point 1)
1 1 (enter point 2)
SHOW TABLE
SHOW MODEL (back to the model view)
RETURN
FIXED DISPLACEMENT
DISPLACEMENT Y
-1 (enter -1)
TABLE
table1- (select time table)
OK
599
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Assign it to the attached surIace:
II you show all boundary conditions now,
You should have the view similar to Figure 2.17-24.
Figure 2.17-24 Boundary Conditions
Global Remeshing Criteria
We need to deIine global remeshing criteria so that the initial hexahedral mesh is converted to tetrahedral mesh and
remeshing is done whenever the strain change level is reached.
SURFACES
ADD (select the surface)
END LIST (#)
RETURN
ID BOUNDARY CONDS
MAIN
MESH ADAPTIVITY
GLOBAL REMESHING CRITERIA
PATRAN TETRA
IMMEDIATE (on)
600 Marc Users Guide: Part I
CHAPTER 2.17
Loadcase
DeIine a loadcase to push the rubber seal.
ADVANCED
STRAIN CHANGE
0.4
OK
#ELEMENTS
SET
1000
ADVANCED
CURVATURE CONTROL
10
CHANGE ELEMENT TYPE
157
OK
OK
REMESH BODY
rubber (select rubber)
MAIN
MECHANICAL
STATIC
LOADS (select all boundary conditions)
GLOBAL REMESHING (select remeshing criterion)
SOLUTION CONTROL
MAX # RECYCLES
20
MIN # RECYCLES
2
Contribution of initial stress
TENSILE STRESS
OK
CONVERGENCE
RESIDUALS OR DISPLACEMENTS
RELATIVE FORCE TOLERANCE
601
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Job Submission
It is important to select NEW STYLE TABLE Iormat Ior this analysis.
0.1
RELATIVE DISPLACEMENT TOLER.
0.01
OK
TOTAL LOADCASE TIME
0.5
CONSTANT TIME STEP
0.01
#STEPS
50
OK
MAIN
MECHANICAL
Lcase1 (select load case 1)
INITIAL LOAD (select all boundary conditions)
ANALYSIS OPTIONS
LARGE DISPLACEMENT (on)
Rubber elasticity procedure
LARGE STRAIN-UPDATED LAGRANGE
OK
JOB RESULTS
Select Cauchy stress for element output
OK
OK
RUN
NEW-STYLE TABLES - (select new format)
SUBMIT(1) (run job)
602 Marc Users Guide: Part I
CHAPTER 2.17
Results
Results can be viewed and compared with others using the contact bodies.
First, we can check iI all boundary conditions are transIerred to the new mesh aIter each remeshing step.
Figure 2.17-25 shows the tetrahedral mesh aIter immediate remeshing and Figure 2.17-26 shows the mesh at the Iinal
step. You can see that boundary conditions are transIerred correctly aIter about nine remeshing steps.
Figure 2.17-25 After Converting to Tetrahedral Mesh in the First Remeshing
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY - (show deformed shape)
Note: The attached element Iaces are shown in red.
603
CHAPTER 2.17
3-D Tetrahedral Remeshing with Boundary Conditions
Figure 2.17-26 Mesh at the Last Increment
The Cauchy equivalent stress can be seen in Figure 2.17-27 and Figure 2.17-28.
Figure 2.17-27 Equivalent Cauchy Stress at Increment 25
604 Marc Users Guide: Part I
CHAPTER 2.17
Figure 2.17-28 Equivalent Cauchy Stress at Increment 50
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
tet_remeshbc.proc Mentat procedure Iile to run the above example
initial_setup.mfd Associated model Iile
Chapter 2.18: Induction Heating of a Tube
2.18
Induction Heating of a Tube

Chapter Overview 606

Heating of a Tube 606

Input Files 618

606 Marc Users Guide: Part I
CHAPTER 2.18
Chapter Overview
This chapter describes the use oI coupled magnetodynamic-thermal analysis in Marc. With this coupling, procedure
induction heating type analyses can be perIormed. The implementation in Marc Iollows a staggered approach. First, a
harmonic magnetodynamic analysis is perIormed Iollowed by a thermal analysis. The harmonic magnetodynamic
Iield generates induction currents in the workpiece. From these induced currents a heat Ilux is computed, which is then
used in the thermal analysis. The thermal analysis can be either a time dependent, or a steady state solution.
Heating of a Tube
In this example, an iron tube is heated by six coils. The upper part oI the workpiece is placed inside the coils.
Figure 2.18-1 shows the model where axisymmetry is considered around the x-axis. A complete description oI this
example can be Iound in |ReIerence 1|.
Figure 2.18-1 Axisymmetric Representation of the Tube Surrounded by Coils
Mesh Generation
The geometry was generated previously, and is read in as an mfd Iile. The model contains the tube and the coils.
SuIIicient space around the tube and coils is meshed to capture the magnetodynamic Iield correctly. The mesh oI the
tube coils and air directly surrounding it is reIined. The surrounding air is meshed more coarsely in order to increase
computational eIIiciency. The curves in the geometry already have divisions assigned to them.
FILE
NEW
OK
RESET PROGRAM
OPEN
tube_geom.mfd
OK
coils
tube
607
CHAPTER 2.18
Induction Heating of a Tube
RETURN
MESH GENERATION
AUTO MESH
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT): QUAD MESH
1 to 11 #
QUADRILATERALS (ADV FRNT): QUAD MESH
12 to 35 #
SELECT
ELEMENTS
1 to 2080 #
ELEMENTS: STORE
tube
OK
ALL: SELECTED
CLEAR SELECT
ELEMENTS
2081 to 2320 #
ELEMENTS: STORE
coils
OK
ALL: SELECTED
RETURN
QUADRILATERALS (ADV FRNT): QUAD MESH
1 to 9 12 to 35 37 to 40 46 47 #
MESH COARSENING PARAMETER: TRANSITION
0.9
QUADRILATERALS (ADV FRNT): QUAD MESH
38 to 45 #
RETURN (twice)
SWEEP
SWEEP: ALL
REMOVE UNUSED: NODES
RETURN (twice)
608 Marc Users Guide: Part I
CHAPTER 2.18
Material Properties
The tube is made oI a non-Ierromagnetic stainless steel X5CrNi 18/9 (1.4301). Temperature dependent material
properties Ior this steel were taken Irom |ReIerence 1|, which are permeability Hm
-1
, permittivity
Fm
-1
, electrical conductivity , with , ,
, and .
Then the thermal conductivity Wm
-1
K
-1
, the mass density , and the
speciIic heat Jkg
-1
K
-1
.
For the surrounding air permeability Hm
-1
, permittivity Fm
-1
, and electrical
conductivity . The thermal conductivity Wm
-1
K
-1
, mass density , and the speciIic
heat Jkg
-1
K
-1
. For the coil, the same material properties as air are taken. An emissivity oI 0.4 is taken Ior
the tube.
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NAME
air
HEAT TRANSFER
CONDUCTIVITY
0.024
SPECIFIC HEAT
1000
MASS DENSITY
1.3
OK
MORE
MAGNETODYNAMIC
PERMEABILITY
1.25e-6
PERMITTIVITY
8.854e-12
OK
PREVIOUS
ELEMENTS ADD

0
1.25 10
6

c 1 o T ( )
1
a b T c T
2
d T
3

-------------------------------------------------------------- O
1
m
1
a 4.9659 10
7
b 8.4121 10
10

c 3.7246 10
13
d 6.196 10
17

T ( ) 100 0.11215 1.4087 10
4
( ) 7900
C T ( ) 1000 0.3562 0.988 10
4
( )

0
1.25 10
6
c 8.854 10
12

o 0 O
1
m
1
0.024 1.3
C 1000
609
CHAPTER 2.18
Induction Heating of a Tube
ALL EXISTING
NEW
NAME
steel
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
tcond
TYPE
temperature
INDEPENDENT VARIABLE V1: MAX
1000
FORMULA
ENTER
100e0*(0.11215+1.4087e-4*v1)
VARIABLES: FIT
NEW
1 INDEPENDENT VARIABLE
NAME
htcap
TYPE
temperature
INDEPENDENT VARIABLE V1: MAX
1000
FORMULA
ENTER
1e3*(3.562e-1+0.988e-4*v1)
VARIABLES: FIT
NEW
1 INDEPENDENT VARIABLE
NAME
sigma
TYPE
temperature
INDEPENDENT VARIABLE V1: MAX
610 Marc Users Guide: Part I
CHAPTER 2.18
1000
FORMULA
ENTER
1./(4.9659e-7+8.4121e-10*v1-3.7246e-13*v1^2+6.196e-17*v1^3)
VARIABLES: FIT
RETURN
HEAT TRANSFER
CONDUCTIVITY
1
CONDUCTIVITY: TABLE
tcond
SPECIFIC HEAT
1
SPECIFIC HEAT: TABLE
htcap
MASS DENSITY
7900
EMISSIVITY
0.4
OK
MORE
MAGNETODYNAMIC
PERMEABILITY
1.25e-6
PERMITTIVITY
1
CONDUCTIVITY
1
CONDUCTIVITY: TABLE
sigma
OK
ELEMENTS ADD
ALL: SET
tube
OK
RETURN (twice)
611
CHAPTER 2.18
Induction Heating of a Tube
Radiation
In this example, heat loss due to radiation is also taken into account. This is activated in Marc by deIining an open
cavity. In Mentat, the cavity is deIined in MODELING TOOLS, and activated in BOUNDARY CONDITIONS.
Initial Conditions and Boundary Conditions
The initial temperature oI all the nodes in the model, and the sink temperature oI the radiating cavity are set to 20C.
At the outer boundary, the magnetic potential and the electric potential are set to zero. In |ReIerence 2|, a total current
oI 1293 A Ilows in each coil. This is a net or eIIective current in a coil and can be represented as a coil current density
in the Iollowing way. The cross-sectional area oI each oI the coils is . Then the magnitude oI the current
density is,
.
In the axisymmetric model, this current points in the z-direction. It is also possible to apply a point current to all the
nodes oI the coils. The magnitude oI the point current Ior each node is then
where is the total number oI nodes in any oI the coils.
MODELING TOOLS
CAVITIES
NEW
CURVES: ADD
1 TO 9 #
RETURN (twice)
INITIAL CONDITIONS
THERMAL
TEMPERATURE
TEMPERATURE (TOP)
20
OK
NODES: ADD
ALL: EXIST
RETURN (twice)
BOUNDARY CONDITIONS
NAME
A 5 10
5
m
2

J
1293
5 10
5

-------------------- 2.586 10
7
Am
2

I
1293
n
----------- -
n
612 Marc Users Guide: Part I
CHAPTER 2.18
fix_A
MAGNETODYNAMIC
HARMONIC BCs
FIX MAGNETIC POTENTIAL
POTENTIAL X
POTENTIAL Y
POTENTIAL Z
OK
CURVES: ADD
43 44 45 #
NEW
NAME
load
VOLUME CURRENT
CURRENT Z
2.586e7
OK
ELEMENTS: ADD
ALL: SET
coils
OK
NEW
NAME
fix_E
FIX ELECTRIC POTENTIAL
POTENTIAL
OK
CURVES: ADD
43 44 45 #
RETURN (twice)
NEW
NAME
radiation
THERMAL
CAVITY RADIATION
RADIATION
613
CHAPTER 2.18
Induction Heating of a Tube
Loadcases and Job Parameters
A transient analysis is perIormed with a Iixed time step. The loading consists oI two stages, in the Iirst 25 seconds the
workpiece is heated, and in the second 10 seconds a temperature relaxation takes place without heating. The time step
used is 0.5 seconds, and the excitation Irequency is 10 kHz. The axisymmetric magnetodynamic element 112 is
selected Ior all the elements.
CAVITY STATUS: OPEN
SINK TEMPERATURE
20
VIEWFACTORS: CALCULATE
OK
CAVITIES: ADD
cavity1
OK
RETURN (twice)
LOADCASES
MAGNETODYNAMIC-THERMAL
NAME
heating
TRANSIENT
FREQUENCY
10000
TOTAL LOADCASE TIME
25
OK
COPY
NAME
relaxation
TRANSIENT
LOADS
load (deselect)
O
K
FREQUENCY
0
614 Marc Users Guide: Part I
CHAPTER 2.18
TOTAL LOADCASE TIME
10
PARAMETERS
# STEPS
20
OK (twice)
RETURN (twice)
JOBS
NAME
induction
ELEMENT TYPES
MAGNETODYNAMIC-THERMAL
AXISYM
112
OK
ALL: EXIST
RETURN (twice)
MOR
E
MAGNETODYNAMIC-THERMAL
heating
relaxation
INITIAL LOADS
icond1
OK
JOB RESULTS
1st Real Component Magnetic Induction
2nd Real Component Magnetic Induction
3rd Real Component Magnetic Induction
1st Imag Component Magnetic Induction
2nd Imag Component Magnetic Induction
3rd Imag Component Magnetic Induction
1st Real Comp Current Density
2nd Real Comp Current Density
3rd Real Comp Current Density
1st Imag Comp Current Density
615
CHAPTER 2.18
Induction Heating of a Tube
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened.
2nd Imag Comp Current Density
3rd Imag Comp Current Density
Temperature
Generated Heat
Electric Current
OK (twice)
FILE
SAVE AS
tube.mud
OK
RETURN
RUN
NEW-STYLE TABLES
SUBMIT(1)
OK
RETURN
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODES
221 161 #
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Time
Temperature
FIT
RETURN
SHOW IDS
0
YMAX
616 Marc Users Guide: Part I
CHAPTER 2.18
Figure 2.18-2 shows the contour plot oI the temperature at the end oI the heating period, and at the and oI the relaxation
period. In |ReIerence 2| at two points, the temperature is measured during the analysis. One point is located at 0.005
m Irom the tip oI the tube, and the other point is located at 0.035 m Irom the tip oI the tube. Figure 2.18-4 shows a
history plot oI the temperatures oI these two points, where a comparison is made with the measured data taken Irom
|ReIerence 2|.
Figure 2.18-2 Contour Plot of the Temperature at the End of the Heating Period,
and After the Relaxation Period
1200
YSTEP
6
617
CHAPTER 2.18
Induction Heating of a Tube
Figure 2.18-3 Contour Plot of the Temperature at the End of the Heating Period Expanded about the Axis
of Revolution - Elements Representing the Air are not Drawn,
Figure 2.18-4 History Plot of the Temperature at Two Nodes on the Workpiece
References
1. C. Chaboudez, S. Clain, R. Glardon, J. Rappaz, M Swierkosz, and R. Touzani, 'Numerical Modelling oI
Induction Heating oI Long Workpieces, IEEE Trans. Magn.,Vol 30, 5026-5037, 1994
618 Marc Users Guide: Part I
CHAPTER 2.18
2. C. Chaboudez, S. Clain, R. Glardon, D.Mari, J. Rappaz, and M Swierkosz, 'Numerical Modeling oI Induction
Heating oI Axisymmetric Geometries, IEEE Trans. Magn.,Vol 33, 739-745, 1997
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
nrepeat.proc Mentat procedure Iile to run the above example
repeat.proc Mentat procedure Iile to run the above example
setup.proc Mentat procedure Iile to run the above example
tube_heating.proc Mentat procedure Iile to run the above example
tube_heating_post.proc Mentat procedure Iile to run the above example
tube_heating_run.proc Mentat procedure Iile to run the above example
tube_geom.mfd Associated model Iile
Chapter 2.19: Magnetostatics with Tables
2.19
Magnetostatics with Tables

Chapter Overview 620

Nonlinear Analysis of an Electromagnet Using Tables 620

Input Files 628

620 Marc Users Guide: Part I
CHAPTER 2.19
Chapter Overview
This chapter describes the use oI tables in a magnetostatic analysis. With this option, a magnetization curve (B-H
relation) can be entered in diIIerent ways. It is set through the ISOTROPIC or ORTHOTROPIC material option, which
contains either the permeability, the inverse permeability, the H-B relation, or the B-H relation. For the H-B relation
and the B-H relation, a table has to be given, where Ior the H-B relation B and B-H relation H is the independent
variable. A magnetization can also be prescribed using the permeability or inverse permeability, where a table has to
be given which depends on either B, or H. A table can be either a set oI data points, or a Iunction. The diIIerent ways
oI deIining a magnetization curve will be illustrated in this example. The magnetic Iield oI an electromagnet is
computed, where magnetization curves are used Ior the material inside oI the magnet.
Nonlinear Analysis of an Electromagnet Using Tables
In this example, an electromagnet is modeled. This is a planar analysis. Figure 2.19-1 shows the model and
its dimensions.
Figure 2.19-1 View of the Electromagnet (Dimensions in m)
The material properties oI the material inside the conductors Iollows magnetization curves, which are prescribed with
tables. In this example, a number oI ways on how this can be done will be demonstrated. A quarter section oI
Figure 2.19-1 is modeled using proper boundary conditions to take care oI the symmetry.
Reading the Model and Adding Material Properties
The mesh was generated previously, and is read in as an mfd Iile. The magnetization relation Ior the material inside
the electromagnet is deIined using the ORTHOTROPIC model deIinition option with tables. The Iollowing equations
are used Ior the magnetization
,
,
Ior part oI the elements,
0.06
0.02 0.04 magnetic material
electric conductor
H
x
B
x
5
150 B
x

H
y
B
y
B
y
1.5 B
y

x
B
x
( )
B
x
H
x
------
B
x
B
x
5
150 B
x

-----------------------------------
1
B
x
4
150
------------------------
621
CHAPTER 2.19
Magnetostatics with Tables
,
Ior another part oI the elements, and
,
.
Ior the remaining elements inside the conductors. In these equations, is the independent variable. The permeability
Ior the electric conductor and the air . Figure 2.19-2 shows the menu Ior selecting the diIIerent
magnetization methods.
Figure 2.19-2 New Menu Layout for Magnetostatic Material Properties
FILE
RESET PROGRAM
OPEN
elmag.mfd
OK
RETURN

y
B
y
( )
B
y
H
y
------
B
y
B
y
B
y
1.5 B
y

---------------------------------------------
1
B
y
1.5
-----------------------
1

x
----- B
x
( ) B
x
4
150
1

y
----- B
y
( ) B
y
1.5
B
1.2566 10
6
Hm
1

Select The Dependent

Variable For Magnetization
622 Marc Users Guide: Part I
CHAPTER 2.19
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NAME
air
MORE
MAGNETOSTATIC
PERMEABILITY
1.2566E-6
OK
ELEMENTS ADD
ALL EXISTING
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
hx
TYPE
magnetic induction
FORMULA
ENTER
v1^5+150.0*v1
NEW
1 INDEPENDENT VARIABLE
NAME
hy
TYPE
magnetic induction
FORMULA
ENTER
abs(v1)*v1+1.5*v1
NEW
1 INDEPENDENT VARIABLE
NAME
mu_Bx
TYPE
magnetic induction
623
CHAPTER 2.19
Magnetostatics with Tables
FORMULA
ENTER
1.0/(v1^4+150.0)
NEW
1 INDEPENDENT VARIABLE
NAME
mu_By
TYPE
magnetic induction
FORMULA
ENTER
1.0/(abs(v1)+1.5)
NEW
1 INDEPENDENT VARIABLE
NAME
invmu_Bx
TYPE
magnetic induction
FORMULA
ENTER
v1^4+150.0
NEW
1 INDEPENDENT VARIABLE
NAME
invmu_By
TYPE
magnetic induction
FORMULA
ENTER
abs(v1)+1.5
RETURN
NEW
NAME
iron_a
624 Marc Users Guide: Part I
CHAPTER 2.19
MAGNETOSTATIC
ORTHOTROPIC
MAGNETIZATION 11 : MAGNETIC FIELD INTENSITY
11 TABLE
hx
MAGNETIZATION 22 : MAGNETIC FIELD INTENSITY
22 TABLE
hy
OK
ELEMENTS ADD
541 to 588
NEW
NAME
iron_b
MAGNETOSTATIC
ORTHOTROPIC
MAGNETIZATION 11 : PERMEABILITY
11 TABLE
mu_Bx
MAGNETIZATION 22 : PERMEABILITY
22 TABLE
mu_By
OK
ELEMENTS ADD
589 to 684
NEW
NAME
iron_C
MAGNETOSTATIC
ORTHOTROPIC
MAGNETIZATION 11 : INVERSE PERMEABILITY
11 TABLE
invmu_Bx
625
CHAPTER 2.19
Magnetostatics with Tables
Boundary Conditions
The potential is set to zero on the outer boundary oI the model, and along the x-axis to support the inverse symmetry.
The potential is leIt Iree along the y-axis. The current density in the electric conductor is .
MAGNETIZATION 22 : INVERSE PERMEABILITY
22 TABLE
invmu_By
OK
ELEMENTS ADD
685 to 732
RETURN (twice)
BOUNDARY CONDITIONS
NAME
current
MAGNETISTATIC
VOLUME CURRENT
CURRENT
2.5e8
OK
ELEMENTS ADD
733 to 924
NEW
NAME
fix
FIXED POTENTIAL
POTENTIAL
OK
NODES ADD
274 to 286 561 to 572 12 325 338 351 1 364 377 390 403 429
442 455 468 481 494 507 520 533 546 559 1725 1779 1 to 13 636
645 654 663 672 681 690 699 708 717 726 735 288 to 299 #
RETURN (twice)
2.5 10
8
Am
2

626 Marc Users Guide: Part I

CHAPTER 2.19
Loadcases and Job Parameters
A steady state analysis is perIormed, the relative convergence tolerance is set to 1e-4.
LOADCASES
MAGNETOSTATIC
STEADY STATE
CONVERGENCE TESTING
RELATIVE CURRENT TOLERANCE
1e-4
OK (twice)
RETURN (twice)
JOBS
ELEMENT TYPES
MAGNETOSTATIC
PLANAR
39
OK
ALL : EXIST
RETURN (twice)
MORE
MAGNETOSTATIC
lcase1
JOB RESULTS
1st Comp Magnetic Induction
2nd Comp Magnetic Induction
3rd Comp Magnetic Induction
1st Comp Magnetic Field Intensity
2nd Comp Magnetic Field Intensity
3rd Comp Magnetic Field Intensity
OK (twice)
627
CHAPTER 2.19
Magnetostatics with Tables
Save Model, Run Job, and View Results
New style tables is selected so that the tables describing the magnetization is used. AIter saving the model, the job is
submitted and the resulting post Iile is opened. Figure 2.19-3 shows the contour plot oI the Iirst component oI the
magnetic induction. Note that the diIIerent magnetization curves used here should all give the same results; so, in this
example, the magnetization curves used Ior the elements can be interchanged, and the computed magnetic induction
will be the same. Only small diIIerences can occur due to the diIIerent ways the magnetization curves are handled.
FILE
SAVE AS
electromagnet.mfd
OK
RETURN
RUN
NEW-STYLE TABLES
SUBMIT(1)
OK
RETURN
RESULTS
OPEN DEFAULT
NEXT
CONTOUR BANDS
SCALAR
1st Comp Magnetic Induction
OK
628 Marc Users Guide: Part I
CHAPTER 2.19
Figure 2.19-3 Contour Plot of the First Component of the Magnetic Induction
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
elmag.proc Mentat procedure Iile to run the above example
elmag.mfd Associated model Iile
elmag.vw View used by procedure Iile
Chapter 2.20: Delamination Crack Propagation
2.20
Delamination and Crack
Propagation

Summary 630

Model Review 631

Results 634

Input Files 636

630 Marc Users Guide: Part I
CHAPTER 2.20
Summary
Title Delamination Crack Propagation
Problem Ieatures Composite ply delamination.
Crack propagation using Virtual Crack Closure Technique (VCCT) and cohesive
zone model
Geometry
Material properties Composite layup: |0 45 90 -45 0|
Orthotropic material:

VCCT:
Cohesive:
Analysis type Quasi-static analysis
Boundary conditions Clamped ends
Applied loads Compression oI the clamped ends, Iixed displacement oI 0.1
Element type Solid composite element type 149. Cohesive element 188
Contact properties Glued contact, deact glue
FE results 1. Plot oI updated crack Iront aIter growth
2. Plot oI damage zone aIter growth
r = 0.1
1x2x0.2 Plate
Units: m
P
P
0.5
E
1
100GPa E
2
E
3
50GPa
G
12
G
31
7GPa G
23
8GPa v
12
v
31
0.3 v
23
0.4
G
c
5
6
10 N m
G
c
7
5
10 N m v
c
0.001 m
Click to play animation (ESC to stop)
631
CHAPTER 2.20
Delamination and Crack Propagation
The delamination in a thick composite structure is studied. Two approaches are used: crack propagation by Virtual
Crack Closure Technique, VCCT, and damage evolution with a cohesive zone model using interIace elements. The
composite has Iour layers, and there is an initial deIect between layers 3 and 4. The structure is loaded in compression
causing buckling oI the part at the initial deIect. The deIect is then allowed to grow using VCCT Ior the Iirst example
and using a cohesive zone model Ior the second. The bottom part oI the structure is modeled with three-layered solid
element with a single element through the thickness. The top part has a single layer and also one element through
the thickness.
The VCCT model deIines the initial deIect by means oI the DEACT GLUE option. The nodes at the deIect should do
regular contact (to avoid penetration). By identiIying them as part oI a DEACT GLUE region, we tell the program to
let them do regular contact even though they are part oI a glued interIace.
In the cohesive zone model, the top and bottom parts do not touch each other directly. A layer oI interIace elements
is placed between the two parts. These elements have the same topology as standard eight-noded bricks. Here, they
have zero thickness in order to model the inIinitely thin region between the composite parts. The top part oI the
composite is glued to the top part oI the interIace elements and the bottom part oI the composite to the bottom part oI
the interIace elements.
In the VCCT case, the two parts are rigidly connected until crack growth occurs. With the interIace elements, there is
an elastic layer between the parts.
The two diIIerent methods are not expected to give the same results. Although both methods can be used Ior studying
this type oI problem, they use quite diIIerent approaches. With VCCT, the parts have a perIect bonding until crack
growth occurs. The user enters a crack growth resistance (G
c
) to indicate when the crack should grow. The cohesive
zone model uses an elastic layer in the interIace. This also inIluences the deIormation oI the structure beIore any
damage occurs. The cohesive energy (also denoted as G
c
) that is input Ior the cohesive material law is related to the
crack growth resistance in the VCCT case in that both have to do with the energy required to split up material. The
way this quantity is used, though, is diIIerent in the two approaches and, Ior this chapter, diIIerent values are used Ior
the VCCT case and the cohesive zone case.
Model Review
The Mentat model Ior the VCCT variant is available in the Iile delam_vcct.mfd. Figure 2.20-1 shows the model with
the diIIerent contact bodies identiIied. The top part has a Iiner mesh in order to accurately describe the deIect region
and allow the crack to grow.
632 Marc Users Guide: Part I
CHAPTER 2.20
Figure 2.20-1 Model for VCCT Calculation
In Figure 2.20-2, we show the bottom side oI the top part, with the DEACT GLUE region and the crack Iront.
Figure 2.20-2 Deact Glue Region and Initial Crack Front
The DEACT GLUE setting can be Iound in Mentat under the menu CONTACT -> CONTACT AREAS and the option to
use is GLUE DEACTIVATION. The setting Ior the crack is in MODELING TOOLS -> CRACKS. Here, we select the
application to be VCCT, and we Iill out the settings Ior the crack propagation in the CRACK PROPAGATION menu
(see Figure 2.20-3). We make sure to set direct growth, release constraints, and enter the crack growth resistance Ior
when crack growth should occur.
Deact Glue Region Initial Crack Front
633
CHAPTER 2.20
Delamination and Crack Propagation
Figure 2.20-3 Mentat menu for crack propagation settings
The model Ior the cohesive zone variant is shown in Figure 2.20-4. The top part is using the same mesh as the interIace
element part. Although this is, strictly speaking, not necessary Ior the contact glue part, we still need a Iine mesh Ior
the deIormation oI the deIect region. It also allows a Iair comparison with the VCCT case. From the Iigure, it is also
clear how the initial deIect is modeled: as a hole in the part with interIace elements.
Figure 2.20-4 Exploded View of Cohesive Zone Model
634 Marc Users Guide: Part I
CHAPTER 2.20
The material properties Ior the cohesive zone model are given in the menu shown in Figure 2.20-5.
Figure 2.20-5 Mentat Menu for Cohesive Zone Model
Both models use the same settings Ior the load stepping: 20 Iixed steps with a maximum oI 20 recycles. DeIaults are
used Ior other control settings.
Results
Figure 2.20-6 shows the deIormed shape Ior the two models. The plate bends down and the region with the deIect
buckles. The VCCT and cohesive variants show a little diIIerence in how much the deIect grows. This, oI course,
changes with the selected values Ior VCCT and the cohesive material.
Figure 2.20-7 shows how much the crack has grown in the VCCT case. The arrows show the x-coordinate oI the local
crack tip system Ior each crack tip. It starts out as a circular crack, and the crack Iront is allowed to grow nonuniIormly.
The VCCT evaluation and crack growth stops as soon as the crack reaches an outer boundary. Note that no
speciIication is needed about the sequence oI the crack nodes as it grows; this is automatically determined by the
program. The user only provides the initial crack Iront and an interIace along which the crack can grow. It is important
to give a reasonable regular mesh in order to get accurate results.
635
CHAPTER 2.20
Delamination and Crack Propagation
Figure 2.20-6 Overall Deformed Shape
Figure 2.20-7 Crack Front at Final Load (Arrows show X-coordinate of Crack Tip System)
VCCT Cohesive
636 Marc Users Guide: Part I
CHAPTER 2.20
The damaged zone Ior the cohesive zone model is shown in Figure 2.20-8. The yellow portion in the middle indicates
where Iull damage occurs. The damage occurs in the interIace elements, and there is no sharp crack Iront as in the
VCCT case. Here, we have larger Ireedom in designing the mesh.
Figure 2.20-8 Extent of Damage Zone for Interface Elements at Final Load
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
delam_vcct.proc Mentat procedure Iile to run the above example
delam_interface.proc Mentat procedure Iile to run the above example
delam_vcct.mfd Associated model Iile
delam_interface.mfd Associated model Iile
delam_vcct_job1.dat Associated model Iile
delam_interface_job1.dat Associated model Iile
Chapter 2.21: Progressive Failure Analysis of Lap Joint
2.21
Progressive Failure Analysis of
Lap Joint

Summary 638

Model Review 639

Results 640

Modeling Tips 642

Input Files 643

638 Marc Users Guide: Part I
CHAPTER 2.21
Summary
Title Progressive Failure Analysis oI Lap Joint
Problem Ieatures Progressive Iailure analysis
Composite material
Geometry
Material properties Composite layup: 5 layers oI equal thickness |0 45 90 -45 0|
Orthotropic material:

See Figure 2.21-2 Ior Failure Data.
Analysis type Quasi-static analysis
Boundary conditions Clamped end, rigid body contact
Applied loads Forced motion oI rigid body
Element type Layered solid shell
Contact properties Rigid body contact, bilinear Iriction,
FE results 1. Plot oI damage zone
2. Forcedisplacement curves
Y X
Z Z
Y X
X
Y
Z
1
X Y
Z
4
0.5
Units: m
Plate 2x1x0.1 Hole 0.1 Radius
E
1
100GPa E
2
E
3
50GPa
G
12
G
31
7GPa G
23
8GPa v
12
v
31
0.3 v
23
0.4
0.3
0.000 0.005 0.010 0.015 0.020
0
300000
600000
900000
1200000
1500000
gradual no friction
immediate
gradual
Pin Force (N)
Pin Displacement (m)
Inc 1
639
CHAPTER 2.21
Progressive Failure Analysis of Lap Joint
This chapter studies progressive Iailure oI a lap joint. A composite plate has a hole with a bolt in it, where the bolt is
modeled as a rigid body. One end oI the plate is clamped, and the rigid pin has a Iorced motion associated with it. This
simulates a symmetric lap joint with a pin which is much stiIIer than the composite plate.
The composite material can suIIer damage due to excessive loading. The Puck Iailure criterion is used to indicate
Iailure. The progressive Iailure option is used Ior degrading the material properties as Iailure occurs. The Puck method
is used here since it allows separate degradation due to Iiber and matrix Iailure. The matrix material oI the composite
is weaker than the Iiber. When Iailure occurs in the matrix in a certain layer, the structure can still carry a substantial
load due to the undamaged Iibers in other layers.
Two oI the supported options Ior degrading the material properties are studied in this chapter: the gradual and
immediate reduction methods with selective degradation. In the gradual method, aIter damage has occurred, the
material is assumed to also sustain loading. The stiIIness is reduced only so much that the largest Iailure index stays
below 1.0. With immediate stiIIness reduction variant, the material stiIIness is set to the minimum value as soon as
damage is indicated. This corresponds to a brittle behavior oI the material.
Model Review
The base model is available in the Mentat database pinplate.mfd. The model used is shown in Figure 2.21-1. The
Iigure also shows the material orientation.
Figure 2.21-1 Model used for PFA analysis
The settings Ior the progressive Iailure are done in the menus as shown in Figure 2.21-2. There we see that we have
chosen the GRADUAL SELECTIVE option Ior progressive Iailure, and it also shows the material parameters used Ior
the Puck Iailure criterion. We deIine all Iour Iailure envelope slopes since we are using solid elements. For plane stress
shells, we only speciIy the Iirst two and let the program calculate the others. We use the deIaults Ior the residual
stiIIness Iactor and the options Ior stiIIness reduction. The latter are used iI we want to control in more detail how the
diIIerent moduli are degraded. We also have the option oI using the UPROGFAIL user subroutine Ior speciIying the
reduction Iactors, but it is not used here.
X Y
Z
F
i
b
e
r

D
i
r
e
c
t
i
o
n
P
640 Marc Users Guide: Part I
CHAPTER 2.21
Figure 2.21-2 Menus for the Progressive Failure Settings
The analysis is done in two load cases. The Iirst load case does one increment up to a load level which is close to where
the Iirst Iailure occurs. The second load case has 100 increments, and the step size is small in order to properly capture
the damage oI the material.
Results
Figure 2.21-3 shows a plot oI the third Iailure index Ior the Iourth layer at the end oI the Iirst load case. This Iailure
index corresponds to matrix tension. This is the largest Iailure index and it indicates where the Iirst Iailure will occur,
as can also be seen in Figure 2.21-4. This layer is in the -45 direction. With increasing load, there is more Iailure.
Figure 2.21-5 shows the Iailure in the mid (90) layer. The Iailure in this layer starts out in Iront oI the pin and spreads
out with increasing load. At some point, there is a drastic stiIIness reduction in the structure as several elements Iail at
the same time. The Iorce-displacement curve Ior the rigid body is shown in Figure 2.21-6. Here, we clearly see the
sudden drop in structural stiIIness but that the structure can continue to carry load. We see that the Iailed elements tend
to bulge up along the rigid pin.
Figure 2.21-6 illustrates the case oI using the immediate stiIIness degradation method. The sudden drop in structural
stiIIness comes much earlier as compared with the gradual degradation. This shows the brittle eIIect oI this method.
The structure can still continue to carry an increased load since there are layers with intact Iibers.
Figure 2.21-6 also illustrates the eIIect oI Iriction on the example. The gradual stiIIness reduction option is used, and
here we clearly see the eIIect oI Iriction. Without Iriction, the Iailure is more localized to the compressed part oI the
structure next to the pin, and the largest load is substantially lower. The maximum load is closer to the case with
immediate stiIIness reduction.
641
CHAPTER 2.21
Progressive Failure Analysis of Lap Joint
Figure 2.21-3 Largest Failure Index Before Failure Occurs
Figure 2.21-4 First Occurrence of Failure
0.000e+000
9.770e-002
1.954e-001
2.931e-001
3.908e-001
4.885e-001
5.862e-001
6.839e-001
7.816e-001
8.793e-001
9.770e-001
lcase2
3rd Failure Index Layer 4
Inc: 10
Time: 2.356e-001
X Y
Z
4
0.000e+000
1.365e-002
2.730e-002
4.094e-002
5.459e-002
6.824e-002
8.189e-002
9.553e-002
1.092e-001
1.228e-001
1.365e-001
lcase2
Damage Layer 4
Inc: 12
Time: 2.524e-001
X Y
Z
4
642 Marc Users Guide: Part I
CHAPTER 2.21
Figure 2.21-5 Damage in Mid-layer at Final Load
Figure 2.21-6 Pin Force versus Pin Displacement for all Three Cases
Modeling Tips
When a lot oI damage takes place, it is oIten diIIicult to obtain convergence. II unstable growth oI the damage zone
occurs, it may even not be possible to obtain a solution Ior a certain load level. It is important to allow Ior more recycles
than one would normally do in a geometrically nonlinear analysis.
Using automatic time stepping with damping may stabilize the solution. This should be used with care, since too much
damping easily destroy the real solution.
0.000e+000
9.900e-002
1.980e-001
2.970e-001
3.960e-001
4.950e-001
5.940e-001
6.930e-001
7.920e-001
8.910e-001
9.900e-001
lcase2
Damage Layer 3
Inc: 101
Time: 1.000e+000
X Y
Z
4
0.000 0.005 0.010 0.015 0.020
0
300000
600000
900000
1200000
1500000
gradual no friction
immediate
gradual
Pin Force (N)
Pin Displacement (m)
Inc 1
643
CHAPTER 2.21
Progressive Failure Analysis of Lap Joint
The elements that suIIer severe damage tend to cause convergence problems as they have a low stiIIness. This can be
avoided by using the deactivation option. Elements that have Iull Iailure in all integration points can be deactivated,
and this can be chosen Ior selected Iailure modes.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
pinplate_immediate.proc Mentat procedure Iile to run the above example
pinplate_gradual.proc Mentat procedure Iile to run the above example
pinplate.mfd Associated model Iile
pinplate_immediate.dat Associated Marc Iile
pinplate_gradual.dat Associated Marc Iile
644 Marc Users Guide: Part I
CHAPTER 2.21
Chapter 2.22: Sheet Metal Forming With Solid Shell Element
2.22
Sheet Metal Forming With
Solid Shell Elements

Summary 646

Model Review 647

Results 649

Input Files 651

646 Marc Users Guide: Part I
CHAPTER 2.22
Summary
Title Sheet Metal Forming With Solid Shell Elements
Problem Ieatures Solid shell elements Iacilitate blank holders, shell elements cannot.
Mentat reading the Marc input Iile history deIinition.
Geometry
Material properties
, ,
Analysis type Quasi-static analysis
Boundary conditions Symmetric displacement constraints along two symmetry planes.
Applied loads Upper die is displaced 4.5 mm in the y-direction.
Element type Solid shell element type 185. Compare to stacked bricks type 7.
Contact properties CoeIIicient oI Iriction and
FE results 1. Contour plot oI plate thickness, equivalent plastic strain
2. History plot oI die load versus stroke
Upper Die Upper Die
Blank Holder
Lower Die
Lower Die
53x27x0.45
Plate
Units: mm
Y
X
Z
4.50 mm
Z X
Y
E
cylinder
210x10
9
Pa v 0.3
c
p
d
do
6.5x10
7
Pa
0.0 0.3
Z X
Y
Double Sided Contact
Caused By Blank Holder
647
CHAPTER 2.22
Sheet Metal Forming With Solid Shell Elements
Higher Iormability and accuracy Ior sheet metal Iorming can be achieved by using blank holders and varying the
holder Iorce during the Iorming process. UnIortunately, shell elements cannot sustain double-sided contact where
the shell is approached by the dies Irom both sides. Solid elements can sustain double-sided contact but require
stacking in the thickness direction in order to improve bending characteristics. The stacked bricks generate many
degrees oI Ireedom and requires computer times Irom days to weeks instead oI hours. The solid-shell element,
type 185, is specially Iormulated to overcome this situation by simulating a shell type element while Iacilitating
double-sided contact.
Model Review
The mesh building oI a similar structure can be Iound in Chapter 3.16, similar steps are not repeated here other than
mention that the Ilat plate is made oI solid shells, element type 185 instead oI the 3-D membrane elements used in
Chapter 3.16. The solid-shell element designed Ior plate or shell type geometry (where the thickness, t, is much smaller
that the in-plane length, L) begins to loose accuracy when the element`s aspect ratio (t/L) drops below 1/100. Another
Ieature oI element type 185 is that it deIines the thickness direction as normal to the Iirst Iace created by the Iirst Iour
nodes in the elements connectivity (Figure 2.22-3), and the thickness oI the element can be contour plotted.
Figure 2.22-1 Sheet and Die Geometry
Let`s suppose this model has already been built and exists as a Marc input Iile named sheetform.dat, and then read
this Iile into Mentat. Unlike previous versions oI Mentat, the history deIinition oI the model is obtained. The model is
saved, and the job is submitted as shown below.
FILES
MARC INPUT FILE READ
sheetform.dat
OK
Upper Die
Upper Die
Blank Holder
Lower Die
Lower Die
53x27x0.45
Plate
Units: mm
Y
X
Z
4.50 mm
Z X
Y
648 Marc Users Guide: Part I
CHAPTER 2.22
You can review the model in Mentat or preview the assembly in Figure 2.22-2.
Figure 2.22-2 3D Assembly View - Click Above to Activate 3-D (ESC to stop)
SAVE AS
sheetform
OK
MAIN
649
CHAPTER 2.22
Sheet Metal Forming With Solid Shell Elements
\
Figure 2.22-3 Visualization of Thickness Orientation for Solid-Shell Elements
Now the job can be submitted, and as the results become available they can be viewed by opening the post Iile as:
and we begin examining the results.
Results
The model read Irom disk already had Iriction selected. The no Iriction case is a new job that simply turns oII the
Iriction. Comparing the Irictionless and Iriction results, we see that the Iriction case gives a substantially thinner sheet,
particularly in the corner as shown in Figure 2.22-4.
JOBS
RUN
SUBMIT
OPEN POST FILE (RESULTS MENU)
650 Marc Users Guide: Part I
CHAPTER 2.22
Figure 2.22-4 Thickness Contours With and Without Friction
Figure 2.22-5 3-D Thickness Contours - Click Above to Activate 3-D.
Upon closer examination oI the two jobs using the path plot Ior a section Irom point A to B in Figure 2.22-6, the
Iriction drops the thickness Irom 0.45 mm to below 0.35 mm, whereas there is only a small drop in thickness Ior the
Irictionless case to 0.43 mm.
3.325e-004
3.367e-004
3.408e-004
3.450e-004
3.491e-004
3.532e-004
3.574e-004
3.615e-004
3.657e-004
3.698e-004
3.740e-004
3.781e-004
3.823e-004
3.864e-004
3.906e-004
3.947e-004
3.989e-004
4.030e-004
4.072e-004
4.113e-004
4.155e-004
4.196e-004
4.237e-004
4.279e-004
4.320e-004
4.362e-004
4.403e-004
4.445e-004
4.486e-004
4.528e-004
4.569e-004
full_load
Thickness of Element
Inc: 50
Time: 1.000e+000
= 0 = 0.3
3.325e-004
3.367e-004
3.408e-004
3.450e-004
3.491e-004
3.532e-004
3.574e-004
3.615e-004
3.657e-004
3.698e-004
3.740e-004
3.781e-004
3.823e-004
3.864e-004
3.906e-004
3.947e-004
3.989e-004
4.030e-004
4.072e-004
4.113e-004
4.155e-004
4.196e-004
4.237e-004
4.279e-004
4.320e-004
4.362e-004
4.403e-004
4.445e-004
4.486e-004
4.528e-004
4.569e-004
Thickness of Element
Inc: 50
Time: 1.000e+000
Once the 3D content is activated by
a left click on the figure to the left,
you can navigate the 3D content e.g.
rotate - left mouse drag
zoom - right mouse drag up/down
A right click and release of the
mouse will open a menu and you can
stop by choosing Disable Content
651
CHAPTER 2.22
Sheet Metal Forming With Solid Shell Elements
Figure 2.22-6 Thickness Profile
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
sheetform.proc Mentat procedure Iile to run the above example
sheetform.mud Associated model Iile
sheetform.dat Associated Marc Iile
0 5 10 15 20 25 30
0.30
0.35
0.40
0.45
0.50
Friction
Frictionless
Thickness (mm)
Length (mm)
B
B
A
A
652 Marc Users Guide: Part I
CHAPTER 2.22
Chapter 2.23: Plastic Limit Load Analysis of a Simple Frame Structure
2.23
Plastic Limit Load Analysis of a
Simple Frame Structure

Summary 654

Detailed Marc Input Description 655

Detailed Mentat Session Description 657

Results 658

Modeling Tips 659

Input Files 659

654 Marc Users Guide: Part I
CHAPTER 2.23
Summary
Title Plastic Limit Load Analysis oI a Simple Frame Structure
Problem Ieatures Beam element type 52 is used with a rectangular cross-section that can account Ior
plastic deIormation through the section by perIorming numerical integration oI the
nonlinear material behavior across the cross section.
Geometry
Material properties E 200 GPa, 0.3, Initial yield strength, 200 MPa,
Analysis type Nonlinear static analysis
Boundary conditions Fully clamped in the two supports
Applied loads Apply a load oI 32 kN in x-direction halI way the 2m high vertical member
Element type Beam element type 52
Cross-section properties Rectangular cross-section, 0.1 m x 0.04 m, using a 5x5 Simpson numerical integration rule
FE results DeIormed Irame structure and Iorce displacement response oI the loaded point
x
y
2m
1m
Cross-section data
x
loc

1m
1m
0.1m
0.04m
Global vector defining
local x: (0,0,1)
Material data:
E =2x10
11
N/m
2

=0.3

y
=2x10
8
N/m
2

E
p
=2x10
7
N/m
2

F=32kN
y
loc

v o
Y
do dc
p
20 Mpa
lcase1
Inc: 100
Time: 1.000e+000
X
Y
Z
F = 32 kN
0.00 0.05 0.10 0.15 0.20
0
5000
10000
15000
20000
25000
30000
35000
External Force X (N) Node 15
Displacement X (m) Node 15
655
CHAPTER 2.23
Plastic Limit Load Analysis of a Simple Frame Structure
The simple Irame structure in the Summary is loaded by a horizontal load oI 32 kN with its associated geometry,
cross-section properties and material data. The material is elastic-plastic with a small isotropic work hardening
slope in the plastic range. The maximum load oI 32 kN is close to the plastic limit load oI the structure. The material
data relevant in this analysis are: Young's modulus, Poisson's ratio, initial yield strength, and the plastic work
hardening slope.
Detailed Marc Input Description
This section describes the Marc input Iile where tables are used in the old input Iormat. The input Iile is:
limit_load_old_job1.dat.
The beam element type used in this analysis is type 52, and it is used with numerical integration over its cross section.
The LARGE STRAIN parameter has been activated, since it is anticipated that when the cross section turns Iully plastic
in some locations, the deIlections oI the Irame may become larger than acceptable in a geometrically linear analysis.
The cross-section properties are deIined in the BEAM SECT parameter and its deIinition is shown in the block below.
beam sect
rectangle
0,-2.0,0.1,0.04
...
...
last
Beam Sect Parameter Definition
The line Iollowing the beam sect keyword deIines the title oI this section. The line Iollowing the title speciIies the
type oI section and its dimensions. The Iirst Iield is zero, meaning a standard section is used. The second Iield speciIies
the section type is rectangular. The third and Iourth Iields speciIy the dimensions oI the rectangle. Then Iollow two
blank lines (represented by the two dotted lines). The Iirst blank line means that the deIault 5 x 5 Simpson rule is used
Ior the cross-section integration and that numerical integration is used throughout the analysis. This is important,
because the cross section will Iirst develop plasticity in its outer Iibers and the plastic zone will gradually grow inward.
The next blank line has no meaning in this analysis. The BEAM SECT parameter deIinition is concluded with the
keyword last.
The material data is deIined through the ISOTROPIC and WORK HARD options and is listed in the block below.
isotropic
...
1,von mises,isotropic
2e11,0.3,1.0,0.0,2.0e8
4 to 53
work hard,data
2,0,1,
2.0e8,0.0
2.2e8,1.0
656 Marc Users Guide: Part I
CHAPTER 2.23
Material Definition
The ISOTROPIC option deIines a material with id=1 using the von Mises yield surIace and isotropic hardening. Its
data line deIines the elastic material properties and the initial yield stress. The WORK HARD option deIines the plastic
hardening data. It deIines two yield limits: one at zero plastic deIormation and one at a plastic strain value oI 1.
The GEOMETRY option input assigns the cross-section properties to all the elements. In this case, it reIers to the Iirst
(and only) section deIined in the BEAM SECT parameter by entering a zero in the 1st Iield and the beam section
number in the 2nd Iield. Through the 4th, 5th, and 6th Iields, it speciIies a vector in the global coordinate system that
deIines the local x-direction oI the cross section. The input is listed in the block below.
geometry
...
0.0,1.0,0.0,0.0,0.0,1.0
4 to 53
Geometry Definition
The two nodes in the supports (nodes 1 and 4) are Iully clamped. These boundary conditions are deIined through the
FIXED DISP and the input is listed in the block below.
fixed disp
...
0.0,0.0,0.0,0.0,0.0,0.0
1,2,3,4,5,6
1,4
Boundary Conditions at Frame Supports
The POST option deIines the element quantities that are desired Ior Iurther postprocessing. The element quantities
requested in this case are the equivalent von Mises stress (code 17) in layers 3, 8, 13, 18, and 23. These represent the
integration points at x0 and y-0.02, y-0.01, y0.0, y0.01 and y0.02, respectively, in the cross section which can
be veriIied Irom the BEAM SECT output written to the analysis output (.out) Iile. Post code 265 is the generalized
bending moment about the local x-axis oI the cross section. The POST option input is listed in the block below.
post
6,
17,3
17,8
17,13
17,18
17,23
265,0
Output Quantities Requested for Postprocessing
The END OPTION input concludes the model deIinition. All input Iollowing this line is part oI the history input. The
total load oI 32 kN is subdivided into 100 equal steps and is deIined through the AUTO LOAD, TIME STEP, and POINT
LOAD options in the history input. The load deIinition input is listed in the block below.
auto load
100,0,10
657
CHAPTER 2.23
Plastic Limit Load Analysis of a Simple Frame Structure
time step
0.01,
point load
...
3.2e2,
15
continue
Load history applied to the frame structure
Detailed Mentat Session Description
This section describes the Mentat menus that are used to deIine a solid cross section that employs numerical integration
and, thus, can account Ior nonlinear material behavior. The complete input Ior the Irame model, including the
cross-section deIinition described here, can be generated in Mentat by running the procedure Iile:
limit_load_new.proc.
Beam cross-sections are deIined in the GEOMETRIC PROPERTIES main menu. Beam element types 52 and 98 are
three-dimensional beam elements; thereIore we enter the 3-D submenu as shown in Figure 2.23-1. For 3-D beam
elements, we have the choice between solid cross-sections or thin-walled cross-sections. For element types 52 or 98,
we enter the SOLID SECTION BEAM menu. We can set the desired type through the ELEMENT TYPES menu. The
PROPERTIES switch determines iI the properties are CALCULATED by numerical integration or iI they are ENTERED
directly. Here, we choose CALCULATED, because we need to account Ior plastic deIormations in the cross section. In
the SHAPE menu, we can toggle through a number oI standard cross-section geometries. We choose RECTANGULAR
Ior our purpose and deIine the dimensions oI the rectangle through the DIMENSION A (=0.1) and the DIMENSION B
(=0.04) inputs. The MATERIAL BEHAVIOR in the section is set to GENERAL to allow Ior plastic deIormations. It can
also be set to LINEAR ELASTIC ONLY, in which case the cross-section properties like area A and moments oI area I
xx

and I
yy
are computed by numerical integration prior to the start oI the analysis (i.e., pre-integrated) and no Iurther
numerical integration over the cross section is carried out during the analysis. The section uses the deIault integration
scheme which Ior a rectangular section is a 5x5 Simpson scheme. Finally, the three components oI the VECTOR
DEFINING LOCAL ZX-PLANE are entered which deIine the orientation oI the cross section in space. In this example,
it is expedient to use a vector pointing in global z-direction Ior all beam elements in the model.
658 Marc Users Guide: Part I
CHAPTER 2.23
Figure 2.23-1 Creating a Solid Rectangular Cross-section Definition
Results
The results oI the analysis are shown in Figure 2.23-2. It displays the deIormed Irame aIter application oI the Iull load
and the Iorce-displacement curve oI the external load versus the displacement oI the loaded point. Initially, the
behavior oI the Irame is elastic. At a certain stage, some locations develop plasticity and the stiIIness is reduced. At a
later stage, some locations completely turn plastic. At this stage, the Irame structure loses its stiIIness considerably
and some cross sections have practically developed a plastic hinge. At this stage, the load carrying capacity oI the
Irame is almost exhausted. The maximum horizontal displacement oI the loaded point is 0.159 m.
Figure 2.23-2 Deformed Frame Structure and Force Displacement Response of the Loaded Point

GEOMETRIC PROPERTIES
3-D
SOLID SECTION BEAM
RECTANGULAR


lcase1
Inc: 100
Time: 1.000e+000
X
Y
Z
F = 32 kN
0.00 0.05 0.10 0.15 0.20
0
5000
10000
15000
20000
25000
30000
35000
External Force X (N) Node 15
Displacement X (m) Node 15
659
CHAPTER 2.23
Plastic Limit Load Analysis of a Simple Frame Structure
Modeling Tips
II the material behavior is linear elastic only, pre-integrated sections can be used to save storage and analysis time. In
this case, the section properties like area and moments oI area are computed only once prior to the analysis by
numerical integration. During the analysis, no numerical section integration is carried out and the beam behaves as iI
its section properties were entered directly in the input. The solid section beam Iormulation that allows Ior numerical
cross-section integration to account Ior the nonlinear material behavior is also available Ior element type 98.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
limit_load.proc Mentat procedure Iile to run the above example using the new style table
input Iormat.
limit_load_old_job1.dat Associated Marc Iile using the old style table input Iormat.
660 Marc Users Guide: Part I
CHAPTER 2.23
Chapter 2.24: Directional Heat Flux on a Sphere from a Distance Source
2.24
Directional Heat Flux on a
Sphere from a Distance Source

Summary 662

Model Review 663

Results 666

Input Files 667

662 Marc Users Guide: Part I
CHAPTER 2.24
Summary
Title Directional heat Ilux on a sphere Irom a distance source
Problem Ieatures Directional heat Ilux
Heat transIer membrane elements
Radiation boundary condition
Geometry Sphere radius 1.0 It
Membrane thickness 0.01 It
Material properties
Thermal conductivity 204.0 BTU/hr It
o
F
Emissivity Absorption 1.0
Analysis type Steady state heat transIer analysis at a series oI angles oI incidence Ior the directional
heat Ilux
Boundary conditions Radiation to the environment, T

60
o
F
Applied loads Directional heat Ilux magnitude 300.0 BTU/hr It
2

Element type 4 node heat transIer membrane elements type 198
Contact properties None
FE results Temperature variation with directional heat Ilux angle oI incidence

-90 -60 -30 0 30 60 90
80
100
120
140
160




Angle (deg)
Temperature ( F)
o
1
1
1
11
11 11
362
362
362
663
CHAPTER 2.24
Directional Heat Flux on a Sphere from a Distance Source
This chapter demonstrates the use oI the directional heat Ilux thermal loading. Heat Ilux Irom a distant source can be
treated in a directional sense with the QVECT model deIinition input option. The Ilux is applied on a spherical shell
modeled using heat transIer membrane elements. For illustrative purposes, the angle oI incidence oI the directional
heat Ilux is varied to create a plot Ior the temperature versus the angle oI incidence. A simple radiation boundary
condition to space represents the loss mechanism and keeps the sphere in a state oI radiative equilibrium.
Model Review
A model Ior the sphere is available in the Iile directional_heat_flux.mfd shown in Figure 2.24-1. The table
driven input style must be used Ior directional heat Ilux thermal loading. The spherical shell is modeled using 720 heat
transIer membrane elements type 198. The membrane thickness is 0.01 It. The initial temperature oI all the nodes in
the model is equal to the environment temperature 60
o
F. The thermal conductivity oI the material is 204.0 BTU/hr
It
o
F. The surIace emissivity and absorption are both equal to 1.0.
Figure 2.24-1 Model for a Sphere with Directional Heat Flux.
The directional heat Ilux is deIined with the QVECT model deIinition option. The directional heat Ilux settings can be
Iound in Mentat menus under BOUNDARY CONDITIONS ->THERMAL -> HEAT FLUX as shown in Figure 2.24-2. The
LOAD TYPE should be set to DIRECTED. The magnitude oI the heat Ilux is 300.0 BTU/hr It
2
.
664 Marc Users Guide: Part I
CHAPTER 2.24
Figure 2.24-2 Mentat Menu for Directional Heat Flux.
The angle oI incidence oI the heat Ilux is to be varied through 180
o
. The vector is initially pointing in the negative x-
direction and rotates counterclockwise via 10
o
increments to be Iinally aligned with the positive x-axis. Nineteen
steady state load cases are used to vary the angle oI incidence which is measured with respect to the positive y-axis.
The direction cosines oI the heat Ilux vector are varied through Iormula-type tables. The expressions Ior the x and y
components oI the direction cosines are
and ,
respectively, where is the increment number. The table entries in Mentat are shown in Figure 2.24-3.
Figure 2.24-3 Formula-type Tables used to change the Heat Flux Angle of Incidence
90 10 v1 1 ( ) ( )
t
180.00
----------------
\ .
| |
sin 90 10 v1 1 ( ) ( )
t
180.0
-------------
\ .
| |
cos
v1
665
CHAPTER 2.24
Directional Heat Flux on a Sphere from a Distance Source
The second boundary condition controls the radiation back to the environment. A radiating cavity is deIined and the
environment temperature is set to 60
o
F. Mentat menu Ior the cavity radiation is shown in Figure 2.24-4.
Figure 2.24-4 Radiation Boundary Condition
The convergence tolerance Ior each loadcase is set such that recycling occurs iI the diIIerence between the temperature
calculated and estimated is greater than 1F as shown in Figure 2.24-5. Because the English unit system is used, it is
necessary to deIine the absolute temperature to be 459.67 and the SteIan-Boltzmann constant to be 1.714 x 10
-9
Btu/hr
It
2o
R
4
.
Figure 2.24-5 Steady State Heat Transfer Loadcase with Convergence Testing
666 Marc Users Guide: Part I
CHAPTER 2.24
Results
Steady state temperatures at three nodes on the surIace oI the sphere versus angle oI incidence are shown in the history
plot oI Figure 2.24-6. The same results are given in Table 2.24-1 versus angle oI incidence oI the directional heat Ilux.
Figure 2.24-6 Temperature versus Increment Number.
Table 2.24-1 Temperature versus Angle of Incidence of Directional Heat Flux .
Increment #
Angle of Incidence
(deg) Node 1 (
o
F) Node 11 (
o
F) Node 362 (
o
F)
1 90 157.687 120.347 102.805
2 80 157.144 125.691 103.053
3 70 155.197 131.437 103.597
4 60 152.041 137.274 104.493
5 50 147.851 142.86 105.799
6 40 142.84 147.885 107.566
7 30 137.25 152.067 109.832
8 20 131.351 155.163 112.629
9 10 125.586 157.095 116.138
10 0 120.387 157.778 120.505
11 -10 116.029 157.06 125.709
12 -20 112.535 155.102 131.481
13 -30 109.751 151.98 137.378
-90 -60 -30 0 30 60 90
80
100
120
140
160




Angle (deg)
Temperature ( F)
o
1
1
1
11
11 11
362
362
362
u
667
CHAPTER 2.24
Directional Heat Flux on a Sphere from a Distance Source
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
14 -40 107.497 147.776 142.959
15 -50 105.744 142.738 147.957
16 -60 104.453 137.146 152.128
17 -70 103.57 131.307 155.258
18 -80 103.041 125.565 157.176
19 -90 102.938 120.442 157.878
File Description
directional_heat_flux.proc Mentat procedure Iile to run the above example
directional_heat_flux.mfd Associated model Iile
directional_heat_flux.dat Associated Marc Iile
Table 2.24-1 Temperature versus Angle of Incidence of Directional Heat Flux (continued).
Increment #
Angle of Incidence
(deg) Node 1 (
o
F) Node 11 (
o
F) Node 362 (
o
F) u
668 Marc Users Guide: Part I
CHAPTER 2.24
Chapter 2.25: Deep Drawing of A Sheet With Global Remeshing
2.25
Deep Drawing of A Sheet With
Global Remeshing

Summary 670

Model Review 671

Results 673

Modeling Tips 674

Input Files 674

670 Marc Users Guide: Part I
CHAPTER 2.25
Summary
Title Deep Drawing oI A Sheet With Global Remeshing
Problem Ieatures A punch and die Iorm a deep box with Iriction
The sheet is remeshed when either one oI two criteria is met
Geometry
Material properties E 210 GPa, 0.3, Initial yield strength, 188.66 MPa
Analysis type Quasi-static analysis
Boundary conditions Symmetric displacement constraints are applied through symmetric contact surIaces.
Applied loads The punch is pushed opposite to Z-direction a stroke oI 96 mm
Element type Shell element type 75 with Iull integration with 7 layers
Contact Properties CoeIIicient oI Iriction
FE results 1 Contour plot oI plate thickness, equivalent plastic strains
2 History plot oI die load versus stroke
Units: mm
84
151.5
210
sheet
punch
die
X Y
Z
Sheet
255x220x1.2
v o
Y
0.4
0 20 40 60 80 100
0
10000
20000
30000
40000
50000
60000
70000
80000
Punch Force (N)
Punch Stroke (mm)
671
CHAPTER 2.25
Deep Drawing of A Sheet With Global Remeshing
Severe deIormation oIten occurs in sheet metal Iorming processes. When applying the technique oI Iinite element
analysis to such processes, the shell elements deIormed so severely that the FE analysis is not able to continue with
the distorted mesh. This is because the distorted elements are unable to provide stable solution due to the extremely
small singularity ratio oI the equation system or negative Jacobian. Besides, during sheet Iorming processes, the
contact condition between die surIaces and blank sheet changes so Iast, the old mesh easily penetrates into die surIaces
without remeshing. ThereIore, remeshing is necessary in order to obtain accurate results.
This chapter describes the usage oI the new extension oI the 3-D remeshing technique oI Marc to the shell elements.
In the current release oI Marc, the shell remeshing is activated by adding the REZONE parameter and ADAPT GLOBAL
model and history deIinition options into the input data Iile. The 3-D surIace meshers (quadrilateral/triangular
elements) are used to generate new shell elements. This is applicable to both quadrilateral and triangular shell
type elements.
Model Review
The sheet plate is initially subdivided as 636 4-node shell elements Ior the Iinite element analysis. Element type 75
(thick shell element with Iull integration) is chosen Ior the analysis. The initial geometry oI die surIaces and blank
sheet are shown in the Summary. Two criteria are used to control the Irequency oI mesh rezoning Ior the Iinite element
model. They are the number oI increment and the amount oI strain change. For example, iI the incremental Irequency
Ior remeshing is set as 5, then new mesh will be regenerated aIter every 5 incremental steps. To set the criteria oI strain
change as 0.3, the new mesh will be created iI the strain change reaches 0.3. In this example, these two criteria are
combined to control the regeneration oI new meshes in order to properly control the Irequency oI remeshing. The mesh
size oI the new meshes can be controlled by setting the element edge length or setting the upper bound oI maximum
number oI elements allowed Ior the new mesh. Here, it is given the value oI 3000. In this case, the mesh size will be
determined automatically by the 3D surIace mesher. As shown below, the global adaptive remeshing is deIined by the
ADAPT GLOBAL option, the mesher 19 is entered Ior quadrilateral shell mesher. For triangular shell element, mesher
12 is needed. The two criteria Ior remeshing are deIined as criterion 1(to generate new mesh aIter every 5 incremental
steps) and criterion 5 (which stands Ior strain change).
ADAPT GLOBAL
1 0 0
19 1 1 0 2 0 0
1 5 0.000000000000000+0 0.000000000000000+0
5 0 3.000000000000000-1 0.000000000000000+0
0.000000000000000+0 0.000000000000000+0 1.000000000000000+2
6.000000000000000+1 1.500000000000000+0
The analysis takes 32 steps in total, within which remeshing is conducted aIter increment step 5, 10, 20, 23, 25, 26,
27, 28, 29, 30, 31. Initially, the remeshing becomes necessary due to criterion oI increment Irequency, like increments
5, 10, 15, 20, and 25.
When the deIormation becomes large, the remeshing is more oIten triggered by strain changes, such as the steps 23,
26, 27, 28, 29, 30, and 31.The Iinal mesh and die surIace are shown in Figure 2.25-1.
672 Marc Users Guide: Part I
CHAPTER 2.25
Figure 2.25-1 Sheet and Die Geometry at the Final Stage of Deep Drawing
Let's suppose this model has already been built and exists as a Marc input Iile named ug_shlremesh.dat, and then
read this Iile into Mentat. Unlike previous versions oI Mentat, the history deIinition oI the model is obtained. The
model is saved, and the job is submitted as shown below.
FILES
MARC INPUT FILE READ
ug_shlremesh.dat
OK
SAVE AS
ug_shlremesh.dat
OK
MAIN
JOBS
RUN
SUBMIT
9.372e-001
9.527e-001
9.681e-001
9.835e-001
9.989e-001
1.014e+000
1.030e+000
1.045e+000
1.061e+000
1.076e+000
1.091e+000
1.107e+000
1.122e+000
1.138e+000
1.153e+000
1.169e+000
1.184e+000
1.199e+000
1.215e+000
1.230e+000
1.246e+000
1.261e+000
1.277e+000
1.292e+000
1.307e+000
1.323e+000
1.338e+000
1.354e+000
1.369e+000
1.385e+000
1.400e+000
Thickness of Element
Inc: 32
Time: 3.200e+001
X Y
Z
673
CHAPTER 2.25
Deep Drawing of A Sheet With Global Remeshing
As the results become available, they can be viewed by opening the post Iile as:
And we begin examining the results.
Results
The model read Irom disk already set the option oI global adaptive remeshing. Once the job is run, we just need to see
the results and check how the remeshing is conducted during the analysis. Figure 2.25-2 (a) and (b) shows the
equivalent stress at increment 5 and 32, respectively. The number oI elements has increased Irom 636 at increment 5
to 2866 at increment 32.
The thickness contour oI the part aIter the deep drawing is shown in the Figure 2.25-3 (a). Upon closer examination
oI the two jobs using remeshing and without using remeshing, the thickness contour with remeshing shows less
thinning at the punch corner area than the job without using remeshing (see Figure 2.25-3 (b)).
Figure 2.25-2 Equivalent Stress at Increment 5 (a) and 32 (b)
OPEN POST FILE (RESULTS MENU)
5
21
37
53
69
86
102
118
134
150
167
183
199
215
231
247
264
280
296
312
328
344
361
377
393
409
425
441
458
474
490
Equivalent of Global Stress Layer 1
Time: 3.200e+001
(a) (b)
674 Marc Users Guide: Part I
CHAPTER 2.25
Figure 2.25-3 Thickness Distributions With Remeshing (a) and Without Remeshing (b)
Modeling Tips
In the current release, three criteria are available Ior both quadrilateral and triangular shell remeshing which include:
increment Irequency, strain change, and penetration. Additionally, when using triangular elements, curvature-based
criteria is also supported.
The sheet was glued to the die to simulate a binder since shell elements were used and cannot support
double-sided contact.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
ug_shlremesh.proc Mentat procedure Iile to run the above example
ug_shlremesh.dat Associated Marc Iile
ug_shlremeshn.dat Associated Marc Iile - no remeshing
9.200e-001
9.360e-001
9.520e-001
9.680e-001
9.840e-001
1.000e+000
1.016e+000
1.032e+000
1.048e+000
1.064e+000
1.080e+000
1.096e+000
1.112e+000
1.128e+000
1.144e+000
1.160e+000
1.176e+000
1.192e+000
1.208e+000
1.224e+000
1.240e+000
1.256e+000
1.272e+000
1.288e+000
1.304e+000
1.320e+000
1.336e+000
1.352e+000
1.368e+000
1.384e+000
1.400e+000
Thickness of Element
Time: 3.200e+001
X Y
Z
(a) (b)
Chapter 2.26: Artery Under Pressure
2.26
Artery Under Pressure

Summary 676

Material Modeling 677

Results 679

Modeling Tips 681

Input Files 682

References 681
676 Marc Users Guide: Part I
CHAPTER 2.26
Summary
Title Artery Under Pressure
Problem Ieatures NLELAST option to model nonlinear elastic material
UELASTOMER user subroutine to deIine same material
Geometry
Material properties NLELAST option uses experimental data
UELASTOMER user subroutine uses data Iit to Fung`s Model as;
Analysis type Quasi-static analysis
Boundary conditions Symmetric displacement constraints are applied at the leIt end oI the model. Axial
displacements at right end oI the tube are Iixed.
Applied loads The internal and external pressures are shown above.
Element type 4-node axisymmetric element type 10 with a Iine gradient at the center.
FE results 1 Stress versus strain plots Ior both models
2 Radial displacement plots during loading Ior both models
3 DeIormed model with the distribution oI equivalent stresses

mm 8
mm 8
mm 20 mm 20
mm 4 . 0
x
kPa P
i
2 =
( ) kPa x P
e
2
4 =
Mesh
-0.3 -0.2 -0.1 -0.0 0.1 0.2 0.3
-400
-200
0
200
400
600 (kPa)
(1)
W
a
b
---
e
b
2
--- I
1
3
( )
1 =
a 44.25 kPa =
b 16.73 =
677
CHAPTER 2.26
Artery Under Pressure
This chapter is to demonstrate the use oI the UELASTOMER user subroutine and the NLELAST option to model
nonlinear elastic behaviors oI soIt tissue materials.
Material Modeling
SoIt tissue materials exhibit a highly nonlinear behavior. Fung's (Fung, 1967) model is one oI the most commonly used
models Ior such materials. Fung`s material model assumes the strain energy density can be expressed as an exponential
oI the Iirst strain invariant, namely, where the material constants and are Irom (MoIrad, 2003).
With the help oI the user subroutine, this model can be implemented with a Iew lines oI code. The new code in
uelastomer.I will look like the Iollowing:
subroutine uelastomer(iflag,m,nn,matus,be,x1,x2,x3,detft,
$ enerd,w1,w2,w3,w11,w22,w33,w12,w23,w31,
$ dudj,du2dj,dt,dtdl,iarray,array)
#ifdef _IMPLICITNONE
implicit none
#else
implicit logical (a-z)
#endif
c ** Start of generated type statements **
real*8 array, be, detft, dt, dtdl, du2dj, dudj, enerd
integer iarray, iflag, m, matus, nn
real*8 w1, w11, w12, w2, w22, w23, w3, w31, w33, x1, x2, x3
real*8 aa,bb,ccc
c ** End of generated type statements **
dimension m(2),be(6),dt(*),dtdl(*),iarray(*),array(*),matus(2)
c implement Fungs model for bio-materials
c W = a/b * { exp[0.5*b*(I_1-3)] - 1 }
c define material parameters
aa=44.25
bb=16.73
ccc=exp(0.5d0*bb*(x1-3.d0))
c w1 is the derivative of the strain energy with respect to
c the first invariant
w1=0.5*aa*ccc
c w11 is the second derivative of the strain energy with
c respect to the first invariant
w11=0.25*aa*bb*ccc
enerd=aa/bb*(ccc-1)
return
end
To activate the user subroutine, simply click MATERIAL PROPERTIES -> MOONEY in Mentat and deIine a list oI
elements associated and submit the user subroutine with the run.
The NLELAST option provides an even simpler way to simulate nonlinear elastic materials. In such a case, the
experimentally obtained data can be used directly as the material input in a table. The eIIort oI curve Iitting to get the
material parameters is no longer needed.
W
a
b
--- e
b
2
--- I
1
3 ( )
1 a b
678 Marc Users Guide: Part I
CHAPTER 2.26
To deIine NLELAST:
The eIIective stress versus strain material data Ior the soIt tissue contained in the table tab_mod_nlelast selected
in Figure 2.26-1 is plotted Figure 2.26-2 connected by dashed lines.
Figure 2.26-1 Define NLELAST
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
HYPOELASTIC
SIMPLIFIED NONLINEAR ELASTIC
Choose stress model
Define the table for effective stress-strain curve (tab_mod_nlelast)
Define the Poisson's ratio (0.49)
OK
679
CHAPTER 2.26
Artery Under Pressure
Figure 2.26-2 Plot of Soft Tissue Material Stress - Strain Behavior
Job Parameters
For the Mooney/uelastomer model, large strains are automatically activated. For the NLELAST model, it needs to be
activated using JOB -> ANALYSIS OPTIONS -> LARGE STRAIN
Results
The cross plots oI the maximum equivalent stress versus the maximum equivalent strain, occurring on the inner surIace
at the center oI the artery, are illustrated in Figure 2.26-3. It can be observed that the results Irom both models are very
close up to the level oI 25 strain. Fung`s model is smooth because oI its analytical description, whereas NLELAST
is piece-wise linear between experimental data points shown in Figure 2.26-2. The maximum stresses reached at Iull
loading are 155 kPa and 154 kPa Ior Fung's model and NLELAST, respectively. The stress diIIerence between the two
material models is smaller than the strain diIIerence. This is reasonable because it is a load-controlled problem.
The history plots oI the change oI tube radius at the center and the end oI the tube are shown in Figure 2.26-4. It can
be observed that the results Irom both models are very close. It is particularly true when the strain is less than 5.
-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3
-400
-200
0
200
400
600 (kPa)
(1)
680 Marc Users Guide: Part I
CHAPTER 2.26
Figure 2.26-3 Stress versus Strain at Node 1
Figure 2.26-4 Radial Displacements at End versus at Center
Figure 2.26-5 shows the deIormed model with the distribution oI equivalent stresses, obtained using Fung`s
material model.
0.00 0.05 0.10 0.15 0.20 0.25 0.30
0
50
100
150
200
NLELAST
Fung
(kPa)
(1)
(0.2598, 155.4) (0.2557, 154.0)
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.1
0.2
0.3
0.4
0.5
NLELAST
Fung's
Inward Radial Displacement
At Center (mm)
Outward Radial Displacement
At End (mm)
681
CHAPTER 2.26
Artery Under Pressure
Figure 2.26-5 Equivalent Stress Contours on Deformed Artery
Modeling Tips
A relatively large bulk modulus is required to enIorce incompressibility oI the materials in deIining Fung's
model using the UELASTOMER user subroutine. This can be done under the MOONEY option.
Because the deIormation is large and the updated Lagrange Iormulation is used in the analysis, the
stress-strain curve must reIer to the true (Cauchy) stress and true (logarithmic) strain.
References
1. Fung, Y. C. (1967) Elasticity oI soIt tissues in simple elongation. Am. J. Physiol. 28, 1532-1544.
2. MoIrad, (2003) et al. Computers and Structures 81(2003) 715726
5.633e+000
1.062e+001
1.562e+001
2.061e+001
2.560e+001
3.059e+001
3.558e+001
4.058e+001
4.557e+001
5.056e+001
5.555e+001
6.054e+001
6.554e+001
7.053e+001
7.552e+001
8.051e+001
8.550e+001
9.049e+001
9.549e+001
1.005e+002
1.055e+002
1.105e+002
1.155e+002
1.204e+002
1.254e+002
1.304e+002
1.354e+002
1.404e+002
1.454e+002
1.504e+002
1.554e+002
lcase1
Equivalent of Stress
Inc: 10
Time: 1.000e+000
X
R
682 Marc Users Guide: Part I
CHAPTER 2.26
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
tube_nlelast.proc Mentat procedure Iile to run the above example
tube_uelastomer.proc Mentat procedure Iile to run the above example
tube.mud Mentat model Iile Ior geometry
tube_nlelast.dat Marc input Iile using NLELAST
tube_uelastomer.dat Marc input Iile using the UELASTOMER user subroutine
uelastomer.f User subroutine to deIine Fung's model
Chapter 2.27: Modeling Riveted Joint with CBUSH or CFAST
2.27
Modeling Riveted Joint with
Bushing, CFAST, or CWELD

Summary 684

Model Review 685

Results 693

Modeling Tips 694

Input Files 695

684 Marc Users Guide: Part I
CHAPTER 2.27
Summary
Title Modeling Riveted Joint with Bushing, CFAST or CWELD
Problem Ieatures Using empirical Iormulation to characterize the rivet.
Using the so-called point-wise and patch-wise connection.
Geometry The joint has 3 rows oI rivets in the loading direction. For analysis purpose only a slice
(one rivet-pitch wide) oI the joint is analyzed with a proper symmetric boundary
condition along the edges oI the plates.
Material properties E 60000 MPa, 0.3
Analysis type Quasi-static with geometrical non-linear analysis
Boundary conditions Clamped on the leIt side oI the joint. Symmetric displacement constraints along two
symmetry lines.
Applied loads Axial load oI 2400 N in the x-direction is applied on the right side oI the joint.
Element type Shell element type 75 and bushing element type 195 or beam element type 98.
FE results 1. DeIormed plot and Contour plot oI equivalent stress
2. Load transIer through the rivets
1 2 3
rivet diameter = 4
Units: mm
rivet pitch = 20
plate length = 160
plate overlap = 60
plate thickness = 1.2
v
4
13
22
30
39
48
56
65
74
83
91
100
109
118
126
135
144
152
161
170
179
187
196
205
214
222
231
240
248
257
266
lcase1
Equivalent of Stress Layer 1
Inc: 10
Time: 1.000e+000
X Y
Z
685
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
This example demonstrates modeling and analysis oI a lap joint. Two plates are joined using riveted connection. The
rivet is modeled with bushing element since its Ilexibility is determined empirically. The bushing elements are
connected with the plates using the so-called point-wise and patch-wise connection. The Iirst way requires that the
nodes oI the plates that need to be connected must be predeIined, since these nodes must belong both to the bushing
element and the plates. Thus, it puts a limitation on how the plates should be meshed. Moreover, this type oI connection
creates a nearly stress singularity in the plate around the rivet position.
The second way, patch-wise connection, is demonstrated using the CFAST model deIinition option. This method does
not require that the bushing nodes have to be congruent with the nodes oI the plate. Internally, CFAST creates a bushing
element and a set oI tying that connect the bushing nodes with a set oI nodes (this set oI nodes Iorm patches) oI the
plates. This type oI connection does not have a singularity as it does Ior point-wise connection.
For the patch-wise connection, another model using CWELD/PWELD model deIinition option is setup to simulate the
rivet connection. Internally, CWELD creates a beam element and a set oI tying that connect the beam nodes with a set
oI nodes (this set oI nodes Iorm patches) oI the plates In this case, the stiIIness oI the rivet is derived using the standard
Iormulation oI beam element by giving the geometry and the material properties oI the rivet.
Model Review
The plates will be meshed using standard Iinite elements. The rivets will be modeled using bushing element in which
their Ilexibility/stiIIness is calculated using an empirical or simple Iormula. The shear Ilexibility (see Vlieger, H.,
Broek, D., 'Residual Strength oI Cracked StiIIened Panels, Built-up Sheet Structure, Fracture Mechanics oI AircraIt
Structure, AGARD-AG-176, NATO, London, 1974) is calculated as Iollows:
The axial rivet stiIIness is calculated using a simple Iormula:
The rotational stiIIness' are assumed to be zero. For model with point-wise connection, small torsion stiIIness is added
to avoid system matrix singularity.
The geometry oI the model is quite simple. Here are the steps that should be Iollowed:
Step 1: Create the Iinite element mesh Ior the lower plate. There should be nodes at the location oI the rivets.
Step 2: Create the Iinite element mesh Ior the upper plate. There should be nodes at the location oI the rivet.
Step 3: DeIine GEOMETRY and MATERIAL Ior both plates
For point-wise connection with bushing element
Step 4: Create bushing elements that connect nodes oI the upper plate with the lower plate
Step 5: Create PBUSH
C
s
1
E
rv
d
----------- 5 0.8
E
rv
d
E
pl
t
pl
--------------
E
rv
d
E
pu
t
pu
----------------
\ .
| |
4.3x10
5

mm
N
---------
K
a
AE
L 2.4 mm ( )
---------------------------------- 314159
N
mm
---------
686 Marc Users Guide: Part I
CHAPTER 2.27
For patch-wise connection with CFAST/PFAST or CWELD/PWELD
Step 6: Create POINTS at the location oI the rivets
For CFAST/PFAST
Step 7: Create CFAST and PFAST
For CWELD/PWELD
Step 8: Create CWLED and PWELD
Step 9: Create boundary conditions and loading,. then create load case
Step 10: Submit the jobs
Step 11: Postprocessing the results
Step 1 to Step 3 is ending up with the creation oI the mesh Ior the plates as shown in Figure 2.27-1. Please run the
procedure Iile, step by step, until the MATERIAL deIinition
Figure 2.27-1 Finite Element Meshes for the Plates
For point-wise connection: using bushing element
Step 12: Create bushing elements that connect nodes oI the lower and upper plates at the rivet location. The created
bushing elements are shown in Figure 2.27-2.
MAIN
MESH GENERATION
ELEMENT CLASS: LINE (2)
ELEMS: ADD
83
216
97
Y
X
Z
687
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
Figure 2.27-2 Created Bushing Elements that Connect Lower- and Upper-plate
Step 13: Creating PBUSH by stepping through the Iollowing menus and Iilling in the requested value Ior stiIIness
properties as shown in Figure 2.27-3. And then assign this property Ior all bushing element created in Step 4.
230
111
244
MAIN
GEOMETRIC PROPERTIES
MECHANICAL ELEMETNS: 3-D
BUSHING (please note: CONNECTION toggle must be OFF)
STIFFNESS/DAMPING PROP.: VALUES
VALUE X = 3.14159e5
VALUE Y = 2.3226e4
VALUE Z = 2.3226e4
VALUE RX = 100.
VECTOR
0
1
0
Y
X
Z
688 Marc Users Guide: Part I
CHAPTER 2.27
Figure 2.27-3 Menus of PBUSH for Stiffness Input
For patch-wise connection: using CFAST/PFAST or CWELD/PWELD
Step 14: Create POINTS at the location oI the rivets
OK
ELEMENTS: ADD
All bushing elements
MAIN
MESH GENERATION
PTS: ADD
110 10 2.4
130 10 2.4
150 10 2.4
689
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
Figure 2.27-4 Created POINTS at the Location of the Rivets
Step 15: Creating CFAST and PFAST
MAIN
LINKS
CONNECTIONS
NEW
NAME: rivets
TYPE: FASTENER
CREATE AND SET
DIAMETER: 4
CREATE AND SET (please note: CONNECTION toggle must be ON)
STIFFNESS/DAMPING PROP.: VALUES
VALUE X = 3.14159e5
VALUE Y = 2.3226e4
VALUE Z = 2.3226e4
OK
2ND DIRECTION OF COORDINATE SYSTEM
COORDINATE SYSTEM: Global
OK
METHOD AND LOCATIONS
MASTER PATCH: FROM PATCH SET
ELEMENT END NODES: GENERATED
3
2
1
Y
Z
X
690 Marc Users Guide: Part I
CHAPTER 2.27
Figure 2.27-5 Menus to Create PFAST
PROJ. POINT'S: POINT
LOCATIONS: ADD
1 2 3 #
PATCH SETS
A: FACES: ADD
All faces belong to the lower plate
B: FACES: ADD
All faces belong to the upper plate
691
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
Figure 2.27-6 Menus to Create CFAST
Step 16: Creating CWELD and PWELD
MAIN
LINKS
CONNECTIONS
NEW
NAME: rivets
TYPE: WELD
CREATE AND SET
DIAMETER: 4
MATERIAL: aluminum
CREATE AND SET
VECTOR DEFINING LOCAL ZX-PLANE
COMPONENT IN GLOBAL SYSTEM
VECTOR
0 1 0
OK
METHOD AND LOCATIONS
METHOD: PATCH TO PATCH
MASTER PATCH: FROM PATCH SET
ELEMENT END NODES: GENERATED
692 Marc Users Guide: Part I
CHAPTER 2.27
The remaining three steps are creating boundary condition and loading, running the analysis, and postprocessing
the results.
Figure 2.27-7 Menus to Create PWELD
Figure 2.27-8 Menus to create CWELD
PROJ. POINTS: POINT
LOCATIONS: ADD
1 2 3 #
PATCH SETS
A: FACES: ADD
All faces belong to the lower plate
B: FACES: ADD
All faces belong to the upper plate
693
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
Results
The deIormed plot and the contour oI the von Misses stresses oI the lower plate Ior model with bushing, CFAST, and
CWELD are shown in Figure 2.27-9, Figure 2.27-10, and Figure 2.27-11, respectively. Comparing the stress contour oI
the model with bushing and CFAST, as expected, the point-wise connection shows a greater stress concentration
around the Iirst rivet.
Figure 2.27-9 Deformed Plot and Stress Contour of the Lower-plate for Model with Bushing/PBUSH
Figure 2.27-10 Deformed Plot and Stress Contour of the Lower-plate for Model with CFAST/PFAST
4
13
22
30
39
48
56
65
74
83
91
100
109
118
126
135
144
152
161
170
179
187
196
205
214
222
231
240
248
257
266
lcase1
Equivalent of Stress Layer 1
Inc: 10
Time: 1.000e+000
X Y
Z
694 Marc Users Guide: Part I
CHAPTER 2.27
Figure 2.27-11 Deformed Plot and Stress Contour of the Lower-plate for Model with CWELD/PWELD
The load transIer through the rivets using all types oI connection is shown in the Iollowing table. The load transIer
through the Iirst and third rivets Ior model with PBUSH are slightly less than that oI the model with CFAST. This is
obviously due to singularity condition which causes the eIIective stiIIness oI the rivet Ior the model with bushing
elements is less than that oI the model with CFAST.
The load transIer through the Iirst and third rivets using CWELD/PWELD is much greater than that oI using
CFAST/PFAST. This indicates that the stiIIness oI beam, given the geometry and material oI the rivet, is much greater
than that given by the empirical Iormula.
Modeling Tips
For geometrically complicated structures, modeling rivet joint with point-wise connection using bushing elements will
be a laborious task since it will need meshes with hard points at the rivet location. Moreover, this type oI connection
will create singularity at the point oI connection. CFAST and CWELD eliminate these drawbacks. For rivet connection,
F
Rivet-1
(N) F
Rivet-2
(N) F
Rivet-3
(N)
Point-wise
(CBUSH/PBUSH)
825 745 825
Patch-wise
(CFAST/PFAST)
843 711 843
Patch-wise
(CWELD/PWELD)
923 553 923
695
CHAPTER 2.27
Modeling Riveted Joint with Bushing, CFAST, or CWELD
CFAST has more Ilexibility to deIine the mechanical properties oI the rivet, normally deIined by using empirical
Iormula, compared to that with CWELD.
As extra exercises, please try the Iollowing variation oI the analysis:
Using scaled beam stiIIness with CWELD/PWELD to meet the value given by the empirical Iormula
Using noncongruent meshes with CFAST/PFAST or CWELD/PWELD
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
lapjoint_cbush.proc Mentat Procedure File Ior point-wise connection
lapjoint_cfast.proc Mentat Procedure File Ior patch-wise connection with CFAST
lapjoint_cweld.proc Mentat Procedure File Ior patch-wise connection with CWELD
lapjoint_cbush.dat Associated Marc Iile
lapjoint_cfast.dat Associated Marc Iile
lapjoint_cweld.dat Associated Marc Iile
696 Marc Users Guide: Part I
CHAPTER 2.27
Chapter 2.28: Speed and Memory Improvements
2.28
Speed and Memory
Improvements

Summary 698

Fast Integrated Composite Shells 698

Combined Multi Frontal Sparse and Iterative Solver 700

Storage of Element Data 702

Input Files 702

698 Marc Users Guide: Part I
CHAPTER 2.28
Summary
This chapter presents a number oI areas in which speed and memory improvements have been obtained. For composite
shell elements, diIIerent methods Ior integration through the thickness are available when elastic material is used; this
improves both speed and memory. A combined multi-Irontal sparse and iterative solver is available where the
decomposition oI the direct solver is used as a preconditioner Ior the iterative solver; this can improve speed Ior mildly
nonlinear problems. Introduced in MSC.Marc 2005r3 is a modiIication in the storage oI element data quantities, which
can lead to savings in memory usage when diIIerent types oI elements and layers are used.
Fast Integrated Composite Shells
The Iast integration option leads to signiIicant speed improvements Ior composite shell structures with a large number
oI layers. The improvements occur due to a diIIerent method oI integration through the thickness. The layers need to
have elastic material properties, either ISOTROPIC, ORTHOTROPIC, or ANISOTROPIC. When large displacements
occur the Total Lagrange (LARGE DISP) Iormulation should be used. The method can be set globally on the SHELL
SECT parameter and locally Ior each group oI elements with the COMPOSITE option. Three integration methods are
available:
Title Speed and Memory Improvements
Enhancement area`s Fast integration schema Ior elastic composite shells
Combined multi Irontal sparse and iterative solver
Storage oI element data
Speed improvements Ior
the presented models
Fast integrated composite shells
No thermal eIIects: speedup oI 4
With thermal eIIects: speedup oI 2
Combined multi Irontal sparse and iterative solver
Speedup oI 1.8 and 5
Memory improvements
Ior the presented models
Fast integrated composite shells
Memory reduction Iactor oI 2
Storage oI element data
Memory reduction Iactor ranging Irom 0 to 2 to 13
FULL Original method, can be used with all the material models
FAST NO THERMAL Integration method Ior elastic material without temperature eIIects
FAST THERMAL Integration method Ior elastic material with temperature eIIects
699
CHAPTER 2.28
Speed and Memory Improvements
The method can be selected in Mentat Ior composite materials as Iollows (see also Figure 2.28-1)
Figure 2.28-1 Define Integration Method
Analyzed is a plate supported by a rigid and deIormed by another rigid, see Figure 2.28-2. The plate has 30 layers,
consisting oI two diIIerent materials, and consists oI 7168 elements. Both a structural and a coupled thermal-structural
analysis is perIormed to compare the diIIerent integration methods.
*
Figure 2.28-2 Analyzed Plate Showing Deformation
MATERIAL PROPERTIES
MATERIAL PROPERTIES
LAYERED MATERIALS
NEW COMPOSITE
INTEGRATION METHOD: DEFAULT
700 Marc Users Guide: Part I
CHAPTER 2.28
shows the results oI the speed and memory improvements oI the diIIerent examples.
Combined Multi Frontal Sparse and Iterative Solver
This procedure can reduce the total solver time signiIicantly Ior nonlinear analyses. This is reached by using the
decomposition obtained Irom the multi-Irontal sparse solver as a preconditioner Ior the iterative solver. The Iirst
solution is obtained with the multi-Irontal sparse solver, and then in the next cycles or increments, the decomposition
is used as the preconditioner Ior the iterative solver. This is done until the iterative solver needs too many cycles to
Iind a solution, or when it Iails to Iind a solution. Then, the solution is again obtained with the multi-Irontal sparse
solver so that a new decomposition is available Ior the iterative solver Ior upcoming cycles and/or increments. II,
repeatedly, no solution is Iound by the iterative solver, this procedure is switched oII, and only the multi-Irontal sparse
solver is used. This procedure is useIul Ior mildly nonlinear problems. In Mentat, this solver can be set as Iollows
The number oI iteration can be controlled by
Table 2.28-1 Speed and Memory Improvements for Fast Integrated Composite Shells
Structural Coupled Thermal Structural
Full Fast no thermal Full Fast thermal
CPU time (s)
Total 202.6 54.9 265.5 136.5
Assembly 78.2 11.4 97.7 37.5
Stress Recovery 88.5 6.5 114.5 40.7
Memory (Mb)
Total 692 236 728 386
Incremental backup 222 20 186 113
Element storage 283 29 354 185
JOBS
MECHANICAL (or other analysis class)
JOB PARAMETERS
SOLVER
MIXED DIRECT-ITERATIVE
SERIAL ITERATIVE
MAX # ITERATION
40
701
CHAPTER 2.28
Speed and Memory Improvements
Note that the number oI iteration Ior the iterative solver should be low. The time needed Ior this amount oI iterations
should be much less than the time needed Ior the multi-Irontal sparse solver to Iind a solution. Marc has some logic to
come up with a number based on wall times. It will reduce the maximum number when it is to large. Note that since
the estimation Irom Marc is based on wall time, it can change when the analysis is repeated.
As an example, a rectangular block oI elastic-plastic material and a rubber top layer is analyzed. A rigid cylinder is
pressed into this block, and in the second loading stage, this cylinder is rolled. Large plastic deIormation is anticipated
in this analysis. The rubber is analyzed with a Mooney material model and Hermann elements since these elements
can handle the incompressibility oI the Mooney Material better. The elastic-plastic material is modeled with normal
brick elements. Figure 2.28-3 shows deIormation and the total plastic strain oI the rectangular block at the end oI
the analysis.
.
Figure 2.28-3 Contourplot of the Plastic Strain of the Deformed block
This model consists oI 8192 elements and needs about 400 Mbytes oI memory during the analysis. The total analysis
time is 3981 seconds Ior the multi-Irontal sparse solver versus 2236 seconds Ior the combined method. This a speedup
oI 1.8. Comparing only the solver times the speedup is even 2.7. Note that no solution is Iound when only the iterative
solver is used. A Iew increments oI this example are repeated Ior an ever Iiner meshed model, where each element was
subdivided into 8 new elements. This model consists oI 65536 brick elements and needs about 3.9 Gbytes oI memory
during the analysis. For this example, the total analysis time is 36467 seconds Ior the multi Irontal sparse solver versus
7276 seconds Ior the combined method. This is a speedup oI 5. Comparing only the solver times the speedup is 7.
702 Marc Users Guide: Part I
CHAPTER 2.28
Storage of Element Data
Released in MSC.Marc 2005r3 is a modiIication in the storage oI element data quantities like stresses and strains.
Previously, the amount oI memory allocated per element was the maximum needed Ior any element type. Now, the
elements are internally subdivided in groups, so that the memory allocation can be optimized with respect to the
number oI integration points and layers etc. in each group. The amount oI memory saved with this new element data
storage depends on the model, extreme savings are obtained when, e.g., a model consists oI a Iew composite bricks
with many layers and a lot oI tetrahedral elements. As an example, the disk drive head Irom the MSC.Marc 2005r3
users guide is taken (see Figure 2.28-4). This model is a combination oI shell and brick elements. Interestingly, in this
conIiguration, the model does not show a memory improvement. However, when the number oI layers oI the shell
elements is increased Irom 3 to 11, the memory saving becomes clear. Note that 3 layers is less then the deIault Ior
shell elements. For 11 layers, the memory needed to store the element data is Ior MSC.Marc 2005r2 134 Mbytes, and
the new version 67.9 Mbytes. When the brick elements are converted to tetrahedral elements, we see a memory
reduction Irom 2.87 Gbytes Ior MSC.Marc 2005r2 to 210 Mbytes Ior MSC.Marc 2005r3 oI the memory needed to
store the element data.
Figure 2.28-4 Model of Disk Drive Heads
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
703
CHAPTER 2.28
Speed and Memory Improvements
File Description
For Fast Composite Shell Integration
composite_plate.proc Mentat Procedure File
composite_plate.mud Associated Mentat model Iile
composite_plate.dat Associated Marc Iile
For Fast Composite Shell Integration, Coupled Analysis
composite_plate_cpl.proc Mentat Procedure File
composite_plate_cpl.mud Associated Mentat model Iile
composite_plate_cpl.dat Associated Marc Iile
For Mixed Solver
block_rolling.proc Mentat Procedure File
block_rolling.mud Associated Mentat model Iile
block_rolling.dat Associated Marc Iile
For Element Storage Example
disk_drive_head_tet.proc Mentat Procedure File with tetrahedral elements
disk_drive_head_tet.mud Associated Mentat model Iile with tetrahedral elements
disk_drive_head_tet.dat Associated Marc Iile with tetrahedral elements
disk_drive_head.proc Mentat Procedure File
disk_drive_head.mud Associated Mentat model Iile
disk_drive_head.dat Associated Marc Iile with 3 layers
disk_drive_head_11layer.dat Associated Marc Iile with 11 layers
704 Marc Users Guide: Part I
CHAPTER 2.28
Chapter 2.29: Implicit Viscoplastic Creep Analysis of Solder
2.29
Implicit Viscoplastic Creep
Analysis of Solder

Summary 706

Introduction 707

Flow Equation 707

Requested Solutions 708

Modeling Details 708

Loading and Boundary Conditions 711

Solution Procedure 713

Result and Plots 716

Conclusion 718

Input Files 718

Video 718
706 Marc Users Guide: Part 2
CHAPTER 2.29
Summary
Title Implicit Viscoplastic Creep Analysis oI Solder
Features Viscoplasticity Anand solder material model.
FE Mesh
Material properties Material Ior
Copper Block: E 1.30e6Mpa; ;Alpha1.78e-5
Ceramic Block: E 3.75e5Mpa; ;Alpha5.36e-6
Temperature dependent material properties are used Ior solder.
Analysis characteristics Nonlinear Elastoplastic Creep Analysis
Boundary conditions and
Applied loads
1. Copper Block is Iixed in all degrees oI Ireedom at the bottom.
2. Repetitive Symmetric boundary conditions are applied at the edges.
3. Cyclic Thermal loads applied on model with temperature ranging Irom 0C to
125C.
4. Sinusoidal structural displacement is applied on two middle nodes oI the ceramic
block varying Irom 0.5 to -0.5mm.
Element type 2-D 4-noded isoparametric, plain strain elements. (Element 11)
FE results 1. Total Equivalent Creep Strain.
2. Equivalent Creep Strain.
V 0.344
V 0.22
707
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
Introduction
Thermal cycling causes thermo-mechanical deIormations in electronic assemblies, which results in damage oI solder
connections. Accurate prediction oI stresses in the solder connections is a critical step in design Ior reliability oI
interconnected parts which requires a well deIined material model Ior solder. Currently Anand Solder model is used
in industry to model solders response to loading.
The Anand Solder material model is implemented under Viscoplastic Material Model in Marc/Mentat. The model is
implemented using a semi implicit (implicit in deIormation resistance and stress, explicit in temperature) approach,
and a backward Euler time integration scheme is adopted Ior solving the set oI constitutive viscoplastic equations.
Flow Equation
Anand Solder material model uses a single scalar internal variable, s, which denotes the averaged isotropic resistance
to macroscopic plastic Ilow oIIered by the underlying isotropic strengthening mechanisms such as dislocation density,
solid solution strengthening, subgrain, and grain size eIIects, etc. The deIormation resistance s is consequently
proportional to the equivalent stress. The Ilow equations and the evolution equations are give below:
Flow Equation:
where
where the evolution equation is expressed as:
Evolution Equations:
Note: The viscoplastic creep solder model needs a creep load case Irom the outset. It will not work in case a static
load case is Iollowed by a creep load case.
Single internal variable representing deIormation resistance
Pre-exponential Iactor
Multiplier oI stress
Strain rate sensitivity oI stress
Activation Energy/Boltzmann`s Constant
dc
p
dt
-------- A
o
s
-------
\ .
| |
sinh
1
m
----
Q
kT
------
\ .
| |
exp
s s
0
A

m
Q k
ds
dt
----- h
0
B
a B
B
-------
)
`

dc
p
dt
--------
B 1
s
s
*
----
s
*
s

1
A
---
dc
p
dt
--------
Q
kT
------
\ .
| |
exp
n

708 Marc Users Guide: Part 2

CHAPTER 2.29
where
represents the saturation value oI associated with a set oI given temperatures and strain rates. Thus the Anand
Solder model has nine material parameters: , , , , , , , , where is the initial value oI the deIormation
resistance needed to determine the evolution oI the deIormation resistance.
Requested Solutions
Numerical analysis is perIormed to Iind Equivalent Creep Strain and Total Equivalent Creep Strain values in the solder
material.
Modeling Details
The analysis model assembly is made oI ceramic and copper blocks bonded together at the ends and center by a
eutectic tin-lead solder and is subjected to cyclic thermal and structural loads. A 2D Plain strain modeling is done by
creating points, surIaces and then by creating mesh Irom surIaces using 'convert surIace to mesh option in Mentat.
The bottom oI the PCB is Iixed in all directions (ux,uy0) and the edges are subjected to repetitive symmetry boundary
conditions (ux0).Material properties and thermal loading are applied shown above.
Figure 2.29-1 Finite Element Model of the Structure
Hardening constant
DeIormation resistance saturation coeIIicient
Strain rate sensitivity oI saturation
Strain rate sensitivity oI hardening
Note: The equivalent Plastic strain in case oI the Solder model can be visualized by plotting oI the equivalent Creep
strain.
h
0
s

n
a
s
*
s
A Q m h
0
s

n a s
0
709
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
Element Modeling
The 4-noded plane strain elements (Element 11) have been used to mesh the model.The assumed strain option and
constant dilatation have been Ilagged on using geometry option.
Constant dilatational option is used Ior element 11 to avoid volumetric locking oI elements.
Assumed Strain Formulation is used Ior element 11 Ior improved results; especially in bending as it better captures the
linear variation in shear strain.
For more inIormation on these options, please reIer Marc Volume A: Theory and User Information.
Material Modeling
Elastic Plastic Isotropic material with viscoplastcity and is selected Ior Solder material. Method used is Anand Solder
with secant approximation material tangent. Temperature dependent material properties (Young's Modulus, Poissons
ratio and coeIIicient oI thermal expansion) are used Ior solder material and are deIined using Tables.
For copper and ceramic blocks, Elastic Plastic isotropic material without Viscoplasticity option is used.
To Create Tables in Mentat use the Iollowing steps.
Table and Co-ordinate Systems ~~New
Select the Independent Variable Type ~~Time (Ior loads) or Temp (Material Properties)
To Create Material Properties in Mentat use the Iollowing steps.
Material properties~~New~~Standard~~Structural~~Viscoplasticity.
710 Marc Users Guide: Part 2
CHAPTER 2.29
Anand Solder~~ Secant Approximation.
Select the appropriate tables in the Structural Properties widget.
Table 2.29-1 Constants used for Anand Solder Model
Anand
Solder
Constants
in Mentat Parameter Value Units
C1 Single internal variable (s0)
representing deIormation resistance
12.41 Stress
C2 pre-exponential Iactor (A) 400000
0
1/time
C3 the multiplier oI stress ( 1.5 Dimensionless
C4 strain rate sensitivity oI stress (m) 0.303 Dimensionless
C5 hardening constant (h0) 1379 Stress
C6 deIormation resistance saturation
coeIIicient (s)
13.79 Stress
C7 strain rate sensitivity oI saturation (n) 0.07 Dimensionless
C8 strain rate sensitivity oI hardening (a) 1.3 Dimensionless
Activation Energy / Boltzmann`s Constant 9400 Energy/Volume
Table 2.29-2 Temperature Dependent Material Properties used for Solder
Temperature
(C)
Young's
Modulus
(Mpa)
Mu
(Poissons
Ratio)
Thermal Expansion
Coefficient
(C)
0.00E00 3.56E04 3.58E-01 2.46E-05
5.00E00 3.49E04 3.58E-01 2.46E-05
2.50E01 3.19E04 3.58E-01 2.48E-05
5.00E01 2.82E04 3.60E-01 2.50E-05
7.50E01 2.44E04 3.61E-01 2.52E-05
1.00E02 2.07E04 3.64E-01 2.55E-05
1.25E02 1.70E04 3.67E-01 2.57E-05
711
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
Loading and Boundary Conditions
Figure 2.29-1 shows the loading and boundary conditions applied on the Iinite element model oI the assembly. The
total duration oI analysis is 120 seconds and the both the thermal and structural loads are applied as varying with
respect to time with the help oI tables.
To Create Tables in Mentat use the Iollowing steps.
Table and Coordinate Systems>New
Select the Independent Variable Type>Time (Ior loads) or Temp (Ior Material Properties) plotted in X-axis oI Graph.
Use data points option to add points or use Iormula to create the dependent variable plotted in Y-axis oI Graph.
Defining Temperature Dependent Material Properties in Mentat
712 Marc Users Guide: Part 2
CHAPTER 2.29
Thermal and Structural Loads Applied
Creating Tables for Sinusoidal Structural Loading using Formula Option
Creating Tables for Thermal Loading with Datapoints
713
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
To Create Boundary Conditions in Mentat use the Iollowing steps
Boundary Conditions>New>Type>Structural>Fixed Displacement
Boundary Conditions>New>Type>State Variable>Nodal Temperature
Select the appropriate tables in the Apply Properties widget.
Solution Procedure
The problem is analyzed in Marc with an implicit creep procedure.
Control parameters Ior the nonlinear solution scheme are described through the CONTROL option and AUTO STEP
option. This can be done using Mentat by the Iollowing steps
To create a new Creep loadcase in Mentat use the Iollowing steps
Loadcase>New>Creep
Creating and Applying Structural Loads by using Tables
Creating and Applying Thermal Loads by using Tables
714 Marc Users Guide: Part 2
CHAPTER 2.29
To create the Autostep option in Mentat use the Iollowing steps
Loadcases>Properties>Multicriteria
To Create the Control option in Mentat use the Iollowing steps
Loadcases>Properties>Solution Control
To set up a Job in Mentat use the Iollowing steps
JobsNew >Properties>Select Required Load case (under Available)
Creating Creep Load Case
715
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
Select the required results, Submit and post processing in Mentat can be done using
Jobs>New >Properties>Job Results tab
Set the Analysis DeIinition as plane strain.
Jobs>Run>Sumbit(1)
Jobs>Run>Open Post File
Selecting Required Load Case
716 Marc Users Guide: Part 2
CHAPTER 2.29
Result and Plots
Equivalent Creep Strain at Time t=30
Total Equivalent Creep Strain at Time t=30
717
CHAPTER 2.29
Implicit Viscoplastic Creep Analysis of Solder
Result Plot of Equivalent Creep Strain at Node = 232 (on Solder)
Result Plot of Total Equivalent Creep Strain at Node = 232 (on Solder)
718 Marc Users Guide: Part 2
CHAPTER 2.29
Conclusion
From the graph oI the results it is recognized that the creep strain is inIluenced more by the structural loading as
compared to thermal loads.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Video
Click on the link below to view a streaming video oI this problem.
File Description
SOLDER_TEST_pstrain.mud Mentat model Ior creep analysis oI solder.
create_points1.proc Procedure Iiles to generate points
mentat-sur-create1.proc Procedure Iiles to generate surIace
mentat-meshdone1.proc Procedure Iiles to generate mesh
File Description
ch02-29.swf Video Ior implicit viscoplastic creep analysis oI solder.
Chapter 2.30: Crack Propagation Capability in Shells
2.30
Crack Propagation Capability
in Shells

Summary 720

Introduction 721

Requested Solution 721

Modeling Details 721

Geometric Properties 721

Material Properties 722

Crack Modeling 722

Loading and Boundary Conditions 723

Solution Procedure 724

Results 725

Input Files 725

Video 725
720 Marc Users Guide: Part 2
CHAPTER 2.30
Summary
Title Crack propagation capability in shells
Features 2-D crack growth
Geometric Model with
Crack
FE Mesh
Material properties Isotropic material with,
E 2e11 N/mm2;
Analysis characteristics Nonlinear static analysis
Boundary conditions and
Applied loads
On end oI plate is completely Iixed. On the other end cyclic Iatigue load oI 0.2 is
applied over a time period oI 7.
Element type 4-noded quad shell element type 75 is used
FE results VCCT energy release
V 0.3
721
CHAPTER 2.30
Crack Propagation Capability in Shells
Introduction
Marc has excellent 'Iracture mechanics capabilities which include both 'J-integral techniques (LORENZI method)
as well as 'VCCT techniques. In Marc 2012 new Ieatures are added to enhance the current VCCT crack propagation
capabilities Ior 2-D and shell structures to
Allow crack growth through elements similar to XFEM
Allow growth in shell structures to biIurcate cracks to grow through shell stiIIeners and similar.
Support crack propagation with elasto-plastic materials
Improve accuracy in crack growth in shells with remeshing by providing a reIined mesh near the crack tip
The present chapter demonstrates Iew oI these capabilities.
Requested Solution
A numerical analysis will be perIormed to Iind the VCCT energy release & to see the crack propagation.
Modeling Details
As shown in Figure 2.30-1, the model comprises oI plate with stiIIener. There is a hole at the center oI plate. The plate
is Iixed on one end and, on the other end, cyclic Iatigue load oI 0.2 is applied over the time period oI seven.
Figure 2.30-1 Plate with Stiffener
Geometric Properties
Four-noded quad shell elements are used Ior both plate and stiIIener. Shell thickness oI 0.01 is deIined in the geometric
properties oI the elements as shown in Figure 2.30-2.
Initial Crack at Node 206
722 Marc Users Guide: Part 2
CHAPTER 2.30
Figure 2.30-2 Geometric Properties of Elements
Material Properties
Linear isotropic material properties are deIined as shown in Figure 2.30-3.
Figure 2.30-3 Linear Isotropic Material Properties
Crack Modeling
Crack is modeled using VCCT techniques as shown in Figure 2.30-4.
723
CHAPTER 2.30
Crack Propagation Capability in Shells
Figure 2.30-4 Crack Modeled with VCCT
Fatigue mode is selected as initial crack propagation mode. Crack propagation will be based upon cutting through
elements is selected as crack growth mode. Direction vector oI -1,-2.5, 0 is used Ior deIining crack growth direction.
Fixed crack growth increment oI 0.1 with time period oI 1 is used.
Loading and Boundary Conditions
One end oI plate is completely constrained is as shown in Figure 2.30-5.
Figure 2.30-5 Loading and Boundary Conditions
724 Marc Users Guide: Part 2
CHAPTER 2.30
On the other end, cyclic load oI 0.02 is applied over the time period oI seven as shown in Figure 2.30-6.
Figure 2.30-6 Cyclic Load
Following table is used Ior applying the cyclic load.
Figure 2.30-7 Cyclic Load Graph
Solution Procedure
Problem is solved with auto-stepping procedure with total time oI seven. Large strain option is turned on in the Jobs
menu.
725
CHAPTER 2.30
Crack Propagation Capability in Shells
Results
Figure 2.30-8 shows the results Ior VCCT energy release. During the course oI crack propagation, one can observe the
Ieatures such as crack through elements and crack biIurcation at stiIIener.
Figure 2.30-8 VCCT Energy Release Results
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Video
Click on the link below to view a streaming video oI this problem.
File Description
UG_2D_Crack.mud Mentat model Ior crack propagation in shells
UG_2D_Crack_job1.dat Marc input Iile Ior crack propagation in shells
File Description
ch02-30.swf Video Ior crack propagation capability in shells.
726 Marc Users Guide: Part 2
CHAPTER 2.30
Chapter 2.31: Segment-to-Segment with Friction
2.31
Segment-to-Segment with
Friction

Summary 728

Introduction 729

Contact Enhancements 729

Requested Solutions 729

Modeling Details 729

Contact 732

Solution Procedure 732

Result and Plots 733

Input Files 733

Video 734
728 Marc Users Guide: Part 2
CHAPTER 2.31
Summary
Title Segment-to-Segment with Friction
Features Segment-to-segment with Iriction
FE Mesh
Material properties Material Ior Ilexible body
Neo-Hookean material model deIined through the Mooney property menu
with
Analysis characteristics Nonlinear static analysis
Boundary conditions and
Applied loads
Moving: Rigid body is moved downward by 200 mm in -Y direction
All the loads have been applied to moving rigid body by activating position controlled
method in contact body creation option
Element type 3-D 4-noded Tetrahedral Herrmann (Element 157) elements
Contact properties 1. Flexible body is selI touching
2. Moving rigid body is glued to Ilexible body
3. Fixed rigid body is glued to Ilexible body
Bilinear Coulomb Iriction has been activated
CoeIIicient oI Iriction0.1 has been used Ior Ilexible body
FE results Comparison oI contact status between node-to-segment and segment-to-segment
C
10
100
729
CHAPTER 2.31
Segment-to-Segment with Friction
Introduction
The segment-to-segment contact algorithm introduced in Marc 2010 had a limitation oI small sliding without Iriction.
In Marc 2012, introduction oI the large sliding with support Ior Iriction has been implemented. Bilinear Coulomb and
the bilinear shear Iriction models have been introduced which were already available within node-to-segment contact
algorithm. In addition to these two Iriction models, the user is allowed to select non-symmetric stiIIness matrix which
is not available in node-to-segment algorithm. Non-symmetric stiIIness matrix capability would reduce the number oI
iterations but requires more memory and more cpu time per iteration. Furthermore, this method can only use the
solvers that allow non-symmetric matrices.
Contact Enhancements
Segment-to-segment with Iriction contact is applicable Ior
2-D and 3-D Ior linear and quadratic solid elements
2-D and 3-D Ior linear and quadratic shell elements
Small and Iinite sliding contact
With Bilinear Coulomb and Bilinear Shear
non-symmetric matrix is given to increase the speed oI convergence, but requires more memory
Analysis with diIIerent augmentation values
Rigid deIormable contact
DeIormable-to-deIormable contact
No Master-Slave Concept
Requested Solutions
A numerical analysis will be perIormed to compare contact status between node-to-segment and segment-to-segment.
Modeling Details
The model shown in Figure 2.31-1 is a structure having a deIormable body and two rigid bodies. Both rigid bodies are
glued to Ilexible body. Loading on the Ilexible body has been applied via moving rigid body by activating position
controlled method in moving rigid contact body creation option by 200 mm in -Y direction.
730 Marc Users Guide: Part 2
CHAPTER 2.31
Figure 2.31-1 Finite Element Model of the Structure
Element Modeling
Four-noded tetrahedral Herrmann elements (Element 157) have been used Ior Ilexible body.
Material Modeling
Neo-Hookean material model deIined through the Mooney property have been used Ior the Ilexible body.
Friction Modeling
Two diIIerent Iriction types are available: one is Bilinear coulomb; the other is Shear Bilinear Coulomb. Bilinear
Coulomb Iriction has been activated Ior both node-to-segment and segment-to-segment.
731
CHAPTER 2.31
Segment-to-Segment with Friction
Loading Conditions
As showing in Figure 2.31-1, the moving rigid body is coming down by 200 mm in the -Y direction. All the loading
on to the Ilexible body have been applied via rigid body by activating position controlled method in Moving rigid
contact body creation option as show below.
732 Marc Users Guide: Part 2
CHAPTER 2.31
Contact
The contact table is used to deIine the Iollowing three contact pairs:
SelI contact Ior Flexible body with coeIIicient oI Iriction0.1
Moving rigid body is glued to Ilexible body
Fixed rigid body is glued Ilexible body.
Solution Procedure
The problem is analyzed using both node to segment and segment to segment contact simulations with adaptive time
stepping. The contact status plots Ior these two models are given in Figures 2.30-2 and 2.30-3.
Figure 2.30-2 Contact Status of Flexible Body at a Distance = 200 mm in Node-to-Segment Contact
Analysis
733
CHAPTER 2.31
Segment-to-Segment with Friction
Figure 2.30-3 Contact Status of Flexible Body at a distance = 200 mm in Segment-to-Segment Contact
Analysis
Result and Plots
In node-to-segment contact simulation, penetrations are observed at many places as shown in Figure 2.30-2. Whereas
in segment-to-segment contact, no penetrations were observed and the contact results are better and much more
accurate compared to node-to-segment as shown in Figure 2.30-3. Apart Irom this, segment-to-segment contact shows
better perIormance with less number oI cycles and CPU time compared to node-to-segment simulation.
Input Files
# of increments # of Cycles Total CPU time
Node-To- Segment 28 240 111.20
Segment-To- Segment 29 160 84.59
File Description
nod2seg_fric.mud Mentat model Ior node-to-segment contact
nod2seg_fric_job1.dat Marc input Iile Ior node-to-segment contact
reference_sgmt2sgmt.mud ReIerence Iile used to end-to-end video demonstration
seg2seg_fric.mud Mentat model Ior segment-to-segment comtact
seg2seg_fric_job1.dat Marc input Iile Ior segment-to-segment contact
734 Marc Users Guide: Part 2
CHAPTER 2.31
Video
Click on the link below to view a streaming video oI this problem.
File Description
ch02-31.swf Video Ior segment-to-segment with Iriction.
Chapter 2.32: Directionally Dependent Friction Summary
2.32
Directionally Dependent
Friction Summary

Summary 736

Introduction 737

Requested Solutions 737

Modeling Details 737

Loading and Boundary Conditions 742

Contact 745

Solution Procedure 746

Result and Plots 747

Input Files 751

Video 751
736 Marc Users Guide: Part 2
CHAPTER 2.32
Summary
Title Directionally Dependent Friction
Features Anisotropic Friction
FE Mesh
Material properties Material Ior deIormable bodies, plate and block
Plate: E 50000 N/mm2 ;
Block: Neo-Hookean material model deIined through the Mooney property menu
with C
10
100
Analysis characteristics Nonlinear static analysis
Boundary conditions and
Applied loads
1. Block is moving Iorward using rigid cylinder by 100 mm in X and Z direction
respectively and then moving back by 100 mm in -Z and -X direction respectively.
2. Block is subjected to -100N Iorce in -Y direction using rigid cylinder when moving
Iorward and then -200N Iorce in same direction while moving back to original
position.
All the loads have been applied to rigid body cylinder with load control option
activated.
Element type 3D- 8 noded hexahedron (Elem 7) elements.
Contact properties 1. block is glued to cylinder
2. block is touching plate
3. plate is glued to support.
Anisotropic Iriction has been activated Ior contact pair block-plate. CoeIIicient oI
Iriction0.1 and 0.01 have been applied using local coordinate system whose X and Y
axes is parallel to global X and Z axes respectively.
FE results 1. Body Iorce on Cylinder vs. time plot.
2. Validate Isotropic Iriction Iormula
v 0.3
737
CHAPTER 2.32
Directionally Dependent Friction Summary
Introduction
For certain applications such as pipe lying, dragging oI chains, Bowden cable (cable in tube) the Irictional behavior is
dependent on the direction. The eIIective coeIIicient oI Iriction may be less in the direction along the cable or pipe and
greater in the perpendicular direction. In general there can be many applications where the coeIIicient oI Iriction
diIIers signiIicantly in two orthogonal directions.
Orthotropic Iriction law can be expressed in mathematical and graphical Iorms as below.
Requested Solutions
A numerical analysis will be perIormed to Iind body Iorce, Irictional Iorce in Iirst and second slip direction and normal
Iorce.
Modeling Details
The model shown in Figure 2.32-1 is a structure having a deIormable block moving on a deIormable plate which is
resting over a rigid support. Vertical load 100N is applied to the block and then the block is displaced over the plate
by 100 mm in X direction and Z direction respectively. AIter completion oI this motion, the vertical Iorce is
increased to 200 N and the block is moved in the reverse path so to reach the original position. All the loading on the
block have been applied via a rigid cylinder using the load controlled motion where a control node is deIine to apply
load and displacement to the cylinder which is blued to the block.
and Friction coeIIicients in Iirst and second slip direction
and Frictional Iorces in Iirst and second slip direction
Normal Iorce acting on touching node
f
t 1

1
-----
\ .
| |
2
f
t 2

2
-----
\ .
| |
2

1 2
f
n
s

1

2
f
t 1
f
t 2
f
n
738 Marc Users Guide: Part 2
CHAPTER 2.32
Figure 2.32-1 FE model of the structure
Element Modeling
Eight-noded hexahedral elements (Element 7) have been used Ior both the block and plate. Support and cylinder are
modeled using surIaces. For plate, assumed strain option and Ior block, both the assumed strain and constant dilatation
have been Ilagged using the geometry option.
739
CHAPTER 2.32
Directionally Dependent Friction Summary
Material Modeling
The plate is a linear isotropic material. The block is considered a Neo-Hookean material model deIined through the
Mooney menu.
Anisotropic Friction Modeling
For the anisotropic Iriction model, one has to both deIine two coeIIicients oI Iriction and also the orientation.
There are Iive ways to deIine orientations Ior the anisotropic Iriction.
1. Coordinate System
2. Curve
3. 1-D Element direction
4. Sharp edge direction
5. User subroutine
In this exercise, orientation Ior the anisotropic Iriction has been deIined using ''Coordinate System'' method. One local
coordinate system (crdsyst1) has been created such that X and Y axes is parallel to the global X and Z axes
respectively. CoeIIicient oI Iriction0.1 and 0.01 have been applied in 1st slip and 2nd slip direction.
740 Marc Users Guide: Part 2
CHAPTER 2.32
741
CHAPTER 2.32
Directionally Dependent Friction Summary
Activating the Anisotroic Friction model and prescribing the local orientation system.
DeIining the two coeIIicients oI Iriction on the Contact Table Entry Properties menu.
742 Marc Users Guide: Part 2
CHAPTER 2.32
Loading and Boundary Conditions
Figure 2.32-1 shows the loading and boundary conditions applied on the Iinite element model oI the solid structure.
Analysis is done in Iour load cases as explained below.
Load and displacement have been applied via Table
Loadcase-1 (motionX) Vertical load oI 100N is instaneously applied and displacement in X direction is
ramped to 100 mm
Loadcase-2 (motionZ) Vertical is held constant at 100N and displacement in Z direction is ramped to 100
mm.
Loadcase-3 (motion-Z) Additional vertical load oI 100N is instaneously applied and displacement in -Z
direction is ramped to 100 mm
Loadcase-4 (motion-X) Vertical load is held constant at 200N and displacement in -X direction is ramped to
100 mm
743
CHAPTER 2.32
Directionally Dependent Friction Summary
Table for Displacement in X Direction
Table for Displacement in Z Direction
744 Marc Users Guide: Part 2
CHAPTER 2.32
Table for Applying the Point Load in -Y direction
745
CHAPTER 2.32
Directionally Dependent Friction Summary
Contact
In total, Iour contact bodies are used. Block and plate are deIormable contact body while cylinder and support are rigid
body. There are three contact pair deIined.
1: block is touching plate
For this contact pair, contact detection method chosen is ''Iirst to second'' to ensure that block acts as slave body and
contact status on the its base, Iirst slip direction vector and second slip direction vector can be seen during post -
processing.
2: block is glued to cylinder
3: plate is glued to support.
746 Marc Users Guide: Part 2
CHAPTER 2.32
Solution Procedure
The problem is analyzed in Marc which is an implicit nonlinear solution procedure. Control parameters Ior the
nonlinear solution scheme are described through the Loadcase.Convergence control. Four load cases are created using
a Iixed time stepping procedure, where the time period oI each load case is one second and ten increments are taken
per load case.
747
CHAPTER 2.32
Directionally Dependent Friction Summary
Result and Plots
Position of Block and Contact Status at Base of the Block at time t=0.0
Position of Block and Contact Status at Base of the Block at time t=1.0
Position of Block and Contact Status at Base of the Block at time t=2.0
748 Marc Users Guide: Part 2
CHAPTER 2.32
Calculation oI the Iorce on the block and so on cylinder:
During the 1st loadcase block in moving in X direction
Force required to move the block would be * Reaction Iorce (R1) 0.1*10010 N
Here, R1 is the vertical point load applied on the block 100 N
During the 2nd loadcase block in moving in Z direction
Force required to move the would be * R10.01*1001 N
During the 3rd loadcase block in moving in -Z direction
Force required to move the would be * R20.01*2002 N
Here, R2 is the vertical point load applied on the block 200 N
During the 4th loadcase block in moving in -X direction
Force required to move the would be * Reaction Iorce (R2) 0.1*20020 N
Position of Block and Contact Status at Base of the Block at time t=3.0
Position of Block and Contact Status at Base of the Block at time t=4.0

1
749
CHAPTER 2.32
Directionally Dependent Friction Summary
Force in Cylinder in X-direction
Force in Cylinder in Y-direction
750 Marc Users Guide: Part 2
CHAPTER 2.32
Checking the result with anisotropic Iriction Iormula
The Iormula Ior anisotropic Iriction
,
At node number 3969 (end oI 3rd load case
Contact Frictional Iorce local X 4.01E-05N
Contact Frictional Iorce local Y 0.03729
Contact Normal Iorce RHS 3.72956
LHS oI the anisotropic Iormula 3.729
LHS RHS
Hence, anisotropic Iriction Iormula is satisIied
At node number 3765 (end oI 4th load case)
Contact Frictional Iorce local X 0.40085
Contact Frictional Iorce local Y 0.0000000197
Force in Cylinder in Z-direction
f
t 1

1
-----
\ .
| |
2
f
t 2

2
-----
\ .
| |
2

1 2
f n s

1
0.1
2
0.01
f
t 1
( )
f
t 2
( )
f
n
( )
f
t 1
( )
f
t 2
( )
751
CHAPTER 2.32
Directionally Dependent Friction Summary
Contact Normal Iorce RHS 4.0085
LHS oI the anisotropic Iormula 4.0085
LHS RHS
Hence, anisotropic Iriction Iormula is satisIied
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Video
Click on the link below to view a streaming video oI this problem.
File Description
dir_fric.mud Mentat model Ior direction dependent Iriction
dir_fric.dat Marc input Iile Ior direction dependent Iriction
reference_dir_fric.mud ReIerence Iile used in end-to-end video demonstration
File Description
ch02-32.swf Video Ior directionally dependent Iriction.
f
n
( )
752 Marc Users Guide: Part 2
CHAPTER 2.32
Chapter 2.33: Improved Accuracy with Remeshing of Herrmann Elements
2.33
Improved Accuracy with
Remeshing of Herrmann
Elements

Summary 754

Introduction 755

Simulation of Elastomeric Seal 755

Mesh 755

Material Properties 755

Contact Body Definitions 756

Contact Table 757

Mesh Adaptivity 758

Boundary Conditions 758

Solution Procedure 760

Results 761

Input Files 761

Video 762
754 Marc Users Guide: Part 2
CHAPTER 2.33
Summary
Title Improved accuracy with remeshing oI Herrmann elements
Problem Features The chapter demonstrates the improved accuracy such as more accurate results and improved
convergence associated with remeshing oI Herrmann integration elements 155, 156 and 157.
Model
Material Properties The deIormable rubber seal is modeled with Mooney constitutive model. The material
parameters are given as C18N/cm2 and C22N/cm2. The bulk modulus is 10000N/cm2.
Boundary
Conditions
The vertical Iaces oI rubber seal block are constrained by applying suitable boundary
conditions to restrict the lateral movement oI seal during deIormation.
The Iixed displacement oI one ramped over time period oI one is applied in negative
Y-direction on the surIace which is attached to the top elements Iaces oI rubber seal.
Analysis Type Static analysis with hyperelastic material behavior
Element Type Low-order Tet4 elements with Herrmann integration
FE Results History plot oI body Iorce & its comparison with that oI older version oI Marc
755
CHAPTER 2.33
Improved Accuracy with Remeshing of Herrmann Elements
Introduction
The chapter demonstrates the improved accuracy with associated with remeshing oI Herrmann integration elements.
Marc has the capability to perIorm the remeshing/rezoning automatically with the GLOBAL ADAPT option. However,
diIIiculties had been observed in the Marc remeshing analysis using elements 155, 156, and 157. Those diIIiculties
were inaccurate results and Iailure in achieving convergence immediately aIter remeshing/rezoning. With the new
implementations, these diIIiculties have been overcome.
Simulation of Elastomeric Seal
Model consists oI rubber seal is squeezed between top and bottom rigid plates by applying a pressure on the surIace
which is attached to the top element Iaces oI rubber seal block. Suitable boundary conditions are applied on the vertical
Iaces oI rubber seal to restrict the lateral movement oI seal during deIormation.
Mesh
The rubber seal is meshed with Tet4 elements with Herrmann integration oI type element 157.
Material Properties
The rubber seal is modeled with Mooney constitutive model. The material parameters are given as C18N/cm2 and
C22N/cm2. The bulk modulus is 10000N/cm2.
The material properties in Mentat are entered in the Iollowing Iigure.
756 Marc Users Guide: Part 2
CHAPTER 2.33
Contact Body Definitions
The rubber seal is deIined as 'deIormable body by adding the elements to the selection as shown in the Iollowing
Iigure.
The two surIaces (top & bottom) are deIined as rigid surIaces by adding the surIaces to the selection as shown in the
Iollowing Iigure.
757
CHAPTER 2.33
Improved Accuracy with Remeshing of Herrmann Elements
Contact Table
The contact relationship between the bodies is deIined via contact table. The deIormable rubber seal is deIined to be
in touching contact with two other rigid bodies as well as with itselI.
758 Marc Users Guide: Part 2
CHAPTER 2.33
Mesh Adaptivity
'Patran Tetra type mesh adaptivity was deIined by adding deIormable body 'rubber to the selection. Remeshing
Irequency oI Iive increments with immediate is deIined in remeshing criteria. Maximum strain change oI 0.4 is deIined
as advanced remeshing criterion. 800 numbers oI elements are deIined as remeshing parameters.
Boundary Conditions
The diIIerent boundary conditions applied on the model are as shown in the Iollowing Iigure
The vertical Iace which is on the opposite side oI the 'top rigid body is constrained in X and Z direction. The other
two larger vertical surIaces oI rubber seal block are constrained in Z-direction.
759
CHAPTER 2.33
Improved Accuracy with Remeshing of Herrmann Elements
The enIorced displacement one in ramped over time period oI one is applied in negative Y-direction on the surIace
which is attached to the element Iaces at the top oI rubber block seal.
760 Marc Users Guide: Part 2
CHAPTER 2.33
Solution Procedure
Problem is solved with Iixed-time stepping criterion 50 steps over time period oI 0.5. The large strain option is turned
on in Jobs menu. For large strain analysis procedure, the 'automatic option is selected as shown in the Iollowing
Iigure.
761
CHAPTER 2.33
Improved Accuracy with Remeshing of Herrmann Elements
The 'automatic option oI large strain analysis procedure writes out the LARGE STRAIN, 4 parameter in the input
deck. With the LARGE STRAIN, 4 parameter Ior each element and material combination, the optimal choice oI
Iormulator Ilags is automatically determined by the program. In this case, with the combination oI element 157 with
Mooney material model, Marc uses updated Lagrangian procedure with multiplicative decomposition (F
e
F
p
).
Results
The history plot oI body Iorce in Y direction on 'bottom rigid body is plotted against time. The plot oI Marc 2012 is
as shown in the Iigure below.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
Herrmann_Remesh_2011.mud Mentat model Ior remeshing with Herrmann elements
Herrmann_Remesh_2011_job1.dat Marc input Iile Ior remeshing with Herrmann elements
762 Marc Users Guide: Part 2
CHAPTER 2.33
Video
Click on the link below to view a streaming video oI this problem.
File Description
ch02-33.swf Video Ior improved accuracy with remeshing oI Herrmann elements.
Section 3: Mechanical Analysis
Section 3: Mechanical Analysis
Marc Users Guide 764
Chapter 3.1: Solid Modeling and Automatic Meshing
3.1
Solid Modeling and Automatic
Meshing

Chapter Overview 766

Background Information 766

Detailed Session Description 767

About HexMesh 783

Using HexMesh Parameters and Commands 787

Using HexMesh Example 797

Input Files 803

766 Marc Users Guide: Part 2
CHAPTER 3.1
Chapter Overview
The sample session described in this chapter demonstrates the process oI solid modeling and automatic meshing. A
simple bolt structure is modeled. The goal oI the analysis is to demonstrate:
Solid modeling, entering simple building blocks.
Using Boolean operations and blending techniques to complete the solid model.
Use oI symmetry to reduce the solid model.
Convert solid Iaces into surIaces and use oI automatic surIace meshers.
Use oI the automatic tetrahedral mesh generator to generate the mesh.
Use oI the symmetry and duplicate options to complete the model.
Background Information
This example demonstrates how to generate an element mesh Ior a simple bolt structure as shown in Figure 3.1-1.
Figure 3.1-1 Simple Bolt Structure
As can be seen Irom this Iigure, the model globally consists oI three simple geometrical components: two cylinders
with diIIerent radii and a 6-sided prism. A solid model oI the structure can be created using two Boolean operations.
Two cylinders must be united and Irom the resulting solid a prism must be subtracted.
X Y
Z
4
767
CHAPTER 3.1
Solid Modeling and Automatic Meshing
AIter these operations, a complex solid is obtained. Some oI the edges oI this solid must be given a speciIic curvature.
This is achieved with the BLEND operator.
BeIore using the mesh generators however, the solid model is reduced. The model is symmetrical with respect to a
segment oI 30. By subtracting two solid blocks Irom the solid model, the 30 segment is obtained.
Solid meshes are generated with a three step approach. First the Iaces oI a solid are converted to surIaces. These
surIaces are used to generate a surIace mesh. This surIace mesh is used to create a solid mesh.
AIter speciIying an average edge length Ior all edges oI the solid, the segment is automatically meshed, creating
tetrahedral elements. The resulting mesh subsequently expands to a mesh Ior the complete bolt by use oI the
SYMMETRY and DUPLICATE processors.
Overview of Steps
Step 1: Input of Basic Solids
Step 2: Refining the Solid Model
Step 3: Reducing the Solid Model to the Smallest Segment with Symmetry
Step 4: Surface Meshing on the Reduced Solid Model
Step 5: Meshing of the Solid Model Based on the Generated Surface Mesh
Step 6: Use of Symmetry and Duplication Operations to Complete the Mesh
Detailed Session Description
Step 1: Input oI Basic Solids
The approach used in generating the solid model is to start with three simple building blocks. The building blocks are
two cylinders and one prism.
MAIN
MESH GENERATION
SOLID TYPE
CYLINDER
RETURN
solids ADD
0 0 0 (solid cylinder origin coordinates)
0 0 1 (solid cylinder axis coordinates)
0.4 0.4 (solid cylinder radii)
0 0 1 (solid cylinder origin coordinates)
768 Marc Users Guide: Part 2
CHAPTER 3.1
In the above VIEW process, view 4 has been activated and set to a perspective projection. View 1 has been deactivated
to prevent switching on perspective plotting Ior this view.
Figure 3.1-2 Two Cylindrical Solids
The two basic cylinders will now be united.
0 0 2 (solid cylinder axis coordinates)
0.7 0.7 (solid cylinder radii)
VIEW
activate 4 (on)
activate 1 (off)
PERSPECTIVE
show 4
FILL
RETURN
MAIN
MESH GENERATION
SOLIDS
UNITE
1 2 (Pick solid 1 and 2)
769
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-3 Result of UNITE Operation
Next, the prism is deIined and subtracted Irom the current solid.
END LIST (#)
RETURN
MAIN
MESH GENERATION
SOLID TYPE
PRISM
RETURN
solids ADD
0 0 1.4 (solid prism base coordinates)
0 0 2.1 (solid prism axis coordinates)
0.4 (solid prism radius)
6 (number of solid prism faces)
SOLIDS
SUBTRACT
1 2 (Pick solids)
END LIST (#)
770 Marc Users Guide: Part 2
CHAPTER 3.1
The basic solid model is now completed and the result is shown in Figure 3.1-4. Note that at this stage only one solid
exists. The basic building blocks are no longer present in the database aIter the Boolean operations.
Figure 3.1-4 Result of SUBTRACT Operation
Step 2: ReIining the Solid Model
In this step, two blending operations are applied to the outer edges oI top cylinder oI the basic solid model. The
blending process consists oI speciIying a radius and indicating to which solid edge the blending operation is applied.
BeIore perIorming the blending operations, the solid edges is labeled. This is not strictly necessary Ior the process,
since all edges are graphically picked. For describing the process, however, it is useIul to indicate the edge labels.
Note that aIter each blending operation the edge numbering changes.
MAIN
MESH GENERATION
PLOT
solid SETTINGS (on)
edges LABELS
REGEN
RETURN
SOLIDS
blend RADIUS
0.1
blend EDGE
1:12 (Pick the solid edge)
771
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-5 Activating the Labeling of Solid Edges
The result aIter the Iirst blending is shown in Figure 3.1-6. The second edge will now be blended and the labelling oI
the solid edges will be switched oII.
MAIN
MESH GENERATION
SOLIDS
blend EDGE
1:20
PLOT
solids SETTINGS (off)
edges LABELS
REGEN
RETURN (twice)
772 Marc Users Guide: Part 2
CHAPTER 3.1

Figure 3.1-6 First Edge Blended
The solid modeling phase is completed as shown in Figure 3.1-7.
Figure 3.1-7 Completed Solid Model
Step 3: Reducing the Solid Model to the Smallest Segment with Symmetry
By looking at the model, we can observe that the solid model has certain symmetry planes. Since the complete model
(360) has a prism with six edges, the model can be considered as a duplication oI six segments.
773
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Furthermore, this 60 segment is symmetric. Thus, a 30 segment is the smallest section Ior which a mesh is required.
In this step, the solid model is reduced to a 30 segment. This is achieved by adding two solid blocks and subtracting
them Irom the solid model.
Figure 3.1-8 First Solid Block Added
MAIN
MESH
GENERATION
VIEW
show 1
FILL
RETURN
SOLID TYPE
BLOCK
RETURN
solids ADD
-1 0 -1 (solid block origin coordinates)
2 2 4 (solid block X, Y, and Z dimensions)
774 Marc Users Guide: Part 2
CHAPTER 3.1
The second block is duplicated Irom the Iirst one.
Figure 3.1-9 Duplicating the Solid Block
Finally, solid 3 and solid 2 is subtracted Irom body 1.
MAIN
MESH GENERATION
DUPLICATE
ROTATION ANGLES
0 0 150 (duplicate rotations in X, Y, and Z)
SOLIDS
2 (Pick the solid block)
END LIST (#)
MAIN
MESH GENERATION
SOLIDS
SUBTRACT
1 (Pick solid to subtract from)
2 (Pick solid to be subtracted)
3 (Pick solid to be subtracted)
END LIST (#)
775
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-10 Completed Solid Segment
Step 4: SurIace Meshing on the Reduced Solid Model
In this step, the surIace oI the reduced model is automatically meshed using one oI the automatic surIace meshers. First
all Iaces oI the solids are converted into surIaces; then plotting oI surIaces and points is switched oII.
FILL
VIEW
show 4
RETURN
FILL
MAIN
MESH GENERATION
SOLIDS
convert SOLID FACES TO SURFACES
all: EXIST.
PLOT
draw SOLIDS (off)
draw POINTS (off)
REGEN
RETURN (twice)
776 Marc Users Guide: Part 2
CHAPTER 3.1
Next, a target element edge length is speciIied by using the curve division command.
Here, Ior all curves, an average element edge length oI 0.1 is used. In order to ensure that elements generated Ior
speciIic surIace match, the curve divisions along particular curves is matched.
In Figure 3.1-11, the resulting seed points are shown.
Figure 3.1-11 Seed Points for Automatic Meshing
Next, the automatic surIace mesh generator is used to generate a surIace mesh. AIter the mesh generation, the drawing
oI the nodes is switched oII Ior easier veriIication oI the generated elements (the elements are drawn in solid mode
instead oI wireIrame mode).
MAIN
MESH GENERATION
AUTOMESH
CURVE DIVISIONS
AVG LENGTH
0.1 (enter length for curve divisions)
APPLY CURVE DIVISIONS
all: EXIST (for all curves)
MATCH CURVE DIVISIONS
0.005 (enter tolerance for vertex points)
all: EXIST. (for all curves)
777
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-12 Completed Surface Mesh for the Segment
Note that Ior each surIace, a surIace mesh has been created. Due to the use oI match curve divisions, it is ensured that
nodes along the curves are close to each other. However, there are still duplicate nodes at these points.
Step 5: Meshing oI the Solid Model Based on the Generated SurIace Mesh
The results oI the surIace mesh generator can be used as input Ior the solid mesh generators. It is required that a closed
surIace is present. Duplicate nodes present to the use oI the surIace mesher Ior each surIace are removed with the
SWEEP processor. This can be veriIied by plotting the outline edges only Ior the structure.
MAIN
MESH GENERATION
AUTOMESH
SURFACE MESHING
triangles (delaunay) SURFACE TRI MESH!
all: EXIST (for all surfaces)
PLOT
draw NODES (off)
draw SURFACES (off)
draw CURVES (off)
elements SOLID
REGEN
RETURN
778 Marc Users Guide: Part 2
CHAPTER 3.1
Subsequently, it is speciIied that the automatic tetrahedral mesher is used Ior all triangular elements generated in the
previous step.
MAIN
MESH GENERATION
PLOT
elements SETTINGS
OUTLINE
RETURN (twice)
SWEEP
TOLERANCE
.0001
NODES
all: EXIST
RETURN
PLOT
elements SETTINGS
REGEN
RETURN (twice)
MAIN
MESH GENERATION
AUTOMESH
SOLID MESHING
tetrahedra TET MESH!
all: EXIST (list of triangular elements)
779
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-13 Completed Tetrahedral Mesh
Step 6: Use oI Symmetry and Duplication Operations to Complete the Mesh
The mesh Ior the solid segment has been generated automatically. As discussed beIore, the complete model consists
oI six identical parts each with a symmetry plane. First, the SYMMETRY processor is used to generate a 60 segment.
Note that a Iormula is used to enter the symmetry plane normal.
MAIN
MESH GENERATION
SYMMETRY
NORMAL
sin(30*pi/180)
cos(30*pi/180)
0
ELEMENTS
all: EXIST.
780 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-14 Mesh after use of the SYMMETRY Operator
The generated 60 segment is now duplicated Iive times to generate the complete model.
MAIN
MESH GENERATION
DUPLICATE
ROTATION ANGLES
0 0 60
REPETITIONS
5
ELEMENTS
all: EXIST.
FILL
781
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-15 Element Mesh after the DUPLICATE Process
Both the SYMMETRY and the DUPLICATE processor generate new elements but do not check iI duplicate nodes are
created which have to be removed in order to make the connection between the diIIerent parts. This removal oI
duplicate nodes is achieved with the SWEEP process using the tolerance oI 0.001.
The mesh Ior the bolt structure is now completed. The model is saved in a Mentat model Iile.
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
SAVE
782 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-16 The SWEEP Processor
783
CHAPTER 3.1
Solid Modeling and Automatic Meshing
About HexMesh
Advantages of HexMesh
HexMesh generates a hexahedral mesh automatically Irom your CAD geometry enabling you to move rapidly Irom a
CAD model to a Iinite element model oI even the most complex shapes.
A model generated with HexMesh allows you to perIorm linear and nonlinear Iinite element analyses and to achieve
the kind oI quality results associated with Iinite element models composed oI hex elements.

Figure 3.1-17 Hexahedral Elements Generated by HexMesh
Advantages of Hexahedral Elements
A mesh with hexahedral elements is generally more accurate and requires Iewer elements than meshes that contain
tetrahedral elements. For complex geometries, hexahedral meshes are easier to visualize and edit than tetrahedral
meshes.
784 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-18 Interior of Model Meshed with HexMesh
Activating the HexMesh Feature
HexMesh is an add-on Ieature. II you purchased HexMesh with your original purchase or license renewal, your license
Iile includes the Ieature line, HEXMESH which activates HexMesh.
II you wish to purchase HexMesh Ior an existing license, contact your local MSC.SoItware oIIice. You will receive an
additional Ieature line Ior your license Iile Irom [email protected].
About the HexMesh Menu in Mentat
Use the HexMesh menu in Mentat to deIine the parameters and apply the commands Ior the HexMesh. To display the
HexMesh menu, choose MESH GENERATION-> AUTOMESH-> SOLID MESHING.
785
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-19 Automesh Solids Menu
About the Input for HexMesh
The HexMesh takes a description oI a surIace that is based on 3- or 4-node elements and perIorms an edge detection
and a hexahedral mesh generation on that surIace.
BeIore you apply the HexMesh, you should create a surIace mesh oI the volume to be meshed. This volume should be
totally enclosed with no Iree edges or torn seams`. The surIace mesh serves as a bounding surIace oI the volume to
be meshed.
parameters
commands
786 Marc Users Guide: Part 2
CHAPTER 3.1
Key Steps in the Meshing Process
Here are the key steps in the meshing process:
Figure 3.1-20 Key Steps in Meshing Process
You can regulate the accuracy and speed oI the hexmeshing operations by speciIying the diIIerent parameters and
applying the HexMesh commands in Mentat.
Create a surface
mesh with 3- or 4-
node elements
Check to ensure
that the outline
length of the
surface mesh is
zero
Select the
appropriate
parameters for the
hexahedral mesh
Need to verify
edges?
Invoke edge
detection
command and edit
edges if necessary
Apply MARC/Hexmesh
Yes
No
Note:
You may need to use the
SWEEP command to remove
coincident nodes.
Note:
You can use any of the
following surface meshers in
Mentat:
Delaunay
Advancing Front
Overlay
Note:
Use the Hexmesh menu in
Mentat to define the
parameters.
Note:
Selected edges
are considered to
be "real" edges.
787
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Using HexMesh Parameters and Commands
Specifying Element Size
Use the Element Size parameter to speciIy the sizes oI hexahedral elements generated in the x-, y-, and z-directions.
The deIault element sizes Ior the x-, y-, and z-directions are 0.1.
Figure 3.1-21 Element Size Parameter
The size oI the element determines the number oI resulting hexahedral elements. The Iollowing table demonstrates
how element size aIIects the meshing process:
To set the element size:
1. Click ELEMENT SIZE.
2. Type the element sizes in the x, y, and z-directions. You must speciIy an element greater than zero.
3. Press Enter.
Specifying Edge Sensitivity
Use the Edge Sensitivity parameter to speciIy when, in the edge detection process, the shared edge between two
input elements represents a 'real edge. The mesher uses these real edges to maintain the geometric representation
oI the model.
Figure 3.1-22 Edge Sensitivity Parameter
If you specify... then...
smaller elements, the quality oI the mesh is better. However, since there are
more elements, the meshing process is slower. Also, iI you
speciIy too small an element size, the meshing grid may
become too large and the mesher may Iail.
a large element size (in comparison to the object size), meshing might Iail.
Mesh Generation-> Automesh-> Solid Meshing
Mesh Generation-> Automesh-> Solid Meshing
788 Marc Users Guide: Part 2
CHAPTER 3.1
A higher value oI edge sensitivity makes the HexMesh more sensitive during the edge detection process. The deIault
value oI edge sensitivity is 0.5. The range oI values is 0 x 1.
How the Value of Edge Sensitivity Affects the Edge
Detection Process
The Iollowing illustrations show how the value oI edge sensitivity aIIects the edge detection process:
Edge sensitivity = 0:
Figure 3.1-23 Edge Detection Process with Fewer Edges Detected
789
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Edge sensitivity = 1:
Figure 3.1-24 Edge Detection Process with More Edges Detected
To speciIy edge sensitivity:
1. Click EDGE SENSITIVITY and type in a value.
2. Press Enter.
Specifying Gap
Use the Gap parameter to speciIy the size oI the gap that is initially leIt between the inner hexahedral elements and
the surIace during mesh generation.
AIter the mesher creates the overlay grid, it removes elements that are either close to or outside the surIace mesh
depending on the value oI the gap that you speciIy. The mesher then meshes the gap area.
The range oI values Ior the Gap parameter is -1 to 1. Negative values result in a smaller gap and can even result in
mesh penetration.
To set the value oI the Gap parameter:
1. Click GAP and type in a value.
2. Press Enter.
790 Marc Users Guide: Part 2
CHAPTER 3.1
How the Value of Gap affects the Mesh
Figures 3.1-25, 3.1-26, and 3.1-27 demonstrate how the value oI gap aIIects the mesh.
Figure 3.1-25 Gap Set to -0.3
Figure 3.1-26 Gap Set to 0
791
CHAPTER 3.1
Solid Modeling and Automatic Meshing
:
Figure 3.1-27 Gap Set to 1
Specifying the Number of Shakes
Shaking is a process oI global mesh enhancement where the nodes tend to move to places oI less potential energy. This
has a relaxing eIIect on the nodes and oIten results in a better mesh quality.
Higher values oI the Shakes parameter take up greater computing resources. Here are some guidelines Ior setting the
values oI the Shakes parameter Ior test and Iinal meshes:
To speciIy the number oI shakes:
1. Click SHAKES and type in a value.
2. Press Enter.
Using the Runs Parameter
II the HexMesh does not produce a valid mesh, it can automatically run again with a smaller element size. Using the
Runs parameter, you can speciIy the maximum number oI reruns perIormed by the HexMesh.
Figure 3.1-28 Runs Parameter
Situation Suggested Value
Test mesh 10
Final mesh 100
Mesh Generation-> Automesh-> Solid Meshing
792 Marc Users Guide: Part 2
CHAPTER 3.1
To speciIy the number oI runs:
1. Click RUNS and type in a value.
2. Press Enter.
To prevent reruns, type in the value, 1.
Using the Coarsening Parameter
Use the Coarsening parameter to speciIy a diIIerence in size between the elements in the interior and the elements in
the surIace. This may reduce the overall number oI elements generated.
Figure 3.1-29 Coarsening Parameter
You can speciIy one oI three levels oI coarsening - 0,1, or 2. A value, 0, indicates that there is no coarsening. A value,
2, speciIies that the elements in the interior can be up to Iour times larger on each side than the elements on the surIace.
To speciIy a level oI coarsening, click on the radio button next to the level.
How the Level of Coarsening affects the Elements
The Iollowing illustrations represent two diIIerent levels oI coarsening.
Coarsening set to 0:
Figure 3.1-30 Interior Elements of the Model are Uniform
Mesh Generation-> Automesh-> Solid Meshing
793
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Coarsening set to 2:
Figure 3.1-31 Interior Elements are larger than the Elements on the Surface
Using the Allow Wedges Parameter
Use the Allow Wedges parameter to create wedge elements iI an edge crosses the diagonal oI a Iace oI the hexahedral
element. This improves the quality oI the resulting mesh.
Figure 3.1-32 Allow Wedges Parameter OFF
no wedges
794 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-33 Allow Wedges Parameter ON
About the Coons Patches Parameter
Use the Coons Patches parameter to reduce the loss oI volume while meshing regions with curved surIaces. This
results in a smoother representation oI the input surIaces and a better approximation oI the input geometry. However,
this parameter consumes greater CPU resources.
The deIault setting Ior the Coons Patches parameter is OFF.
Using the Detect Edges Command
Use the Detect Edges command to automatically select geometric edges Irom an input list oI triangular and
quadrilateral elements that enclose the volume to be meshed. These detected edges help deIine the input geometry Ior
the HexMesh.
Figure 3.1-34 Detect Edges Command
The elements in the input list should be oriented with their tops Iacing outward and there must not be any Iree edges
or holes in the list.
wi t h wedge el ements
Mesh Generation-> Automesh-> Solid Meshing
795
CHAPTER 3.1
Solid Modeling and Automatic Meshing
To apply the Detect Edges command:
1. Click EDGE SENSITIVITY and speciIy a value other than 0.
2. Click DETECT EDGES and enter a list oI triangular or quadrilateral elements.
3. Press Enter.
When you apply the Detect Edges command, the detected element edges are automatically included in the list oI
selected edges. However, you can modiIy this list by selecting (or deselecting) edges beIore applying the HexMesh
command.
Selecting Edges
To select edges:
1. Choose Mesh Generation-~ Select.
2. Choose the select mode, AND.
3. Enter a list oI edges.
4. Press Enter.
Deselecting Edges
To deselect edges:
1. Choose Mesh Generation-~ Select.
2. Choose the select mode, EXCEPT.
3. Enter a list oI edges.
4. Press Enter.
Checklist for the HexMesh Command
BeIore you apply the HexMesh command you should ensure that:
the input list oI triangular and quadrilateral elements enclose the volume to be meshed.
there are no Iree edges or holes in the input list.
the elements are oriented with their tops Iacing outward
the length assigned to the element edges does not exceed the thickness oI geometry to be meshed (a good rule-
oI-thumb is: edge length 1/3 thickness oI the smallest section)
.
Any element edges that you select using the Detect Edges command are considered to
be real edges.
To remove these edges, clear them from the selection list using the Select menu
options in Mentat (see Deselecting Edges).
796 Marc Users Guide: Part 2
CHAPTER 3.1
Applying the HexMesh Command
To apply the HexMesh command:
1. Click HEXMESH
.
Figure 3.1-35 HexMesh Command
2. SpeciIy a list oI triangular and quadrilateral elements.
3. Press Enter.
About the Meshing Tools
The Iollowing table describes the operations supported by meshing tools available Ior the hexmesher:
Rectifying an Unsuccessful Hexmeshing Operation
II your hexmeshing operation is unsuccessIul, here are some measures that you can take beIore running the
operation again:
In the static menu area, click UNDO to return to the input mesh.
Check the detected edges and edit them iI necessary (see Using the Detect Edges Command).
Select a gap parameter value other than 0 (see Specifying Gap).
SpeciIy a diIIerent element size (see Specifying Element Size).
ModiIy the input list.
Check the Mentat shell window Ior any status, warning, and error messages:
Tool Operation
Outline Edge Length Computes the outline edge length. A value, 0, signiIies that there are:
no Iree edges
all elements have the same orientation
Sweep Outline Nodes Removes coincident nodes on the outline.
Tolerance SpeciIy tolerance Ior sweeping operation.
Align Shells Make all the surIace elements to have same orientation.
Check Mesh Checks mesh Ior distorted, upside-down, or inside-out elements. Reverses the orientation oI
elements, curves, and surIaces.
Clear Mesh Removes the entire mesh leaving the geometry intact.
Mesh Generation-> Automesh-> Solid Meshing
797
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-36 Mentat Shell Window
Using HexMesh Example
About the Example
The meshing example in this chapter demonstrates the steps in meshing a solid model with HexMesh. The example is
a procedure Iile, hexmesh.proc, and uses the model, hexmesh.mfd.
The procedure Iile and the model are located in the Mentat directory, examples/marc_ug.
Example Overview
The key stages in this example are:
Stage 1: Running the Procedure File.
Stage 2: Prepare the Input Model Ior SurIace-Meshing using the Delaunay SurIace Tri-Mesh.
Stage 3: Applying the Delaunay Tri-Mesh
Stage 4: Prepare the Input List Ior HexMesh using the HexMesh Parameters
Stage 5: Applying HexMesh
798 Marc Users Guide: Part 2
CHAPTER 3.1
Running the Procedure File
To run the procedure Iile, hexmesh.proc:
1. Choose UTILS-> PROCEDURES.
2. In the Mentat Procedure Control window, click LOAD.
3. In the Mentat Procedure Files window, locate the Iile, hexmesh.proc, in the directory, examples/marc_ug.
4. Click OK. The procedure Iile appears in the Mentat Procedure Control window.
5. Use one oI the Iollowing options to run the procedure Iile:
To run the procedure Iile without interruptions, click START/CONT.
To run the procedure Iile step by step, click STEP.
Figure 3.1-37 Procedure File Menu
Preparing the Model for Surface Meshing
To prepare the model Ior surIace-meshing using the Delaunay surIace tri-mesh:
1. Click FILL to make the entire model visible.
2. Click DRAW and turn the drawing oI nodes and points to OFF.
3. Choose VIEW-> VIEW STATUS-> SHOW VIEW 4.
run procedure file
without interruptions
press to stop or
pause file
run procedure file
step by step
799
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-38 Displaying View 4
4. Click MESH GENERATION-> AUTOMESH-> REMOVE FREE CURVES to remove curves not attached to
the surIace.
5. Click BREAK CURVES and speciIy:
a vertex tolerance (0.5)
a list oI curves (all existing)
6. Clean the surIace geometry by speciIying the Iollowing tolerance settings:
minimum tolerance (.01)
surIace parametric space tolerance (.01)
7. Click CLEAN SURFACE LOOPS and speciIy a list oI surIaces (all existing).
8. Click CHECK SURFACES and speciIy a list oI surIaces (all existing).
9. Choose AUTOMESH-> CURVE DIVISIONS-> TYPE.
10. SpeciIy a curve division with Iixed average length (1).
11. Click APPLY CURVE DIVISIONS and speciIy a list oI curves (all existing).
12. Click MATCH CURVES and speciIy:
a vertex tolerance (.05)
a list oI curves (all existing)
800 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-39 Displaying Matched Curves
Applying the Delaunay Tri-Mesh
To apply the Delaunay tri-mesh to the model:
1. Choose SurIace MESHING-> SURFACE TRI MESH.
2. SpeciIy a list oI curves (all existing).
Figure 3.1-40 Surface Tri-mesh Applied
801
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Preparing the Input List for HexMesh
To prepare the input list Ior HexMesh:
1. Sweep any extra nodes by speciIying:
a sweep tolerance (.05)
a list oI nodes to sweep (all existing)
2. Click PLOT and change the Iollowing plot settings to view the mesh more clearly:
Set the drawing oI curves and surIaces to OFF.
In the Elements areas, click SOLID to display the element Iaces in solid color.
3. Click REDRAW to redraw the model with the new settings.
4. Choose MAIN-> MESH GENERATION-> AUTOMESH-> SOLID MESHING.
5. In the Hexmesh area, CLICK EDGE SENSITIVITY and speciIy a value (.5).
6. Click DETECT EDGES to identiIy the geometric edges in the triangular elements and speciIy a list oI edges
(all existing).
7. Choose MAIN-> VISUALIZATION-> COLORS-> SELECT EDGES.
8. Change the selected edge color by speciIying a colormap number (23 1 0.6 0).
9. Choose MAIN-> MESH GENERATION-> AUTOMESH-> SOLID MESHING.
10. Click EDGE SENSITIVITY and speciIy a higher edge sensitivity (.6).
11. Click DETECT EDGES again and speciIy a list oI edges (all existing).
12. Click EDGE SENSITIVITY and set the edge sensitivity even higher (.7) to detect more edges.
13. SpeciIy a list oI edges (all existing).
14. Zoom in on the model and pick a Iew more edges.
15. Click FILL VIEW to make the entire model visible.
16. Rotate the model, zoom in, and pick a Iew more edges.
17. Rotate the model again to ensure that you picked all the edges.
18. Click RESET VIEW to reset the view to its original state.
19. Click FILL VIEW to make the entire model visible again.
802 Marc Users Guide: Part 2
CHAPTER 3.1
Figure 3.1-41 Edges Picked
Applying HexMesh
To apply HexMesh:
1. Choose MAIN-> AUTOMESH-> SOLID MESHING.
2. In the HexMesh area, click ELEMENT SIZE and speciIy the element sizes in x, y,
and z-directions (.25, .25, .25).
3. Click HEXMESH! and speciIy a list oI edges (all existing).
803
CHAPTER 3.1
Solid Modeling and Automatic Meshing
Figure 3.1-42 HexMesh Applied
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
solid_modeling_auto_meshing.proc Mentat procedure Iile to run the above example
hexmesh.proc Mentat procedure Iile to run the above example
hexmesh.mfd Associated model Iile
804 Marc Users Guide: Part 2
CHAPTER 3.1
Chapter 3.2: Manhole
3.2
Manhole

Chapter Overview 806

Background Information 806

Detailed Session Description 809

Conclusion 833

Input Files 833

806 Marc Users Guide: Part 2
CHAPTER 3.2
Chapter Overview
This chapter demonstrates the analysis oI a region where one cylinder penetrates another cylinder oI a larger radius
and where the structure is loaded by an internal pressure. The radius thickness ratio oI the structure warrants the use
oI shell theory instead oI a Iull three-dimensional analysis using hexahedral elements.
The objective oI this chapter is to highlight the Iollowing three Mentat capabilities.
Generation oI a cylinder-cylinder intersection.
Application oI Iace loads.
Display oI results in a contour plot.
Background Information
Description
In this session, you analyze a cylindrical pressure vessel penetrated by an oII-centered manhole. The diameter oI the
vessel is 168 inches and the plate thickness is 0.54 inches. The manhole has a diameter oI 48 inches and a plate
thickness oI 1.0 inches. The dimensions oI the structure are shown in Figure 3.2-1.
Figure 3.2-1 Vessel Dimensions
MANHOLE
VESSEL
Thickness = 1.0 inches
Radius = 24.0 inches
Thickness = 0.54 inches
Radius = 84.0 inches
Internal pressure = 100 psi
Length = 116 inches
45 inches off center
807
CHAPTER 3.2
Manhole
The manhole is positioned 45 inches Irom the center line as indicated by Figure 3.2-2.
Figure 3.2-2 Side View of Vessel
Idealization
Only a portion oI the vessel needs to be modeled due to symmetry and to the way arched structures respond to localized
loads. The thickness/radius ratio is small enough that it allows you to use the shell approximation instead oI a Iull
three-dimensional analysis. As the Iocus oI this analysis is on the response in the vicinity oI the penetration, the mesh
can be limited to the portion oI the structure shown in Figure 3.2-3.
It can be theoretically proven that Ior a material with a Poisson`s ratio oI 0.3, when measured at a distance
removed Irom the edge, the inIluence oI the penetration is reduced to 4 oI the value at the edge. Here r is the radius
and t the thickness. For this particular case, the decay distance is 16.84 inches.
ThereIore, the boundary conditions can be applied at the shell edge without aIIecting the stresses at the vessel-manhole
intersection. Due to symmetry, it is suIIicient to analyze halI the vessel section shown in Figure 3.2-3.
Vessel
Manhole
45 Inches
2.5 rt
808 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-3 Section to be Analyzed
Requirements for a Successful Analysis
The analysis is considered successIul iI the localized stresses are known at the intersection oI the two cylinders. The
decay distance oI 16.84 inches is assumed to be valid, and thereIore, the stresses at the edge oI the analyzed structure
should be less than 4 oI the peak value.
Full Disclosure
Analysis Type
Linear static.
Element Type
Marc Element Type 75, Iour-noded shell element.
Material Properties
Steel
Young's Modulus 30e6 psi
Poisson's Ratio 0.3
Overview of Steps
Step 1: Create two cylindrical surfaces: one for the vessel and one for the manhole.
Step 2: Convert the surface of the vessel into a finite element mesh.
Step 3: Remove the elements in the vicinity of the manhole, creating a hole in the surface of the vessel.
809
CHAPTER 3.2
Manhole
Step 4: Attach the nodes of the circumference of the hole to the intersection of the vessel and the
manhole surfaces.
Step 5: Redistribute the nodes on the perimeter of the hole.
Step 6: Add line elements to the circumference of the circular hole.
Step 7: Drag the line into shell elements thus creating the manhole.
Step 8: Attach the rim of the manhole to a flat surface.
Step 9: Subdivide the top row of elements of the manhole to improve the element aspect ratio of
these elements.
Step 10: Sweep the entire mesh to remove duplicate nodes.
Step 11: Apply boundary conditions.
Step 12: Apply material properties.
Step 13: Apply geometrical properties.
Step 14: Submit the analysis.
Step 15: Postprocessing: contour the equivalent von Mises stress on the structure.
Detailed Session Description
As mentioned in earlier sample sessions, the Iirst step is to establish a coordinate system. It seems natural to orient the
z-axis oI the global coordinate system in the direction oI the axial axis oI the vessel.
Step 1: Create two cylindrical surIaces: one Ior the vessel and one Ior the manhole.
Choose an origin that lies in the plane oI the end cap oI the vessel. This way the x-y axes oI the global coordinate
system span a plane that coincides with the plane oI the end cap. Idealization mentions the need to model only a quarter
section in circumIerential direction. It is in this quarter section oI the hull oI the vessel that the manhole is modeled.
Make use oI the ruled surIace to create the quarter section oI the vessel. The two curves necessary Ior ruled surIaces
are arcs oI equal radius extending 90 at a z-coordinate oI 0 and 116, respectively. Click on the Iollowing button
sequence to use the Center/Radius/begin Angle/end Angle arc type (CRAA) and to enter the data Ior the two curves.
MAIN
MESH GENERATION
CURVE TYPE
CENTER/RADIUS/ANGLE/ANGLE
RETURN
crvs ADD
0 0 0 (Center point)
810 Marc Users Guide: Part 2
CHAPTER 3.2
To make the two arcs visible, you need to deviate Irom the deIault viewpoint oI 0 0 1. There are two ways to do this:
you can change the view (and the viewpoint) by clicking the appropriate view number on the view menu, or you can
rotate the picture by an increment oI 45 about the y-axis.
Use the latter method and set the rotate increment in the VIEW menu using the Iollowing button sequence:
Now that both curves can be distinguished, add the surIace by Iirst speciIying the surIace type:
84 (Radius)
0 (Beginning angle)
90 (Ending angle)
0 0 116 (Center point)
84 (Radius)
0 (Beginning angle)
90 (Ending angle)
PLOT
curves SETTINGS
LABEL (on)
RETURN
REGEN
FILL
MAIN
VIEW
VIEW SETTINGS
model increments ROTATE
45
RETURN
RY+ (in the static menu next to RX+)
FILL
MAIN
MESH GENERATION
SURFACE TYPE
RULED
RETURN
srfs ADD
811
CHAPTER 3.2
Manhole
To pick the two previously deIined curves, use the <ML> with the in the vicinity oI the curve. The program
displays the surIace. Similar to the button sequence outlined above, set the surIace type to CYLINDER to add the
surIace oI the manhole. The surIace oI the manhole is only used here to determine the intersecting curve; it is not used
as a primitive entity to be converted to elements.
The basic geometry oI the model is now complete. Rotate the picture about the y-axis over -45. Switch oII the drawing
oI points and display Iour views oI the model. Fill the graphic area Ior all views aIter activating them.
1 (Pick first curve)
2 (Pick second curve)
MAIN
MESH GENERATION
SURFACE TYPE
CYLINDER
RETURN
srfs ADD
45 40 58 (1st point on the axis of the cylinder)
45 120 58 (2nd point on the axis of the cylinder)
24 24 (Radii at 1st and 2nd point)
MAIN
RY- (in the static menu next to RX -)
PLOT
draw POINTS (off)
RETURN
VIEW
SHOW ALL VIEWS
ACTIVATE ALL
FILL
< >
812 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-4 Four Views of Completed Model Geometry
Step 2: Convert the surIace oI the vessel into a Iinite element mesh.
Use the CONVERT processor to convert the surIace oI the vessel to Iinite elements. Click on the Iollowing button
sequence to create a mesh oI 20x20 elements on the quarter cylinder surIace.
To get a better overview oI the model, change the view option to 2 and deactivate the Iace identiIication option to
clariIy the picture. Display curves and surIaces using a high accuracy. Figure 3.2-5 illustrates where the manhole
cylinder penetrates the surIace oI the vessel.
MAIN
MESH GENERATION
CONVERT
DIVISIONS
20 20
SURFACES TO ELEMENTS
1 (Pick the ruled surface)
END LIST (#)
PLOT
elements SETTINGS
FACES (off)
RETURN
curves SETTINGS
813
CHAPTER 3.2
Manhole
Figure 3.2-5 Elements Generated on Vessel Surface
Step 3: Remove the elements in the vicinity oI the manhole, creating a hole in the surIace oI the vessel.
Remove a group oI 6x6 elements Irom the vessel surIace that occupy the hole caused by the penetrating manhole.
Next, all unused nodes must be removed.
HIGH
RETURN
surfaces SETTINGS
predefined settings HIGH
RETURN
REGEN
VIEW
show 2 (Below SHOW ALL VIEWS)
RETURN
MAIN
MESH GENERATION
elems REM
(Box pick the elements)
END LIST (#)
SWEEP
remove unused NODES
814 Marc Users Guide: Part 2
CHAPTER 3.2

Figure 3.2-6 Vessel with Elements Removed
Step 4: Attach the nodes oI the circumIerence oI the hole to the intersection oI the vessel and the manhole surIaces.
The surIace oI the vessel now has a square hole. The nodes on the perimeter oI the square hole must now be attached
to the intersection oI the vessel and manhole surIaces which is done using the Iollowing button sequence:
Relax the nodes while keeping the outline oI the mesh Iixed, using the button sequence given below. The resulting
mesh is shown in Figure 3.2-7.
MAIN
MESH GENERATION
MOVE
MOVE TO GEOMETRIC ENTITIES
move nodes INTERSECT
2 (Pick the manhole surface)
1 (Pick the vessel surface)
(Box pick the nodes on the perimeter of the hole)
END LIST (#)
MAIN
MESH GENERATION
RELAX
NODES
all: EXIST.
815
CHAPTER 3.2
Manhole
Figure 3.2-7 Peripheral Nodes Attached to Cylinder
Step 5: Redistribute the nodes on the perimeter oI the hole.
Figure 3.2-7 clearly indicates that the mesh pattern around the hole is not optimal. The primary cause oI this is the
irregular node distribution around the hole. In order to redistribute the nodes, you must STRETCH the nodes in groups
so that the nodes are evenly distributed as indicated in Figure 3.2-8.
The Iollowing button sequence is used to stretch the nodes:
Repeat this operation Ior the other nodes on the perimeter oI the hole as indicated in Figure 3.2-8.
MAIN
MESH GENERATION
STRETCH
NODES
226 (Pick the first node of the stretch node path)
161 (Pick the last node of the stretch node path)
END LIST (#)
816 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-8 Evenly Distributed Nodes
It is obvious Irom the picture, the stretch operation no longer preserves the requirement that the perimeter oI the hole
is on the intersection oI the two main surIaces.
To re-attach the nodes, use a directed attach method which guarantees that the nodes are moved to the intersection
along a speciIied direction. The Iollowing button sequence demonstrates the steps required to apply the directed
attach method.
Repeat this process Ior all Iour sides that have been stretched using a diIIerent direction Ior each side. The result oI
the Iirst attach operation is shown in Figure 3.2-9.
MAIN
MESH GENERATION
MOVE
MOVE TO GEOMETRIC ENTITIES
move nodes INTERSECT
2 (Pick the surface)
1 (Pick the surface)
247 268 289 288 (Pick the nodes)
END LIST (#)
817
CHAPTER 3.2
Manhole
Figure 3.2-9 Using the Directed Attach Method to Re-Attach the Nodes
As noted beIore, it is suIIicient to create only halI oI the model shown in Figure 3.2-10 due to symmetry. The reason
Ior generating the entire model is that the nodes on the boundary oI a mesh remain at their location during relax
operation and only interior nodes are moved. Had we generated only halI the model, the nodes on the line oI symmetry
(in the XY plane) would have required a manual redistribution.
MAIN
MESH GENERATION
elems REM
(Box pick all elements below the symmetry line)
END LIST (#)
SWEEP
remove unused NODES
818 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-10 Removing Unused Nodes
Step 6: Add line elements to the circumIerence oI the circular hole.
There are several ways to create the manhole. The user is to Iollow the same steps used Ior the vessel. The cylindrical
surIace is Iirst converted to elements. The bottom edge oI the manhole elements is then attached to the intersecting
line oI the vessel and manhole surIace. Instead, you use a diIIerent approach that involves the use oI line elements.
The edge oI the existing hole in the vessel is plated with line elements that serve as the generating elements
in an expand operation. Use the Iollowing button sequence to create line elements to the exposed side oI the
quadrilateral elements:
Use the Iollowing button sequence to select and store the line elements generated by this operation into a set name Ior
later reIerence.
MAIN
MESH GENERATION
CONVERT
EDGES TO ELEMENTS
(Pick the edges at the perimeter of the hole)
END LIST (#)
MAIN
MESH GENERATION
SELECT
SELECT BY
819
CHAPTER 3.2
Manhole
The EXPAND processor drags line elements into shell elements and shell elements into volume elements, eIIectively
increasing the dimensionality oI the element type by one. Use the EXPAND operation to drag the line elements
equidistantly over 10 inches Ior three layers. The rim oI the manhole created in this manner has the same shape as the
intersection line oI the two cylinders.
Step 7: Drag the line into shell elements thus creating the manhole.
Use the Iollowing button sequence to perIorm the expand operation.
Use the SWEEP processor and click on NODES Irom the SWEEP panel to eliminate the duplicate nodes created by
the expand operation. Click on the all: EXIST. button oI the static menu to indicate that you want to rid the entire mesh
oI duplicate nodes. You can veriIy the elimination oI the nodes by only drawing the outline oI the mesh.
elements by CLASS
LINE(2)
OK
RETURN
elements STORE
sticks
all: SELECT.
CLEAR SELECT
MAIN
VIEW
SHOW ALL VIEWS
RETURN
MESH GENERATION
EXPAND
TRANSLATIONS
0 10 0
REPETITIONS
3
ELEMENTS
sticks
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
820 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-11 Manhole with Nearly Correct Coordinates
Step 8: Attach the rim oI the manhole to a Ilat surIace.
Attach the doubly curved rim oI the manhole to a patch. To create the patch, select QUAD as the current surIace type.
Use the coordinates given below to add the patch in a local coordinate system. Create the local coordinate system by
rotating 90 about the global x-axis and translating it 112 inches in the global y-direction.
MAIN
MESH GENERATION
SURFACE TYPE
QUAD
grid ON (on)
RETURN
coordinate system SET
U DOMAIN
-100 100
U SPACING
10
V DOMAIN
-100 100
V SPACING
10
grid ON (on)
821
CHAPTER 3.2
Manhole
Figure 3.2-12 Creating the Patch
ROTATE
90 0 0
TRANSLATE
0 112 0
RETURN
PLOT
draw POINTS (on)
RETURN
pts ADD
-10 0 0
100 0 0
100 100 0
-10 100 0
FILL
srfs ADD
29 30 31 32 (pick the four points generated above)
GRID (off)
FILL
822 Marc Users Guide: Part 2
CHAPTER 3.2
The nodes on the cut-oII boundary oI the manhole need to be moved to the intersection oI the two surIaces. Use the
Iollowing button sequence to move the nodes:
The results oI the move operation are shown in Figure 3.2-13.
Figure 3.2-13 Moved Nodes to Patch
Step 9: Subdivide the top row oI elements oI the manhole to improve the element aspect ratio oI these elements.
Subdivide the top row oI elements in the second direction oI connectivity to improve the aspect ratio. Once again, it
is most convenient to use the Polygon Pick Method to select the elements.
MAIN
MESH GENERATION
MOVE
MOVE TO GEOMETRIC ENTITIES
move nodes INTERSECT
2 (Pick the cylinder)
3 (Pick the quad)
(Use the Polygon Pick Method to
pick the nodes from view 1)
MAIN
MESH GENERATION
SUBDIVIDE
823
CHAPTER 3.2
Manhole
Figure 3.2-14 Improved Aspect Ratio for Top Row Elements
Step 10: Sweep the entire mesh to remove duplicate nodes.
Sweep the mesh to eliminate duplicate nodes aIter each operation that generates elements.
You have now completed the topological part oI the mesh. For subsequent tasks, it is no longer required to use the
geometric entities points, curves, and surIaces, and thereIore, the plotting oI these items is switched oII.
DIVISIONS
1 2 1
ELEMENTS
(Pick the top row of elements)
END LIST (#)
MAIN
MESH GENERATION
SWEEP
NODES
all: EXIST.
PLOT
draw POINTS (off)
draw CURVES (off)
824 Marc Users Guide: Part 2
CHAPTER 3.2
Step 11: Apply boundary conditions.
There are two types oI symmetry conditions across the edge that cut the vessel and manhole in halI:
1. Zero displacement in z-direction,
2. Zero local rotations along the edge.
The Iirst boundary condition (1) is expressed in global coordinates. To apply the second boundary condition, it is
necessary to apply a transIormation to the nodes oI the vessel such that the boundary conditions can be expressed as
a Iunction oI the global degrees oI Ireedom. Use the Iollowing button sequence to create the transIormations.
The boundary conditions (1) and (2) mentioned on the previous page are then applied through the Iollowing
button sequence:
draw SURFACES (off)
REGEN
FILL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
VIEW
show 2
RETURN
TRANSFORMS
CYLINDRICAL
0 0 0
0 0 100 (center line)
(Pick the nodes along the curved and straight edges
of the vessel; not those of the manhole)
END LIST (#)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT Z (on)
ROTATION Y (on)
OK
825
CHAPTER 3.2
Manhole
Figure 3.2-15 Boundary Conditions Applied
The other curved edge oI the vessel has an edge load applied in the direction oI the center line.
nodes ADD
(Pick the nodes along the symmetry
plane of the vessel and manhole)
END LIST (#)
VIEW
show 4
RETURN
FILL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
EDGE LOAD
PRESSURE
-4200
OK
edges ADD
826 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-16 Edge Loads Applied in Direction of Center Line
The vessel is under internal pressure which is applied through the Iollowing button sequence:
(Pick the edges on the curved side
of the vessel)
END LIST (#)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FACE LOAD
PRESSURE
-100.0
OK
faces ADD
all: EXIST.
Note: The deIinition oI a positive pressure is one that is directed towards the Iace oI the element.
827
CHAPTER 3.2
Manhole
Figure 3.2-17 shows that the pressure on the manhole is applied as an external pressure. Two methods can be used to
correct the direction in which the load is applied: either the sign oI the applied pressure load Ior the manhole elements
is changed or the direction oI the connectivity oI the elements oI the manhole is changed.
Figure 3.2-17 Internal Pressure Applied
The latter method used in the container sample session described in Chapter 3.30 is also used in this session and
invoked with the Iollowing button sequence:
MAIN
MESH GENERATION
CHECK
VIEW
show 2
RETURN
FLIP ELEMENTS
(Use the Polygon Pick Method to select
the elements of the manhole)
END LIST (#)
VIEW
show 4
RETURN
828 Marc Users Guide: Part 2
CHAPTER 3.2
Figure 3.2-18 Corrected Load Direction
The Iollowing two types oI symmetry conditions exist along the straight edges oI the vessel: displacement in tangential
direction is zero, rotation in axial direction is zero.
Due to the previously deIined transIormations, these boundary conditions can be applied using the Iollowing
button sequence:
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
ROTATION Z (on)
OK
nodes ADD
(Pick the nodes on the straight
edges of the vessel)
END LIST (#)
829
CHAPTER 3.2
Manhole
Figure 3.2-19 Boundary Conditions Applied to Vessel Edges
Finally, an edge load is applied to the top perimeter oI the manhole.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
VIEW
show 1
RETURN
EDGE LOAD
PRESSURE
-1200
OK
edges ADD
(Pick the edges at the top rim of the manhole)
END LIST (#)
VIEW
show 4
RETURN
830 Marc Users Guide: Part 2
CHAPTER 3.2

Figure 3.2-20 Edge Load Applied to Top Rim of Manhole
Step 12: Apply material properties.
The material properties Ior both the vessel and the manhole are speciIied in Material Properties. Use the Iollowing
button sequence to apply steel properties to the two structures.
Step 13: Apply geometrical properties.
The manhole is manuIactured out oI a steel plate with a thickness oI 1 inch. The thickness oI the vessel is 0.54 inches.
Click on the GEOMETRIC PROPERTIES button oI the main menu and go to the 'mechanical elements 3-D submenu.
Enter the SHELL pop-up menu, click on the THICKNESS button and type in 0.54. To conIirm the correctness, click on
the OK button. Assign the thickness to the elements oI the vessel only. Repeat this process Ior the manhole using the
Iollowing button sequence:
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
30.0e6
POISSON'S RATIO
0.3
OK
elements ADD
all: EXIST.
831
CHAPTER 3.2
Manhole
ConIirm the correctness oI the thickness application using the ID GEOMETRIES button.
You have now completed the modeling process, Step 2 oI the Analysis Cycle. Continue with the preparatory steps Ior
the Iinite element analysis.
As this is a linear static analysis, you do not need to create a loadcase. The INITIAL LOADS option in the JOBS menu
is used to speciIy the loading pattern.
Step 14: Submit the analysis.
Use the Iollowing button sequence to deIine the Marc element type, to veriIy that the appropriate initial loads have
been activated, to speciIy the desired result variables, and to submit the job.
MAIN
GEOMETRIC PROPERTIES
mechanical elements 3-D
SHELL
THICKNESS
.54
O
K
VIEW
show 2
RETURN
elements ADD
(Pick the elements of the manhole)
END LIST (#)
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
3-D MEMBRANE/SHELL
75
OK
all: EXIST.
RETURN
RETURN
MECHANICAL
JOB RESULTS
832 Marc Users Guide: Part 2
CHAPTER 3.2
Step 15: Postprocessing: contour the equivalent von Mises stress on the structure.
The screen is updated periodically to report the progress oI the job. II the job has been successIully completed, the exit
message on the panel will be 3004.
To display the results oI the analysis Ior interpretation, use the Iollowing button sequence:
available element tensors
Stress
layers: OUT & MID
scalars
Equivalent Von Mises Stress
layers: OUT & MID
OK
INITIAL LOADS
OK
JOB PARAMETERS
#SHELL/BEAM LAYERS
3
OK (twice)
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
NEXT
PLOT
draw NODES (off)
RETURN
SCALAR
Equivalent von Mises Stress Layer 1
OK
CONTOUR BANDS
DEF & ORIG
FILL
833
CHAPTER 3.2
Manhole
Figure 3.2-21 shows the resulting model contoured with von Mises stresses.
Figure 3.2-21 Model with von Mises Stress Contours
Conclusion
Due to reproduction constraints, Figure 3.2-21 does not give you a clear representation oI the actual resulting stress
distribution that appears in the graphics area on your screen. The results indicate that, due to the penetration oI the
manhole into the vessel, the localized stress concentrations occur near the intersection.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
manhole.proc Mentat procedure Iile to run the above example
834 Marc Users Guide: Part 2
CHAPTER 3.2
Chapter 3.3: Contact Modeling of Pin Connection Joints with Higher-Order Elements
3.3
Contact Modeling of Pin
Connection Joints using
Higher-Order Elements

Chapter Overview 836

Pin Connection 837

Input Files 850

836 Marc Users Guide: Part 2
CHAPTER 3.3
Chapter Overview
This chapter is intended to show the Marc capability oI modeling contact using quadratic elements. You have a choice
as to whether the boundaries are to be linearized or genuine quadratic contact is to be used. Quadratic contact takes
into account the curved geometry and shape Iunctions oI such elements. This implies that both the corner and mid-side
nodes may come into contact. In case oI deIormable contact, searching Ior contact is done with respect to curved
elements and the multi-point constraint equations to enIorce the contact conditions are based on the complete quadratic
displacement Iield.
Due to the nature oI equivalent nodal Iorces Iollowing Irom a uniIormly distributed pressure, the separation behavior
is based on nodal stresses and not the nodal Iorces. These stresses are based on extrapolated and averaged integration
point values.
Contact with quadratic elements is demonstrated using a pin connection. A number oI pin connections is used to mount
a thick polymer insulation layer on a perIorated steel plate (see Figure 3.3-1). Plastic 'bolts with grooves are
positioned in the holes oI the insulation layer and the steel plate. They are pressed into the insulation layer, so that the
steel pins can be moved into the grooves. The bolt heads and the pins hold the insulation layer Iixed to the steel plate.
Figure 3.3-1 Perforated Steel Plate with Insulation Layer
Assuming symmetry conditions, the analysis is carried out using halI a bolt and pin and corresponding parts oI the
steel plate and insulation layer. In total, eight contact bodies will be used: Iour deIormable bodies (bolt, pin, insulation
layer, and steel plate) and Iour rigid bodies (all symmetry planes).
The analysis consists oI two loadcases. During the Iirst loadcase, the bolt is inserted into the insulation layer by
prescribed displacements. The pin is initially modeled to be in contact with the bolt, but contact between the pin and
the steel plate is not allowed. Preventing separation between the bolt and the pin yields a small gap between the pin
and the steel plate at the end oI the Iirst loadcase. During the second loadcase, the displacement constraint in axial
837
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
direction on the bolt is removed and the pin comes into contact with the steel plate, thus causing the insulation layer
to be Iixed to the steel plate. Friction between the various contact bodies is neglected.
The material behavior oI all materials is assumed to be linear elastic with the Iollowing Young`s moduli and
Poisson`s ratios:
N/cm
2
, ;
N/cm
2
, ;
N/cm
2
, .
The analysis is geometrically nonlinear, but materially linear. The Iinite elements used are 10-node tetrahedral
elements Ior the bolt and the pin and 20-node hexahedral elements Ior the other parts.
Pin Connection
The analysis oI the pin connection is done using the standard steps: deIine Iinite element mesh and geometric entities
(symmetry planes), apply boundary conditions, assign material properties, deIine contact bodies and contact tables,
deIine loadcases and collect them in a job with the proper job settings. AIter running the analysis, some postprocessing
is perIormed.
Finite Element Mesh and Geometric Entities
The Iinite element mesh and geometric entities are available in an Mentat model Iile, called pin_fe_model.mfd. AIter
resetting, the program deIaults activating view 1, and this Iile is opened. The various parts oI the model are stored in
element sets called bolt, pin, insulation, and steel_plate, where the elements are selected using the select
method Ilood. In order to easily apply displacement boundary conditions on the bolt, the nodes on the top oI the bolt
head are stored in a set called bolt_top_nodes; they are selected using the select method box.
FILES
NEW
OK
RESET PROGRAM
VIEW
SHOW VIEW 1
RESET VIEW
RETURN
OPEN
pin_fe_model.mfd
OK
FILL
E
steel plate
E
pin
2100 v
steel plate
v
pin
0.3
E
bolt
20 v
bolt
0.28
E
insulation
0.7 v
insulation
0.2
838 Marc Users Guide: Part 2
CHAPTER 3.3
MAIN
MESH GENERATION
SELECT
METHOD
FLOOD
RETURN
ELEMENTS
3861
STORE
bolt
OK
ALL SELECTED
CLEAR SELECT
ELEMENTS
1066
STORE
pin
OK
ALL SELECTED
CLEAR SELECT
ELEMENTS
119
STORE
insulation
OK
ALL SELECTED
CLEAR SELECT
ELEMENTS
690
STORE
steel_plate
OK
ALL SELECTED
CLEAR SELECT
METHOD
BOX
839
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
Figure 3.3-2 Finite Element Model Used
Boundary Conditions
As mentioned beIore, the analysis is carried out using two loadcases. During the Iirst loadcase, the bolt is inserted into
the insulation layer by prescribing the displacement component in global x-direction as a Iunction oI time Ior the nodes
on the top oI the bolt head. Moreover, the displacement in global y-direction oI one oI these nodes is suppressed to
RETURN
NODES
-0.01 0.01 (define range in x-direction)
-2 2 (define range in y-direction)
-2 2 (define range in z-direction)
STORE
bolt_top_nodes
OK
ALL SELECTED
CLEAR SELECT
METHOD
SINGLE
RETURN
MAIN
840 Marc Users Guide: Part 2
CHAPTER 3.3
prevent a rigid body motion oI the bolt. Two nodes at the end Iace the pin are constrained similarly. For the assembly,
the displacement component in global x-direction oI two corner nodes oI the steel plate is also suppressed.
BOUNDARY CONDITIONS
MECHANICAL
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
displacement_time
TYPE
time
ADD
0 0
1 1
RETURN
NEW
NAME
press_bolt
FIXED DISPLACEMENT
DISPLACEMENT X
0.7
TABLE
displacement_time
OK
NODES ADD
bolt_top_nodes
END LIST (#)
NEW
NAME
suppress_rigid_body_motion
FIXED DISPLACEMENT
DISPLACEMENT Y
NODES ADD
3557
4297
4056
841
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
An overview oI the boundary conditions is shown in Figure 3.3-3.
Figure 3.3-3 Overview of Applied Boundary Conditions
Material Properties
The elastic material properties oI the bolt, pin, insulation layer, and steel plate are easily entered using the previously
introduced element sets.
END LIST (#)
NEW
NAME
hold_steel_plate
FIXED DISPLACEMENT
DISPLACEMENT X
NODES ADD
664 670
END LIST (#)
MAIN
MATERIAL PROPERTIES
NEW
NAME
steel
842 Marc Users Guide: Part 2
CHAPTER 3.3
Contact Bodies and Contact Tables
The contact body deIinition Ior quadratic elements is similar to that Ior linear elements. The main diIIerence occurs in
the deIinition oI separation, where either relative or absolute testing on stresses has to be selected. Although this is
done in the JOBS menu, it is necessary to recognize at this stage the method that is chosen, since separation threshold
ISOTROPIC
YOUNGS MODULUS
2100
POISSONS RATIO
0.3
OK
ELEMENTS ADD
steel_plate
pin
NEW
NAME
bolt
ISOTROPIC
YOUNGS MODULUS
20
POISSONS RATIO
0.28
O
K
ELEMENTS ADD
bolt
NEW
NAME
insulation
ISOTROPIC
YOUNGS MODULUS
0.7
POISSONS RATIO
0.2
OK
ELEMENTS ADD
insulation
MAIN
843
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
values are entered within the contact tables. It obviously makes a diIIerence iI this threshold value is interpreted as a
stress (absolute testing) or as a percentage (relative testing).
Four deIormable contact bodies and Iour symmetry planes are deIined. DiIIerent contact tables are needed to easily
move the pin together with the bolt during the Iirst loadcase by avoiding contact between the pin and the steel plate.
During this Iirst loadcase, separation between the pin and the bolt is not allowed by deIining a large separation
threshold. Then, during the second loadcase, contact between the pin and the steel plate is allowed, and the separation
behavior between all bodies is based on realistic values: 10 percent oI the maximum contact normal stress in the
corresponding contact body. Since the insulation layer is signiIicantly soIter than the bolt, it is numerically preIerable
that nodes oI the insulation layer contacts the bolt. This is achieved by setting the contact detection method Irom the
second to the Iirst body Ior this body combination. To maintain contact at the boundary oI the bolt head, the delayed
slide oII option is invoked.
CONTACT
CONTACT BODIES
NEW
NAME
bolt
DEFORMABLE
OK
ELEMENTS ADD
bolt
NEW
NAME
pin
DEFORMABLE
OK
ELEMENTS ADD
pin
(repeat for deformable contact bodies insulation and steel_plate)
NEW
NAME
symmetry_1
SYMMETRY
OK
SURFACES ADD
1
END LIST (#)
NEW
844 Marc Users Guide: Part 2
CHAPTER 3.3
NAME
symmetry_2
SYMMETRY
OK
SURFACES ADD
2
END LIST (#)
(repeat for symmetry bodies symmetry_3 and symmetry_4)
RETURN
CONTACT TABLES
NEW
NAME
table_press_bolt
PROPERTIES
1 2
CONTACT TYPE: TOUCHING
PROJECT STRESS-FREE
SEPARATION THRESHOLD
1e30
1 3
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: SECOND ->FIRST
PROJECT STRESS-FREE
DELAY SLIDE OFF
1 5
CONTACT TYPE: TOUCHING
PROJECT STRESS-FREE
1 6
CONTACT TYPE: TOUCHING
PROJECT STRESS-FREE
1 7
CONTACT TYPE: TOUCHING
PROJECT STRESS-FREE
1 8
CONTACT TYPE: TOUCHING
PROJECT STRESS-FREE
845
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
Figure 3.3-4 Settings of First Contact Table
COPY
NAME
table_depress_bolt
PROPERTIES
1 2
CONTACT TYPE: TOUCHING
SEPARATION THRESHOLD
0.1
2 4
CONTACT TYPE: TOUCHING
OK
Loadcases
The analysis is perIormed using two mechanical static loadcases, both with a loadcase time oI one. The Iirst is carried
out in two equally sized steps, the second in one step. During the Iirst loadcase, all boundary conditions are active and
contact table table_press_bolt is selected. During the second loadcase, boundary condition press_bolt is
deactivated and contact table table_depress_bolt is selected. The control settings are leIt deIault Ior the Iirst
loadcase, while relative displacement checking with a tolerance oI 0.05 is selected Ior the second loadcase, since the
maximum reaction Iorce drops due to removal oI boundary conditions.
(repeat for body combinations 2-6, 3-4, 3-5, 3-6, 3-7, 3-8, 4-5, 4-6, 4-7, 4-8)
OK
846 Marc Users Guide: Part 2
CHAPTER 3.3
LOADCASES
NEW
NAME
press_bolt
MECHANICAL
STATIC
CONTACT
CONTACT TABLE
table_press_bolt
OK
# OF STEPS
5
OK
NEW
NAME
depress_bolt
STATIC
LOADS
press_bolt (deselect)
CONTACT
CONTACT TABLE
table_depress_bolt
OK
CONVERGENCE TESTING
DISPLACEMENTS
RELATIVE DISPLACEMENT TOLERANCE
0.05
OK
# OF STEPS
1
OK
MAIN
RETURN
847
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
Jobs
A mechanical job is created in which the two previously deIined load cases are selected. The analysis is geometrically
nonlinear, so the large displacement option is selected. During increment 0, the Iirst contact table is selected to avoid
wrong contact detection between the pin and the steel plate and to get stress-Iree initial contact. The contact tolerance
and the bias Iactor are set to 0.005 and 0.9, respectively. In this way, a small contact tolerance due to the small elements
in the bolt is avoided, while the outside contact tolerance remains small due to the bias Iactor. The separation method
is set to relative stress-based, using the deIault tolerance oI 0.1, and single-sided contact is activated. Notice, that the
quadratic segment button 'genuine is turned on, indicating true quadratic contact. In addition to the deIault nodal post
Iile variables, the equivalent von Mises stress is selected as an element variable. The element types used are 127
(10-node tetrahedral) Ior the bolt and 21 (20-node hexahedral) Ior the pin, insulation, and steel plate. AIter saving the
model, the job is submitted Ior analysis.
JOBS
MECHANICAL
press_bolt (select)
depress_bolt (select)
CONTACT CONTROL
ADVANCED CONTACT CONTROL
RELATIVE SEPARATION STRESS
OK
OK
ANALYSIS OPTIONS
LARGE DISPLACEMENT
OK
JOB RESULTS
EQUIVALENT VON MISES STRESS
CENTROID
OK
ELEMENT TYPES
MECHANICAL
3-D SOLID
127
bolt
127
pin
21
insulation
21
848 Marc Users Guide: Part 2
CHAPTER 3.3
Results
AIter running the job, the post Iile is opened and some characteristic results are examined. Figure 3.3-5 shows the
deIormations at the end oI the Iirst loadcase. Clearly, there is a gap between the pin and the steel plate. Figure 3.3-6
shows the deIormation at the end oI the second loadcase and illustrates how contact between the pin and the steel plate
is established. Finally, Figure 3.3-7 shows the stress concentrations around the bolt-pin and pin-plate contact areas.
steel_plate
OK
RETURN (twice)
FILES
SAVE AS
pin_complete
OK
RETURN
RUN
SUBMIT 1
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Contact Status
OK
NEXT
NEXT
NEXT
SCALAR
Equivalent Von Mises Stress
OK
849
CHAPTER 3.3
Contact Modeling of Pin Connection Joints using Higher-Order Elements
Figure 3.3-5 Contact Status and Deformations at End of First Loadcase
Figure 3.3-6 Contact Status and Deformations at End of Second Loadcase
850 Marc Users Guide: Part 2
CHAPTER 3.3
Figure 3.3-7 Equivalent von Mises Stress Around Bolt-pin and Pin-plate Contact Areas
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
higher_order.proc Mentat procedure Iile to run the above example
pin_fe_model.mfd Associated model Iile
Chapter 3.4: Beam Contact Analysis of an Overhead Power Wire of a Train
3.4
Beam Contact Analysis of an
Overhead Power Wire of
a Train

Chapter Overview 852

Pantograph of a Train Touching the Overhead Power Wire 852

Input Files 868

852 Marc Users Guide: Part 2
CHAPTER 3.4
Chapter Overview
This chapter demonstrates the beam-to-beam contact Ieature oI Marc. The options required Ior a beam-to-beam
contact analysis are discussed in detail. The capability is illustrated by the analysis oI a pantograph oI a train that
touches the overhead wire to extract electrical power, while the train moves with a velocity oI 40 m/s, or 144 km/h.
Pantograph of a Train Touching the Overhead Power Wire
The model consists oI the pantograph oI a train and the overhead power wire (see Figure 3.4-1)
.
Figure 3.4-1 Model of the Pantograph and the Overhead Wire
The wire is located 5.5 m above the railway tracks and is suspended Irom a system oI catenary wires and vertical
droppers, that are Iastened to seven mast poles (see Figure 3.4-2). Stabilizers restrict the horizontal movement oI the
wire. They are connected to the mast poles by pin joints. The masts are 70 m apart, resulting in a total track length oI
420 m, and are positioned in such a way that the overhead power wire Iollows a zig-zag pattern between the masts
with a maximum horizontal deIlection oI 60 cm. This ensures that any damage oI the pantograph, that may occur due
to Iriction with the power wire, is spread out over a large portion oI the pantograph head.
853
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
Figure 3.4-2 The Different Components of the Overhead Wire and the Mast Poles
The pantograph is a mechanism that consists oI three parts: the lower Irame and the thrust, the main Irame, and the
head (see Figure 3.4-3). The latter contains the horizontal bars that are pushed upwards to the overhead power wire to
extract electrical power. The three parts are connected by hinges and nonlinear springs. The hinges allow only relative
rotation oI the connected parts around the global x-axis, while the nonlinear rotational springs add a certain amount oI
stiIIness to this relative rotation. The pantograph head is pushed upwards by moving the end point B oI the thrust in
the negative z-direction, towards end point A oI the lower Irame. Once the head is in its Iinal position, the hinges are
locked. This is simulated by speciIying the stiIIness oI the springs as a Iunction oI time: the stiIIness is zero when the
head is pushed upwards and is set to a large value once the head is its Iinal position.
854 Marc Users Guide: Part 2
CHAPTER 3.4
Figure 3.4-3 The Pantograph Example
Boundary Conditions
The analysis consists oI three loadcases. In the Iirst loadcase, a static pre-tension Iorce oI 30 kN is applied to the
overhead power wire and the catenary wires. Simultaneously, a gravity load is applied to all elements in the model. In
this stage, six oI the seven mast poles are allowed to move Ireely in the z-direction while the other degrees oI Ireedom
are suppressed. The resulting boundary conditions on the overhead wire (except Ior the gravity load) are depicted in
Figure 3.4-4.
Hinges
A
B
855
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
Figure 3.4-4 The Boundary Conditions (except the gravity load) on the Overhead Wire
during the Static Preloading of the Wire
In the second stage oI the analysis, the pantograph head is moved up towards the overhead wire. This is achieved by
moving the end node (B) oI the thrust in the negative z-direction, towards the end node oI the lower Irame (A). The
displacements oI the latter are suppressed in this loadcase. The rotation oI the pantograph head and its displacement
in the z-direction are suppressed by the boundary condition fix_panto_head (see Figure 3.4-5). The loads on the
overhead wire are retained and the mast poles are all Iixed to the ground.
In the Iinal loadcase, the motion oI the pantograph is prescribed: the nodes A and B are moved 400 m in the positive
z-direction. The rotation oI the pantograph head is no longer suppressed and the loads on the overhead wire are the
same as in the second loadcase.
856 Marc Users Guide: Part 2
CHAPTER 3.4
Figure 3.4-5 The Boundary Conditions (except the gravity load) on the Pantograph
Initial Conditions
The initial velocity oI the all the nodes oI the pantograph is set to 40 m/s.
Links
The stabilizers that restrict the horizontal movement oI the overhead power wire are connected to the mast poles by
means oI beam pin joints (tying type 52). The hinges between the lower Irame and the main Irame, and between the
main Irame and the head oI the pantograph consist oI tying types 103 and 506, to suppress all relative displacements
and the relative rotations about the y- and the z-axis. The stiIIness oI the nonlinear spring that acts on the relative
rotation about the x-axis oI the connected parts (the Iourth degree oI Ireedom) is deIined as a Iunction oI time by means
oI a table: in the Iirst two loadcases (0-2s), the stiIIness is 0 Nm and in the last loadcase (2-12s), the stiIIness is
1000000 Nm. The table is subsequently selected in the MECHANICAL PROPERTIES menu oI the spring (see
Figure 3.4-6).
LINKS
SPRINGS/DASHPOTS
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
857
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
spring_stiffness
TYPE
time
ADD
0 0
1 0
2 0
2.001 1
12 1
FIT
RETURN
NEW
PROPERTIES
STIFFNESS
SET
1e6
TABLE
spring_stiffness
OK
BEGIN NODE
542
DOF 4
END NODE
517
DOF 4
RETURN (twice)
858 Marc Users Guide: Part 2
CHAPTER 3.4
Figure 3.4-6 The SPRINGS/DASHPOTS Menu
Material Properties
The overhead wire, the catenary wire, and the vertical droppers are made oI copper and the mast poles and the
pantograph are composed oI steel. Young`s moduli oI these materials are 120 GPa and 210 GPa, respectively,
Poisson`s ratios are given by 0.33 and 0.3, and the mass densities are equal to 8900 kg/m
3
and 7850 kg/m
3
. No material
nonlinearities are taken into account in this example.
Geometry Properties
The beam-to-beam contact option assumes that the beam elements are cylinders with a circular cross-section. The
radius oI these cylinders, the contact radius, must be entered via the GEOMETRIC PROPERTIES menu (see
Figure 3.4-7), along with the other parameters that deIine the actual shape oI the cross-section Ior the stiIIness
computation oI the beam elements. The contact radius is used Ior the detection oI contact and in the multi-point
constraint when contact is Iound. It must be deIined Ior all beam elements that belong to a contact body. Furthermore,
the contact radius must be the same Ior all elements in a contact body.
In the present example, the elements oI the overhead wire and the horizontal bars oI the head oI the pantograph are
part oI a contact body (see below). Consequently, the contact radius must be deIined Ior these elements.
GEOMETRIC PROPERTIES
3-D
NEW
NAME
overhead_wire
859
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
ELASTIC BEAM
AREA
pi*0.006*0.006
Ixx
pi*0.006*0.006*0.006*0.006/4
Iyy
pi*0.006*0.006*0.006*0.006/4
VECTOR DEFINING LOCAL X-AXIS
1 0 0
CONTACT RADIUS
0.006
OK
ELEMENTS ADD
SET
overhead_wire
OK
NEW
NAME
panto_bars
ELASTIC BEAM
AREA
pi*0.01*0.01
Ixx
pi*0.01*0.01*0.01*0.01/4
Iyy
pi*0.01*0.01*0.01*0.01/4
VECTOR DEFINING LOCAL X-AXIS
0 0 1
CONTACT RADIUS
0.01
OK
ELEMENTS ADD
SET
panto_bars
OK
860 Marc Users Guide: Part 2
CHAPTER 3.4
Figure 3.4-7 The GEOMETRIC PROPERTIES Menu for 3-D Elastic Beams
Contact
The Iirst contact body consists oI the beam elements oI the overhead power wire. The second contact body contains
the beam elements that constitute the horizontal bars oI the pantograph head. Coulomb Iriction is taken into account
in the analysis and the Iriction coeIIicients are set to 0.2 Ior both bodies.
CONTACT
CONTACT BODIES
NEW
overhead_wire
NAME
DEFORMABLE
FRICTION COEFFICIENT
0.2
OK
ELEMENTS ADD
overhead_wire
NEW
NAME
pantograph
DEFORMABLE
861
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
Loadcases
As mentioned beIore, the analysis consists oI three loadcases. In the Iirst loadcase, the overhead wire and the catenary
wires are loaded by the static preload, while, simultaneously, the gravity load is applied to all elements in the model.
The ground2_preload boundary condition is chosen over the ground2 boundary condition to allow six oI the seven
mast poles to move Ireely in the z-direction. The total loadcase time is 1 second and the loading is applied
incrementally in 20 increments.
In the second stage oI the analysis, the pantograph head is pushed upwards until it touches the overhead power wire.
In this static loadcase, all mast poles are Iixed to the ground (using the ground2 boundary condition). The total
loadcase time is again one second and the loading is applied incrementally in 20 increments.
FRICTION COEFFICIENT
0.2
OK
ELEMENTS
panto_bars
LOADCASES
MECHANICAL
NEW
NAME
preload
STATIC
LOADS
ground2 (deactivate)
OK
CONVERGENCE TESTING
DISPLACEMENTS
OK
# STEPS
20
OK
NEW
NAME
thrust
STATIC
862 Marc Users Guide: Part 2
CHAPTER 3.4
The Iinal loadcase oI the analysis simulates the motion oI the train. In this dynamic transient loadcase, the rotation oI
the pantograph head is no longer suppressed and the train moves with a constant velocity oI 40 m/s in the positive
z-direction. The total loadcase time is 10 seconds, so that the total displacement oI the train is 400 m. The displacement
is prescribed in 160 increments.
Note that in a dynamic contact analysis, where the single-step Houbolt dynamic operator is used (this is the deIault
operator and is also employed in this example), nodes and also beam elements that are Iound in contact are not
projected onto the surIace unless the DYNAMIC CONTACT PROJECTION FACTOR in the NUMERICAL
PREFERENCES menu is set to a nonzero value. The reason is that a nonzero projection Iactor may introduce
undesired (artiIicial) inertia eIIects in the analysis.
A zero projection Iactor can lead to a gradual increase oI the amount oI penetration (even though the elements are in
contact). In the case oI beam elements, the amount oI penetration may even grow to such an extent that the elements
move through each other and separate. This can happen iI the relative rotation oI the beam elements is large, as
illustrated in Figure 3.4-8.
Figure 3.4-8 Penetration due to Zero Dynamic Contact Projection Factor
In this Iigure, the beam elements B
1
(oriented in the direction perpendicular to the plane) and B
2
are in contact and B
2

rotates around its current point in contact. Since the latter is continuously updated and since the multi-point constraint
LOADS
ground2_preload (deactivate)
OK
CONVERGENCE TESTING
DISPLACEMENTS
OK
# STEPS
20
OK
B
1
B
2
(1)
B
2
(2)
B
2
(3)
n
(1)
n
(2)
n
(3)
863
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
that suppresses the relative motion oI the beams, acts in the direction oI the current normal vector n, the distance
between the beam elements decreases gradually. II the dynamic contact projection Iactor is set to a nonzero value, a
displacement correction is included in the multi-point constraint that ensures that the distance between the beam
elements remains constant. In this example, the projection Iactor is set to 1.
Job Parameters
The Coulomb Iriction model is selected in the CONTACT CONTROL menu and the relative sliding velocity is set to 1.
Note that Iriction between beam elements is always based on nodal Iorces.
Beam-to-beam contact is activated in the ADVANCED CONTACT CONTROL menu (see Figure 3.4-9). Note that
beam-to-beam contact automatically activates penetration checking per iteration and that separation is always based
on nodal Iorces.
The LARGE DISPLACEMENT option, the UPDATED LAGRANGE PROCEDURE, and the LARGE ROTATION BEAM
option are used. The latter improves the large rotation behavior oI the beam elements.
Element type 52, a 2-node straight Euler-Bernoulli beam element, is chosen Ior all beam elements in the model.
Element type 75, a 4-node thick shell element, is used Ior the two shell elements oI the main Irame oI the pantograph.
NEW
NAME
motion
DYNAMIC TRANSIENT
LOADS
ground2_preload (deactivate)
fix_panto_top (deactivate)
OK
CONVERGENCE TESTING
DISPLACEMENTS
OK
NUMERICAL PREFERENCES
DYNAMIC CONTACT PROJECTION FACTOR
1
OK
TOTAL LOADCASE TIME
10
# STEPS
160
OK
RETURN (twice)
864 Marc Users Guide: Part 2
CHAPTER 3.4
JOBS
NEW
MECHANICAL
preload
thrust
motion
INITIAL LOADS
ground2 (deactivate)
OK
CONTACT CONTROL
FRICTION TYPE
COULOMB
RELATIVE SLIDING VELOCITY
1
ADVANCED CONTACT CONTROL
BEAM TO BEAM CONTACT
ON
OK (twice)
ANALYSIS OPTIONS
LARGE DISPLACEMENTS
ADVANCED OPTIONS
UPDATED LAGRANGE PROCEDURE
LARGE ROTATION BEAM
OK (twice)
JOB RESULTS
1st Comp of Stress
2nd Comp of Stress
3rd Comp of Stress
4th Comp of Stress
OK (twice)
ELEMENT TYPES
MECHANICAL
3-D TRUSS/BEAM
52
OK
ALL EXIST
865
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
Figure 3.4-9 The ADVANCED CONTACT CONTROL Menu
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the post Iile is opened.
Figure 3.4-10 shows the contact status oI the nodes at increment 200, when the train is halIway down the track. Note
that iI two beam elements are in contact, the contact status oI all Iour nodes involved in the contact is set to 1.
3-D MEMBRANE/SHELL
75
OK
ALL EXIST
RETURN (twice)
FILE
SAVE AS
train.mud
OK
RETURN
RUN
SUBMIT(1)
OPEN POST FILE (RESULTS MENU)
866 Marc Users Guide: Part 2
CHAPTER 3.4
Inspection oI the contact status during the motion oI the train reveals that Irom increment 334 through 337, due to
Iriction and dynamic eIIects, contact between the pantograph and the overhead wire is lost.
Figure 3.4-10 Contact Status at Increment 200
In Figure 3.4-11, the Iriction Iorces on the nodes (again at increment 200), due to the contact between the pantograph
and the overhead wire are depicted and Figure 3.4-12 displays the velocity distribution oI the overhead wire. Finally,
the vertical displacement (in the y-direction) oI the overhead wire at three positions (beginning, halIway and end) are
plotted as a Iunction oI time in Figure 3.4-13.
867
CHAPTER 3.4
Beam Contact Analysis of an Overhead Power Wire of a Train
Figure 3.4-11 Friction Force at Increment 200
Figure 3.4-12 Velocity of the Overhead Wire at Increment 200
868 Marc Users Guide: Part 2
CHAPTER 3.4
Figure 3.4-13 Displacement in the Y-direction of the Overhead Wire at the beginning of the Track
Note that the track shows node 134, (green) curve, halIway down the track node 452, (red) curve, and
at the end oI the track node 458, (blue) curve as a Iunction oI time.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
train.proc Mentat procedure Iile to run the above example
Chapter 3.5: Gas Filled Cavities
3.5
Gas Filled Cavities

Chapter Overview 870

Simulation of an Airspring 870

Input Files 884

870 Marc Users Guide: Part 2
CHAPTER 3.5
Chapter Overview
This chapter demonstrates the modeling oI gas Iilled cavities. The cavity option allows the automatic update oI the
cavity pressure as the cavity volume change. The application oI this Iunctionality can be Iound in several places:
airsprings, athletic shoes with pneumatic soles, as well as lungs, etc. The simulation oI an airspring is used as an
example in this chapter. The example also employs the AXITO3D capability Ior automatic transIer oI axisymmetric
data to 3-D.
Simulation of an Airspring
Problem Description
Airsprings are Ilexible containers that inIlate by compressed air and can be used as pneumatic actuators or vibration
isolators. Depending on the inIlation pressure, airsprings can provide variable amounts oI loads and strokes.
Airsprings are known Ior being versatile, robust and easy to maintain. The airspring model discussed in this example
is constructed Irom cord-reinIorced rubber clamped by metal beads.
The airspring is loaded in three stages: Iirst clamping and inIlation, Iollowed by axial compression, and Iinally axial
expansion with lateral deIlection. The Iirst two loading stages can be perIormed using an axisymmetric analysis. The
axisymmetric model is then expanded into a 3-D model where the Iinal loading step can be executed.
Figure 3.5-1 shows a 3-D schematic oI the airspring. The airspring is initially cylindrical in shape with a length and
diameter oI 200 and 95 mm, respectively. The wall thickness is 1.7 mm. The airspring material is taken to be a rubber
matrix with two layers oI positively and negatively oriented skew rebars. The rubber is modeled using the Mooney
constitutive model with C
10
3 MPa and C
01
1 MPa. The rebars are made oI steel with E 210.0 GPa and v 0.3
with a cross-sectional area oI 10
-6
mm
2
and 45 orientations.
The air inside the cavity oI the airspring has a reIerence density oI 1.0 kg/m
3
at a reIerence pressure oI 0.1 MPa and a
reIerence temperature oI 300K and is assumed to Iollow the ideal gas law.
Figure 3.5-1 Schematic of the Airspring
871
CHAPTER 3.5
Gas Filled Cavities
Axisymmetric Analysis
An axisymmetric model Ior the airspring is shown in Figure 3.5-2. The model is constructed oI 100 4-noded
axisymmetric rubber elements (element type 10) and 100 rebar elements (element type 144) sharing the same nodes.
The airspring is Iirst clamped and pressurized to 1.5 MPa. It is then subjected to an axial displacement oI 150 mm by
the leIt clamps. The loading is thus divided into two loadcases. The model Iile airspring_axi.mfd contains the
complete model Ior the problem except Ior the cavity deIinition and the associated pressure load.
Figure 3.5-2 Axisymmetric Model of the Airspring
To deIine the cavity, the user needs to select the rubber element edges Iorming the cavity. The user can select these
edges by Iirst making only the rubber elements (element type 10) visible, then interactively selecting the edges shown
in Figure 3.5-2. For convenience, the cavity edges (edges 6:1 to 95:1) have already been selected and stored into a set
named cavity_edge_list in airspring_axi.mfd. To open the model:
AIter opening the model and examining it, Iollow the steps described below to deIine the cavity and apply the cavity
pressure load:
FILES
OPEN
airspring_axi.mfd
OK
MAIN
MODELING TOOLS
CAVITIES
NEW
EDGES
ADD
ALL
SET
cavity_edge_list
OK
PARAMETERS
872 Marc Users Guide: Part 2
CHAPTER 3.5
REF. PRESSURE
1.0E5
REF. TEMPERATURE
300.0
REF. DENSITY
1.0
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
MORE
CAVITY MASS LOAD
MASS
CLOSED CAVITY
OK
CAVITIES
ADD
ALL
EXISTING
NEW
CAVITY PRESSURE LOAD
PRESSURE
1.5E6
TABLE
table2
OK
CAVITIES
ADD
ALL
EXISTING
MAIN
LOADCASES
MECHANICAL
STATIC
LOADS
apply2
OK (twice)
873
CHAPTER 3.5
Gas Filled Cavities
Figure 3.5-3 through Figure 3.5-5 show the CAVITIES menu, the CAVITY MASS LOAD menu, and the CAVITY
PRESSURE LOAD menu.
Figure 3.5-3 Cavities Menu
NEXT
STATIC
LOADS
apply1
OK (twice)
MAIN
JOBS
MECHANICAL
INITIAL LOADS
apply1
OK
FILES
SAVE AS
airspring_axi_wcav.mfd
OK
874 Marc Users Guide: Part 2
CHAPTER 3.5
Figure 3.5-4 Cavity Mass Load Menu
Figure 3.5-5 Cavity Pressure Load Menu
To run the job:
MAIN
JOBS
RUN
RESET
SUBMIT (1)
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
MONITOR
HISTORY PLOT
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD GLOBAL CRV
875
CHAPTER 3.5
Gas Filled Cavities
The Iinal deIormed shape is shown in Figure 3.5-6. Figure 3.5-7 (a) and (b) show the variation oI cavity pressure with
time and with cavity volume, respectively.
Figure 3.5-6 Deformed Shape of the Axisymmetric Airspring Model
Figure 3.5-7 (a) Cavity Pressure vs. Time (b) Cavity Pressure vs. Volume
GLOBAL VARIABLES
Time
Pressure Cavity 1
FIT
REMOVE CURVE
CLEAR CURVES
ADD GLOBAL CRV
GLOBAL VARIABLES
Volume Cavity 1
Pressure Cavity 1
FIT
RETURN (twice)
CLOSE
(a) (b)
876 Marc Users Guide: Part 2
CHAPTER 3.5
AIter reviewing the axisymmetric results and closing the post Iile, the next step is to transIer the axisymmetric model
into 3-D where the axial expansion with lateral deIlection can be applied. The AXITO3D option is used to perIorm the
transIer. BeIore expanding the model into 3-D, the rigid contact bodies must Iirst be moved to their Iinal position at
the end oI the axisymmetric analysis.
Follow the steps below to expand the axisymmetric model into 3-D.
MAIN
MESH GENERATION
MOVE
TRANSLATIONS Y
0.002
CURVES
1 3 4 6 7 8 #
RESET
TRANSLATIONS X
0.15
CURVES
1 2 6 8 9 10 #
MAIN
MESH GENERATION
EXPAND
AXISYMMETRIC MODEL TO 3D
ANGLE
12
REPETITIONS
30
TIME
SET
2
EXPAND MODEL
MAIN
INITIAL CONDITIONS
MECHANICAL
AXISYMMETRIC TO 3D
POST FILE
airspring_axi_wcav_job1.t16
OK (twice)
877
CHAPTER 3.5
Gas Filled Cavities
Notice that during the model expansion to 3-D, the 2-D cavity has been automatically expanded into a 3-D one as
indicated by the cavities menu in Figure 3.5-8. Also notice that a new table, table2.5, has been created based on
table2 and has been used to apply the cavity pressure load. Contact bodies 2 and 3 are now moved and deIlected
using load control. A control node and an auxiliary node must Iirst be deIined.
Figure 3.5-8 Cavity Menu after AXITO3D Expansion
This creates nodes 6061 and 6062. Contact bodies 2 and 3 are now switched to load control and associated with nodes
6061 and 6062 as control and auxiliary nodes. Contact body 4 position is Iixed.
MAIN
MESH GENERATION
NODES
ADD
0.15 -0.07 0.0
ADD
0.16 -0.07 0.0
MAIN
CONTACT
CONTACT BODIES
NEXT
RIGID
BODY CONTROL
LOAD
PARAMETERS
ROTATION AXIS
Z
1
878 Marc Users Guide: Part 2
CHAPTER 3.5
Figure 3.5-9 shows the load control menu oI bodies 2 and 3. The next step is to apply the displacements and rotations
to the control and auxiliary nodes, respectively. This is accomplished on two loadcases: axial expansion Iollowed by
combined axial expansion and lateral deIlection.
OK (twice)
LOAD CONTROL
CONTROL NODE
6061
AUX. NODE
6062
NEXT
RIGID
BODY CONTROL
LOAD
PARAMETERS
ROTATION AXIS
Z
1
OK (twice)
LOAD CONTROL
CONTROL NODE
6061
AUX. NODE
6062
NEXT
RIGID
BODY CONTROL
POSITION
PARAMETERS
POSITION
Y
0
TABLE
CLEAR
OK (twice)
879
CHAPTER 3.5
Gas Filled Cavities
Figure 3.5-9 Load Control Menu for Contact Bodies 2 and 3
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
-0.075
TABLE
table3
DISPLACEMENT Y
-0.01
TABLE
table1
DISPLACEMENT Z
0.0
OK
NODES
ADD
6061 #
NEW
FIXED DISPLACEMENT
DISPLACEMENT X
0.0
DISPLACEMENT Y
0.0
DISPLACEMENT Z
0.1
880 Marc Users Guide: Part 2
CHAPTER 3.5
The Iinal step is to set up the loadcases and job parameters beIore submitting the job.
TABLE
table1
OK
NODES
ADD
6062 #
MAIN
LOADCASES
PREV
MECHANICAL
STATIC
LOADS
apply1 (add)
apply2 (remove)
apply3 (add)
apply4 (add)
OK
CONSTANT TIME STEP
# STEPS
10
OK
NEXT
STATIC
LOADS
apply3 (add)
apply4 (add)
OK
CONSTANT TIME STEP
# STEPS
20
OK
881
CHAPTER 3.5
Gas Filled Cavities
MAIN
JOBS
MECHANICAL
ANALYSIS DIMENSION
3-D
INITIAL LOADS
apply1 (remove)
apply2 (add)
apply3 (add)
apply4 (add)
icond1 (add)
OK (twice)
FILES
SAVE AS
airspring_axito3d_wcav.mfd
OK
MAIN
JOBS
RUN
RESET
SUBMIT (1)
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
MONITOR
HISTORY PLOT
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD GLOBAL CRV
GLOBAL VARIABLES
Time
Pressure Cavity 1
FIT
882 Marc Users Guide: Part 2
CHAPTER 3.5
Figure 3.5-10 (a) and (b) shows the initial and Iinal conIigurations oI the airspring Ior the 3-D analysis.
Figure 3.5-10 (a) Initial 3-D Configuration (b) Final 3-D Configuration
The loads applied to the control and auxiliary nodes as well as the number oI Iixed time steps used were selected such
that the total solution time is reasonably small. For more expansion and deIlection oI the airspring, one can use the
Iollowing:
table4 instead oI table3 Ior the x-displacement oI node 6061
-0.05 m Ior the y-displacement oI node 6061
0.4 radians Ior the rotation applied to node 6062
50 time steps Ior loadcase 1
80 time steps Ior loadcase 2
The Iinal shape oI the 3-D airspring using the above loading parameters is displayed in Figure 3.5-11. Figure 3.5-12
(a) and (b) shows the variation oI cavity pressure with time and with cavity volume, respectively.
REMOVE CURVE
CLEAR CURVES
ADD GLOBAL CRV
GLOBAL VARIABLES
Volume Cavity 1
Pressure Cavity 1
FIT
RETURN (twice)
CLOSE
(a) (b)
883
CHAPTER 3.5
Gas Filled Cavities
Figure 3.5-11 Final Configuration for 3-D Analysis at Higher Loads
Figure 3.5-12 (a) Cavity Pressure vs. Time (b) Cavity Pressure vs. Volume
Figure 3.5-13 combines the curve in Figure 3.5-12 (a) Ior the 3-D analysis with that oI Figure 3.5-7 (a) Ior the
axisymmetric analysis and compares the combined curve with the case where the cavity option is not used. The
comparison clearly shows the eIIect oI using the cavity option on the airspring internal pressure and demonstrates
its signiIicance.

(a) (b)
884 Marc Users Guide: Part 2
CHAPTER 3.5
Figure 3.5-13 Cavity Pressure vs. Time with and without Cavity Option
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
airspring.proc Mentat procedure Iile to run the above example
airspring_axi.mfd Associated model Iile

Chapter 3.6: Tube Flaring
3.6
Tube Flaring

Chapter Overview 886

Background Information 886

Detailed Session Description 888

Conclusion 909

Input Files 909

886 Marc Users Guide: Part 2
CHAPTER 3.6
Chapter Overview
The sample session described in this chapter analyzes the process oI Ilaring. A cone-shaped Ilaring tool is pushed into
a cylindrical tube to permanently increase the diameter oI the tube end. Both the steel tube and Ilaring tool are modeled
as deIormable contact bodies. The goal oI the quasi-static analysis described in this chapter is threeIold:
to determine whether the Iinal shape oI the tube meets the objective oI the analysis
to study whether residual stresses are present in the steel tube and Ilaring tool
to determine the magnitude oI the residual stresses (iI present)
Background Information
Description
This session demonstrates the analysis oI a contact problem involving two deIormable contact bodies, multiple
materials, kinematic constraints and loads. The nonlinear nature oI the problem along with the irreversible
characteristics make it impossible to determine in advance the load required to drive the tool into the tube. As a result,
multiple runs through the analysis cycle are necessary to determine the maximum load required to meet the objective
oI the analysis.
The diameter oI the tube is 8 inches, the thickness is 0.3 inches and the length is 8 inches. The Ilaring tool is modeled
as a hollow cone with an apex angle oI 30, a wall thickness oI 0.6 inches, and a length suIIicient to model the process.
Figure 3.6-1 Cylindrical Tube and Flaring Tool
Idealization
The loading and geometry oI the structure are symmetrical about the center line oI the cylindrical tube. Due to the
nature oI the analysis, you are only required to analyze an axisymmetric model oI the structure. II the appropriate
boundary condition is prescribed, the tube is prevented Irom moving in the axial direction but is Iree to move in a radial
direction at one end. A load is applied to the rim oI the Ilaring tool to push it into the Iree end oI the pipe.
887
CHAPTER 3.6
Tube Flaring
Figure 3.6-2 Axisymmetric Model of Tube and Flaring Tool
Requirements for a Successful Analysis
The analysis is considered successIul when the Ilaring tool expands the tube diameter by 10. You can plot the tool
load versus the radial displacement at the tube end Ior several load increments to adjust the maximum load and repeat
the analysis cycle until you reach the objective.
Full Disclosure
The steel tube is modeled by Iour-noded axisymmetric elements with a Young`s Modulus oI 30.0e6 psi and a Poisson`s
Ratio oI 0.3. It is assumed that the tube material with an initial yield stress oI 3.6e4 psi will not harden during the
process. The tube diameter is 8.0 inches, and the thickness is 0.3 inches.
The Ilaring tool is modeled as a hollow cone with an apex angle oI 30 degrees and a thickness oI twice that oI the tube
with a suitable length to model the working area. The Ilaring tool is modeled as a case hardened steel object with a
Young's Modulus oI 40.0e6 psi, a Poisson's ratio oI 0.3, an initial yield stress oI 6.0e4 psi. The larger diameter end oI
the Ilaring tool is loaded to drive the smaller end oI the Ilaring tool into the steel tube.
Overview of Steps
Step 1: Create a model of two patches and convert to finite elements. Apply kinematic constraints to the
tube and add a low stiffness spring to avoid rigid body motions of the tool.
Step 2: Apply material properties to the tube and flaring tool.
Step 3: Create contact bodies.
Step 4: Apply edge loads to the larger diameter end of the tool to push it into the steel tube and
create a loadcase.
Step 5: Create a job and activate appropriate large strain plasticity procedure.
Axis of Revolution
Tube
Flaring Tool
888 Marc Users Guide: Part 2
CHAPTER 3.6
Step 6: Submit the job.
Step 7: Postprocess the results by looking at the deformed structure and a history plot of the tip deflection
of the tube.
Detailed Session Description
Step 1: Create a model oI two patches and convert to Iinite elements. Apply kinematic constraints to the tube and add
a low stiIIness spring to avoid rigid body motions oI the tool.
The approach used in this session to generate the model is the geometric meshing technique which involves converting
geometric entities to Iinite elements. ReIer to Mesh Generation in Chapter 1 Ior a detailed discussion on mesh
generation techniques.
As in the sample session described in Getting Started in Chapter 1, the Iirst step Ior building a Iinite element mesh is
to establish an input grid. Click on the MESH GENERATION button oI the main menu.
Next click on the SET button to access the coordinate system menu where the grid settings are located. Use the
Iollowing button sequence to set a grid in the u domain between 0 and 8 with spacing 1 and in the v domain between
0 and 5 with spacing 1.
The geometric entity used in this session is a patch. The surIace type used to enter a patch is a QUAD which is the
deIault setting Ior surIace types. Use the Iollowing button sequence to deIine the Iirst patch:
MAIN
MESH GENERATION
SET
U DOMAIN
0 8
U SPACING
1
V DOMAIN
0 5
V SPACING
0.5
GRID (on)
RETURN
FILL
MAIN
MESH GENERATION
889
CHAPTER 3.6
Tube Flaring
These points are the Iour corners oI the Iirst patch deIined as the cross-section oI the cylindrical tube. Next, move the
top two points oI the patch to their exact location.
The next step is to deIine a second patch Ior the cross-section oI the Ilaring tool. Due to the conical shape oI tool, it is
best to use the cylindrical coordinate system to enter a set oI points Ior the cone angle oI 15 Irom the horizontal axis.
PLOT
label POINTS (on)
RETURN
srfs ADD (Pick the following corner points from the grid)
point(0,4,0)
point(8,4,0)
point(8,4.5,0)
point(0,4.5,0)
MAIN
MESH GENERATION
MOVE
TRANSLATIONS
0 -0.2 0
POINTS
3 4 (Pick top 2 points)
END LIST (#)
MAIN
MESH GENERATION
RECTANGULAR (to switch to CYLINDRICAL)
pts ADD
5 15 0
16 15 0
GRID (to switch off the grid)
FILL
890 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-3 First Patch (Cross Section of Cylindrical Tube)
BeIore duplicating the newly generated points, it is important to realize that all operations are done in the local
coordinate system. For now, simply change the coordinate system back to rectangular. This can be done by clicking
on the CYLINDRICAL button twice.
Duplicate the just entered points and translate them 0.6 in the y-direction using the Iollowing button sequence to Iorm
the upper-corners oI the Ilaring tool.
MAIN
MESH GENERATION
CYLINDRICAL (to switch to SPHERICAL)
SPHERICAL (to switch to RECTANGULAR)
DUPLICATE
TRANSLATIONS
0 0.6 0
POINTS
5 6 (Pick the two points generated above)
END LIST (#)
891
CHAPTER 3.6
Tube Flaring
The Iour points Ior the second patch have now been deIined. Use the srfs ADD command to enter the second patch.
Figure 3.6-4 Second Patch (Cross Section of Flaring Tool)
Use the Iollowing button sequence to translate the second patch until it almost meets the cylindrical tube.
The two patches that outline the cylindrical tube and conical Ilaring tool respectively are shown in Figure 3.6-5.
MAIN
MESH GENERATION
srfs ADD
5 6 8 7 (Pick points in counter-clockwise direction)
MAIN
MESH GENERATION
MOVE
TRANSLATIONS
0 1.25 0
SURFACES
2 (Pick the surface to move)
END LIST (#)
892 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-5 Tube and Flaring Tool Patches Defined
The two patches are converted to elements. The number oI subdivisions is set to 8 x 3 Ior the cylindrical tube and to
14 x 6 Ior the conical Ilaring tool. Use the Iollowing button sequence to convert the two patches.
MAIN
MESH GENERATION
CONVERT
DIVISIONS
8 3
SURFACES TO ELEMENTS
1 (Pick the surface to convert)
END LIST (#)
DIVISIONS
14 6
SURFACES TO ELEMENTS
2 (Pick the surface to convert)
END LIST (#)
PLOT
draw POINTS (off)
label POINTS (off)
draw SURFACES (off)
REGEN
RETURN
893
CHAPTER 3.6
Tube Flaring
In this sample session, the CONVERT option is used instead oI the AUTOMESH option. Although both options would
create a Iinite element mesh, the CONVERT processor allows Ior better control oI the element distribution.
Figure 3.6-6 shows the results oI the conversion process.

Figure 3.6-6 Tube and Flaring Tool Patches Converted
Once you have converted the two patches to elements, you can assign a Marc element type to the elements. Although
Mentat assigns a deIault type to elements, based on the dimensionality oI the problem, it is advised to explicitly set
the element type. The element type selected Ior this analysis is Marc element type 10, a Iour-noded axisymmetric
quadrilateral element. Use the Iollowing button sequence to select this element type Ior all existing elements (pick
Irom row FULL INTEGRATION and column QUAD(4)).
MAIN
JOBS
ELEMENT TYPES
mechanical elements AXISYMMETRIC SOLID
10 (FULL INTEGRATION / QUAD(4))
OK
all: EXIST.
894 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-7 Select Marc Element Type
The displacement degree oI Ireedom in the x-direction Ior the nodes at the Iar leIt end oI the cylindrical tube is Iixed.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
OK
nodes ADD
1 10 19 28 (Pick left row of nodes)
END LIST (#)
895
CHAPTER 3.6
Tube Flaring
Figure 3.6-8 Fixed Nodes of Cylindrical Tube in X-Direction
The deIormable tool will be loaded by a pressure load. II there is no contact between tube and tool, a rigid body mode
is present. This rigid body mode can be removed by entering a week spring between tube and tool. Enter a spring by
using the Iollowing button sequence below:
MAIN
LINKS
SPRING/DASHPOT
ZOOM BOX
(create a zoom box by moving
while keeping <ML> depressed)
STIFFNESS
10.0e3
NODE 1
131
DOF 1
1
NODE 2
9
DOF 2
1
< >
896 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-9 Specifying the Spring Between the Deformable Bodies
Step 2: Apply material properties to the tube and Ilaring tool.
Apply material properties to both the tube and Ilaring tool. The properties Ior the tube are diIIerent Irom those oI the
Ilaring tool. Use the Iollowing button sequence to assign the material to all elements oI the tube.
MAIN
FILL
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
30.0e6
POISSON'S RATIO
0.3
PLASTICITY
INITIAL YIELD STRESS
3.6e4
OK (twice)
elements ADD (use the Box Pick Method to
select all tube elements)
END LIST (#)
ID MATERIALS (on)
897
CHAPTER 3.6
Tube Flaring
Figure 3.6-10 Material Properties Applied to all Tube Elements
Apply the material properties Ior all elements oI the Ilaring tool using the Iollowing button sequence.
MAIN
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNG'S MODULUS
40.0e6
POISSON'S RATIO
0.3
PLASTICITY
INITIAL YIELD STRESS
6.0e4
OK (twice)
elements ADD (Use the Polygon Pick Method to
select all tool elements)
END LIST (#)
898 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-11 Material Properties Applied to all Elements
Step 3: Create contact bodies.
IdentiIy the two contact bodies by storing the elements oI each deIormable body in a set using the Iollowing
button sequence:
MAIN
CONTACT
CONTACT BODIES
NAME
tube
DEFORMABLE
OK
elements ADD (Use the Box Pick Method to
select all tube elements)
END LIST (#)
NEW
NAME
tool
DEFORMABLE
OK
elements ADD (Use the Polygon Pick Method to
select all tool elements)
END LIST (#)
899
CHAPTER 3.6
Tube Flaring
The easiest way to identiIy the contact bodies is to request the program to draw the bodies in diIIerent colors.
Figure 3.6-12 Identifying the Tube and Flaring Tool by Color
Although maybe not apparent in Figure 3.6-12, the color oI the tube is diIIerent Irom the Ilaring tool and is indicated
in the key that appears in the upper leIt-hand corner oI the graphics area. Click on ID CONTACT once again to switch
oII the PLOT IDENTIFY mode.
Step 4: Apply edge loads to the larger diameter end oI the tool to push it into the steel tube and create a loadcase.
The Iollowing button sequence deIines a table to speciIy the loading oI the Ilaring tool. Figure 3.6-13 gives a graphical
representation oI the Ilaring tool being loaded.
MAIN
CONTACT
CONTACT BODIES
ID CONTACT (on)
MAIN
CONTACT
CONTACT BODIES
ID CONTACT (off)
MAIN
BOUNDARY CONDITIONS
900 Marc Users Guide: Part 2
CHAPTER 3.6
MECHANICAL
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
loading
TYPE
time
OK (Select OK button only if type time was typed in)
independent variable v1: MAX
87
independent variable v1: STEPS
87
function value f: MAX
2400
ADD
0 0
9 900
39 2400
8 0
FILLED
SHOW TABLE
SHOW MODEL (Select SHOW MODEL from list)
901
CHAPTER 3.6
Tube Flaring
Figure 3.6-13 Loading of Flaring Tool
Apply the load to all edges at the Iar right end oI the Ilaring tool.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
EDGE LOAD
pressure TABLE
loading
OK (twice)
edges ADD (Pick edges)
38:1 52:1 66:1 80:1 94:1 108:1
END LIST (#)
902 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-14 Loading Applied to Far Right End of Flaring Tool
The Iollowing button sequence creates a loadcase with the deIault name lcase1.
Step 5: Create a job and activate appropriate large strain plasticity procedure.
Once you have deIined the loadcase, activate the constant dilatation procedure Ior all elements to avoid numerical
problems due to the incompressible plasticity and activate the large strain plasticity procedure based on the mean
normal plasticity solution procedure.
Step 6: Submit the job.
Select the result to be written on the post Iile, switch to axisymmetric analysis, and submit the job.
MAIN
LOADCASES
mechanical STATIC
LOADS (select all loads - done by default)
OK
TOTAL LOADCASE TIME
87
# STEPS
87
OK
903
CHAPTER 3.6
Tube Flaring
Step 7: Postprocess the results by looking at the deIormed structure and a history plot oI the tip deIlection oI the tube.
The Iinal phase oI the analysis cycle (shown in Figure 1.1-1 oI Introduction) is postprocessing. Postprocessing
involves viewing and evaluating the results oI an analysis.
In order to evaluate analysis results with Mentat, you must have a post Iile which consists oI analysis results Irom the
Iinite element analysis program Marc.
A typical postprocessing session may consist oI the Iollowing steps:
Reading the post Iile created by submitting the job
Creating a history or path plot oI the model
MAIN
JOBS
MECHANICAL
loadcases available
lcase1
ANALYSIS OPTIONS
ADVANCED OPTIONS
CONSTANT DILATATION (on)
OK
plasticity procedure SMALL STRAIN
(Switch to large-strain additive procedure)
OK
JOB RESULTS
available element tensors
Stress
Plastic Strain
available element scalars
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
OK
AXISYMMETRIC
OK
SAVE
RUN
SUBMIT 1
MONITOR
904 Marc Users Guide: Part 2
CHAPTER 3.6
Displaying a plot oI the model at speciIic increments
Viewing diIIerent levels oI stress types on the model
The results oI the Ilaring process analysis have been saved in a post Iile. Use the Iollowing button sequence to open
the Iile:
Zoom in on the area oI contact Ior better access oI the node that represents the tip deIlection oI the tube. The resulting
close-up oI the contact area shown in Figure 3.6-15 should now appear in the graphics area.
Figure 3.6-15 Close up of Contact Area
The objective oI the analysis, stated in Requirements for a Successful Analysis, requires a plot that demonstrates the
tip displacement versus the load. Since the loading pattern is given in Figure 3.6-13, a displacement versus the
increment plot can also be used. The tip displacement in y-direction on the inner diameter oI the tube is collected and
displayed using the HISTORY PLOT option.
MAIN
RESULTS
OPEN DEFAULT
FILL
MAIN
RESULTS
ZOOM BOX
905
CHAPTER 3.6
Tube Flaring
The 0 is the Iirst history increment, 100 the last history increment, and 1 is the increment step size. The program reads
the increments indicated by the message Collected increment (number) in the dialogue area.
Once all the data Ior a plot has been collected, it can be displayed in a diagram where the increment number is the
x-axis variable and the displacement in the y-direction is the y-axis variable. The FIT option allows you to view the
history plot in the graphics area. Use the Iollowing button sequence to display the graph:
Recall the objective oI our analysis: to expand the tube diameter by 10. The Y-axis variable, displacement y, has to
reach a value oI 0.4 in the unloaded conIiguration to meet the objective.
Click on the YMAX button and enter 0.5. Set the Iollowing plot settings to label the history graph.
MAIN
RESULTS
HISTORY PLOT
SET NODES
9
END LIST (#)
COLLECT DATA
0 100 1
NODES/VARIABLES
ADD 1-NODE CURVE
9 (from the NODES panel)
Increment (from the GLOBAL VARIABLES panel)
Displacement Y (from the VARIABLES AT NODES panel)
FIT
MAIN
RESULTS
HISTORY PLOT
SHOW IDS
10
XSTEP
20
YSTEP
20
YMAX
0.5
906 Marc Users Guide: Part 2
CHAPTER 3.6
The maximum value Ior y-displacement is obtained in increment 39. AIter this increment, the Ilaring tool is unloaded.
The overshoot is necessary to obtain a 10 permanent diameter increase in the load-Iree state.
Figure 3.6-16 History Plot of Node 9 over 87 Increments
To better understand the process, it is helpIul to look at an animation oI the deIormation oI the tip oI the tube. Return
to the postprocessing results panel and click on the DEF & ORIG button to view both the original and the deIormed
structure. At this point, drawing the nodes and the internal edges oI the mesh is no longer necessary.
Use the Iollowing button sequence to change the plot settings so that only the outline edges oI the model are displayed.
FILL
SHOW TABLE
SHOW MODEL (Switch to SHOW MODEL to view the model)
MAIN
PLOT
elements SETTINGS
OUTLINE (on)
FACES (off)
RETURN
draw NODES (off)
REGEN
FILL
907
CHAPTER 3.6
Tube Flaring
Once the nodes and element Iaces oI the interior mesh have been suppressed, leaving only an outline oI the two
structures, you get a much clearer picture oI the extent oI the deIormation that has taken place.
For animation purposes, the data that is processed needs to be condensed. The data is automatically condensed and
written to disk Ior each Irame oI animation. Once this process has been completed, the Irames can be traversed when
shown in playback mode. Use the Iollowing button sequence to condense the data and activate the playback.
The 100 increments is a user-deIined upper limit oI the number oI Irames that are to be created Ior the animation. The
numeral 1 on the same line represents the interval at which to create a Irame. In this case, each increment is deIined
to be a Irame. The SHOW MODEL command is selected to switch Irom the animation view back to the model view.
The von Mises stresses induced by the Ilaring process on the model can be viewed using the Iollowing
button sequence:
The resulting model shown in Figure 3.6-17 clearly indicates that the von Mises stress is concentrated in two areas:
the tip oI deIlection, where the tube made contact with the tool, and in the area where the tube is deIormed.
MAIN
RESULTS
REWIND
NEXT
DEF & ORIG
MORE
ANIMATION
create INCREMENTS
100 1
FILL
PLAY
SHOW MODEL (To display the model)
MAIN
RESULTS
SKIP TO INC
39
SCALAR
Equivalent von Mises Stress
OK
CONTOUR BANDS
908 Marc Users Guide: Part 2
CHAPTER 3.6
Figure 3.6-17 Plot of Original & Deformed Tube showing von Mises Stresses at Increment 39
Next, you can check the model Ior the plastic strain. Since you have already speciIied the increment and have contour
bands selected, you only need to click on SCALAR and Total Equivalent Plastic Strain Irom the pop-up menu to check
Ior permanent deIormation. The resulting model, shown in Figure 3.6-18, indicates where plastic strain is Iound.
Figure 3.6-18 Plot of Original & Deformed Tube showing Plastic Strain at Increment 39
909
CHAPTER 3.6
Tube Flaring
II you are interested in viewing the von Mises stresses over the course oI 87 increments, make sure you have
CONTOUR BANDS selected under the SCALAR PLOT panel, and von Mises as scalar quantity, prior to animating the
model using the button sequence shown beIore.
Conclusion
As mentioned in the chapter overview, the goal oI the analysis described in this sample session was threeIold.
1. To determine whether the Iinal shape oI the tube meets the objective oI the analysis.
2. To determine whether residual stresses are present in the steel tube and Ilaring tool.
3. II residual stresses are present, to determine what are the residual stresses.
The results oI the analysis demonstrate that the goals oI the analysis have been met.
1. You have seen that the Ilaring tool expands the diameter oI the tube by 10.
2. Residual stresses are present in the steel tube; however, there are not any noticeable stresses in the Ilaring tool.
3. Figure 3.6-18 shows the equivalent plastic strain just beIore the tool is released.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
tube_flaring.proc Mentat procedure Iile to run the above example
910 Marc Users Guide: Part 2
CHAPTER 3.6
Chapter 3.7: Punch
3.7
Punch

Chapter Overview 912

Background Information 912

Detailed Session Description 915

Input Files 932

912 Marc Users Guide: Part 2
CHAPTER 3.7
Chapter Overview
The sample session described in this chapter analyzes the process oI punching. A tool with a rigid dimple is pushed
into a circular plate. The object oI this process is to produce a circular plate with spherical indentation. The goal oI the
static analysis described in this chapter is to determine the residual stresses and plastic strains in the workpiece aIter
the operation.
Background Information
This problem demonstrates the preparation oI a contact analysis involving multiple rigid bodies (the tool) and a
deIormable body (the workpiece). The top oI the tool is a sphere blended in with a Ilat rigid plate. The workpiece is
supported such that radial displacements are constrained at the outer diameter while axial displacements are
constrained at the node positioned at the corner oI the outer diameter and the backing plate. The bottom part oI the tool
is a Ilat backing plate with a hole at the same location as the dimple oI the top part oI the tool. The plate oI the tool
supports the entire workpiece, except Ior the region oI the hole.
Idealization
Because oI the axisymmetric nature oI the geometry and the loading, this process can be idealized to an axisymmetric
model. The edge oI the workpiece is clamped which prevents rigid body motion oI the workpiece. The backing plate
that backs the workpiece is modeled as a rigid body and remains in place during the analysis. The punch is modeled
as a rigid body and moves during the analysis towards the static backing plate, while indenting the workpiece.
The tool is stopped when the Ilat surIaces oI both parts oI the tool are in Iull contact with the workpiece. This occurs
when the total displacement oI the punch is 0.1488 inches, which is reached in 0.4 seconds. Hence, the velocity oI the
top part oI the tool (i.e. punch) is 0.372 inch per second. The Iriction between tool and workpiece is assumed to be
negligible and is thereIore not taken into consideration in this analysis.
913
CHAPTER 3.7
Punch
Figure 3.7-1 Punch, Workpiece, and Backing Plate
Requirements for a Successful Analysis
The analysis is considered successIul when the punch becomes Ilush with the workpiece and is released aIterwards to
determine the residual stresses.
Full Disclosure
The workpiece is constructed out oI steel with a Young's Modulus oI 30.0e6 psi and a Poisson's Ratio oI 0.3. It has a
yield stress oI 39,000 psi. The material exhibits workhardening. The workpiece has a radius oI 0.7874 inches and a
thickness oI 0.117 inches.
Z
X
Y
PUNCH
WORKPIECE
BACKING PLATE
4
PUNCH
WORKPIECE
BACKING
PLATE
914 Marc Users Guide: Part 2
CHAPTER 3.7
Figure 3.7-2 Dimensions of Punch, Workpiece, and Backing Plate
The punch is a sphere oI radius 0.24 with a Iillet oI radius 0.109 that brings it tangent to a horizontal piece. It will move
over a total distance oI 0.1488 inches in a period oI 0.4 seconds. The backing plate has a cylindrical hole oI radius 0.25
inches into which the workpiece is Iorced. Both punch and backing plate are considered to be rigid during the analysis.
X
Y Z
cbody1
cbody2
cbody3
none
Center Line
Workpiece
Width = 0.7874 Inch
Height = 0.117
Radius = 0.25 Inches
Rigid die or backing plate
Punch
Radius 0.24
Angle 55 deg
Radius 0.109
Angle 55 deg
1
Punch
Radius Radius
cbody1
cbody2
cbody3
none
Workpiece
Width = 0.7874 inch
Height = 0.117
Rigid die or backing plate
Radius = 0.25 inches
Center Line
915
CHAPTER 3.7
Punch
Figure 3.7-3 The Workhardening Curve for the Workpiece Material
Overview of Steps
Step 1: Create a model of a rectangular patch and convert it to finite elements.
Step 2: Create the curves required for the punch & backing plate.
Step 3: Apply the required fixed displacements to the rim of the workpiece. Apply the material data.
Step 4: Identify the contact bodies and create the table that defines the motion of the rigid die, representing
the punch.
Step 5: Define the incremental steps and convergence testing parameters.
Step 6: Activate the large strain parameters and submit the job.
Step 7: Postprocess the results by displaying the deformed structure and the residual stresses and strains.
Detailed Session Description
Step 1: Create a model oI a rectangular patch and convert it to Iinite elements.
The approach used in this session to generate the model is the geometric meshing technique.
2
3
4
5
1
7.215
WorkHardening
Plastic strain (x.01)
Yield stress (x10000)
0 1
3.900
6
1
916 Marc Users Guide: Part 2
CHAPTER 3.7
The Iirst step is to create the workpiece. The recommended method is to create a point and expand it to a line curve,
Iollowed by expanding this curve to a quad surIace. Use the Iollowing button sequence to create the Iirst point.
MAIN
MESH GENERATION
pts ADD
0.24 0 0
Next, expand the point using a translation oI 0.117 inches in the x-direction and then expand the resulting curve using
a translation oI 0.7874 in the y-direction. Use the Iollowing button sequence to create the quad surIace.
MAIN
MESH GENERATION
EXPAND
TRANSLATIONS
0.117 0 0
POINTS
all: EXIST.
TRANSLATIONS
0 0.7874 0
CURVES
all: EXIST.
FILL
The next step is to convert the geometric entities to Iinite elements. This is done using the CONVERT processor. Five
divisions will be used through the thickness and 20 along the radius. Use the Iollowing button sequence to mesh the
surIace.
MAIN
MESH GENERATION
CONVERT
DIVISIONS
5 20
SURFACES TO ELEMENTS
all: EXIST.
PLOT
draw SURFACES (on)
REGEN
RETURN
917
CHAPTER 3.7
Punch
Figure 3.7-4 Result of the Convert Command
An important portion oI the analysis requires that a sharp corner will be developed at the lip oI the cylinder. To do this,
the mesh must be reIined in that area. The user will need to zoom in on that area. The nodes near the radius oI 0.25
will be moved to exactly that location. The y-coordinates oI these nodes can be determined by the SHOW command
on the NODES panel. The move operation is done using the Iollowing button sequence:
MAIN
MESH GENERATION
coordinate system SET
V DOMAIN
0 .25
V SPACING
.25
GRID (on)
RETURN
MOVE
FORMULAS
x
0.25
z
NODES (Box pick the 8th row of nodes)
37 38 39 40 41 42
END LIST (#)
RETURN
918 Marc Users Guide: Part 2
CHAPTER 3.7
The next step is to subdivide the sixth row oI elements. Use the Iollowing button sequence to subdivide the elements.
MAIN
MESH GENERATION
SUBDIVIDE
DIVISIONS
1 2 1
ELEMENTS (Box pick the 7th row of elements)
26 27 28 29 30
END LIST (#)
AIter subdividing, it is usually necessary to remove all the duplicate nodes. It is also advisable to renumber the
elements because there is a gap in the numbering Irom the subdivide operation. This can be done with the Iollowing
button sequence.
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
RETURN
RENUMBER
ALL
Step 2: Create the curves required Ior the punch & backing plate.
The next step is to create the dies. The dies will be represented by geometric entities. These entities are a combination
oI curves.
For the punch, the Iirst step is to put a point at the center oI the sphere. Then use that point to create an arc. It is easier
to have a rigid body almost touching the deIormable body. That is why the center point will be created by duplicating
the top center point oI the workpiece at a distance equal to the sphere radius. The Iollowing button sequence will create
the center point and arc.
MAIN
MESH GENERATION
DUPLICATE
TRANSLATIONS
-0.24 0 0
POINTS
1 (Pick the lower left point)
END LIST (#)
RETURN
CURVE TYPE
CENTER/RADIUS/ANGLE/ANGLE
RETURN
crvs ADD
0 0 0 (Pick the point just created (Center))
0.24 (Radius)
0 55 (Beginning angle, ending angle)
919
CHAPTER 3.7
Punch
The next curve must be tangent to the one just created. ThereIore, the curve type must be changed to arc type
tangent/radius/angle beIore creating the curve. The radius oI the arc is 0.109 inches and the angle will be a negative
55 degrees. The negative sign makes the arc go clockwise. The Iollowing button sequence will create the arc.
MAIN
MESH GENERATION
CURVE TYPE
TANGENT/RADIUS/ANGLE
RETURN
crvs ADD
11 (Pick the end point of the arc
just created (Tangent point))
0.109 (Radius)
-55 (Angle)
Figure 3.7-5 The Spherical Part for the Punch
The next step is to Iinish the rigid body. There is one line required to Iinish the punch. This is a horizontal line tangent
to the second arc. The Iollowing button sequence will create the line.
MAIN
MESH GENERATION
EXPAND
TRANSLATIONS
0 0.6 0
POINTS
14 (Pick the end point of the last arc)
END LIST (#)
FILL
920 Marc Users Guide: Part 2
CHAPTER 3.7
Figure 3.7-6 The Geometry of the Punch
The next step is to create the backing plate. First, a point will be added at the bottom oI the workpiece at a y location
oI 0.25. It will then be expanded in the x and y-direction creating the two lines required Ior the rigid body. The
Iollowing button sequence will generate these curves.
MAIN
MESH GENERATION
pts ADD
0.357 0.25 0
EXPAND
POINTS
16 (Pick the point just created)
END LIST (#)
TRANSLATIONS
0.4 0 0
POINTS
16 (Pick the corner point)
END LIST (#)
FILL
921
CHAPTER 3.7
Punch
Figure 3.7-7 Punch, Workpiece, and Backing Plate
Step 3: Apply the required Iixed displacements to the rim oI the workpiece. Apply the material data.
The Iirst step is to create a table with the stress versus plastic strain table. The Iollowing button sequence creates
the table.
MAIN
MATERIAL PROPERTIES
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
eq_plastic_strain
OK (Select OK button only if type eq_plastic_strain was typed in)
ADD
0 39000
0.7e-3 58500
1.6e-3 63765
2.55e-3 67265
3.3e-3 68250
10e-3 72150
FIT
NAME
work-hard
MORE
independent variable v1: LABEL
plastic strain
function value f: LABEL
922 Marc Users Guide: Part 2
CHAPTER 3.7
yield stress
RETURN
FILLED
SHOW TABLE
SHOW MODEL (select SHOW MODEL to go back to model view)
The next step is to input the material properties and assign them to the elements. The table must be assigned to the
yield stress value to include the workhardening. The Iollowing button sequence will assign the material properties.
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNGS MODULUS
30.0e6
POISSONS RATIO
0.3
PLASTICITY
ELASATIC-PLASTIC
INITIAL YIELD STRESS
1
initial yield stress TABLE
work-hard
OK (twice)
elements ADD
all: EXIST.
Figure 3.7-8 Workhardening Curve
The next step is to clamp the end oI the workpiece. The model is axisymmetric and thereIore, has only 2 degrees oI
Ireedom at each node. The Iirst set oI boundary conditions will clamp the node on the top right in both axial and radial
923
CHAPTER 3.7
Punch
direction. The second set oI boundary conditions will constrain the radial motion oI both the nodes on the axis oI
symmetry and the nodes on the outer radius oI the workpiece.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
OK
nodes ADD
126 (Pick the node at the top right point)
END LIST (#)
NEW
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
OK
nodes ADD
(Pick the bottom edge nodes)
(Pick the top edge nodes)
END LIST (#)
Step 4: IdentiIy the contact bodies and create the table that deIines the motion oI the rigid die, representing the punch.
This step assigns the elements and curves to the correct contact bodies. Rigid bodies must always Iollow all
deIormable bodies. The Iollowing button sequence will assign all the elements to deIormable body 1.
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE
NAME
workpiece
elements ADD
all: EXIST.
The next step is to assign the curves to rigid bodies. By deIault, analytical curves will be used Ior rigid bodies
composed oI curved entities. ThereIore, no manual interIerence is required to speciIy the number oI subdivisions used
to discretize the curves.
The Iollowing button sequence will create the 2 rigid bodies.
MAIN
CONTACT
CONTACT BODIES
NEW
NAME
punch
crvs ADD
1 2 3 (Pick curves of punch)
924 Marc Users Guide: Part 2
CHAPTER 3.7
END LIST (#)
NEW
NAME
back
crvs ADD
5 4 (Pick curves of backing plate)
END LIST (#)
At this point, it is advisable to check the correctness oI the deIinition direction oI the curves used in the rigid bodies.
MAIN
CONTACT
CONTACT BODIES
PLOT
elements SOLID
REGEN
RETURN
ID CONTACT (on)
Figure 3.7-9 Incorrect Definition Direction of Curves in Back-Plate
The ID CONTACT button will show the rigid bodies and their direction. II either oI the curves is deIined such that the
rigid body is on the same side as the deIormable body, the curve can be Ilipped by using the FLIP CURVES button.
MAIN
CONTACT
CONTACT BODIES
FLIP CURVES
4 (Pick curve)
END LIST (#)
925
CHAPTER 3.7
Punch
ID CONTACT (off)
Figure 3.7-10 Corrected Definition
The punch will move during the analysis. To deIine the motion, a table oI time versus velocity must be deIined. The
axial distance oI the straight section oI the punch and the workpiece is 0.1488 inches. This value can be determined
with the DISTANCE command on the second page oI the UTILITIES menu (use MORE).
As stated beIore, this gap will be closed in 0.4 seconds. As soon as the horizontal part oI the punch touches the
workpiece, the motion will be reversed and the release option will be switched on. In order to accomplish separation
within this single increment, the punch will be withdrawn at high velocity.
The Iollowing button sequence will deIine the table.
MAIN
CONTACT
CONTACT BODIES
TABLES
NEW
1 INDEPENDENT VARIABLE
NAME
punch_motion
TYPE
time
OK (Select OK button only if type time was typed in)
ADD
0 0.1488/0.4 (0.1488/0.4 = velocity)
0.4 0.1488/0.4
0.4 -10*0.1488/0.4 (-10*velocity)
0.5 -10*0.1488/0.4
926 Marc Users Guide: Part 2
CHAPTER 3.7
FIT
SHOW TABLE
SHOW MODEL (Select SHOW MODEL to go back to model view)
Figure 3.7-11 Velocity as a Function of Time
The Iollowing button sequence will assign the table to the punch motion (please notice that the current and thereIore
active body is body 3, the backing plate!).
MAIN
CONTACT
CONTACT BODIES
body control velocity PARAMETERS (to activate body 2)
CONTACT BODY PROPERTIES
velocity X
1
velocity x TABLE
punch_motion
OK (twice)
Step 5: DeIine the incremental steps and convergence testing parameters.
The loadcases describe the Iirst and second part oI the loading history and the loads used during those parts. The
Iollowing button sequence will create the loadcases.
MAIN
LOADCASES
NAME
indent
mechanical STATIC
927
CHAPTER 3.7
Punch
LOADS
OK
TOTAL LOADCASE TIME
0.4
# STEPS
100
SOLUTION CONTROL
MAX # RECYCLES
20
OK
OK
NEW
NAME
release
mechanical STATIC
LOADS
OK
TOTAL LOADCASE TIME
0.1
# STEPS
1
SOLUTION CONTROL
MAX # RECYCLES
20
OK
CONTACT
CONTACT RELEASES
SELECT
punch
OK
928 Marc Users Guide: Part 2
CHAPTER 3.7
Figure 3.7-12 Specify Loadcase for Indentation
Step 6: Activate the large strain parameters and submit the job.
The Iinal preprocessing step is to create the job and submit it to run in the background. The job menu deIines the
special analysis options, the results saved, and other global parameters. This is also where the loadcases can be selected
in the desired order. The Iollowing button sequence below will create the job and submit it.
Figure 3.7-13 Specify Release of Contact Body
929
CHAPTER 3.7
Punch
MAIN
JOBS
MECHANICAL
available loadcases
indent
release
ANALYSIS OPTIONS
ADVANCED OPTIONS
CONSTANT DILATATION (on)
OK
plasticity procedure SMALL STRAIN
(Switch to large strain additive procedure)
JOB RESULTS
available element tensors
Stress
available element scalars
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
OK
AXISYMMETRIC (This makes sure that the
default element type 10 is used)
OK
SAVE
RUN
SUBMIT 1
MONITOR
The MONITOR option will continually update the log and return the program control to the user when the analysis is
complete. II the user wishes, there is a monitor capability in the RESULTS menu which will allow the user to watch
the deIormations and stresses during the analysis.
930 Marc Users Guide: Part 2
CHAPTER 3.7
Figure 3.7-14 Select Job Options
Step 7: Postprocess the results by displaying the deIormed structure and the residual stresses and strains.
The analysis requires the Iinal deIormed shape and the stresses at that time. The Iollowing button sequence will present
the results.
MAIN
RESULTS
OPEN DEFAULT
FILL
PLOT
draw NODES (off)
elements SETTINGS
draw OUTLINE
RETURN
RETURN
DEF & ORIG
SCALAR
Equivalent Von Mises Stress
OK
CONTOUR BANDS
MONITOR
The deIormed shape shows that the 90 degree lip is well developed. It also shows the Iinal stresses aIter the punch
operation. The last increment (101) shows the residual stresses aIter the punch has been withdrawn.
931
CHAPTER 3.7
Punch
Figure 3.7-15 Select Quantities to be Written on Post Tape
Figure 3.7-16 Deformations and Stresses after Springback
932 Marc Users Guide: Part 2
CHAPTER 3.7
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
punch.proc Mentat procedure Iile to run the above example
Chapter 3.8: Torque Controlled Dies with Twist Transfer
3.8
Torque Controlled dies with
Twist Transfer

Chapter Overview 934

Belt and Pulley Assembly 934

Preprocessing 934

Results 940

Input Files 944

934 Marc Users Guide: Part 2
CHAPTER 3.8
Chapter Overview
Belt and Pulley Assembly
Rotational motion (twist) is prescribed on a rigid drive pulley. This twist is transIerred to another rigid pulley driven
by an elastic belt. The belt is stretched between the two rigid pulleys as shown in Figure 3.8-1. The drive pulley on the
right has a constant rotational velocity about its center, whereas the driven pulley on the leIt rotates about its center
constrained by a torsional spring. The rotational velocity oI the driven pulley is unknown and depends upon the
stiIIness oI the torsional spring and Iriction. Friction between the belt and pulleys transIers power Irom the drive to
the driven pulley. As the driven pulley rotates, the spring torque continues to increase until it overcomes the Iriction
between the pulley and belt. Then the leIt pulley begins to slip transmitting a constant torque. A small nub on the outer
portion oI the belt helps visualize the belt rotation. Units are MNewtons, meters, seconds, and radians.
Figure 3.8-1 Belt and Pulleys
Preprocessing
Let`s tour the existing model Iile by opening it in Mentat.
FILES
OPEN ch23
RETURN
CONTACT
CONTACT BODIES
ID CONTACT
935
CHAPTER 3.8
Torque Controlled dies with Twist Transfer
Figure 3.8-2 Existing Model
Nodes A and B above are associated with the driven pulley. Node A is called the control node which controls the
location oI the center oI rotation and the translational degrees oI Ireedom (DOF) that may be applied to this rigid body.
In 2-D, the Iirst DOF is rotation about the Z-axis, whereas in 3-D, DOF 1,2,3 are rotations about the X-, Y-, and Z-axis,
respectively. Note the link (spring) that grounds the Z-moment oI auxiliary node B to ground, node C. Select the driven
contact body and look at the control inIormation shown in Figure 3.8-3.
The belt is a deIormable body containing all oI the elements. The drive contact body is the curve on the right and is a
velocity controlled rigid body as shown in Figure 3.8-4.
The tables translate the drive to the right pre-tensioning the belt and then begin rotating counter clockwise. The
coeIIicient oI Iriction oI 0.5 is entered by using contact table.
The boundary condition is shown in Figure 3.8-2 where the control node oI the driven pulley is pinned along with node
C that is used to ground a torsional spring to the auxiliary node B. Figure 3.8-5 shows the link used to ground the
torsional spring with a spring constant oI 0.002 |kN-m/radian|.
The belt is a Mooney material with C
10
1 MPa. There are two loadcases, the Iirst translates the drive pulley to the
right pre-tensioning the belt. The second loadcase begins rotating the drive pulley.
In JOBS, the loadcases are chosen, element type 118 is selected, and the Iriction type is selected as shown in
Figure 3.8-6. The relative sliding velocity is set to 0.01 or 30 times slower than the tangential velocity oI the drive
pulley as per the guide lines using this Iriction model as outlined in Marc Volume A: Theory and User Information.
This value oI the relative sliding velocity will insure adequate Iriction resistance without causing numerical
convergence problems with Iriction.
Note: Node B is called the auxiliary node and it controls the rotational DOF that may be applied to this rigid body. The auxiliary
node is special because it has either one DOF (as in this problem) or three DOF`s in 3-D. Either rotations or moments
can be prescribed on the auxiliary node.
A
B
C
936 Marc Users Guide: Part 2
CHAPTER 3.8
Figure 3.8-3 Details of Rigid Body-Driven, Control, and Auxiliary Nodes
Figure 3.8-4 Details of Rigid Body-Drive
Select NEXT twice to go to
the driven contact body.
Pick left circle in Figure 3.8-2.
Click CONTROL NODE and pick Node A
(# 242 here). Its coordinates become the
center of rotation, then pick Auxiliary Node B
(# 244 here).
937
CHAPTER 3.8
Torque Controlled dies with Twist Transfer
Figure 3.8-5 Link (Torsional Spring) from Driven Pulley to Ground
Figure 3.8-6 Select Friction Type
In JOB RESULTS, custom nodal quantities are requested ( Figure 3.8-7). They include the normal and tangential
(Iriction) contact Iorce, along with a user-deIined nodal vector that is the vector sum oI the normal and tangential
contact Iorce. The UPSTNO user subroutine is used to add this combined contact nodal vector to the post Iile.
Figure 3.8-7 Select Custom Nodal Quantities
A
B
C
938 Marc Users Guide: Part 2
CHAPTER 3.8
This vector will be the sum oI the normal and Iriction components oI the belt contact Iorces. It is in a Iile called ch23.f
and is listed below:
subroutine upstno(nqcode,nodeid,valno,nqncomp,nqtype,
* nqaver,nqcomptype,nqdatatype,
* nqcompname)
implicit real*8 (a-h,o-z)
dimension valno(*)
character*24 nqcompname(*)
c
c input: nqcode user nodal post code , e.g. -1
c nodeid node id
c nqcompname not used (future expansion)
c
c output: valno() nodal values
c real/imag valno( 1: nqncomp) real
c valno(nqncomp+1:2*nqncomp) imag
c magn/phas valno( 1: nqncomp) magn
c valno(nqncomp+1:2*nqncomp) phas
c nqncomp number of values in valno
c nqtype 0 = scalar
c 1 = vector
c nqaver only for DDM 0 = sum over domains
c 1 = average over domains
c nqcomptype 0 = global coordinate system (x,y,z)
c 1 = shell (top,bottom,middle)
c 2 = order (1,2,3)
c nqdatatype 0 = default
c 1 = modal
c 2 = buckle
c 3 = harmonic real
c 4 = harmonic real/imaginary
c 5 = harmonic magnitude/phase
c
c to obtain nodal values to be used in this subroutine from
c the marc data base the general subroutine NODVAR is available:
c
c call nodvar(icod,nodeid,valno,nqncomp,nqdatatype)
c
c output: valno
c nqncomp
c nqdatatype
c
c input: nodeid
c icod
c 0='Coordinates '
c 1='Displacement '
c 2='Rotation '
c 3='External Force '
939
CHAPTER 3.8
Torque Controlled dies with Twist Transfer
c 4='External Moment '
c 5='Reaction Force '
c 6='Reaction Moment '
c 7='Fluid Velocity '
c 8='Fluid Pressure '
c 9='External Fluid Force '
c 10='Reaction Fluid Force '
c 11='Sound Pressure '
c 12='External Sound Source '
c 13='Reaction Sound Source '
c 14='Temperature '
c 15='External Heat Flux '
c 16='Reaction Heat Flux '
c 17='Electric Potential '
c 18='External Electric Charge '
c 19='Reaction Electric Charge '
c 20='Magnetic Potential '
c 21='External Electric Current'
c 22='Reaction Electric Current'
c 23='Pore Pressure '
c 24='External Mass Flux '
c 25='Reaction Mass Flux '
c 26='Bearing Pressure '
c 27='Bearing Force '
c 28='Velocity '
c 29='Rotational Velocity '
c 30='Acceleration '
c 31='Rotational Acceleration '
c 32='Modal Mass '
c 33='Rotational Modal Mass '
c 34='Contact Normal Stress '
c 35='Contact Normal Force '
c 36='Contact Friction Stress '
c 37='Contact Friction Force '
c 38='Contact Status '
c 39='Contact Touched Body '
c 40='Herrmann Variable '
c
dimension valno1(3),valno2(3)
c
if (nqcode.eq.-1) then
c... pick up contact normal force
call nodvar(35,nodeid,valno1,nqncomp,nqdatatype)
c... pick up contact friction force
call nodvar(37,nodeid,valno2,nqncomp,nqdatatype)
c... add normal and friction force
valno(1)=valno1(1)+valno2(1)
valno(2)=valno1(2)+valno2(2)
c... indicate that valno represents a vector
940 Marc Users Guide: Part 2
CHAPTER 3.8
nqtype=1
end if
c... only use nodes on belt, zero otherwise
if(nodeid.ge.242.and.nodeid.le.244) then
valno(1)=0.0d0
valno(2)=0.0d0
end if
c
return
end
The job is then submitted including the user subroutine above.
Results
Open the post Iile and skip to increment 50 which is the end oI the pre-tensioning and contour plot equivalent Cauchy
stress as shown in Figure 3.8-8.
Figure 3.8-8 Pre-tensioning the Belt
Between the pulleys, the tensile stress is about 0.80 MPa and is very uniIorm.
941
CHAPTER 3.8
Torque Controlled dies with Twist Transfer
Skip to increment 203, and plot component 11 oI the Cauchy stress. The belt has revolved halI way around the pulley
assembly and the nub is on the leIt. The diIIerence in belt tension between the upper and lower parts is due to Iriction.
The coeIIicient oI Iriction can be computed Irom the belt tension as:
OI course this is very close to the actual value used, hence the ratio oI the above stresses are correct, and Iriction is
correctly simulated.
Figure 3.8-9 Belt Stresses and Friction

1
t
---
o
b
11
o
t
11
----------
\ .
|
| |
ln 0.51
o
t
11 0.27 MPa d ( )
o
b
11 1.33 MPa
942 Marc Users Guide: Part 2
CHAPTER 3.8
Figure 3.8-10 plots the contact Iorces on the belt, Irom the user subroutine. The presence oI the Iriction shiIts the
contact Iorce vector oII normal by an angle , where . Since the coeIIicient oI Iriction is 0.5, this angle is
about 27
o
.
Figure 3.8-10 Contact Forces on Belt
u u tan
2
7
o
943
CHAPTER 3.8
Torque Controlled dies with Twist Transfer
Figure 3.8-11 is a history plot oI the pulley`s velocity. The driven pulley initially rotates quickly; however, because oI
the load induced by the torsional spring, it quickly slows down and oscillates slightly.
Figure 3.8-11 Angular Velocity of Pulleys
This oscillation occurs when the driven pulley slips and then sticks again.
Clearly the driven pulley`s angular velocity magnitude and oscillation Irequency is a Iunction oI the Iriction and
the torsional spring stiIIness. The torsional spring is used here to represent a load to the system. Remember that
this is only a statics problem and that inertial eIIects oI the system can be important during start/stop transients as
well as high angular velocities. Although a dynamic simulation would be interesting, it is beyond the intent oI
this demonstration.This simulation represents a steady quasi-static condition and time is only used to move the
drive pulley.
20.0 10.0 0.0
-0.5
0.0
0.5
1.0
Angular Velocity History
A
n
g
u
l
a
r

V
e
l
o
c
i
t
y

|
r
a
d
i
a
n
/
s
e
c
.
|
Input: Drive Twist |radians|
Drive Pulley
Driven Pulley
944 Marc Users Guide: Part 2
CHAPTER 3.8
Figure 3.8-12 is a history plot oI the input twist versus the output twist Ior various torsional spring constants. Clearly
as the load (torsional spring stiIIness, k) increases, the output twist drops. For k = 0.020, the driven pulley rotates
less. Finally, iI the was no load (k = 0), the driven pulley would be Iree wheeling and this quasi-static problem would
become singular.
Figure 3.8-12 Output Versus Input Twist of Pulleys (click right figure for animation, ESC to stop)
Using load controlled rigid surIaces one can apply a torque to the pulleys. This is better than using velocity controlled
rigid surIaces because the angular velocity oI the driven pulley is initially unknown.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
torque_ctrl_dies.mfd Mentat model Iile oI above example
torque_ctrl_dies.f Mentat model Iile oI above example
O
u
t
p
u
t
:

D
r
i
v
e
n

T
w
i
s
t

|
r
a
d
i
a
n
s
|6.0
0.0
2.0
4.0
0.0 5.0 10.0 15.0 20.0
Drive
Driven
k
Input Versus Output Twist
[k]=[kN-m/radian]
k = 0.002
k = 0.020
Input: Drive Twist |radians|
Chapter 3.9: Break Forming
3.9
Break Forming

Summary 946

Detailed Session Description of Break Forming 948

Run Job and View Results 954

Discussion 958

Modeling Tips 959

Input Files 960

Animation 962
946 Marc Users Guide: Part 2
CHAPTER 3.9
Summary
Title Break Iorming oI a metal bracket
Problem Ieatures Contact and metal Iorming - click here Ior interactive preview
Geometry
Material properties E 30x10
6
Psi, 0.3,
Analysis type Static with elastic plastic material behavior
Boundary conditions Ux 0 at center nodes, cylindrical rigid body bends metal sheet
Element type Plane strain element type 11
FE results Punch load verses stroke
u=1
v=0.4 Grid Spacing 0.1 in
X
Y
Z
v o 5x10
4
1 c
p
0.6
( )
job1
70 69 68 67 66 65 64 63 62 61
Pos Y cbody2 (x.1)
Force Y cbody2 (x1000)
-3 0
60
1.289
0
59 57
40
39
38
37
36
35
34
33
32
31
30
41
42
43
44
56 55 54 53 52 58 51
49
48
47
46
45
50
29
28
26
9
8
7
6
5
4
32
1
0
10 11 12 13
25
2423222120
18171615
14
19
27
947
CHAPTER 3.9
Break Forming
A Ilat sheet is Iormed into an angled bracket by punching it though a hole in a table using the contact option.
Figure 3.9-1 Punching Examples
The cylindrical punch drives the sheet down into the hole oI the table to a total stroke oI 0.3. The punch then returns
to its original position. The material is elastic plastic with workhardening.
Figure 3.9-2 (A) Vertical Punch Load versus Stroke (B) Stress versus Plastic Strain
At the bottom oI the stroke, the total plastic strain is nearly 45. The vertical punch Iorce is plotted versus its vertical
position. This Iorce rises quickly, hardens though about halI oI the stroke, then soItens near the end oI the stroke. Upon
liIting the punch, the punch Iorce drops rapidly and the sheet has very little springback.
The stress-plastic strain response oI a point in the sheet under the punch is plotted and shown to overlay the material
data. This workshop problem exempliIies how every point in the sheet must Iollow the material`s constitutive behavior
as well as being in equilibrium throughout the deIormation. The vertical line in the history plot to the right is the elastic
unloading oI this point in the sheet.
This is a break Iorming problem where a punch indents a sheet over a table to make an bracket. The problem geometry
is shown below:
X
Y
Z
-2.149e-004
4.581e-002
9.184e-002
1.379e-001
1.839e-001
2.299e-001
2.759e-001
3.220e-001
3.680e-001
4.140e-001
4.600e-001
lcase1
Total Equivalent Plastic Strain
Inc: 50
Time: 5.000e-001
X
Y
Z
1
-2.149e-004
4.581e-002
9.184e-002
1.379e-001
1.839e-001
2.299e-001
2.759e-001
3.220e-001
3.680e-001
4.140e-001
4.600e-001
lcase2
Total Equivalent Plastic Strain
Inc: 70
Time: 1.000e+000
X
Y
Z
1
job1
70 69 68 67 66 65 64 63 62 61
Pos Y cbody2 (x.1)
Force Y cbody2 (x1000)
-3 0
60
1.289
0
59 57
40
39
38
37
36
35
34
33
32
31
30
41
42
43
44
56 55 54 53 52 58 51
49
48
47
46
45
50
29
28
26
9
8
7
65
4
32
1
0
10 11 12 13
25
24232221201817161514
19
27
1
0.0 0.2 0.4 0.6 0.8 1.0
0
20000
40000
60000
80000
100000
Strength
Equivalent Von Mises Stress Node 63
Total Equivalent Plastic Strain
Elastic Unload
A B
948 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-3 Break Forming Geometry Problem
Detailed Session Description of Break Forming
MESH GENERATION
COORDINATE SYSTEM SET: GRID ON
V DOMAIN
-.7 .4
FILL
RETURN
CURVES ADD (pick indicated points on grid)
POINT (1,0,0), POINT(.3,0,0)
POINT(.3,0,0), POINT(.3,-.6,0)
POINT(.3,-.6,0), POINT(-.3,-.6,0)
POINT(-.3,-.6,0), POINT(-.3,0,0)
POINT(-.3,0,0), POINT(-1,0,0)
CURVE TYPE
FILLET
RETURN
CURVES ADD
radius (right horizontal curve, right vertical curve)
0.1
radius (left vertical curve, left horizontal curve)
Grid spacing 0.1" X 0.1"
u=1
v=0.4 Grid Spacing 0.1 in
Y
Z
949
CHAPTER 3.9
Break Forming
0.1
CURVE TYPE
CIRCLES: CENTER/RADIUS
RETURN
CURVES ADD
0 .2 0
.1
ELEMENTS ADD (pick points on grid)
NODE (-.9,0,0), NODE(.9,0,0)
NODE(.9,.1,0), NODE(-.9,.1.0)
SUBDIVIDE
DIVISIONS
30, 3, 1
ELEMENTS
ALL:EXISTING
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
COORDINATE SYS: SET GRID OFF
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISP
X=0
OK
NODES:ADD (pick nodes along x=0,)
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NEW
STANDARD
STRUCTURAL
950 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-4 Flow Stress
E = 3E7
= .3
OK
ELEMENTS ADD
ALL: EXISTING
TABLES
NEW
1 IND. VARIABLE
TABLE TYPE
eq_plastic_strain
OK
FORMULA
ENTER
5E4*(1+V1^.6) The equation describing the flow
FIT stress is
u
6
5
4
3
2
1
1
0.5
1 0
F (x1e5)
V1
table1
7
8
9
10
11
oy 5x10
4
1. c
p
.6
( )
951
CHAPTER 3.9
Break Forming
Figure 3.9-5 Punch Position
NEW
1 IND. VARIABLE
TABLE TYPE time
OK
ADD POINT
0, 0, .5, -.3, 1, 0
FIT
SHOW MODEL
RETURN
STRUCTURAL
PLASTICITY (twice)
INITIAL YIELD STRESS
1
TABLE1
table1 (eq_plastic_strain)
table2
V1
F (x.1)
-3
0
1
2
3
952 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-6 Identify Contact Bodies
OK (twice),
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE
OK
ELEMENTS ADD
ALL:EXISTING
NEW
RIGID
POSITION PARAMS
Y=1
TABLE
table2 (time),
OK (twice)
CURVE ADD (pick cylinder)
END LIST
ID CONTACT
NEW
CONTACT BODY TYPE RIGID
OK
CURVES ADD (pick all remaining curves)
END LIST
MAIN
cbody1
cbody2
cbody3
X
Y
Z
953
CHAPTER 3.9
Break Forming
LOADCASES
MECHANICAL
STATIC
LOADCAS
E TIME
.5
# OF STEPS
50
CONVERGENCE TESTING
DISPLACEMENTS
RELATIVE DISPLACEMETN TOLERANCE
0.001
OK (twice)
COPY
STATIC
LOADCASE TIME
.5
# OF STEPS
20
OK
MAIN
JOBS
NEW
MECHANICAL
PROPERTIES
ANALYSIS OPTIONS
LARGE STRAIN
OK
lcase1
lcase2
ANALYSIS DIMENSION: PLANE STRAIN
JOB RESULTS
EQUIVALENT VON MISES STRESS
TOTAL EQUIVALENT PLASTIC STRAIN
OK
954 Marc Users Guide: Part 2
CHAPTER 3.9
Run Job and View Results
Figure 3.9-7 Run Job Menu
CONTACT
CONTROL
ADVANCED CONTACT CONTROL
SEPARATION FORCE
.1
OK
OK (thrice)
SAVE
RUN
SUBMIT
MONITOR
OPEN POST FILE (RESUTLS MENU)
DEF ONLY
SCALAR
955
CHAPTER 3.9
Break Forming
Figure 3.9-8 Plastic Strain Plot at Increment 50
TOTAL EQUIVALENT PLASTIC STRAIN
CONTOUR BANDS
SKIP TO INCREMENT 50
RESULTS
SKIP TO INCREMENT 70
-2.149e-004
4.581e-002
9.184e-002
1.379e-001
1.839e-001
2.299e-001
2.759e-001
3.220e-001
3.680e-001
4.140e-001
4.600e-001
lcase1
Total Equivalent Plastic Strain
Inc: 50
Time: 5.000e-001
X
Y
Z
1
956 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-9 Plastic Strain Plot at Increment 70
RESULTS
HISTORY PLOT
SET LOCATIONS
n:63 # (pick bottom middle node n:63)
ALL INCS
ADD CURVES
GLOBAL
Pos Y cbody2
Force Y cbody2
FIT
-2.149e-004
4.581e-002
9.184e-002
1.379e-001
1.839e-001
2.299e-001
2.759e-001
3.220e-001
3.680e-001
4.140e-001
4.600e-001
lcase2
Total Equivalent Plastic Strain
Inc: 70
Time: 1.000e+000
X
Y
Z
1
957
CHAPTER 3.9
Break Forming
Figure 3.9-10 History Plot of Punch Force versus Stroke
RESULTS
HISTORY PLOT
CLEAR CURVES
ALL INCS
ADD CURVES
ALL LOCATIONS
Total Equivalent Plastic Strain
Equivalent Von Mises Stress
FIT
job1
70 69 68 67 66 65 64 63 62 61
Pos Y cbody2 (x.1)
Force Y cbody2 (x1000)
-3 0
60
1.289
0
59 57
40
39
38
37
36
35
34
33
32
31
30
41
42
43
44
56 55 54 53 52 58 51
49
48
47
46
45
50
29
28
26
9
8
7
6
5
4
32
1
0
10 11 12 13
25
2423222120
1817
16
15
14
19
27
958 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-11 Stress versus Plastic Strain Node 63
Discussion
Since the sheet completely wraps around the rigid cylinder an the end oI the bending, we can estimate the strain
assuming that the sheet completely surrounds the cylinder, and by knowing the strain, the stress can also be estimated
as shown in Figure 3.9-12.
Figure 3.9-12 Estimating the Bending Strain and Stress in the Center of the Sheet
0.0 0.2 0.4 0.6 0.8 1.0
0
20000
40000
60000
80000
100000
Strength
Equivalent Von Mises Stress Node 63
Total Equivalent Plastic Strain
Elastic Unload
A B
C D
A B
O

C D
( )
= =
( ) ( ) = 75 Ksi

CD
CD C D
C D
--------------------------- =
C D A B =
A B AB =

CD
CD AB
AB
-----------------------
r
OC
r
OA

r
OA

---------------------------------
r
OC
r
OA

r
OA
------------------------- =

CD
0.20 0.15
0.15
---------------------------
0.05
0.15
----------
1
3
-- - = = =

CD
5x10
3
1 1 3
0.6
+
/
r = 0.1
t = 0.1
959
CHAPTER 3.9
Break Forming
With the stresses estimated, we can assume that a Iully plastic hinge Iorms in the center oI the sheet and estimate the
punch load as shown in Figure 3.9-13.
Figure 3.9-13 Estimating the bending moment and maximum punch load
The estimate Ior the maximum punch load, 1250 lbI, is very close to that Iound by the analysis as shown in
Figure 3.9-10 oI 1289 lbI. Finally, although the Iinal angle aIter spring back appears close to 90
o
, its actual value is
84
o
, and the punch stroke should be slightly reduced to Iorm a right angle aIter springback.
Modeling Tips
The punch Iorce in Figure 3.9-10 has an abrupt jump at increment 50 which is caused by a new node entering contact
between the sheet and the rigid table. This can be improved by using segment to segment contact and entered in the
modeling under contact control; the comparison between node to segment and segment to segment contact is shown
in Figure 3.9-14.
JOBS
NEW
MECHANICAL
PROPERTIES
CONTACT CONTROL
METHOD: SEGMENT TO SEGMENT
OK (twice)
RUN, SUBMIT....
0.3
M
M
P/2 P/2
(y)
y
y
x
x
960 Marc Users Guide: Part 2
CHAPTER 3.9
Figure 3.9-14 Comparison of Node to Segment versus Segment to Segment Contact
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
s4.proc Mentat procedure Iile to run the above example
55 54 53 52 51
50
49
48
47
46
45
44
43
42
41
56 57 58 59
1.289
0
0 -3
Y (x1000)
Pos Y cbody2 (x.1)
70
40
69 67 66 64 63 62 61 60 68
Force Y cbody2
39
37 16151413121110
9
8
7
6
5
4
32
17 18 19 20 36
35
34
33
32
31
30
38
29
27
2625
24232221
28
1
0 65 65 68
49
48
47
46
45
44
50
43
41
40
39
38
37
36
42
35
51 53 67 66 64 63 62 52 61 59 58 57 56 55 54 60 69
34
32
13
12
11
10
9
8
14
7
5
4
3
2
1
0
6
33
15
17
31
30
29
28
2726
16 25232221201918 24
Segment to Segment Contact
Segment to Segment Contact
Node to Segment Contact
Node to Segment Contact
0.000e+000
8.587e+001
1.717e+002
2.576e+002
3.435e+002
4.293e+002
5.152e+002
6.011e+002
6.869e+002
7.728e+002
8.587e+002
Contact Normal Force
Inc: 50
Time: 5.000e-001
X
Y
Z
961
CHAPTER 3.9
Break Forming
Also, this problem can automatically be run Irom the HELP menu under DEMONSTRATIONS > RUN A DEMO
PROBLEM > CONTACT as shown below.
962 Marc Users Guide: Part 2
CHAPTER 3.9
Animation
Click on the Iigure below to activate the video; it lasts about 10 minutes and explains
how the steps above are done. Once the video is activated, a right click oI the mouse
will open the menu shown at the right; you can switch to various screen sizes or stop
the video by disabling the content.
Full Screen Multimedia
Close Floating Window
Disable Content
Properties...
Insert c3.9.Ilv here
Chapter 3.10: Hertz Contact Problem
3.10
Hertz Contact Problem

Summary 964

Run Jobs and View Results 970

FEA versus Theoretical Solutions 971

Input Files 972

964 Marc Users Guide: Part 2
CHAPTER 3.10
Summary
Title Hertz contact problem
Problem Ieatures Steel cylinder contacting an aluminum block
Geometry
Material properties E 30x10
6
Psi, 0.3 Ior steel and E 10x10
6
Psi, 0.33 Ior Al
Analysis type Static with elastic material behavior
Boundary conditions Symmetry with point load pushing cylinder into block
Element type Linear and parabolic plane strain quads
FE results Comparison to theory
D_
2
= 5
10
X
Y
Z
(10, 0)
(10, 2)
p = 10,000 psi
P
b
D
Cylinder of length L >> D; p = P/L
K
D
= D
v v
Theory versus FEA
Theory Linear Elements Quadratic Elements
230.9 181.6 225.3
Error () -21.4 -2.4
Max o ksi | |
965
CHAPTER 3.10
Hertz Contact Problem
In this example problem, a steel cylinder with a radius oI 5 is pressed against a 2 deep aluminum base. A small strain
elastic analysis is perIormed. so the only nonlinearity introduced is due to contact. A comparison between using linear
and quadratic elements is shown. The problem is similar to the Hertz contact problem in Timoshenko and Goodier,
Theory of Elasticity, 1951.
Figure 3.10-1 (A) Finite Element Mesh (B) o
yy
using Linear Element (C) o
yy
using Quadratic Elements
In this problem, you will modiIy an existing model and add quadratic elements with contact. The steel material
properties have an elastic modulus oI 30E6 and a Poisson`s ratio oI 0.30 and the aluminum properties have an elastic
modulus oI 10E6 and a Poisson`s ratio oI 0.33. The cylinder and base plate are pressed together with a load oI 100,000
lbI or 10,000 psi applied across the diameter oI the he mi cylinder.
X
Y
Z
1
-1.816e+005
-1.605e+005
-1.395e+005
-1.184e+005
-9.736e+004
-7.630e+004
-5.524e+004
-3.418e+004
-1.312e+004
7.935e+003
2.899e+004
lcase1
Comp 22 of Stress
Inc: 10
Time: 1.000e+000
X
Y
Z
1
-2.253e+005
-2.018e+005
-1.783e+005
-1.548e+005
-1.313e+005
-1.078e+005
-8.432e+004
-6.082e+004
-3.733e+004
-1.384e+004
9.658e+003
lcase1
Comp 22 of Stress
Inc: 10
Time: 1.000e+000
X
Y
Z
(A)
(B) (C)
966 Marc Users Guide: Part 2
CHAPTER 3.10
Figure 3.10-2 Cylinder and Base Pressed Together
FILES
OPEN
hertzbase.mud
OK
MAIN
CONTACT
CONTACT BODIES
ID CONTACT
MAIN
P
b
D
Cylinder of length L >> D; p = P/L
K
D
= D
967
CHAPTER 3.10
Hertz Contact Problem
Figure 3.10-3 Hertz Base Analysis
JOBS
RUN
SUBMIT(1)
MONITOR
OK
OPEN POST FILE (RESULTS MENU)
DEF ONLY
SCALAR
Comp 22 of Stress
OK
CONTOUR BANDS
LAST
D_
2
= 5
10
X
Y
Z
(10, 0)
(10, 2)
p = 10,000 psi
968 Marc Users Guide: Part 2
CHAPTER 3.10
Figure 3.10-4 Hertz Base at Comp 22 of Stress (using Lower-order Elements)
Here, we see that the peak stress using linear elements is around 141 Ksi in compression. We suspect that this is low
due to the Iact that linear elements cannot capture stress concentration as well as quadratic elements. ThereIore, we
will change the element type and rerun the problem.
First, we move the aluminum sheet down one inch, attach edges to the arc, change element types, sweep, and move
the aluminum sheet back to its original position.
CLOSE
FILES
SAVE AS
hertzbasequad.mud
MAIN
MESH GENERATION
MOVE
TRANSLATIONS
0 -1 0
ELEMENTS (select the aluminum elements)
al
END LIST
SELECT
-1.816e+005
-1.605e+005
-1.395e+005
-1.184e+005
-9.736e+004
-7.630e+004
-5.524e+004
-3.418e+004
-1.312e+004
7.935e+003
2.899e+004
lcase1
Comp 22 of Stress
Inc: 10
Time: 1.000e+000
X
Y
Z
1
969
CHAPTER 3.10
Hertz Contact Problem
Figure 3.10-5 Steel Cylinder Nodes N1, N2, and N3
METHOD = PATH
OK
EDGES (pick nodes N1, N2, N3)
END LIST
OK (twice)
ATTACH
EDGES -> CURVE\ (select circular arc)
ALL: SELECTED EDGES
RETURN
CHANGE CLASS
QUAD(8)
ELEMENTS
ALL: EXISTING
RETURN
SWEEP
ALL
RETURN
RENUMBER
N1
N2
N3
970 Marc Users Guide: Part 2
CHAPTER 3.10
Run Jobs and View Results
ALL
RETURN
MOVE
TRANSLATIONS
0 1 0
ELEMENTS (select aluminum elements)
END LIST
MAIN
JOBS
ELEMENT TYPES
ANALYSIS DIMENSION
SOLID
27
OK
ALL: EXISTING
RETURN
PROPERTIES
CONTACT CONTROL
ADVANCED CONTACT CONTROL
QUAD. SEGMENTS GENUINE,
OK (thrice)
SAVE
RUN
SUBMIT(1)
MONITOR,
OK
OPEN POST FILE (RESULTS MENU)
LAST
DEF ONLY
SCALAR (Comp 22 of Stress)
CONTOUR BANDS
SELECT
971
CHAPTER 3.10
Hertz Contact Problem
L
Figure 3.10-6 Results in Steel Cylinder (using Higher-order Elements)
FEA versus Theoretical Solutions
From the 6th Edition oI Roark`s Formulas Ior Stress and Strain (by Warren C. Young, 1989, pg 651), we can derive
the contact patch, b, and the maximum compressive stress, .
Figure 3.10-7 Steel Cylinder in Contact with Aluminium Base
SELECT CONTACT BODY ENTITIES
(steel)
OK
MAKE VISIBLE
FILL
-2.253e+005
-2.018e+005
-1.783e+005
-1.548e+005
-1.313e+005
-1.078e+005
-8.432e+004
-6.082e+004
-3.733e+004
-1.384e+004
9.658e+003
lcase1
Comp 22 of Stress
Inc: 10
Time: 1.000e+000
X
Y
Z
1
Max o
P
b
D
Cylinder of length L >> D; p = P/L
K
D
= D
972 Marc Users Guide: Part 2
CHAPTER 3.10
The contact width Ior model depicted is given by . Where , and the contact area
Ior the halI model becomes, .
The maximum stress becomes .
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Table 2.10-1 Theory versus FEA
Theory Linear Elements Quadratic Elements
230.9 181.6 225.3
Error () -21.4 -2.4
File Description
s8.proc Mentat procedure Iile to run the above example
hertzbase.mud Mentat model Iile read by above procedure Iile
b 1.60 pK
D
C
E
C
E
1 v
1
2

E
1
-----------------
1 v
2
2

E
2
-----------------
b
2
--- 0.80 pK
D
C
E
0.276
Max o 0.798
p
K
D
C
E
--------------- 230.9Ksi
Max o ksi | |
Chapter 3.11: Anisotropic Sheet Drawing using Reduced Integration Shell Elements
3.11
Anisotropic Sheet Drawing
using Reduced Integration
Shell Elements

Chapter Overview 974

Simulation of Earing for Sheet Forming with Planar Anisotropy

975

Boundary Conditions 976

Advanced Topic: Drawbead Modeling using Nonlinear Spring

991

Input Files 996

References 997
974 Marc Users Guide: Part 2
CHAPTER 3.11
Chapter Overview
In many manuIacturing area such as packing, automotive, aerospace, and electronics industries, the control oI sheet
metal Iorming processes has become a key Iactor to reduce the development time and the Iinal cost oI products. In
general, sheet metal Iorming is analyzed on the basis oI stretching, drawing, bending, or various combinations oI these
basic modes oI deIormation, which, Irom the viewpoint oI mechanics, involve nonlinearity resulting Irom geometry,
material, and contact aspects. Numerical simulations oI sheet Iorming processes need to account Ior
those nonlinearities. The Iollowing aspects warrant special attention:
1. Geometric Nonlinearity:
In order to describe the nonlinear geometric behavior, especially shells, three basic approaches can be
identiIied: degenerated shell elements, classical shell elements, and more recently, enhanced strain
Iormulation. The need Ior large-scale computations together with complex algorithms Ior geometrical and
material nonlinear applications motivated Iinite element researchers to develop elements that are simple and
eIIicient. Various signiIicant research has been carried out to develop reduced integration shell element based
on the degenerated shell approach. A one point quadrature shell has been developed in Marc based on the work
oI Cardosa, Reference 1. This is a Iour-node, thick-shell element with global displacements and rotations as
degrees oI Ireedom. Bilinear interpolation is used Ior the coordinates, displacements and the rotations. This
shell introduced the MITC4 shell geometry with the ANS (Assumed Natural Strain) method in conjunction
with the physical stabilization scheme to construct an element with reduced integration, which is Iree oI any
artiIicial correction Ior warping. This procedure improves the accuracy oI one point quadrature shell element
without sacriIicing the computational speed and permits large nonlinear behavior. The nodal Iiber coordinate
system at each node is update by a step-by-step procedure in order to consider the warping oI the element. A
rigid-body projection matrix is applied to extract out rigid-body motion so the element can undergo large
rotations.
2. Material Nonlinearity:
The nonlinear plastic behavior must account Ior the anisotropy exhibited by sheet metals. During cold working,
anisotropic properties change due to the material microstructure evolution. The assumption that the change oI
anisotropic properties during plastic deIormation is small and negligible when compared to the anisotropy
induced by rolling has been widely adopted in the analysis oI sheet metal Iorming. The appropriate anisotropic
yield Iunctions Ior sheet metal Iorming simulations is important to obtain a reliable material response. Barlat,
Reference 2 proposed a general criterion Ior planar anisotropy that is particularly suitable Ior aluminum
alloy sheets. This criterion has been shown to be consistent with polycrystal-based yield surIaces, which oIten
exhibit small radii oI curvature near uniaxial and balanced biaxial tension stress states. An advantage oI this
criterion is that its Iormulation is relatively simple as compared with the Iormulation Ior polycrystalline
modeling and, thereIore, it can be easily incorporated into Iinite element (FE) codes Ior the analysis oI metal
Iorming problems. Mentat provides the automatic calculation oI the anisotropic coeIIicients directly Irom
experimental data.
975
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
3. Drawbead Modeling:
To Iorm complex shaped surIace, drawbeads are used to insure the accuracy oI the Iinal shape, and also to
prevent Iracture and cracks. Extreme caution has to be placed in installing drawbeads, especially in the case oI
an aluminum plate that has low Ilexibility. The design oI drawbeads are determined based on the result oI try-
out, which causes Iorming tool design to be rather diIIicult. ThereIore, there is the need Ior the development oI
a logical numerical method, Ior understanding the quantitative eIIect oI the drawbeads at the stage oI die
design. In Marc, a simple drawbead model based on nonlinear spring concepts has been developed. The
nonlinear drawbead Iorce with displacement is applied to the nodes located on the blank edges.
4. Forming Limit Analysis:
Forming Limit Diagrams (FLD) are used extensively during tool design Ior the manuIacturing oI sheet metal
parts. It is also used Ior trouble shooting during regular shop Iloor production. It is observed that FLD is
strongly dependent on the basic mechanical properties oI sheet metal like the work hardening exponent, initial
sheet thickness, and the strain rate sensitivity. In addition, it is Iound that strain paths have signiIicant inIluence
on the limit strains that develop during sheet metal Iorming. In Marc, the combined method accommodating
localized necking and diIIused necking with Keeler's experimental work (IDDRG, 1976) was adopted to
predict FLD.
The cup drawing example presented here was designed to demonstrate Iour Ieatures: One-point integration shell
element, Barlat's yield Iunction, drawbead modeling with nonlinear springs, and FLD. The tool geometry and material
data was taken Irom NUMISHEET 2002 benchmark, Reference 3. But, process conditions are slightly diIIerent Irom
the original data.
Simulation of Earing for Sheet Forming with
Planar Anisotropy
The cup drawing test simulation with circular punch and blank is one oI most popular tests to veriIy the planar
anisotropic behavior through the prediction oI the earing proIile. In the cylindrical cup drawing test, the material
undergoes compressive deIormation in the Ilange area due to the circumIerential contraction. Some stretching occurs
also in the radial direction oI a cup. This test was simulated Ior a 6111-T4 aluminum alloy sheet based on the new
one-point shell element and Barlat's yield Iunction. Also, FLD prediction and drawbead modeling with nonlinear
spring were investigated. Assuming isotropic hardening, the yield Iunction coeIIicients are kept constant during the
simulation. The schematic view oI the cup drawing process analyzed are shown in Figure 3.11-1.
976 Marc Users Guide: Part 2
CHAPTER 3.11
Figure 3.11-1 Tool for Cylindrical Cup Drawing
Only a quarter section oI the cup was analyzed in the light oI the orthotropic symmetry. The generation oI mesh using
Mentat is straightIorward, so it is not discussed here. The generated mesh was stored in sheet_mesh.mud.
Boundary Conditions
The symmetric boundary conditions were imposed Ior the corresponding symmetric nodes using two boundary node
sets: (1) x-displacement y and z rotations are zero Ior the nodes located in y0 and (2) y-displacement x and z rotations
are zero Ior the nodes located in x0.
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
ROTATION Y
0
ROTATION Z
0
OK
NODES ADD
2 39 57 75 93 111 129 147 165 183 201
219 237 255 273 291 309 327 345 363 381
END LIST
R!
R#
t
0

PUNCH
HOLDER
DIE
R
2

R
3

R
1

R
1
50.0, R
2
51.25, R
3
9.53, R
4
7.14 (Unit: mm)
(Blank size: R
o
90.0, t
o
1.0)
977
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-2 Boundary Condition ID
Material Properties
Two tables are provided to characterize stress vs. strain behavior and to control the motion oI rigid surIace (punch).
The stress-strain law Ior 6111-T4 aluminum alloy sheet is given as Iollows:
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT Y
0
ROTATION X
0
ROTATION Z
0
OK
NODES ADD
2 3 4 5 6 7 8 9 10 11 12
13 14 15 16 17 18 19 20 21 1
END LIST
o 429.8 237.7*exp 8.504c
p
( )
978 Marc Users Guide: Part 2
CHAPTER 3.11
Voce-hardening curve is used to Iit the saturation behavior oI the aluminum alloy.
Figure 3.11-3 Generated Table for Stress-Strain Curve
A second table representing a ramp Iunction Ior control oI rigid-body (especially Ior punch) is generated by simply
adding Iew points.
MATERIAL PROPERTIES
TABLES
NEW
NEW TABLE
1 INDEPENDENT VARIABLE
TYPE
eq_plastic_strain
FORMULA
428.8-237.7*exp(-8.504*v1)
FIT
TABLES
NEW
NEW TABLE
1 INDEPENDENT VARIABLE
979
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-4 Generated Table for Rigid-Body Control
The material Ior all elements is treated as an elasto-plastic properties with Young's modulus oI 70 Gpa, Poisson's ratio
oI 0.3, and the initial yield stress oI 192.1. Anisotropic material data Ior Barlat's yield Iunctions is taken Irom
Numisheet 2002 benchmark:
Anisotropic material data for Barlat's (1991) yield criterion
Marc calculates Barlart's anisotropic coeIIicients (C
1
, C
2
, C
3
, C
4
, C
5
, C
6
) directly Irom raw experimental data (initial
yield stresses along 0,45,90, biaxial directions) by solving a nonlinear equation. II Barlat's anisotropic coeIIicients are
already known, then the calculation is not necessary and direct input oI the coeIIicients is also allowed in Mentat. II
the biaxial yield stress (Y
b
) is not available, Y
b
/Y
0
could be assumed to be 1. The material coeIIicients, C
i1-6
,
TYPE
time
ADD
0 0
1 1
FIT
Yield stresses: Y
0
= 192.1, Y
45
=187.4, Y
90
=181.2, Y
b
=191.4
(Ratio: Y
45
/Y
0
= 0.9755, Y
90
/ Y
0
= 0.9432, Y
B
/ Y
0
= 0.9963)
Exponent: m=8
980 Marc Users Guide: Part 2
CHAPTER 3.11
represent anisotropic properties. When C
i1-6
1, the material is isotropic and Barlat's (1991) yield Iunction reduces to
the Tresca yield condition Ior m 1 or , and the von Mises yield criterion Ior m 2 or 4. The exponent 'm is mainly
associated with the crystal structure oI the material. A higher 'm value has the eIIect oI decreasing the radius oI
curvature oI rounded vertices near the uniaxial and balanced biaxial tension ranges oI the yield surIace, in agreement
with polycrystal models. Values oI m 8 Ior FCC materials (e.g. aluminum) and m 6 Ior BCC materials (e.g. steel)
are recommended.
The yield surIace has been proven to be convex Ior . Figure 3.11-5 shows the yield surIaces obtained Irom
von Mises, Hill, and Barlat`s yield Iunctions Ior an aluminum alloy.
Figure 3.11-5 Comparison of Yield Surfaces Obtained from
von Mises, Hill, and Barlats Yield Functions
The ORIENTATION option is required to assign the initial rolling and transverse direction Ior all elements. In the
simulation, rolling direction vector is (1,0,0) and transverse direction is (0,1,0).
FLD Prediction
In order to accommodate Iailure prediction in the analysis results, the FLD
0
value as shown in Figure 3.11-6 need to
be inserted. The FLD
0
value increases with the strain-hardening exponent, n, and the strain-rate exponent, m.
According to large amount oI experiments, the real FLD curves are also aIIected by the thickness oI the sheet metal.
This phenomenon is reIerred as thickness eIIect and it is characterized as thickness coeIIicient t
c
. Experiments tended
to express this relationship by Keeler:
where , iI n is less than 0.21 otherwise . T is the thickness oI the sheet metal. The thickness
coeIIicient, t
c
is the set as 3.59 iI the unit used to deIine the thickness is 'Inch'. II unit oI 'mm' is used, t
c
is the set as
0.141. For this material, the eIIective value oI n is 0.226.

m 1 >
FLD
0
Q 0.233 t
c
T ( )
Q n 0.21 Q 1.0
981
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-6 Forming Limit Diagram
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNG'S MODULUS
70000
POISSON'S RATIO
0.3
ELASTO-PLASTIC
YIELD SURFACE
BARLAT
INITIAL YIELD STRESS
1
TABLE
table1
EXPERIMENTAL DATA INPUT
EXPERIMENTAL DATA
M
8

Experimental
Iorming limit
Local necking DiIIuse necking
0
Minor Principal Engineering Strain e
2
Major Principal Engineering Strain e
1
FLD
0
982 Marc Users Guide: Part 2
CHAPTER 3.11
Y45/Y0
0.9755
Y90/Y0
0.9432
YB/Y0
0.9963
COMPUTE
COMPUTED DATA
APPLY
OK (twice)
FORMING LIMIT
PREDICTED
STRAIN HARDENING EXP.
0.226
THICKNESS COEFFCIENT
0.141
OK (twice)
ORIENTATIONS
NEW
ZX PLANE
ELEMENTS ADD
ALL EXIST
RETURN
ELEMENTS ADD
ALL EXIST
983
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-7 Calculation of Barlats Anisotropic Coefficients
Figure 3.11-8 Orientation Arrow for Rolling Direction
984 Marc Users Guide: Part 2
CHAPTER 3.11
Figure 3.11-9 FLD (Forming Limit Diagram) Input
Geometric Properties
The sheet thickness is 1 mm and shell elements are also used Ior the analysis
Contact
The Iirst body is the deIormable workpiece; the second, the third, and the Iourth are respectively the rigid punch, rigid
die, and rigid holder deIined with analytical surIaces. Friction coeIIicient was taken as 0.05. The second body (punch)
is moved up to 40 mm with Iixed displacement boundary condition using table2 in CONTACT BODY option. The
gap between die and blankholder is uniIorm.
GEOMETRIC PROPERTIES
NEW
3D
SHELL
THICKNESS
1
FLAT ELEMENT
OK
ELEMENTS ADD
ALL EXIST
985
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
CONTACT
CONTACT BODIES
NEW
DEFORMABLE
FRICTION COEFFICIENT
0.05
OK
ELEMENTS ADD
ALL EXIST
NEW
RIGID
POSITION
PARAMETERS
POSITION (CENTER OF POSITION)
Z
40
OK
TABLE
table2
OK
FRICTION COEFFICIENT
0.05
ANALYTICAL
OK
SURFACES ADD
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23 24 25 26 27
END LIST
NEW
RIGID
VELOCITY
FRICTION COEFFICIENT
0.05
ANALYTICAL
OK
986 Marc Users Guide: Part 2
CHAPTER 3.11
SURFACES ADD
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
43 44 45 46 47 48 49 50 51 52 53 54
END LIST
NEW
RIGID
VELOCITY
FRICTION COEFFICIENT
0.05
ANALYTICAL
OK
SURFACES ADD
56 57 58 59 60 61 62 63
END LIST
RETURN
CONTACT TABLE
NEW
PROPERTIES
cbody1 cbody2
CONTACT TYPE
TOUCHING
DISTANCE TOLERANCE
0.1
SEPARATION THRESHOLD
10.0
OK
cbody1 cbody3
CONTACT TYPE
TOUCHING
DISTANCE TOLERANCE
0.1
SEPARATION THRESHOLD
10.0
OK
987
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-10 Contact ID: (a) ID Contact (b) ID Backface
Load Steps and Job Parameters
A total oI 100 Iixed steps are used Ior the entire analysis with a convergence displacement norm oI 0.1.
cbody1 cbody4
CONTACT TYPE
TOUCHING
DISTANCE TOLERANCE
0.1
SEPARATION THRESHOLD
10.0
OK (twice)
LOADCASES
MECHANICAL
NEW
STATIC
CONTACT
CONTACT TABLE
ctable1
OK
(a) (b)
988 Marc Users Guide: Part 2
CHAPTER 3.11
The analysis is a normal mechanical analysis with one loadcase. COULOMB FOR ROLLING option is selected with a
bias Iactor oI 0.9. New post variable Forming Limit Parameter is selected in this example, besides the Equivalent Von
Mises Stress and Equivalent Plastic Strain. The ADDITIVE DECOMPOSITION option must be chosen Ior plasticity
procedure when the anisotropic yield Iunction is used.
CONVERGENCE TESTING
RELATIVE
DISPLACEMENTS
RELATIVE DISPLACEMENT TOLERANCE
0.1
OK
STEPPING PROCEDURE
CONSTANT TIME STEP
# STEP
100
OK
JOBS
NEW
MECHANICAL
LOADCASES
activate:
lcase1
CONTACT CONTROL
FRICTION TYPE
COULOMB FOR ROLLING
INITIAL CONTACT
CONTACT TABLE
ctable1
OK
ADVANCED CONTACT CONTROL
DISTANCE TOLERANCE
0.1
DISTANCE TOLERANCE BIAS
0.9
989
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
For the analysis oI the cup-drawing, newly developed one-point quadrature shell element oI 140 is being used.
Save Model, Run Job, and View Results
SEPARATION FORCE
10
OK (twice)
ANALYSIS OPTIONS
PLASTICITY PROCEDURE
LARGE STRAIN ADDITIVE
OK
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
Forming Limit Parameter
Major Engineering Strain
Minor Engineering Strain
OK (twice)
ELEMENT TYPES
MECHANICAL
3-D MEMBRANE SHELL
140
OK
ALL: EXIST
FILE
SAVE AS
sheet.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
990 Marc Users Guide: Part 2
CHAPTER 3.11
MONITOR
Figure 3.11-11 shows the top view oI deIormed conIigurations based on simulation and experiment. The experimental
cup shape measured Irom Numisheet 2002 benchmark is used Ior the comparison. It is shown that both results are
compatible. Figure 3.11-12 shows FLD parameter and equivalent plastic strain. Forming Limit Parameter covers the
range oI 0 to1, where '0.0 means no strain and '1.0 means Iailure.
Figure 3.11-11 Top View for Deformed Shape at the Punch Stroke of 40 mm: (a) Simulation (b) Experiment
MAIN
RESULTS
OPEN
sheet.t16
OK
DEF & ORIG
CONTOUR BAND
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
SCALAR
Forming Limit Parameter
OK
MONITOR
SCALAR
Equivalent Plastic Strain
OK

(a) (b)
991
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-12 Deformed Configuration at the Punch Stroke of 40 mm:
(a) Forming Limit Parameter (b) Equivalent Plastic Strain
Advanced Topic: Drawbead Modeling using Nonlinear Spring
Nonlinear springs in Marc is designed Ior multiple purposes. Nonlinear springs can be speciIied using either spring
stiffness or spring force. Spring stiIIness method is usually used Ior heat transIer coeIIicients Ior thermal springs,
electrical conductivity Ior electrical springs. While, spring Iorce method can be used Ior Ilux Ior thermal springs,
current Ior electrical springs and drawbead model in sheet metal Iorming, etc. When nonlinear spring Iorce option is
employed, the use oI a table as a Iunction oI displacement is required, the spring stiIIness based on the table gradient
is then internally calculated.
In order to utilize a simple drawbead model based on nonlinear spring concepts, the nonlinear drawbead Iorce with
displacement is applied to the nodes located on the blank edges. For the implementation, Iorce vs. displacement table,
LINKS and boundary condition need to be added.
Links
For the spring Iorce-displacement table, analytical Iormula using 500*tanh(x) was used.
TABLES
NEW
TYPE
displacement
FORMULA
500*tanh(v1)
FIT
(a) (b)

992 Marc Users Guide: Part 2
CHAPTER 3.11
Figure 3.11-13 Generated Table for Nonlinear Spring Behavior (Force vs. Displacement)
For the creation Ior nonlinear springs, the N TO N SPRING option is used to generate 19 springs combined with
table3 (Iorce vs. displacement).
LINKS
SPRINGS/DASHPOTS
N TO N SPRINGS
TYPE
TRUE DIRECTION
BEHAVIOR
PROPERTIES
FORCE
SET
1
TABLE
table3
OK
ADD SPRINGS
(Enter n to n spring/dashpots begin node path)
364 365 366 367 368 369 370 371 372 373
374 375 376 377 378 379 380 381 1
993
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-14 Nonlinear Springs based on Spring Force Method
Figure 3.11-15 Generated Nonlinear Springs
END LIST
(Enter n to n spring/dashpots begin node path)
383 384 385 386 387 388 389 390 391 392
393 394 395 396 397 398 399 400 401
END LIST
994 Marc Users Guide: Part 2
CHAPTER 3.11
Boundary Conditions
The spring nodes, which are not connected with sheet metal must be constrained in all directions.
Save Model, Run Job, and View Results
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
DISPLACEMENT Y
0
DISPLACEMENT Z
0
ROTATION X
0
ROTATION Y
0
ROTATION Z
0
OK
NODES ADD
383 384 385 386 387 388 389 390 391 392 393
394 395 396 397 398 399 400 401
END LIST
FILE
SAVE AS
sheet_ns.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
995
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
Figure 3.11-16 shows the top views Ior the deIormed conIiguration. As shown in the Iigure, the use oI nonlinear
springs as drawbead constrains the Ilow oI sheet metal into die cavity. Hence, less draw-in is observed compared to
the results based on the simulation without nonlinear springs. Also, Figure 3.11-17 shows Forming Limit Parameter
and Equivalent Plastic Strain contours. Compared with Figure 3.11-12 (simulation without nonlinear spring), the
levels oI Forming Limit Parameter and Equivalent Plastic Strain are larger due to more plastic deIormation.
MAIN
RESULTS
OPEN
sheet.t16
OK
DEF & ORIG
CONTOUR BAND
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
SCALAR
Forming Limit Parameter
OK
MONITOR
SCALAR
Equivalent Plastic Strain
OK
MONITOR
996 Marc Users Guide: Part 2
CHAPTER 3.11
Figure 3.11-16 Top View for Deformed Shape at the Punch Stroke of 40 mm:
(a) with Nonlinear Springs (b) without Nonlinear Springs
Figure 3.11-17 Deformed Configurations at the Punch Stroke of 40 mm with Nonlinear Springs:
(a) Forming Limit Parameter (b) Equivalent Plastic Strain
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
sheetforming_nolink.proc Mentat procedure Iile to run the above example
sheetforming_link.proc Mentat procedure Iile to run the above example
sheet_mesh.mud Mentat model Iile read by above procedure Iiles

(a) (b)

(a) (b)


997
CHAPTER 3.11
Anisotropic Sheet Drawing using Reduced Integration Shell Elements
References
1. Cordosa, R.P.R., Yoon, J.W., Gracio, J.J., Barlat, Fl, and Cesar de Sa, J.M.A., Development oI a one point
quadrature shell element Ior nonlinear applications with contact and anisotropy, Comput. Methods Appl.
Mech. Engrg, 191, 5177-5206 (2002).
2. Barlat, F., Lege, D.J., and Brem, J.C., A six-component yield Iunction Ior anisotropic metals, Int. J. Plasticity,
7, 693-712 (1991).
3. Yang, O.Y., Oh, S.I., Huh, H., and Kim, Y.H., Proceedings oI NUMISHEET 2002, Oct. 21-25, Seju Island,
Korea (2002).
998 Marc Users Guide: Part 2
CHAPTER 3.11
Chapter 3.12: Chaboche Model
3.12
Chaboche Model

Chapter Overview 1000

Blade on a Fan of a Turbine Engine 1000

Input Files 1013

1000 Marc Users Guide: Part 2
CHAPTER 3.12
Chapter Overview
This chapter describes the use oI the Chaboche hardening Ieature in Marc. This option describes the plastic response
under cyclic loading. In this chapter, a blade oI a Ian Irom a turbine engine is simulated under thermal loading. A
coupled analysis is perIormed where diIIerent cyclic variation in temperature is prescribed at the tip and at the root oI
the blade. This time dependent temperature results in a nonsymmetric strain-controlled cyclic loading oI the blade.
Blade on a Fan of a Turbine Engine
This example describes a blade on a Ian oI a turbine engine. The blade is modeled as a wing shaped body mounted on
a surIace (Figure 3.12-1). In this model, the mounting is done with contact using the glue option. The interest is in
analyzing this blade under cyclic loading at short term high output stages on the turbine engine. The Chaboche model
is used to represent the plastic deIormation. The Iocus is on the plastic behavior and not on the complex hot air Ilow
around the blade. The blade is cooled internally and we assume that the part where the blade is mounted gets much
warmer than the tip oI the blade. This temperature diIIerence is applied as the loading.
Figure 3.12-1 Model of the Blade
Mesh Generation
The mesh generation starts Irom an existing model Iile containing the geometry. The curves in this geometry are
converted into the blade and the surIace on which it is mounted. This is perIormed by using a combination oI the
CONVERT curve to element, EXPAND SHELL, EXPAND ELEMENT, and the advancing Iront AUTOMESH commands.
The span oI the blade is 0.05 m and the chord length is about 0.07 m.
blade
surface
1001
CHAPTER 3.12
Chaboche Model
FILE
NEW
RESET PROGRAM
OPEN
blade_geom.mfd
OK
SAVE AS
blade.mud
RETURN
MESH GENERATION
ELEMENT CLASS
LINE(2)
RETURN
CONVERT
DIVISIONS
15 1
CURVES TO ELEMENTS
1 4 #
DIVISIONS
8 1
CURVES TO ELEMENTS
2 3 #
RETURN
EXPAND
SHELL/LINE ELEMENTS EXPAND
THICKNESS
0.003
LINE ELEMENTS
ALL EXIST
RETURN (twice)
SWEEP
ALL
RETURN
EXPAND
TRANSLATIONS
0 0 0.005
1002 Marc Users Guide: Part 2
CHAPTER 3.12
REPETITIONS
10
EXPAND ELEMENTS
ALL EXIST
RETURN
AUTOMESH
CURVE DIVISIONS
AVG LENGTH
0.004
APPLY CURVE DIVISIONS
5 6 7 8 9 10 #
RETURN
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT):QUAD MESH!
5 6 7 8 9 10 #
RETURN (twice)
EXPAND
TRANSLATIONS
0 0 -0.005
REPETITIONS
2
SELECT
METHOD
BOX
RETURN
ELEMENTS
-10 10
-10 10
-1E-6 1E-6
RETURN
EXPAND ELEMENTS
ALL SELECT
RETURN
SWEEP
REMOVE UNUSED:NODES
RETURN (twice)
1003
CHAPTER 3.12
Chaboche Model
Boundary Conditions
Temperature is prescribed at the tip and root oI the blade. At the tip, the temperature increases Irom 300K to 800K in
50 seco nds. and then decreases to 300K in 50 seconds. At the root, the temperature increases Irom 300K to 1300K in
50 seconds. and then decreases to 300K in 50 seconds. This temperature change is repeated Iive times. Displacement
boundary conditions are applied to remove rigid body modes, and the z-displacement is 0 at the root oI the blade.
BOUNDARY CONDITIONS
THERMAL
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 300
50 1300
100 300
150 1300
200 300
250 1300
300 300
350 1300
400 300
450 1300
500 300
FIT
NEW
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 300
50 800
100 300
150 800
200 300
1004 Marc Users Guide: Part 2
CHAPTER 3.12
250 800
300 300
350 800
400 300
450 800
500 300
FIT
RETURN
FIXED TEMPERATURE
TEMPERATURE(TOP)
TABLE
table1
OK
SELECT
CLEAR SELECT
NODES
-10 10
-10 10
-0.01-1e-6 -0.01+1e-6
RETURN
NODES ADD
ALL SELECT
NEW
FIXED TEMPERATURE
TEMPERATURE(TOP)
TABLE
table2
OK
SELECT
CLEAR SELECT
NODES
-10 10
-10 10
0.05-1e-6 0.05+1e-6
RETURN
NODES ADD
1005
CHAPTER 3.12
Chaboche Model
ALL SELECT
RETURN
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Y
DISPLACEMENT Z
OK
NODES ADD
2813 #
NEW
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Z
OK
NODES ADD
3113 #
NEW
FIXED DISPLACEMENT
DISPLACEMENT Z
OK
SELECT
CLEAR SELECT
NODES
-10 10
-10 10
-0.01-1e-6 -0.01+1e-6
RETURN
NODES ADD
ALL SELECT
SELECT
CLEAR SELECT
RETURN (thrice)
1006 Marc Users Guide: Part 2
CHAPTER 3.12
Initial Conditions
The initial temperature oI the blade is 300K.
Material Properties
The blade is made Irom steel, where the Young`s modulus is 210 GPa, the Poisson`s ratio 0.3, the Thermal Expansion
CoeIIicient is 1.8e-5K
-1
, the Conductivity is
80 W/m/K, and the Mass Density is 7800 kg/m
3
. The SpeciIic Heat is taken to be 0 to simulate Iast cooling due to the
internally open structure oI the blade. The initial Yield stress is 100 MPa, and the nonlinear kinematic hardening
coeIIicients C and are taken 100 GPa and 2000 respectively. Figure 3.12-2 shows the Mentat menu to add the
Chaboche properties.
INITIAL CONDITIONS
THERMAL
TEMPERATURE
TEMPERATURE(TOP)
300
OK
NODES ADD
ALL EXIST
RETURN (twice)
MATERIAL PROPERTIES
ISOTROPIC
YOUNGS MODULUS
2.1e11
POISSONS RATIO
0.3
MASS DENSITY
7800
THERMAL EXP.
COEFFICIENT
15e-6
OK
ELASTIC-PLASTIC
METHOD
CHABOCHE
1007
CHAPTER 3.12
Chaboche Model
Figure 3.12-2 Chaboche Properties Menu
INITIAL YIELD STRESS
1e7
COEFFICIENT C
1e11
COEFFICIENT GAMMA
1000
OK (twice)
HEAT TRANSFER
CONDUCTIVITY
80
MASS DENSITY
78
00
OK
ELEMENTS ADD
ALL EXIST
RETURN
1008 Marc Users Guide: Part 2
CHAPTER 3.12
Geometric Properties
The CONSTANT TEMPERATURE option is selected Ior all the elements. A constant temperature is then computed
throughout the element. This improves the stress calculation in the elements.
Contact
The blade and the underlying surIace are taken as separate contact bodies. The glue option is used Ior the interIace
where contact bodies touch each other. The contact heat transIer coeIIicient Ior this interIace is set to 1 MW/m
2
.
GEOMETRIC PROPERTIES
MECHANICAL ELEMENTS 3-D
SOLID
CONSTANT TEMPERATURE
OK
ELEMENTS ADD
ALL EXIST
RETURN (twice)
CONTACT
CONTACT BODIES
DEFORMABLE
OK
SELECT
CLEAR SELECT
ELEMENTS
-10 10
-10 10
-1e-6 10
RETURN
ELEMENTS ADD
ALL SELECT
NEW
DEFORMABLE
OK
SELECT
CLEAR SELECT
ELEMENTS
1009
CHAPTER 3.12
Chaboche Model
Loadcases and Job Parameters
A coupled thermal mechanical analysis are perIormed. The analysis is perIormed in 80 increments oI a constant time
step, where the total analysis time in 500 seconds. The multiIrontal sparse solver is used in this analysis.
-10 10
-10 10
-10 1e-6
RETURN
ELEMENTS ADD
ALL SELECT
RETURN
CONTACT TABLES
NEW
PROPERTIES
1 2
CONTACT TYPE: GLUE
THERMAL PROPERTIES
CONTACT HEAT TRANSFER COEFFICIENT
1e6
OK (twice
RETURN (twice)
LOADCASES
COUPLED
QUASI-STATIC
CONTACT
CONTACT TABLE
ctable1
OK
TOTAL LOADCASE TIME
500
PARAMETERS
# STEPS
80
OK (twice)
RETURN (twice)
1010 Marc Users Guide: Part 2
CHAPTER 3.12
The LARGE STRAIN ADDITIVE Iormulation is selected.
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened.
JOBS
ELEMENT TYPES
COUPLED
3-D SOLID
7
OK
ALL EXIST
RETURN (twice)
COUPLED
lcase1
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK (twice)
ANALYSIS OPTIONS
PLASTICITY PROCEDURE: LARGE STRAIN ADDITIVE
OK (twice)
JOB RESULTS
Stress
Total Strain
Elastic Strain
Plastic Strain
Total Equivalent Plastic Strain
OK
JOB PARAMETERS
SOLVER
MULTIFRONTAL SPARSE
OK
1011
CHAPTER 3.12
Chaboche Model
Figure 3.12-3 shows the equivalent stress in the blade at the maximum temperature diIIerence.
As mentioned in the Volume A: Theory and User Information manual, one oI the material phenomenon that can be
simulated by Chaboche hardening rule is mean-stress-relaxation. This happens when the material is subject to
nonsymmetric strain-controlled cyclic loading.
Figure 3.12-4 shows the time history oI component 33 oI the total strain at Node 308 (It is on the outside oI the blade
close to the surIace). There are Iive cycles and they are nonsymmetric with regard to the zero axis (Please notice that
the strain is a combination oI thermal and mechanical ones). ThereIore, the associated stress strain curve shows
mean-stress-relaxation phenomenon as shown in Figure 3.12-5. A large number oI cycles is necessary to reach the
stabilized one or this model since the plastic strain per cycle is relatively small.
FILES
SAVE AS
blade.mud
RETURN
RUN
SUBMIT(1)
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
SET NODES
308 #
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Comp 33 of Total Strain
Comp 33 of Stress
FIT
RETURN
SHOW IDS
0
1012 Marc Users Guide: Part 2
CHAPTER 3.12
Figure 3.12-3 Equivalent Stress at the First Occurrence of the Maximum Temperature Difference
Figure 3.12-4 Time History Plot of Component 33 of the Total Strain
1013
CHAPTER 3.12
Chaboche Model
Figure 3.12-5 History Plot of Component 33 of the Total Strain versus Component 33 of the Stress
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
blade.proc Mentat procedure Iile to run the above example
blade_geom.mfd Mentat model Iile read by above procedure Iile
1014 Marc Users Guide: Part 2
CHAPTER 3.12
Chapter 3.13: Modeling of a Shape Memory Alloy Orthodontic Archwire
3.13
Modeling of a Shape Memory
Alloy Orthodontic Archwire

Chapter Overview 1016

Simulation of an Archwire with Shape Memory Alloy Models

1016

Input Files 1033

References 1034
1016 Marc Users Guide: Part 2
CHAPTER 3.13
Chapter Overview
Shape-memory properties Ior nickel (Ni) titanium (Ti) alloy were discovered in the 1960s, at the Naval Ordnance
Laboratory (NOL); hence, the acronym NiTi-NOL or Nitinol, which is commonly used when reIerred to Ni-Ti based
shape-memory alloys. Since 1970, Ni-Ti has been widely investigated due to its Irequent use in applications and today
it is probably the shape-memory materials most Irequently used in commercial applications. The amount oI thermally
activated recoverable memory strain and the size oI the hysteresis loop strongly depend on alloy composition, thermo-
mechanical processing, testing direction and deIormation mode (that is, iI the material is in simple tension, simple
compression or shear). For the Iull austenitemartensite phase transIormation, the recoverable memory strain is oI
the order oI 8, while the hysteresis width is typically oI 30-50C. For uniaxial states oI stress and in the usual range
oI applications, the stress-temperature regions in which the phase transIormation may occur are delimited with good
approximation by straight lines with slopes ranging Irom 2.5 Mpa/C to over 15 Mpa/C.
Experimental evidence shows that:
1. Phase transIormations do not exhibit pressure dependence in the case oI long-aged Ni-Ti; Ior short-aged Ni-Ti
the R-phase transition is unaIIected by pressure, while the martensitic transIormation is pressure dependent
2. Phase transIormation are insensitive to temperature rates and to stress rates.
The SMA underlying micro-mechanics is quite complex. Moreover, due to the increasing sophistication oI SMA-
based devices, there is a growing need Ior eIIective computational tools able to support the design process. Two shape
memory models based on the Auricchio |Reference 1| mechanical model and the SaeedvaIa and Asaro |Reference 2|
thermo-mechanical model have been implemented in Marc and are reviewed with an Archwire example.
Simulation of an Archwire with Shape Memory Alloy Models
In order to show how to use two shape memory alloy models available in Marc, we consider the simulation oI an
orthodontic archwire, (Figure 3.13-1). The dimensions are taken Irom Auricchio (Int. J. Plasticity, 2001) and they are
reported in Table 2.13-1 Ior the entire segment indicated in Figure 3.13-1. Moreover, we assume that the archwire is
made out oI a wire with rectangular 0.635 x 0.432 mm cross section.
Figure 3.13-1 Orthodontic Archwire: Geometry Data
A B
Y
C
D
X
19 mm
1 2
3
4
5
6
7
8
9
10
11
12 13
1017
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
.
The right halI oI geometry was modeled considering symmetry. The generation oI mesh using Mentat is
straightIorward. So, it is not discussed here. The generated mesh was stored in sma_mesh.mud.
Boundary Conditions
The boundary conditions are set to reproduce displacement control during loading and unloading. Fixed boundary
condition is applied to the symmetric nodes. Another set oI boundary condition is applied to the right edge nodes to
impose the displacement control in x-direction by inserting table1. The movement in the z-direction is also
constrained Ior the nodes.
Table 2.13-1 Orthodontic Archwire: Details on the Geometry
Segment
Number Segment Type
Length
(mm) Angle (
o
) Radius (mm)
1 Straight 7.5
2 Circular 90 1
3 Straight 2.0
4 Circular 90 1
5 Straight 2.5
6 Circular 180 1
7 Straight 9.0
8 Circular 180 1
9 Straight 2.5
10 Circular 90 1
11 Straight 2.0
12 Circular 90 1
13 Straight 7.5
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
DISPLACEMENT Y
0
1018 Marc Users Guide: Part 2
CHAPTER 3.13
DISPLACEMENT Z
0
ROTATION X
0
ROTATION Y
0
ROTATION Z
0
OK
NODES ADD
11 12 13 14 251 252 253 254
END LIST
NEW
TABLES
NEW
NEW TABLE
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 0
1 1
2 0
10 0
SHOW MODEL
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
5
TABLE
table1
DISPLACEMENT Z
0
ROTATION X
0
1019
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
Figure 3.13-2 Generated Table for Displacement Control
ROTATION Y
0
OK
OK
NODES ADD
185 212 213 214 425 452 453 454
END LIST
1020 Marc Users Guide: Part 2
CHAPTER 3.13
Figure 3.13-3 Boundary Condition ID
Initial Conditions
The option is only active Ior thermo-mechanical shape memory alloy. The temperature was initialized to 19C (room
temperature) by means oI a STATE VARIABLE initial condition.
Material Properties
Material property data Ior mechanical model and thermo-mechanical model are completely diIIerent. Here is the
summary Ior both data.
INITIIAL CONDITIONS
NEW
MECHANICAL
STATE VARIABLE
STATE VARIABLE
19
OK
ELEMENTS ADD
ALL EXIST
1021
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
1. Mechanical Shape Memory Model
Other parameters used:
, , .
In the above summary, superscript ' and '- mean tensile and compression properties, respectively. Also, subscript
's and 'I mean starting and Iinishing points, respectively. In addition, superscript 'AS means austenite-to-
martensite transIormation and 'SA means martensite-to-austenite transIormation. Then, the meaning oI the symbols
are summarized as Iollows:
Noting that, given and , the parameter , which is measured Irom the diIIerence between the response in
tension and compression, can be obtained as Iollows:
When the compression test data is not available, is usually set to be zero. It means that tensile and compressive
responses are the same. is a scalar parameter representing the maximum deIormation obtainable only by
detwinning oI the multiple-variant martensite. Classical value Ior is in the range 0.0 to 0.10. In this example, it was
set to 0.007.
Values with stress dimension
Mpa
5x10
4
500 500 300 300 700
:
Starting tensile stress in austenite-to-martensite transIormation
:
Finishing tensile stress in austenite-to-martensite transIormation
:
Starting tensile stress in martensite-to-austenite transIormation
:
Finishing tensile stress in martensite-to-austenite transIormation
:
Starting compressive stress in austenite-to-martensite transIormation
MATERIAL PROPERTIES
NEW
MORE
SHAPE MEMORY ALLOY
MECHANICAL (AURICCHIOS) MODEL
E
o
s
AS
o
f
AS
o
s
SA
o
s
SA
o
s
AS-
v 0.3 c
L
0.007 o 0.12
o
s
AS
o
f
AS
o
s
SA
o
f
SA
o
s
AS-
o
s
AS
o
s
AS-
o
o
2
3
-- -
o
s
AS-
o
s
AS

o
s
AS-
o
s
AS

------------------------------- 0.12
o
c
L
c
L
1022 Marc Users Guide: Part 2
CHAPTER 3.13
Figure 3.13-4 Material Properties Menu in Mechanical Shape Memory Model
YOUNGS MODULUS
50000
POISSONS RATIO
0.3
sigAS_s
500
sigAS_f
500
sigSA_s
300
sigSA_f
300
alpha (0.0 - 1.0)
0.12
espL (0.0 - 0.10)
0.007
OK (twice)
ELEMENTS ADD
ALL EXIST
1023
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
2. Thermo-Mechanical Shape Memory Model
Austenite properties
Young`s modulus (E): 50000 Mpa, Poisson`s ratio (v): 0.33, Thermal expansion coeIIicient (o) 1.0e-5, Equivalent
tensile yield stress: 10000 Mpa
Martensite properties
Young`s modulus (E): 50000 Mpa, Poisson`s ratio (v): 0.33, Thermal expansion coeIIicient (o) 1.0e-5, Equivalent
tensile yield stress: 10000 Mpa
Austenite to Martensite
Martensite start temperature in stress-Iree condition ( ): -45C, Martensite Iinish temperature in stress-Iree
condition : -90C, Slope oI the stress-dependence oI martensite start-Iinish temperatures ( ): 6.6666
Martensite to Austenite
Austenite start temperature in stress-Iree condition ( ): 5C, Austenite Iinish temperature in stress-Iree condition
( ): 20C, Slope oI the stress-dependence oI austenite start-Iinish temperatures ( ): 8.6667
Transformation strains
Deviatoric part oI transIormation strain ( ): 0.0
Volumetric part oI the transIormation strain ( ): 0.0
Twinning stress ( ): 100 Mpa
Coefficients of g function
, , , , ,
, ,
So, the chosen 'g Iunction is .
MATERIAL PROPERTIES
NEW
MORE
SHAPE MEMORY ALLOYS
THERMO-MECHANICAL MODEL
AUSTENITE PROPERTIES
M
s
0
M
f
0
C
m
A
s
0
A
f
0
C
a
c
eq
T
c
v
T
o
ef f
g
g
o
eq
g
o
--------
\ .
| |
1 g
a
o
eq
g
o
--------
\ .
| |
g
b
g
c
o
eq
g
o
--------
\ .
| |
g
d
g
e
o
eq
g
o
--------
\ .
| |
g
f
exp
g
a
4 g
b
2 g
c
0.0 g
d
2.75 g
e
0.0 g
f
3.0
g
o
1000.0 g
max
1.0 g
max
g
1.0e20
g
o
eq
1000
----------- -
\ .
| |
1 4
o
eq
1000
------------
\ .
| |
2
exp
1024 Marc Users Guide: Part 2
CHAPTER 3.13
YOUNGS MODULUS
50000
POISSONS RATIO
0.3
MASS DENSITY
1
THERMAL EXP. COEF.
1e-5
INITIAL YIELD STRESS
10000
OK
MARTENSITE PROPERTIES
YOUNGS MODULUS
50000
POISSONS RATIO
0.3
MASS DENSITY
1
THERMAL EXP. COEF.
1e-5
INITIAL YIELD STRESS
10000
OK
AUSTENITE TO MARTENSITE
MARTENSITE START TEMPERATURE
-45
MARTENSITE FINISH TEMPERATURE
-90
SLOPE
6.6667
OK
MARTENSITE TO AUSTENITE
AUSTENITE START TEMPERATURE
5
AUSTENITE FINISH TEMPERATURE
20
1025
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
SLOPE
8.6667
OK
TRANSFORMATION STRAINS
DEVIATORIC TRANS. STRAIN
0
VOLUMETRIC TRANS. STRAIN
0
TWINNING STRESS
100
g-A
-4
g-B
2
g-C
0
g-D
2.75
g-E
0
g-F
3
g-0
1000
g-max
1
STRESS AT g-max
1e+020
OK (thrice)
ELEMENTS ADD
ALL EXIST
1026 Marc Users Guide: Part 2
CHAPTER 3.13
Figure 3.13-5 Material Properties in Thermo-mechanical Shape Memory Model
Load Steps and Job Parameters
The job consists oI two mechanical loadcases. The loading histories is given as Iollows:
Total 200 Iixed steps are used Ior the entire analysis with residual norm oI 0.1. Each loadcase consists oI 100 steps.
Time (s) Displ (mm)
0 0.0
1 5.0
2 0.0
LOADCASES
NEW
MECHANICAL
STATIC
SOLUTION CONTROL
MAX # RECYCLES
20
OK
STEPPING PROCEDURE
CONSTANT TIME STEP
# STEP
100
OK
NEW
STATIC
1027
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
The analysis is a normal mechanical analysis in which all two loadcases are perIormed in sequence. INITIAL LOADS
is only active Ior thermo-mechanical shape memory alloy in order to set initial temperature. Also, when mechanical
shape memory alloy is used, LARGE STRAIN MULTIPLICATIVE option is activated in the PLASTICITY PROCEDURE.
II user choose other options, Marc changes the option internally to LARGE STRAIN MULTICATIVE option. New scalar
quantity Ior Volume Fraction oI Martensite was selected in this example, as well as Equivalent von Mises Stress.
1. Mechanical Shape Memory Alloy
SOLUTION CONTROL
MAX # RECYCLES
20
OK
STEPPING PROCEDURE
CONSTANT TIME STEP
# STEP
100
OK
JOBS
NEW
MECHANICAL
LOADCASES
activate:
lcase1
lcase2
ANALYSIS OPTIONS
LARGE DISPLACEMENT
PLASTICITY PROCEDURE
LARGE STRAIN MULTIPLICATIVE
OK
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress
Volume Fraction of Martensite
ELEMENT RESULTS
CENTROID
OK (twice)
1028 Marc Users Guide: Part 2
CHAPTER 3.13
2. Thermo-mechanical Shape Memory Alloy
For the analysis oI the Archwire model, element type 7 is being used. Mechanical shape memory model only supports
3-D, plane strain, and axisymmetric continuum elements.
JOBS
NEW
MECHANICAL
LOADCASES
activate:
lcase1
lcase2
INITIAL LOADS
INITIAL CONDITIONS
activate:
icond1 state_variable
OK
ANALYSIS OPTIONS
LARGE DISPLACEMENT
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress
Volume Fraction of Martensite
ELEMENT RESULTS
CENTROID
OK (twice)
ELEMENT TYPES
MECHANICAL
3-D SOLID
7
OK
ALL:EXIST
RETURN
1029
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
Save Model, Run Job, and View Results
1. Mechanical Shape Memory Alloy
2. Thermo-mechanical Shape Memory Alloy
FILE
SAVE AS
sma_m.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN
sma_m.t16
OK
DEF & ORIG
CONTOUR BAND
SCALAR
Volume Fraction of Martensite
OK
MONITOR
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
FILE
SAVE AS
sma_tm1.mud
OK
RETURN
RUN
SUBMIT 1
1030 Marc Users Guide: Part 2
CHAPTER 3.13
For two shape memory model, the contours Ior Volume Fraction oI Martensite were compared in Figure 3.13-6 and
Figure 3.13-7 and the contours Ior Equivalent Von Mises Stress were compared in Figure 3.13-8 and Figure 3.13-9 at
the step oI 100 and 200, respectively. For thermo-mechanical shape memory alloy, initial temperature were set to 19C
in INITIAL CONDITIONS. As shown in the Iigures, in both models, two scalar properties reach to maximum at the
maximum displacement and come back close to zero at the last step. Both models predicts superelasticity behavior
(shape memory eIIect) well.
Figure 3.13-6 Volume Fraction of Martensite at the Maximum Displacement:
(a) Mechanical Model (b) Thermo-mechanical Model
MONITOR
OK
MAIN
RESULTS
OPEN
sma_tm1.t16
OK
DEF & ORIG
CONTOUR BAND
SCALAR
Volume Fraction of Martensite
OK
MONITOR
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
(a) (b)
1031
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
Figure 3.13-7 Volume Fraction of Martensite at the Last Step:
(a) Mechanical Model (b) Thermo-mechanical Model
Figure 3.13-8 Equivalent Von Mises Stress at the Maximum Displacement:
(a) Mechanical Model (b) Thermo-mechanical Model
(a) (b)
(a)
(b)
1032 Marc Users Guide: Part 2
CHAPTER 3.13
Figure 3.13-9 Equivalent Von Mises Stress at the Last Step:
(a) Mechanical Model (b) Thermo-mechanical Model
In order to show the diIIerent behavior according to initial temperature Ior thermo-mechanical shape memory model,
additional simulation was perIormed with initial temperature oI 5C.
Save Model, Run Job, and View Results
INITIIAL CONDITIONS
NEW
MECHANICAL
STATE VARIABLE
STATE VARIABLE
5
OK
ELEMENTS ADD
ALL:EXIST
FILE
SAVE AS
sma_tm2.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
(a) (b)
1033
CHAPTER 3.13
Modeling of a Shape Memory Alloy Orthodontic Archwire
As shown in Figure 3.13-10, Volume Faction oI Martensite is not decreased Ior the simulation oI thermo-mechanical
model under the temperature oI 5C even at the last step.
Figure 3.13-10 Volume Fraction of Martensite for Thermo-mechanical Shape Memory Alloy
under the Temperature of 5
o
C: (a) step =100 (b) step = 200
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
MAIN
RESULTS
OPEN
sma_tm2.t16
OK
DEF & ORIG
CONTOUR BAND
SCALAR
Volume Fraction of Martensite
OK
MONITOR
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
(a) (b)
1034 Marc Users Guide: Part 2
CHAPTER 3.13
References
1. Auricchio, F., A robust integration-algorithm Ior a Iinite-strain shape-memory-alloy superelastic model, Int. J.
Plasticity, 17, 971-990 (2001).
2. SaeedvaIa, M. and Asaro, R.J., Los Alamaos Report, LA-UR-95-482 (1995).
File Description
sma_mech.proc Mentat procedure Iile to run the above example
sma_mesh.mud Mentat model Iile read by above procedure Iile
Chapter 3.14: Implicit Creep Analysis of a Solder Connection between a Microprocessor and PCB
3.14
Implicit Creep Analysis of
Solder Connection between
Microprocessor and PCB

Chapter Overview 1036

Microprocessor Soldered to a PCB 1036

Input Files 1054

References 1054
1036 Marc Users Guide: Part 2
CHAPTER 3.14
Chapter Overview
This chapter describes the use oI the implicit creep Ieature in Marc. The available options to simulate power law creep
in conjunction with von Mises plasticity are described in detail. The example chosen Ior this purpose is a
Microprocessor-solder-PCB assembly that is subjected to both electrical and thermal loads.
Microprocessor Soldered to a PCB
This example describes a ceramic ball grid array (CBGA), a ceramic substrate package. This is one oI the ways ICs
are packaged in the electronic industry. In a CBGA, the die is glued to a ceramic. This ceramic is soldered with small
solder balls to the PCB (printed circuit board), where the solder balls represent the electric contacts. The stress Iree
temperature is 443 K (170
o
C), so the chip needs to be cooled down to room temperature beIore it can be used.
Figure 3.14-1 shows the CBGA, where the diIIerent components can be seen.
.
Figure 3.14-1 Model of the CBGA Showing the Different Contact Bodies
Mesh Generation
The mesh is generated by Iirst making the solder balls, these are represented as cuboids. Two rows oI three balls are
generated. Then the ceramic is generated, the die, and Iinally the PCB. Note that all parts are generated away Irom
each other. This is done to prevent components Irom being joined when the model is swept to remove double nodes.
AIter sweeping, the parts are brought into contact.
MESH GENERATION
NODES ADD
0.002 0.001 0
0.004 0.001 0
0.004 0.003 0
die
ceramic
solder balls
PCB
1037
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
0.002 0.003 0
ELEMENTS ADD
1 2 3 4
DUPLICATE
TRANSLATIONS
0.007 0 0
REPETITIONS
2
ELEMENTS
ALL EXIST
TRANSLATIONS
0 0.006 0
REPETITIONS
1
ELEMENTS
ALL EXIST
RETURN
EXPAND
TRANSLATION
0 0 0.002
REPETITIONS
1
ELEMENTS
ALL EXIST
RETURN
ADD NODES
0.00 0.00 0.003
0.02 0.00 0.003
0.02 0.01 0.003
0.00 0.01 0.003
ADD ELEMENTS
97 98 99 100
EXPAND
TRANSLATIONS
0 0 0.003
1038 Marc Users Guide: Part 2
CHAPTER 3.14
REPETITIONS
1
ELEMENTS
13 #
RETURN
ADD NODES
0.005 0.003 0.007
0.015 0.003 0.007
0.015 0.007 0.007
0.005 0.007 0.007
ADD ELEMENTS
113 114 115 116
EXPAND
TRANSLATIONS
0 0 0.0012
REPETITIONS
1
ELEMENTS
15 #
RETURN
ADD NODES
-0.005 -0.005 -0.01
0.025 -0.005 -0.01
0.025 0.015 -0.01
-0.005 0.015 -0.01
ADD ELEMENTS
129 130 131 132
EXPAND
TRANSLATIONS
0 0 0.008
REPETITIONS
1
ELEMENTS
17 #
RETURN
1039
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
Other parts oI the model are stored in sets in the same way. Then the elements are subdivided.
SELECT
ELEMENTS
18 #
STORE
pcb
ALL SELECT
CLEAR SELECT
RETURN
SUBDIVIDE
DIVISIONS
16 9 4
ELEMENTS
18 #
DIVISIONS
2 2 5
ELEMENTS
7 8 9 10 11 12 #
DIVISIONS
15 7 4
ELEMENTS
14 #
DIVISIONS
6 3 3
ELEMENTS
16 #
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
1040 Marc Users Guide: Part 2
CHAPTER 3.14
The other sets are moved towards each other in a similar way.
Boundary Conditions
Displacement boundary conditions are applied to prevent the rigid body modes, and the potential is set to 0 V Ior a
node on each solder ball, the ceramic, and the PCB. A temperature drop to 298 K (25
o
C) is prescribed at the bottom
oI the PCB, and a potential diIIerence oI 10 V is applied across the die.
MOVE
SELECT
SELECT SET
pcb
OK
RETURN
TRANSLATIONS
0 0 0.002
ELEMENTS
ALL SELECT
SELECT
CLEAR SELECT
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
NAME
fix
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Y
DISPLACEMENT Z
OK
NODES ADD
67 #
NEW
NAME
fix_xz
1041
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Z
OK
NODES ADD
66 #
NEW
NAME
fix_z
FIXED DISPLACEMENT
DISPLACEMENT Z
OK
NODES ADD
68 #
RETURN
JOULE
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
NAME
temp
ADD
0 443
100 298
3E4 298
FIT
RETURN
NEW
NAME
temp
FIXED TEMPERATURE
TEMPERATURE
1
1042 Marc Users Guide: Part 2
CHAPTER 3.14
TABLE
temp
OK
SELECT
METHOD
BOX
RETURN
NODES
-10 10
-10 10
-8e-3-1e-6 -8e-3+1e-6
RETURN
NODES ADD
ALL SELECT
NEW
NAME
pot_0
FIXED VOLTAGE
VOLTAGE
OK
SELECT
CLEAR SELECT
NODES
0.015-1e-6 0.015+1e-6
-10 10
-10 10
RETURN
NODES ADD
ALL SELECT
NODES ADD
66 8 20 7 19 2 14 50 #
SELECT
METHOD
SINGLE
RETURN
CLEAR SELECT
1043
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
Initial Conditions
The initial temperature Ior the whole model is set to 443 K (170
o
C).
Material Properties
The material properties used are listed in the Iollowing two tables. The ceramic, the die, and the PCB are taken to be
elastic. Solder A consists oI Sn63/Pb37, which are the layers oI the solder balls touching the ceramic and the PCB,
and solder B consists oI Sn10/Pb90, which is the middle part oI the solder balls.

RETURN
NEW
NAME
pot_10
FIXED VOLTAGE
VOLTAGE
10
OK
NODES ADD
1831 1827 1828 1832 #
RETURN (twice)
INITIAL CONDITIONS
THERMAL
TEMPERATURE
TEMPERATURE (TOP)
443
OK
NODES ADD
ALL EXIST
RETURN (twice)
Solder A Solder B Ceramic
Young`s modulus (GPa) 30.2 30.2 300
Poisson`s ratio 0.4 0.4 0.23
Thermal Expansion CoeIIicient (K
-1
) 24 x 10
-6
27.8 x 10
-6
6.7 x 10
-6
Conductivity (W/m/K) 50.6 35.5 16.5
1044 Marc Users Guide: Part 2
CHAPTER 3.14
Creep and plasticity properties are used Ior both solder A and solder B. A perIectly-plastic behavior with no strain
hardening and no temperature dependence has been assumed. Yield stress oI 49.2 GPa is speciIied |Reference 1
and Reference 2|. The creep properties that are available Ior 60Sn-40Pb in Reference 1 are used herein Ior both
solder A and solder B. The creep behavior used herein is an approximation oI GaroIalo`s hyperbolic sine law used Ior
relating steady-state creep rate to stress and temperature. The sine law taken Irom Reference 1 and Reference 3 is
shown below:
where C
1
16.7x10
-6
(K/sec)/(N/m
2
); T temperature in Kelvin; G temperature dependent shear modulus (28388
- 56 T)10
6
(N/m
2
); o 866; n 3.3; Q - activation energy Ior creep deIormation process .548 eV; and k 8.617 x
10
-5
(Boltzmann`s constant).
It should be noted that the deIault implicit creep capability in Marc allows the creep strain rate to be expressed in terms
oI power law expressions oI stress, temperature, creep strain, and time.
Figure 3.14-2 shows the Mentat menu to add the implicit creep properties.
Resistivity (Om)
1 x 10
6
1 x 10
6
1 x 10
6
SpeciIic Heat (J/kg/K) 200 130 1050
Mass Density (kg/m
3
)
0
9000
0
11000
0
2000
Die Die part PCB
Young`s modulus (GPa) 162 162 18.2
Poisson`s ratio 0.28 0.28 0.25
Thermal Expansion CoeIIicient (K
-1
) 23 x 10
-6
23 x 10
-6
15 x 10
-6
Conductivity (W/m/K) 120 120 5
Resistivity (Om)
0 0
600
0
0.25 1 x 10
6
SpeciIic Heat (J/kg/K) 700 700 820
Mass Density (kg/m
3
)
0
2330
0
2330
0
2000
Solder A Solder B Ceramic
c

c
C
1
G
T
----
\ .
| |
o
o
G
----
\ .
| |
sinh
n
Q
kT
------
\ .
| |
exp
c
c
Ao
m
c
c
( )
n
- T
p
- qt
q 1
( ) -
1045
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
Figure 3.14-2 Creep Properties Menu
Use can be made oI the UCRPLW user subroutine to speciIy more complex relationships Ior the creep strain rate.
.
In the current example, the temperature dependence in the GaroIalo law is treated exactly while the hyperbolic
sine Iunction Ior stress is reduced to a power Iunction using a one-term Taylor series expansion. UCRPLW.F is
written below:
SUBROUTINE UCRPLW(CPA,CFT,CFE,CFTI,CFSTRE,CPTIM,TIMINC,
* EQCP,DT,DTDL,MDUM,NN,KC,MAT)
C* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
C user routine to define implicit creep law
C input:
C cptim time at beginning of increment
C timinc time increment
C eqcp creep strain at beginning of increment
C dt temperature at beginning of increment
C dtdl incremental temperature
C mdum(1) user element number
C mdum(2) elsto element number
C nn integration point number
C kc layer number
C mat material number
C output:
C cpa creep constant
c
c
A o
m
g c
c
( ) h T ( )
dk t ( )
dt
------------- - - - -
1046 Marc Users Guide: Part 2
CHAPTER 3.14
C cft temperature factor
C cfe creep strain factor
C cfti time factor
C cfstre stress exponent
C where:
C creep strain rate = cpa*cft*cfe*cfti*(stress**cfstre)
C* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
IMPLICIT REAL*8 (A-H,O-Z)
DIMENSION MDUM(*)
DTEND=DT+DTDL
CFTI=1.D0
CFE=1.D0
CFSTRE=3.3D0
C1=16.7D-6
ALP=866.D0
G=(28388.D6-56.D6*DTEND)
CPA=C1*(ALP/G)**CFSTRE
Q=0.548D0
AK=8.617D-5
CFT=DEXP(-Q/AK/DTEND)*G/DTEND
RETURN
END
MATERIAL PROPERTIES
NAME
solder_A
ISOTROPIC
YOUNGS MODULUS
3.02e10
POISSONS RATIO
0.4
MASS DENSIY
9000
THERMAL EXP
THERMAL EXP COEF
2.4e-5
OK
CREEP
YIELD STRESS
4.92e7
USER SUB. UCRPLW
OK (twice)
1047
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
The material properties Ior the other components are added in a similar way where, Ior the elastic components, the
creep section is omitted.
Contact
The solder balls, the ceramic, the die, and the PCB are taken as separate contact bodies. The glue option is used Ior
each interIace where contact bodies touch each other, and the contact heat transIer coeIIicient is set to 100 W/m
2
Ior
these interIaces.
JOULE HEATING
CONDUCTIVITY
50.6
RESISTIVITY
1e6
SPECIFIC HEAT
200
MASS DENSITY
9000
OK
ELEMENTS
577 582 587 592 597 602 607 612 617 622 627 632 637 642 647
652 657 662 667 672 677 682 687 692 581 586 591 596 601 606
611 616 621 626 631 636 641 646 651 656 661 666 671 676 681
686 691 696 #
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE
OK
ELEMENTS
577 to 596 #
NEW
DEFORMABLE
OK
ELEMENTS
597 to 616 #
1048 Marc Users Guide: Part 2
CHAPTER 3.14
The other solder balls are added as contact bodies in a similar way.
The other components are selected as contact bodies is a similar way.
In this contact table, the same properties are also set Ior the Iollowing combinations: 3 7, 4 7, 5 7, 6 7, 1 9, 2 9,
3 9, 4 9, 5 9, 6 9, 7 8.
NEW
DEFORMABLE
OK
SELECT
CLEAR SELECT
SELECT SET
ceramic
OK
RETURN
ELEMENTS
ALL SELECT
RETURN
CONTACT TABLES
NEW
PROPERTIES
1 7
CONTACT TYPE: GLUE
THERMAL PROPERTIES
CONTACT HEAT TRANSFER COEFFICIENT
100
2 7
CONTACT TYPE: GLUE
CONTACT HEAT TRANSFER COEFFICIENT
100
OK (twice)
RETURN (twice)
1049
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
Loadcases and Job Parameters
A coupled Joule-mechanical creep analysis will be perIormed. The loading is divided in three stages. In the Iirst
loadcase, the temperature is decreased Irom 170C to 25C at the bottom oI the PCB over a period oI 100 seconds.
Only plasticity is allowed Ior the solder balls during this period. In the second loadcase, the temperature is maintained
at 25C and the solder balls are allowed to creep over a period oI 10000 seconds. In the third loadcase, an electric
potential oI 10 V is applied across the die Ior 10000 seconds. The generated heat due to the induced electric currents
causes a temperature increase in the assembly. The Iixed time-stepping scheme TRANSIENT NON AUTO is used Ior
loadcase 1. An adaptive time-stepping scheme based on MULTI-CRITERIA is used Ior loadcases 2 and 3. It should be
noted that the thermal loading in loadcase 1 is linearly ramped over 10 increments, whereas, the electrical loading in
loadcases 3 is instantaneously applied.
Fixed Stepping (TRANSIENT NON AUTO in Volume C: Program Input) uses the time step speciIied by the user. Any
speciIied tolerance Ior allowable temperature change is ignored. The thermal solution is recycled till the tolerance Ior
temperature error in estimate (iI nonzero) is satisIied.
MULTI-CRITERIA (AUTO STEP in Volume C: Program Input) controls the time step based on the convergence
characteristics oI the thermal and mechanical passes oI the loadcases. For the thermal pass, the time step control is
based on the actual temperature change compared to a user-speciIied tolerance on temperature change (deIault is 20).
II the temperature change in any increment exceeds the allowed value, the time step is reduced, and the electrical and
thermal passes are repeated with a smaller time step. For the mechanical pass, by deIault, the time step control is based
on the number oI recycles used to reach convergence compared to a desired number oI recycles (deIault is 3). II the
number oI recycles in the mechanical pass exceeds the number oI desired recycles, the time step oI the increment is
cut back, and the electrical, thermal and mechanical passes are repeated. In addition to these numerical criteria, iI
deemed necessary, one can choose to add user-speciIied or automatic physical criteria to control the time stepping. In
the latter case, the algorithm also keeps track oI the changes in the speciIied physical quantity and cuts back as soon
as the change exceeds allowed tolerances. In the current example, an allowed temperature change oI 100 K is
speciIied. Also, the initial time step Ior loadcase 2 is speciIied as 0.001 oI 10000 10 seconds. This matches the time
step used in the Iirst loadcase. At transition stages where loading changes, it is advisable to use smaller time steps in
order to capture the creep more accurately. Alternately, physical criteria based on creep strain changes can be used.
These are not used in the current example.
LOADCASES
JOULE-MECHANICAL
TRANSIENT
LOADS
pot_10
OK
CONTACT
CONTACT TABLE
ctable1
1050 Marc Users Guide: Part 2
CHAPTER 3.14
CONVERGENCE TESTING
MAX. ERROR IN TEMPERATURE ESTIMATE
5
OK
TOTAL LOADCASE TIME
100
CONSTANT TIME STEP
PARAMETERS
# STEPS
10
OK (twice)
CREEP
LOADS
pot_10
OK
CONTACT
CONTACT TABLE
ctable1
CONVERGENCE TESTING
MAX. TEMPERATURE CHANGE ALLOWED
100
OK
TOTAL LOADCASE TIME
10000
MULTI-CRITERIA
PARAMETERS
INITIAL FRACTION OF LOADCASE TIME
0.001
OK
COPY
CREEP
LOADS
temp
pot_10
OK (twice)
RETURN (twice)
1051
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
The implicit creep analysis option needs to be set Irom the JOBS menu. Also, three choices are provided as to what
kind oI tangent matrix is to be Iormed. The Iirst is using an elastic tangent, which requires more iterations, but can be
computationally eIIicient because re-assembly might not be required. The second is a secant (approximate) tangent
that gives the best behavior Ior general viscoplastic models. The third is an algorithmic tangent that provides the best
behavior Ior small strain power law creep. When implicit creep is speciIied in conjunction with plasticity, the elastic
tangent option is not available.
JOBS
ELEMENT
TYPES
JOULE-MECHANICAL
3-D SOLID
7
OK
ALL EXIST
RETURN (twice)
JOULE-MECHANICAL
lcase1
lcase2
lcase3
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK (twice)
ANALYSIS OPTIONS
CREEP TYPE & PROCEDURE: IMPLICIT MAXWELL
CREEP TYPE & PROCEDURE: SECANT TANGENT
JOB RESULTS
Stress
Plastic Strain
Creep Strain
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
Total Equivalent Creep strain
1st Comp of Heat Flux
2nd Comp of Heat Flux
3rd Comp of Heat Flux
1052 Marc Users Guide: Part 2
CHAPTER 3.14
Save Model, Run Job, and View Results
AIter saving the model, and selecting the user subroutine ucrplw.f, the job is submitted.
Figure 3.14-3 shows the plastic strain in the solder balls. The plastic deIormation occurs in the Iirst Iew increments oI
the analysis, when the temperature change is the highest. Figure 3.14-4 shows the equivalent creep strain as a Iunction
oI time Ior a node on two solder balls. The (blue) curve is Irom a node Irom a solder ball at the outside oI the
grid and the (red) curve is Irom a node Irom a solder ball at the center oI the grid. Figure 3.14-5 shows the
temperature as a Iunction oI time Ior a node at the top oI the die and a node at the bottom oI the PCB.
Electric Current
Generated Heat
Displacement
Temperature
Electric Potential
External Heat flux
External Electric Current
Reaction Force
Reaction heat Flux
Reaction Electric Current
Contact Normal Stress
Contact Normal Force
Contact Friction Stress
Contact Friction Force
Contact Status
Contact Touched Body
OK (twice)
FILE
SAVE AS
die.mud
OK
RETURN
RUN
USER SUBROUTINE FILE
ucrplw.f
SUBMIT(1)
1053
CHAPTER 3.14
Implicit Creep Analysis of Solder Connection between Microprocessor and PCB
Figure 3.14-3 Equivalent Plastic Strain in the Solder Balls
Figure 3.14-4 Equivalent Creep Strain as a Function of Time for a Node on two Solder Balls
1054 Marc Users Guide: Part 2
CHAPTER 3.14
Figure 3.14-5 Temperature as a Function of Time for two Nodes, one at the Top of the Die and one at the
Bottom of the PCB
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
References
1. H.L.J. Pang, C.W. Seetoh, Elasto-Plastic Creep Analysis of Ceramic BGA Solder Joints Subjected to
Temperature Cycling Loading, Proceedings, The 17th MARC Users` Meeting, 1997, pp.183 - 194
2. H.U. Akay, Y. Tong, N. Paydar, Thermal fatigue analysis of a SMT solder joint using non-linear FEM
approach, The Int. Journal oI Microcircuits and Electronics Packaging, Vol. 16, No. 2, pp. 79-88, 1993.
3. R. Darveaux and K. Banerji, Constitutive relations for tin-based solder joints, IEEE Transactions on
Components, Hybrids, and ManuIacturing Technology, Vol. 15, No. 6, pp. 1013-1024, 1992
File Description
die.proc Mentat procedure Iile to run the above example
ucrplw.f User subroutine
Chapter 3.15: Continuum Composite Elements
3.15
Continuum Composite
Elements

Chapter Overview 1056

Background Information 1056

Analysis 1057

Input Files 1065

1056 Marc Users Guide: Part 2
CHAPTER 3.15
Chapter Overview
This chapter demonstrates the use oI the continuum elements. Compared to the composite shell elements, the
continuum composite elements are oIten advantageous especially in the cases that the use oI continuum elements are
unavoidable to achieve an accurate solution. For example, iI the nonlinear deIormation behavior is not negligible
through the thickness because oI either material and/or geometrical nonlinearity, a discretization along thickness
direction is required.
A thick composite cylinder subjected to an inner pressure is considered. The detailed description oI the analysis
procedure is presented using Mentat GUI. Steps on deIining material properties Ior each composite layer and deIining
layer orientations will be highlighted.
Background Information
An inIinitely long thick cylinder with an interior radius oI 60 mm and an exterior radius oI 140 mm is subjected to the
inner pressure oI 50 N/mm
2
. The cylinder consists eight layers with equal thickness. From interior to exterior, the
material layers are numbered Irom 1 to 8. The orthotropic material properties Ior layers 1, 3, 5, and 7 are given as
E
11
250000 N/mm
2
, E
22
E
33
10000 N/mm
2
,
v
12
v
31
0.01, v
23
0.25,
G
12
G
23
5000 N/mm
2
, G
31
2000 N/mm
2
.
The orthotropic material properties Ior layers 2, 4, 6, and 8 are given as
E
11
E
22
10000 N/mm
2
, E
33
250000 N/mm
2
,
v
12
v
31
0.25, v
23
0.01,
G
12
2000 N/mm
2
, G
23
G
31
5000 N/mm
2
.
The material orientations are based on global coordinate system; i.e., axial, radial, and circumIerential directions,
respectively.
See Figure 3.15-1 Ior the cross-section geometry oI the cylinder and the loading conIiguration.
Element type 154 (8-node, isoparametric, axisymmetric continuum composite element) is used in the analysis.
1057
CHAPTER 3.15
Continuum Composite Elements
Figure 3.15-1 Cross-Section Geometry and Loading of a Thick Composite Cylinder
The cylinder is modeled with only Iour axisymmetric elements in the r-z plane, with two elements spanning the
thickness oI the cylinder. Since the cylinder`s thickness spans eight layers oI two orthotropic materials, each element
spans Iour layers oI two orthotropic materials each. Prior to this Ieature, each layer would have been modeled with a
single element, hence requiring more elements. With the composite continuum element, we may span several layers
oI diIIerent materials.
Analysis
Model Generation
There are two elements along the radial (thickness) direction; i.e., each element contains Iour material layers.
C
L
r-z plane r- u plane
Two Elements
through Thickness
1058 Marc Users Guide: Part 2
CHAPTER 3.15
Boundary Conditions and Loads
MESH GENERATION
nodes ADD
0 140 0
0 60 0
80 60 0
80 140 0
FILL
elems ADD
1 2 3 4
SUBDIVIDE
DIVISIONS
2 2 1
ELEMENTS
all: EXIST.
RETURN
CHANGE CLASS
QUAD (8)
ELEMENTS
all: EXIST.
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
OK
nodes ADD
2
3
4 1
Figure 3.15-2 FE-Mesh
1059
CHAPTER 3.15
Continuum Composite Elements
Figure 3.15-3 Mesh for the Thick Composite Cylinder
Material Properties
DeIine two sets oI orthotropic materials, but no elements associated with material sets are given. This is
deIined aIterwards.
(boxes A)
END LIST (#)
NEW
EDGE LOAD
PRESSURE
50
OK
edges ADD
(box B)
END LIST (#)
MAIN
A
B
1060 Marc Users Guide: Part 2
CHAPTER 3.15
Composite Layer Property Definition
DeIinition oI composite layer properties is a key step in perIorming an analysis with continuum composite elements.
This includes the number oI layers within the elements, the material set associated with each layer, the percentage oI
the layer thickness respect to the total element thickness, and the elements using this set oI deIinition.
MATERIAL PROPERTIES
NAME mat1
ORTHOTROPIC
(fill out form to right)
1e4
1e4
25e4
.25
.01
.25
2e3
5e3
5e3
OK
NEW
NAME mat2
ORTHOTROPIC
(fill out form to right)
25e4
1e4
1e4
.01
.25
.01
5e3
5e3
2e3
OK
Figure 3.15-4 (Mat1) Orthotropic Materials
mat1
Figure 3.15-5 (Mat2) Orthotropic Materials
mat2
1061
CHAPTER 3.15
Continuum Composite Elements
LAYERED MATERIALS
NEW COMPOSITE
ADD LAYER
1
mat1
THICKNESS
25
ADD LAYER
2
mat2
THICKNESS
25
ADD LAYER
3
mat1
THICKNESS
25
ADD LAYER
4
mat2
THICKNESS
25
OK
NAME
comp1
elements ADD
all: EXIST.
MAIN
1062 Marc Users Guide: Part 2
CHAPTER 3.15
Figure 3.15-6 Define Composite Material
Composite Layer Orientation Definition
The composite layer orientation is deIined using geometric properties. In our problem, the layers are similar to element
edge 3 deIined by node 4 to node 1.
Figure 3.15-7 Solid Composite Menu
GEOMETRIC PROPERTIES
AXISYMMETRIC
SOLID COMPOSITE
thickness direction
EDGE 3 (1-4)
OK
elements ADD
all: EXIST.
MAIN
1063
CHAPTER 3.15
Continuum Composite Elements
Figure 3.15-8 Definition of Composite Layer Orientations
Define Job Parameters, Save Model, and Run Job
Element type 154 is used. This element type is one oI the special designed continuum composite elements. Stresses
are written into the post Iile.
JOBS
MECHANICAL
OK
ELEMENT TYPES
MECHANICAL
AXISYM
154
OK
all: EXIST.
RETURN (twice)
MECHANICAL
JOB RESULTS
available element tensors
Stress
OK (twice)
mat1
mat2
c
o
m
p
1
Edge 3
mat1
mat2
1064 Marc Users Guide: Part 2
CHAPTER 3.15
View Results
Figure 3.15-9 Radial Displacements of Thick Composite Cylinder
RUN
SUBMIT 1
MONITOR
OK
SAVE
MAIN
RESULTS
OPEN DEFAULT
NUMERICS
SCALAR
Displacement Y
OK
OUT (zoom out)
1065
CHAPTER 3.15
Continuum Composite Elements
Comparison
The analytical solution Ior this thick cylinder problem is available (see S.G. Lekhnitskii, Anisotriopic Plates, 1968).
The radial displacements oI the interior and the exterior surIaces oI the thick cylinder are 7.20e-2 and 7.38e-3,
respectively. These results are in good agreement with our Iinite element solutions which are 7.03e-2 and 7.62e-3,
respectively. Considering the relatively coarse mesh, the results are encouraging.
For the purpose oI comparison, an analysis based on the same mesh but low-order elements (Element type 152) is also
perIormed. The radial displacements oI the interior and the exterior surIaces oI the thick cylinder are 5.84e-2 and
4.18e-3, respectively. Obviously, iI the low-order elements are used, a Iiner mesh is needed to achieve reasonably
good results.
You may wish to run Mentat procedure Iiles that are in the examples/new_features subdirectory under
Mentat. The procedure Iile c17.proc builds, runs, and postprocesses this simulation.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
continuum_composite_elem.proc Mentat procedure Iile to run the above problem
1066 Marc Users Guide: Part 2
CHAPTER 3.15
Chapter 3.16: Super Plastic Forming (SPF)
3.16
Super Plastic Forming (SPF)

Summary 1068

SPF Modeling 1070

Discussion 1087

SPF with Adaptive Remeshing 1087

Discussion of Adaptive Meshing 1091

Input Files 1092

1068 Marc Users Guide: Part 2
CHAPTER 3.16
Summary
Title Superplastic Iorming
Problem Ieatures Flat sheet Iormed into a die by pressure
Geometry
Material properties Rigid plastic Ilow
Analysis type Static
Boundary conditions Edges Iixed with pressure adjusted to maintain proper strain rate
Element type Membrane elements with local mesh adaptivity
FE results Edge thickness proIiles and pressure history
X Z
Y 3.328e-002
3.774e-002
4.219e-002
4.664e-002
5.109e-002
5.555e-002
6.000e-002
6.445e-002
6.890e-002
7.336e-002
7.781e-002
lcase1
Thickness of Element
Inc: 38
Time: 2.943e+003
Z X
Y
Node A
Node B
t 0.080 in
116
115
113
112
117
Fixed Mesh
3.829
1 0
Thickness of Element (x.01)
7.544
Arc Length (x10)
118
120
239
238
237
236
119
235
233
232
122
121 234
240 241
114
118
113
112
241
122
121
120
119
117
116
115
114
232
Adaptive Mesh
661
706
240
239
509
238
504
237
236
235
234
233
33
34
35
36
37
38
lcase1
Time (x1000)
Process Pressure (x10)
0 2.943
0
7.018
32
31
29
1415 16 17 1819202122232425
26
2728
30
13 12
10
0123456789
11
1069
CHAPTER 3.16
Super Plastic Forming (SPF)
A Ilat sheet is Iormed into a rigid die by pressure. The die is three-dimensional and represents a corner oI a pan. This
Iine-grained material is assumed to be a rigid-plastic material with no elasticity and the Ilow stress is only a Iunction
oI the strain rate.
As the sheet contacts the die, Iriction causes the thickness oI the sheet to vary. In addition the pressure must be adjusted
to keep this strain rate sensitive material within a certain target range. This is necessary to maintain the proper Ilow oI
the super plastic material. Prediction oI thinning oI the sheet is very important since the sheet may become too thin Ior
its application.
Figure 3.16-1 Sheet Before and After Forming
The SPF Pressure control in Loadcase (see graphic below) is used to automatically adjust the pressure on the sheet to
keep within the target strain rate. Furthermore, the maximum pressure is limited by the capacity oI the rig used to Iorm
the sheet. Units used are inches, pounds, and seconds.
X Z
Y
3.328e-002
3.774e-002
4.219e-002
4.664e-002
5.109e-002
5.555e-002
6.000e-002
6.445e-002
6.890e-002
7.336e-002
7.781e-002
lcase1
Thickness of Element
Inc: 38
Time: 2.943e+003
Z X
Y
Node A
Node B
33
34
35
36
37
38
lcase1
Time (x1000)
Process Pressure (x10)
0 2.943
0
7.018
32
31
29
1415 16 17 1819202122
23
24
25
26
27
28
30
13 12
10
0123456789
11
1070 Marc Users Guide: Part 2
CHAPTER 3.16
SPF Modeling
This problem does a Super Plastic Forming oI a corner using the Super Plastic Forming loading scheme that keeps
adjusting the applied pressure to maintain an average target strain rate in the material.
Preprocessing
Model Generation consists oI making the die and sheet. Building the die Iirst, we begin in Mentat with a curve:
FILES
SAVE AS spf
RETURN
MESH GENERATION
coordinate system SET
GRID (on)
U DOMAIN
-7 7
U SPACING
.5
V DOMAIN
0 5
V SPACING
.5
FILL
RETURN
crvs: ADD
point( 7.0, 4.5, 0.0)
point( 4.0, 4.5, 0.0)
point( 4.0, 4.5, 0.0)
point( 3.5, 0.0, 0.0)
point( 3.5, 0.0, 0.0)
point( 0.0, 0.0, 0.0)
point( 0.0, 0.0, 0.0)
point(-4.0, 0.0, 0.0)
1071
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-2 Start with Rough Layout
Figure 3.16-3 Trim Lines with Fillets
CURVE TYPE
FILLET
RETURN
crvs: ADD
1
2
.5
8
3
.5
VIEW
SHOW VIEW 2
FILL
RETURN
EXPAND
SHIFT
TRANSLATIONS
0 0 3.5
CURVES
all: EXIST.
u=7
v=5
X
Y
Z
u=7
v=5
X
Y
Z
1072 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-4 Build Surface from Trimmed Curves
Figure 3.16-5 Continue Building Surface
RESET
SHIFT
CENTRIOD
0 0 3.5
ROTATIONS
0 -90/10 0
REPETITIONS
10
CURVES
pick curves shown
FILL
RESET
SHIFT
CENTROID
0 0 3.5
TRANSLATIONS
v=5 u=7
Y X
Z
v=5 u=7
Y X
Z
Pick
1073
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-6 Finish Die Surface
Now, we add nodes that will contain the mesh.
-4.0 0 0
REPETITIONS
1
CURVES (pick curves shown)
END LIST (#)
RETURN
CURVES REMOVE
ALL:EXISTING
SELECT
SURFACES
ALL:EXISTING
MAKE INVISIBLE
RETURN
CURVE TYPE
LINE
RETURN
VIEW 1, RETURN
crvs: ADD
point(-3.5, 5.0, 0.5) (type in coordinates as shown)
point( 0.0, 5.0, 0.5)
point( 0.0, 5.0, 0.5)
point( 6.5, 5.0, 0.5)
v=5 u=7
Y X
Z
Pick
1074 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-7 Add More Lines to Make Mesh
Figure 3.16-8 Expand Lines to make Mesh Surface
EXPAND
RESET
SHIFT
TRANSLATIONS
0 0 3.0
CURVES
all: EXIST.
RESET
SHIFT
CENTRIOD
0 0 3.5
ROTATIONS
0 -90/10 0
REPETITIONS
10
CURVES (pick curve shown)
v=5 u=7
Y X
Z
v=5
Y
Z
1075
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-9 Continue Building Mesh Surface
Figure 3.16-10 Finish Building Mesh Surface
RESET
SHIFT
TRANSLATIONS
-3.5 0 0
REPETITIONS
1
CURVES (pick curve shown)
END LIST (#)
RETURN
v=5 u=7
Y X
Z
Pick
v=5 u=7
Y X
Z
Pick
1076 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-11 Convert Sector Surfaces to Elements
CONVERT
DIVISIONS
10 1
SURFACES TO ELEMENTS (pick those shown)
DIVISIONS
10 10
SURFACES TO ELEMENTS (pick remaining rectangular surfaces)
END LIST (#)
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN (twice)
v=5 u=7
Y X
Z
Pick
1077
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-12 Final Mesh
BOUNDARY CONDITIONS
MECHANICAL
SELECT
ELEMENTS
all: EXIST.
MAKE VISIBLE
RETURN
FIXED DISPLACEMENT
FIX X,Y,Z = 0
OK
SELECT
METHOD PATH
NODES
(pick 1st middle and last node of outer path)
END LIST (#)
RETURN
nodes: ADD
all: SELECTED
X Z
Y
1078 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-13 Fix Displacements on Outer Binding
NEW
FIX X = 0, OK
nodes: ADD
(along x=0)
END LIST (#)
NEW
FIX Z = 0
nodes: ADD
(along z=0)
END LIST (#)
Y X
Z
1079
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-14 Fix Symmetry Displacements
Figure 3.16-15 Turn on SPF Pressure Control
NEW
FACE LOAD
SUPERPLASTICITY CONTROL
ON PRESSURE NEGATIVE
OK
FACES ADD
ALL EXISTING
MAIN
MATERIAL PROPERTIES (twice)
NEW
STANDARD
STRUCTURAL
TYPE: RIGID-PLASTIC
PLASTICITY
METHOD: POWER LAW (fill out as shown in Figure 3.16-16)
apply1
apply2
apply3
Y X
Z
X
Y
Z
1080 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-16 Enter SPF Material Behavior
OK (twice)
ELEMENTS ADD: ALL EXISTING
MAIN
GEOMETRIC PROPERTIES
3-D
MEMBRANE
THICKNESS
.080
OK
ELEMENTS ADD
all: EXIST.
SELECT
MAKE INVISIBLE
MAIN
CONTACT
CONTACT BODIES
NEW
NAME
workpiece
DEFORMABLE
FRICTION COEFFICIENT
.3
o Bc

1081
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-17 Define Contact Bodies
OK
elements ADD
All: EXIST.
NEW
NAME
die
RIGID
VELOCITY PARAMETERS
APPROACH VELOCITY Y
.01
OK
FRICTION COEFFICIENT
.3
OK
surfaces: ADD (pick surfaces forming die)
END LIST (#)
ID BACKFACES
FLIP SURFACES
all: EXIST. (flip die surfaces until gold color will touch the workpiece)
MAIN
LOADCASES
MECHANICAL
STATIC
TOTAL LOADCASE TIME
3000
workpiece
die
none
X
Y
Z
1082 Marc Users Guide: Part 2
CHAPTER 3.16
stepping procedure
MULTI-CRITERIA
PARAMETERS
INITIAL FRACTION
1e-4
MAXIMUM FRACTION
5e-3
OK
CONVERGENCE TESTING
RELATIVE/ABSOLUTE
RESIDUALS AND DISPLACEMENTS
RELATIVE FORCE TOLERANCE = 0.01
MAXIMUM REACTION FORCE CUTOFF
6
MAXIMUM ABSOLUTE RESIDUAL FORCE
6
RELATIVE DISPLACEMENT TOLERANCE = 0.05
MINUMUM DISPLACEMENT CUTOFF
5e-5
MAXIMUM ABSOLUTE DISPLACEMENT
5e-5
OK
SUPERPLASTICITY CONTROL
pressure
MINIMUM
.001
MAXIMUM
300
TARGET STRAIN RATE METHOD
2e-4
TARGET STRAIN RATE METHOD (on)
CONSTANT (on)
PRE_STRESS
50
# INCREMENTS
5 (filled out as shown in Figure 3.16-18)
1083
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-18 Define Loadcase with SPF Parameters
Analysis
Here, we set up the problem to run with Coulomb Irictions using membrane elements. Later, we will run the problem
with adaptive meshing.
OK (twice)
MAIN
JOBS
NEW (MECHANICAL)
PROPERTIES
lcase1
ANALYSIS OPTIONS
LARGE STRAIN (on)
FOLLOWER FORCE (on)
OK
JOB RESULTS
available element scalars
Equivalent Plastic Strain Rate
Thickness of Element
1084 Marc Users Guide: Part 2
CHAPTER 3.16
Results
OK
CONTACT CONTROL
ADVANCED CONTACT CONTROL
COULOMB
BILINEAR
OK (twice)
ELEMENT TYPES, MECHANICAL
3-D MEMBRANE/SHELL
18 (Quad 4)
OK
all: EXIST.
RETURN (twice)
SAVE
RUN
STYLE: OLD
SUBMIT1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
NEXT
DEF ONLY
CONTOUR BAND
SCALAR
Thickness of Element
LAST (Last Increment)
1085
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-19 Thickness Contours
PATH PLOT
SET NODES
(Node A) (Node B)
END LIST (#)
ADD CURVES
ADD CURVE
Arc Length
Thickness
FIT
RETURN
generalized xy plot: COPY TO (Send to XY plotter)
3.328e-002
3.774e-002
4.219e-002
4.664e-002
5.109e-002
5.555e-002
6.000e-002
6.445e-002
6.890e-002
7.336e-002
7.781e-002
lcase1
Thickness of Element
Inc: 38
Time: 2.943e+003
Z X
Y
Node A
Node B
1086 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-20 Thickness Profile along Edge
RESULTS
HISTORY PLOT
ALL INCS
ADD CURVES
GLOBAL
Time
Process Pressure
FIT
0
Time : 2942.77
Inc : 38
Thickness of Element (x.01)
Arc Length (x10)
lcase1
112
113
114
115
116
117
235
234
233
232
122
121
120
119
118
1
4.09
7.544
236
237
239
241 240
238
1
1087
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-21 Pressure Schedule
Discussion
From Figure 3.16-19, we see a minimum thickness oI about 0.034 which is 2.3 Iold decrease in the original sheet
thickness oI 0.080. The path in Figure 3.16-20 shows how rapidly the thickness reduces Irom the binder to the center
oI the sheet. Figure 3.16-21 shows how the pressure is automatically adjusted to keep the average strain rate in the
sheet at the target strain rate speciIied in Figure 3.16-18. Also the procedure stops when 100 oI the nodes come into
contact with the die at a time beIore the allotted 3000 seconds.
You will Iind it very instructive to turn oII the Iriction (JOBS>PROPERTIES>CONTAT CONTROL>FRICTION TYPE:
NONE). Then, compare the thickness proIile with Figure 3.16-19 to see how much Iriction thins the sheet. Many times
the thinning (thanks to Iriction) is too severe and another Iorming technique may be necessary.
As the mesh Iorms over the die the original element size may be too large to capture local surIace details properly.
Now that the SPF simulation is running, we can use the adaptive remeshing with local reIinement to increase the
number oI elements where they can improve this situation.
SPF with Adaptive Remeshing
Preprocessing consists starting with our original model and adding adaptive remeshing and running a new model.
Starting Irom Mentat, let`s open our previous model and save as a new model.
FILES
OPEN spf
SAVE AS
33
34
35
36
37
38
lcase1
Time (x1000)
Process Pressure (x10)
0 2.943
0
7.018
32
31
29
1415 16 17 1819202122
23
24
25
26
27
28
30
13 12
10
0123456789
11
1
1088 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-22 Elements Selected for Local Refinement
spf_adapt
OK
RETURN
ADAPTIVE REMESHING
LOCAL ADAPTIVITY CRITERIA
NODES IN CONTACT
MAX # LEVELS = 2
OK
ELEMENTS ADD (pick elements shown in Figure 3.16-22)
END LIST
ID LOCAL ADAPTIVITY CRITERIA
MAIN
JOBS
adapt1
none
Y X
Z
1089
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-23 Status of Job
Results
RUN
SAVE MODEL
SUBMIT1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
NEXT
DEF ONLY
CONTOUR BAND
SCALAR
Contact Status
LAST
1090 Marc Users Guide: Part 2
CHAPTER 3.16
Figure 3.16-24 Contact Status (1 = touching, 0 = not touching)
Figure 3.16-25 Adapted Thickness Contours
SCALAR
Thickness of Element
PATH PLOT
NODE PATH
0.000e+000
1.000e-001
2.000e-001
3.000e-001
4.000e-001
5.000e-001
6.000e-001
7.000e-001
8.000e-001
9.000e-001
1.000e+000
lcase1
Contact Status
Inc: 62
Time: 2.611e+003
Z X
Y
3.012e-002
3.477e-002
3.941e-002
4.406e-002
4.870e-002
5.334e-002
5.799e-002
6.263e-002
6.728e-002
7.192e-002
7.656e-002
lcase1
Thickness of Element
Inc: 62
Time: 2.611e+003
Z X
Y
4
1091
CHAPTER 3.16
Super Plastic Forming (SPF)
Figure 3.16-26 Thickness Profile along Edge (with and without Adaptive Meshing)
Discussion of Adaptive Meshing
From Figure 3.16-24 that all oI the nodes touching the binder have a contact status oI 1. Note that with more elements
in the die corners that the minimum thickness is a bit lower that the coarse mesh as shown in Figure 3.16-25. Using
the adaptive remeshing with local reIinement to increase the number oI elements has given a better Iit to the die.
Figure 3.16-26 uses the Generalize XY plotter to compare results between these two runs. The original mesh had 400
elements and the Iinal adapted mesh has 760 elements. More elements can easily be added by changing the remeshing
criteria to continue to improve the results.
(Node A) (Node B)
END LIST
ADD CURVES
ADD CURVE
Arc Length
Thickness Of Element
RETURN
FIT
generalized xy plotter: COPY TO (Send to XY plotter)
FIT
116
115
113
112
117
Fixed Mesh
3.829
1 0
Thickness of Element (x.01)
7.544
Arc Length (x10)
118
120
239
238
237
236
119
235
233
232
122
121 234
240 241
114
118
113
112
241
122
121
120
119
117
116
115
114
232
Adaptive Mesh
661
706
240
239
509
238
504
237
236
235
234
233
1092 Marc Users Guide: Part 2
CHAPTER 3.16
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Also, this problem can automatically be run Irom the HELP menu under DEMONSTRATIONS > RUN A DEMO
PROBLEM > SUPERPLASTIC FORMIN and SPF + ADAPT. MESHING.
File Description
super_plastic_forming.proc Mentat procedure Iile to run the above example
super_plastic_forming_b_help.proc Mentat procedure Iile to run the above example with adaptive
meshing
Chapter 3.17: Gaskets
3.17
Gaskets

Chapter Overview 1094

Simulation of a Cylinder Head Joint 1095

Input Files 1122

1094 Marc Users Guide: Part 2
CHAPTER 3.17
Chapter Overview
Engine gaskets are used to seal joints between the metal parts oI the engine to prevent steam or gas Irom escaping.
They are complex (oIten multi-layer) components, usually rather thin and typically made oI several diIIerent materials
oI varying thickness. The gaskets are careIully designed to have a speciIic behavior in the thickness direction. This is
to ensure that the joints remain sealed when the metal parts are loaded by thermal or mechanical loads. The through-
thickness behavior, usually expressed as a relation between the pressure on the gasket and the closure distance oI the
gasket, is highly nonlinear, oIten involves large plastic deIormations, and is diIIicult to capture with a standard material
model. The alternative, modeling the gasket in detail by taking every individual material into account in the Iinite
element model oI the engine, is not Ieasible as it requires a large number oI elements, making the model unacceptably
large. The GASKET material model addresses these problems by allowing gaskets to be modeled with only one
element through the thickness, while the experimentally or analytically determined complex pressure-closure
relationship in that direction can be used directly as input Ior the material model.
This chapter introduces the GASKET material. For this purpose, a cylinder head joint subjected to a combination oI
mechanical and thermal loadings will be analyzed. The mechanical loading consists oI sequentially Iastening the bolts
oI the joint, Iollowed by application oI an internal pressure to the cylinder head. The Iastening oI the bolts is simulated
by introducing a pre-tension Iorce in the bolts. This is achieved by splitting up the Iinite element mesh oI the bolts in
two parts and connecting them by using the standard options, TYING and SERVO LINK. The thermal loading involves
uniIormly heating the joint, cooling it down, and bringing it back to room temperature.
The procedure Iile to demonstrate this example is called gasket.proc under mentat2010/
examples/marc_ug/s3/c3.17.
Figure 3.17-1 Finite Element Mesh of the Cylinder Head Joint
1095
CHAPTER 3.17
Gaskets
Simulation of a Cylinder Head Joint
The model consists oI the cylinder head cover and a small portion oI the lower part oI the cylinder head (see
Figure 3.17-1). Both the cover and the lower part are made oI steel. A thin gasket layer seals the joint between the
cover and the lower part. The joint is Iastened by two steel bolts.
The assembly is loaded in six stages. In the Iirst two stages, the Iastening oI the joint is simulated by applying a
pre-tension oI 12000 N to each oI the bolts. AIter the bolts have been loaded, the three-stage thermo-mechanical
loading cycle starts. First, the assembly is heated uniIormly to 180C, while simultaneously an interior pressure oI
1.2 MPa is applied to the cover and the lower part oI the assembly. Next, the joint is cooled down uniIormly to -10C,
while retaining the interior pressure. In the IiIth stage, the pressure is removed and the temperature is increased again
to room temperature. The Iinal loadcase oI the analysis consists oI disassembling the joint by loosening the bolts.
Figure 3.17-2 PARAMETERS Menu
Mesh Generation
Since the assembly and the applied loads are symmetric with respect to the zx-plane (see Figure 3.17-1), only one halI
oI the assembly is taken into account in the model, while symmetry conditions are imposed by means oI a contact
symmetry plane.
Note: The PARAMETERS menu and the parameters describes the Iinite element mesh oI the cylinder head joint. The values
oI the parameters are in millimeters.
1096 Marc Users Guide: Part 2
CHAPTER 3.17
The model is setup in a Iully parametric Iashion, allowing diIIerent, but similarly shaped models to be created by
modiIying the parameters. The parameters are deIined in the UTILS-> PARAMETERS menu. Figure 3.17-2 displays
the menu and lists the parameters that govern the dimensions oI the diIIerent components oI the model, as well as their
respective values. The PARAMETERS menu oIIers two methods oI deIining or modiIying parameters. II the
EVALUATION METHOD is DELAYED (the deIault), the parameter becomes an abbreviation Ior the expression that is
assigned to it. II the EVALUATION METHOD is IMMEDIATE, the expression is evaluated Iirst and its value is being
assigned to the parameter. The diIIerence becomes apparent when the expression contains other parameters or calls to
numerical Iunctions that return inIormation about the model. The value oI a parameter deIined using the delayed
evaluation method is the value oI the expression assigned to it at the time the parameter is being used. By contrast, the
value oI a parameter deIined using the immediate evaluation method is the value oI the expression at the time oI the
deIinition oI the parameter.
The generation oI the parametric Iinite element mesh will not be discussed in detail here. Instead, the reader is reIerred
to the procedure Iile that belongs to this chapter and the comments in that Iile. For the set oI parameters used in the
present example, the resulting Iinite element mesh is depicted in Figure 3.17-1.
Tyings and Servo Links
As mentioned earlier, in the Iirst two stages oI the analysis, the Iastening oI the joint is simulated by applying
pre-tension loads oI 12000 N to each oI the bolts. In the Iirst stage, the leIt bolt (see Figure 3.17-1) is pre-tensioned
while the right bolt is locked and in second stage the right bolt is loaded while the length oI the leIt bolt is Iixed. During
the subsequent three-stage thermo-mechanical loading cycle, the bolts are locked and in the Iinal stage oI the analysis,
the joint is disassembled by loosening the bolts.
Figure 3.17-3 Bolts split up into a Top and Bottom Parts connected by Tyings and Servo Links
The pre-tension Iorce in the bolt is simulated using standard options existing in Marc, namely, TYING and SERVO
LINK. During the mesh generation process, the Iinite element meshes oI the bolts have been split up into a top and a
top parts
bottom parts
1097
CHAPTER 3.17
Gaskets
bottom part as depicted in Figure 3.17-3. Corresponding nodes on both sides oI the cut are now connected to each other
and to a special node, called the control node oI the bolt, by means oI a set oI tyings (to prevent relative tangential
motion oI the two parts) and servo links. As is shown below, the servo links can be chosen in such a way that a
pre-tension Iorce can be applied to the bolt simply by applying a POINT LOAD boundary condition to the control node
oI the bolt. Alternatively, the bolt can be tightened, locked, or loosened by applying a FIXED DISPLACEMENT
boundary condition to the control node.
Figure 3.17-4 Applying Pre-tension to a Truss Modeled by Two Truss Elements
Consider a truss that is clamped at both ends and that is divided into two truss elements, as show in Figure 3.17-4. The
elements are not connected to each other; at the center oI the truss, two distinct nodes exist, called the top node and
bottom node. Let
and (3.17-1)
be the vectors with, respectively, the displacements oI these two nodes in the axial direction oI the truss and the
corresponding Iorces. The required continuity oI the displacement Iield in the truss can be ensured by stating that both
displacements are equal. Denoting this common displacement value by u*, the continuity oI the displacement Iield is
expressed by the constraint equation,
, (3.17-2)
Note: Please note that the small gap between the top and bottom parts shown in Figure 3.17-3 is purely Ior visualization
purposes and to allow easy selection oI the nodes on both sides oI the gap. The gap is closed by moving down the
nodes oI the top part just above the gap in the negative z-direction aIter the links have been created, although the duplicate
nodes remain.
top node
bottom node
F
bottom
F
top
u
bottom
u
top
u
u
top
u
bottom
T
F
F
top
F
bottom
T

u Tu
*

1098 Marc Users Guide: Part 2

CHAPTER 3.17
in which T |1 1|
T
. One oI the properties oI a constraint equation is the Iact that the work done by the constraint is
zero. II F* is the Iorce that is work conjugate to u*, the zero-work principle can be stated as Iollows,
. (3.17-3)
Substitution oI equation (3.17-2) into equation (3.17-3) and requiring that the result is valid Ior arbitrary values oI u*,
yield the Iollowing expression Ior the Iorce F*,
. (3.17-4)
Due to the zero-work principle, the displacement constraint equation (3.17-2) is equivalent to the Iorce constraint
equation (3.17-4).
At this point, it is important to realize that the pre-tension Iorce in the truss is nothing but the Iorce on the bottom node
F
bottom
(or minus that on the top node). Prescribing that Iorce basically amounts to stating that,
. (3.17-5)
The latter relation can be viewed as an additional constraint expressed in terms oI Iorces. Introducing a new Iorce
vector F
new
given by,
, (3.17-6)
allows both Iorce constraints (equations (3.17-4) and (3.17-5)) to be combined into the matrix equation
, with . (3.17-7)
To Iind the equivalent displacement variant oI equation (3.17-7), the zero-work principle is again applied. Let
u
pre-tension
be the degree oI Ireedom that is work conjugate to the pre-tension Iorce and deIine,
. (3.17-8)
The zero-work principle now reads,
, (3.17-9)
in which equation (3.17-7) is used. Since equation (3.17-9) must hold Ior all Iorce vectors F, it Iinally Iollows that
(3.17-10)
or,
. (3.17-11)
F
*
u
*
F
T
u
F
*
F
T
T
F
top
F
bottom
1
1
F
top
F
bottom

F
bottom
F
pre-tension

F
new
F
*
F
pre-tension
T

F
new
T
new
F T
new
1 1
0 1

u
new
u
*
u
pre-tension
T

F
T
u F
new
T
u
new
F
T
T
new
T
u
new

u T
new
T
u
new

u
top
u
bottom
1 0
1 1
u
*
u
pre-tension

1099
CHAPTER 3.17
Gaskets
From the Iirst row oI equation (3.17-11), it Iollows that u
top
u*. Substitution oI this relation into the equation given
by the second row then yields
. (3.17-12)
The latter equation is the desired servo link. It is the displacement equivalent oI equation (3.17-5) and relates the
displacements oI the top and bottom nodes to the 'pre-tension displacement. Rewriting equation (3.17-12) as,
. (3.17-13)
shows that the 'pre-tension displacement can be interpreted as the shortening oI the truss.
A pre-tension Iorce can thus be applied to the truss by creating one additional node and by tying the displacement oI
the bottom node to that oI the top node and oI the additional node using the servo link equation (3.17-12). Then, by
virtue oI equation (3.17-5), iI a POINT LOAD is applied to the additional node, a pre-tension Iorce oI that amount is
introduced in the truss. Conversely, iI a FIXED DISPLACEMENT boundary condition is applied to the additional node,
the shortening oI the truss is prescribed.
Since the pre-tension Iorces and the shortening oI the bolts must be controlled separately, two additional nodes (one
Ior each bolt) are introduced in the present example. For each bolt, the nodes oI the bottom part just below the cut (the
node set bolt_bottom_nodes) are tied to the corresponding nodes oI the top part just above the cut (the node set
bolt_top_nodes) and the control node oI the bolt. The servo links are deIined between the Iirst degrees oI Ireedom
oI the nodes. A local coordinate system is deIined in the bolt_bottom_nodes and the bolt_top_nodes such that
the local x-axis coincides with the global z-direction (the axial direction oI the bolts). Since servo links and nodal ties
always act on local degrees oI Ireedom, the servo links between the bolt_bottom_nodes and the bolt_top_nodes
act in the global z-direction (a local coordinate system allows the servo link to act in any desired direction and not just
in one oI the global directions). In addition to the servo links, the bolt_bottom_nodes are tied to the
bolt_top_nodes using tying type 203 (second and third degree oI Ireedom, or global y- and x-directions) to prevent
relative tangential motion between the top and the bottom part.
Multiple servo links and nodal ties are most easily created using the ADD SERVOS command in the N TO N SERVO
LINKS menu, respectively the ADD TIES command in the N TO N NODAL TIES menu. Since these commands require
an equal number oI nodes to be entered Ior each term in the constraint, the nodes oI bolt_top_nodes set are
duplicated Iirst and the copies are put in a set called bolt_control_nodes (and removed Irom the
bolt_top_nodes set). AIter the servo links have been created, the bolt_control_nodes Ior each bolt are merged
into a single node using a sweep operation. The button sequence Ior creating the servo links, the nodal ties and the
local coordinate system reads:
LINKS
SERVO LINKS
N TO N SERVO LINKS
TIED DOF
1
RETAINED # TERMS
2
u
bottom
u
top
u
pre-tension

u
pre-tension
u
bottom
u
top

1100 Marc Users Guide: Part 2
CHAPTER 3.17
TERM 1 DOF
1
TERM 1 COEF.
1.0
TERM 2 DOF
1
TERM 2 COEF.
1.0
CREATE PATHS
(off)
ADD SERVOS
bolt_bottom_nodes
bolt_top_nodes
bolt_control_nodes
RETURN (thrice)
LINKS
NODAL TIES
N TO N NODAL TIES
TYING TYPE
203
OK
CREATE PATHS
(off)
ADD TIES
bolt_bottom_nodes
bolt_top_nodes
RETURN (thrice)
MESH GENERATION
SWEEP
SWEEP_NODES
bolt_control_nodes
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
TRANSFORMS
NEW
1101
CHAPTER 3.17
Gaskets
Boundary Conditions
To load the bolts with a pre-tension oI 12000 N, two POINT LOAD boundary conditions are created. Since only halI
oI the bolts is taken into account in the model, halI oI the pre-tension load is applied to the control nodes oI the bolts.
The locking oI the bolts in the subsequent thermo-mechanical loading cycle and the loosening in the Iinal loadcase oI
the analysis is simulated by applying FIXED DISPLACEMENT boundary conditions to the control nodes. The Iirst
degree oI Ireedom is Iixed to 0 mm in the loading cycle and decreased to -0.2 mm in the Iinal loadcase, to prescribe
an elongation oI 0.2 mm. Tables are being used to control the loading history oI these boundary conditions. For the
leIt bolt, the button sequence is given by:
ROTATE
0 -90 0
ADD NODES
bolt_bottom_nodes
bolt_top_nodes
RETURN (thrice)
BOUNDARY CONDITIONS
MECHANICAL
TABLES
NEW
TABLE TYPE
time
ADD POINT
0 0
1 1
6 1
NAME
left_bolt_load_history
NEW
TABLE TYPE
time
ADD POINT
0 0
5 0
6 1
NAME
left_bolt_lock_history
1102 Marc Users Guide: Part 2
CHAPTER 3.17
For the right bolt, a similar sequence is used, except that the table that deIines the history oI the POINT LOAD is slightly
diIIerent. Since the second bolt is loaded in the second loadcase, the table deIined by the points (0,0), (1,0), (2,1) and
(6,1).
In three-stage thermo-mechanical loading cycle that Iollows the prestressing oI the bolts, the cylinder head joint is
subjected to a combination oI mechanical and thermal loads. The mechanical loading consists oI a pressure oI 1.2 MPa
applied to the interior oI the cylinder head cover and the lower part. The pressure is applied in the Iirst stage oI the
loading cycle and removed in the third stage, using a FACE LOAD boundary condition in which the PRESSURE is set
to 1.2 MPa. The TABLE that deIines the history oI the pressure is oI type time and is deIined by the points (0,0), (2,0),
(3,1), (4,1), (5,0) and (6,0). The pressure is applied to the element Iaces at the interior boundary oI the cover and the
lower part using the FACES ADD button. These Iaces may be picked by clicking each oI them using the leIt mouse
button. However, this is cumbersome. ThereIore, during the mesh generation process, the nodes at the interior
RETURN
NEW
POINT LOAD
X FORCE
6000
TABLE
left_bolt_load_history
OK
NODES ADD
4104
END LIST
NAME
prestress_left_bolt
NEW
FIXED DISPLACEMENT
X DISPLACE
-0.2
TABLE
left_bolt_lock_history
OK
NODES ADD
4104
END LIST
NAME
lock_unlock_left_bolt
1103
CHAPTER 3.17
Gaskets
boundary have been stored in a set called interior_nodes. Using this node set, the Iaces are selected easily by
means oI the SELECT FACES BY NODES operation,
and subsequently added to the boundary condition using the ALL: SELECT. button.
The thermal part oI the loading cycle consists oI a uniIorm increase oI the temperature to 180C in the Iirst stage, a
uniIorm decrease to -10C in the second stage and again a uniIorm increase back to room temperature (20C) in the
third stage oI the loading cycle. This is achieved by applying a NODAL TEMPERATURE boundary condition to all
nodes in the model, setting the TEMPERATURE to 1 and employing a TABLE to a table oI type time deIined by the
points (0,20), (2,20), (3,180), (4,-10), (5,20) and (6,20).
Finally, to suppress rigid body motions, the displacements in the z-direction oI all nodes at the bottom oI the lower-
part oI the cylinder head assembly are suppressed as well as the displacements in the x-direction oI the nodes at the
bottom oI the lower-part that lie in the yz-plane. The applied mechanical loads are depicted in Figure 3.17-1.
Figure 3.17-1 Mechanical Boundary Conditions applied to the Cylinder Head Joint
SELECT
SELECT BY
ALL IN LIST
FACES BY NODES
SET
interior_nodes
RETURN (twice)
1104 Marc Users Guide: Part 2
CHAPTER 3.17
Initial Conditions
The temperature oI the model is initialized to 20C (room-temperature) by means oI a NODAL TEMPERATURE
initial condition.
Material Properties
The new GASKET material allows gaskets to be modeled with only one element through the thickness, while the
analytically or experimentally determined complex pressure-closure relationship can be used directly as input Ior the
material model. The behavior in the thickness direction, the transverse shear behavior, and the membrane behavior are
Iully uncoupled in the model. The transverse shear and membrane behavior are linear elastic, characterized by a
transverse shear modulus and the in-plane Young`s modulus and Poisson`s ratio, respectively. In the thickness
direction, the behavior in tension is also linear elastic and is governed by a tensile modulus, deIined as a pressure per
unit length.
In compression, two types oI gasket behavior can be simulated: fully elastic and elastic-plastic. For the Iully elastic
model, the user only supplies the loading path in the Iorm oI a (nonlinear) relation between the pressure on the gasket
and the closure distance oI the gasket.
In the elastic-plastic model, the user speciIies the loading path, the yield pressure above which plastic deIormation
develops and up to ten unloading paths. The loading and unloading paths must be supplied as (nonlinear) relations
between the pressure on the gasket and the closure distance oI the gasket. The unloading paths deIine the elastic
unloading behavior at diIIerent amounts oI plastic deIormation. II the gasket unloads at an amount oI plastic
deIormation Ior which no unloading path has been given, the unloading path is constructed automatically by
interpolation between the two nearest user supplied paths. The elastic-plastic model allows Ior large plastic
deIormations.
Note: The thermal loading consists oI a NODAL TEMPERATURE boundary condition applied to all nodes oI the model and
is not drawn here.
INITIAL CONDITIONS
MECHANICAL
NEW
NODAL TEMPERATURE
ON
TEMPERATURE
20
OK
NODES ADD
ALL: EXIST.
NAME
initial_temperature
1105
CHAPTER 3.17
Gaskets
Figure 3.17-2 The Finite Element Mesh of Gasket
Since the thickness oI a gasket can vary considerably throughout the gasket, an initial gap may be set Ior the gasket
material to account Ior the Iact that the gasket is actually thinner than the Iinite elements used to model it. As long as
the closure distance is smaller than the initial gap, no pressure is built up in the gasket. The gasket can then be modeled
as a Ilat sheet oI uniIorm thickness and the initial gap parameter can be set Ior those regions where the gasket is thinner
than the mesh.
The gasket material must be used with the gasket element types 149 (3-D solid), 151 (plane strain) or 152
(axisymmetric). Note that these elements currently have no associated heat-transIer element, so gaskets cannot be used
in coupled thermo-mechanical analyses. However, the material can exhibit isotropic thermal expansion, characterized
by a single thermal expansion coeIIicient.
The gasket used in this example is modeled as a Ilat sheet with a thickness oI one millimeter and consists oI two
regions with diIIerent material properties (see Figure 3.17-2). For both regions, the data oI the loading path and one
unloading path are stored in Iour two-column data Iiles. The Iirst column in these Iiles contains the closure distances,
the second column the corresponding pressures. These data are used as input Ior the gasket material model by creating
Iour tables oI type gasket_closure and reading in the Iiles.
MATERIAL PROPERTIES
TABLES
READ
RAW
body_loading.raw
OK
1106 Marc Users Guide: Part 2
CHAPTER 3.17
The loading and unloading paths oI the gasket materials can be displayed in one graph using the Generalized XY
Plotter. The GET PLOTS FROM TABLE operation copies the data Irom every table in the model to the plotter, including
the data Irom the load history tables oI the boundary conditions. The latter are subsequently being removed Irom the
XY plotter. The resulting picture is depicted in Figure 3.17-3.
TABLE TYPE
gasket_closure
NAME
gasket_body_loading
READ
RAW
ch31_body_unloading.raw
OK
TABLE TYPE
gasket_closure
NAME
gasket_body_unloading
READ
RAW
ring_loading.raw
OK
TABLE TYPE
gasket_closure
NAME
gasket_ring_loading
READ
RAW
ring_unloading.raw
OK
TABLE TYPE
gasket_closure
NAME
gasket_ring_unloading
1107
CHAPTER 3.17
Gaskets
Figure 3.17-3 The Loading and Unloading Paths of the Gasket in the two Regions
The deIinition oI the membrane properties and the thermal expansion coeIIicient oI the gasket is separated Irom the
deIinition oI the properties in the thickness direction and the transverse shear behavior. For the Iormer, the GASKET
material reIers to an existing isotropic material. Multiple gasket materials can reIer to the same isotropic material Ior
their membrane properties and thermal expansion. However, in the present example, the membrane stiIIness oI the
body oI gasket is 120 MPa and its thermal expansion coeIIicient is per C, while Ior the ring the membrane
modulus is 100 MPa and the thermal expansion coeIIicient is per C. Poisson`s ratio is 0 Ior both regions.
ThereIore, two isotropic materials are created with the membrane properties and thermal expansion coeIIicient oI the
two gasket regions.
UTILS
GENERALIZED XY PLOT
GET PLOTS FROM
TABLE
FIT
REMOVE
1 1 1 1 1 1
FIT
RETURN (twice)
loading path body
loading path ring
unloading path body
unloading path ring
5 10
5

1 10
4

1108 Marc Users Guide: Part 2

CHAPTER 3.17
The behavior in the thickness direction and the transverse shear behavior oI the gasket is deIined by creating a new
GASKET material and supplying the yield pressure, the tensile modulus, the initial gap parameter (iI necessary), the
tables oI the loading and unloading paths and the transverse shear modulus in the GASKET MATERIAL PROPERTIES
menu as depicted in Figure 3.17-4. In this menu, also the isotropic material Ior membrane behavior has to be selected.
In the example, the thickness oI the gasket ring is 10 larger than the thickness oI the body oI the gasket. Since the
gasket is modeled as a Ilat sheet oI uniIorm thickness, the INITIAL GAP is used Ior the body oI the gasket to take this
into account.
The yield pressure oI the body oI the gasket is 52 MPa, its tensile modulus is 72 MPa/mm, and the transverse shear
modulus is 40 MPa. For the ring, the yield pressure is 42 MPa, the tensile modulus is 64 MPa/mm and the transverse
shear modulus is 35 MPa.
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNGS MODULUS
120
THERMAL EXP.
THERMAL EXP. COEF.
5e-5
OK (twice)
NAME
gasket_body_membrane
NEW
ISOTROPIC
YOUNGS MODULUS
100
THERMAL EXP.
THERMAL EXP. COEF.
1e-4
OK (twice)
NAME
gasket_ring_membrane
1109
CHAPTER 3.17
Gaskets
Figure 3.17-4 The GASKET MATERIAL PROPERTIES Menu
NEW GASKET
YIELD PRESSURE
52
TENSILE MODULUS
72
INITIAL GAP
1/11
LOADING PATH TABLE
gasket_body_loading
UNLOADING PATHS TABLE 1
gasket_body_unloading
TRANSVERSE SHEAR BEHAVIOR MODULUS
40
MEMBRANE BEHAVIOR MATERIAL
gasket_body_membrane
OK
ELEMENTS ADD
1110 Marc Users Guide: Part 2
CHAPTER 3.17
The cylinder head cover, the lower part and the bolts are all made oI steel. Young`s modulus is , Poisson`s ratio
0.3, and thermal expansion coeIIicient .
SET
gasket_body
OK
NAME
gasket_body
NEW GASKET
YIELD PRESSURE
42
TENSILE MODULUS
64
LOADING PATH TABLE
gasket_ring_loading
UNLOADING PATHS TABLE 1
gasket_ring_unloading
TRANSVERSE SHEAR BEHAVIOR MODULUS
35
MEMBRANE BEHAVIOR MATERIAL
gasket_ring_membrane
O
K
ELEMENTS ADD
SET
gasket_ring
OK
NAME
gasket_ring
RETURN
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNGS MODULUS
2.1e5
2.1 10
5

1.5 10
5

1111
CHAPTER 3.17
Gaskets
Geometric Properties
The thickness direction oI the gasket elements has to be speciIied by means oI a geometric property oI type 3-D SOLID
COMPOSITE/GASKET. The Iinite element mesh oI the gasket is created in such a way that Ior all elements in the
gasket, the thickness direction is given by direction Irom FACE 4 (1-2-3-4) to FACE 5 (5-6-7-8).
POISSONS RATIO
0.3
THERMAL EXP.
THERMAL EXP. COEF.
1.5e-5
OK (twice)
ELEMENTS ADD
SET
cover
lower_part
bolts
OK
NAME
steel
GEOMETRIC PROPERTIES
3-D
NEW
SOLID COMPOSITE/GASKET
THICKNESS DIRECTION
FACE 4 (1-2-3-4) TO FACE 5 (5-6-7-8)
OK
ELEMENTS ADD
SET
gasket
OK
NAME
thickness_direction
1112 Marc Users Guide: Part 2
CHAPTER 3.17
Contact
The automatic contact algorithm is used to describe the contact between the gasket and the metal parts oI the joint and
between the bolts and the cylinder head cover. Moreover, a contact symmetry surIace is used to take symmetry
conditions into account.
Since the Iinite element mesh oI the gasket is Iiner than the meshes oI the metal parts, the best results are obtained iI
the gasket touches the latter. ThereIore, the Iirst contact body consists oI the gasket elements. The second contact body
consists oI the lower part and the bolts and the third is the cover. The last contact body is the symmetry plane.
CONTACT
CONTACT BODIES
NEW
DEFORMABLE
OK
ELEMENTS ADD
gasket
NAME
gasket
NEW
DEFORMABLE
OK
ELEMENTS ADD
lower_part
bolts
NAME
lower_part
NEW
DEFORMABLE
OK
ELEMENTS ADD
cover
NAME
cover
NEW
SYMMETRY
OK
1113
CHAPTER 3.17
Gaskets
The gasket is glued to the metal parts and is not allowed to separate. Normal touching contact is used between the
cover and the bolts. To activate GLUE contact Ior the gaskets, a CONTACT TABLE is created.
Load Steps and Job Parameters
The job consists oI six loadcases, each with a total loadcase time oI one second. The Iirst two loadcases,
prestress_left_bolt and prestress_right_bolt, are dedicated to the prestressing oI the bolts. In the
prestress_left_bolt loadcase, the leIt bolt is pre-tensioned while the right bolt remains locked. OI the two
boundary conditions applied to the control node oI the leIt bolt, prestress_left_bolt and
lock_unlock_left_bolt, only the POINT LOAD prestress_left_bolt is active in this loadcase. The FIXED
DISPLACEMENT lock_unlock_left_bolt is deactivated. Conversely, oI the two boundary conditions applied to
the control node oI the right bolt, lock_unlock_right_bolt and prestress_right_bolt, only the FIXED
DISPLACEMENT lock_unlock_right_bolt is active and the POINT LOAD prestress_right_bolt is
deactivated. The Iixed stepping procedure is employed in this loadcase using ten increments.
SURFACES ADD
1
END LIST
NAME
symmetry_plane
CONTACT
CONTACT TABLES
NEW
PROPERTIES
1 2
CONTACT TYPE: GLUE
1 3
CONTACT TYPE: GLUE
1 4
CONTACT TYPE: TOUCHING
2 3
CONTACT TYPE: TOUCHING
2 4
CONTACT TYPE: TOUCHING
3 4
CONTACT TYPE: TOUCHING
1114 Marc Users Guide: Part 2
CHAPTER 3.17
The prestress_right_bolt loadcase is identical to the prestress_left_bolt loadcase, except that the
lock_unlock_left_bolt and prestress_right_bolt boundary conditions are active, while
prestress_left_bolt and lock_unlock_right_bolt are deactivated.
In the next three loadcases, loading, cooling and unloading, the thermo-mechanical loading cycle is simulated
and in the Iinal loadcase, disassemble, the joint is disassembled. Since the bolts are locked (in the loading cycle) or
loosened (in the disassemble loadcase), only the FIXED DISPLACEMENT boundary conditions on the control nodes
oI the bolts are active. The POINT LOADS are deactivated. The Iixed stepping procedure is also employed Ior these
Iour loadcases, now using Iive increments per loadcase. The button sequence Ior the loading loadcase reads:
LOADCASES
MECHANICAL
NEW
STATIC
LOADS
deactivate:
lock_unlock_left_bolt
prestress_right_bolt
OK
STEPPING PROCEDURE
CONSTANT TIME STEP
# STEPS
10
OK
NAME
prestress_left_bolt
NEW
STATIC
LOADS
deactivate:
prestress_left_bolt
prestress_right_bolt
OK
STEPPING PROCEDURE
CONSTANT TIME STEP
# STEPS
1115
CHAPTER 3.17
Gaskets
The remaining three loadcases are similar.
The analysis is a normal mechanical analysis in which all six loadcases are preIormed in sequence. The
lock_unlock_left_bolt and prestress_right_bolt initial loads are deactivated to make sure that the active
initial loads match those oI the Iirst loadcase. The contact table is activated and the LARGE DISPLACEMENT option
and the ASSUMED STRAIN Iormulation are selected. The latter is used to improve the bending behavior oI the lower
order solid elements.
Three new scalar quantities are available Ior postprocessing the gaskets: Gasket Pressure (Marc post code 241),
Gasket Closure (Marc post code 242), and Plastic Gasket Closure (Marc post code 243). All three are selected in this
example, as well as, the Equivalent Von Mises Stress (see Figure 3.17-5).
5
OK
NAME
loading
JOBS
NEW
MECHANICAL
LOADCASES
activate:
prestress_left_bolt
prestress_right_bolt
loading
cooling
unloading
disassemble
INITIAL LOADS
deactivate lock_unlock_left_bolt
deactivate prestress_right_bolt
OK
1116 Marc Users Guide: Part 2
CHAPTER 3.17
Figure 3.17-5 JOB RESULTS Menu and the Selected Quantities for Postprocessing
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK (twice)
ANALYSIS OPTIONS
LARGE DISPLACEMENT
ADVANCED OPTIONS
ASSUMED STRAIN
OK (twice)
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Von Mises Stress
Gasket Pressure
Gasket Closure
Plastic Gasket Closure
OK (twice)
1117
CHAPTER 3.17
Gaskets
For the metal parts oI the model, element type 7 is being used. For the gasket, element type 149 is selected.
Save Model, Run Job, and View Results
ELEMENT TYPES
MECHANICAL
3-D SOLID
7
SET
cover
7
SET
lower_part
7
SET
bolts
149
SET
gasket
OK
FILE
SAVE AS
gasket.mud
OK
RETURN
JOBS
RUN
SUBMIT 1
MONITOR
OK
RETURN
RESULTS
OPEN DEFAULT
1118 Marc Users Guide: Part 2
CHAPTER 3.17
To monitor the pressure distribution on the gasket throughout the analysis, select the gasket elements and make them
visible. Switch oI drawing oI the nodes and isolate the gasket ring elements. Make a contour plot oI the gasket pressure,
set the range and the legend, and monitor the results:
Figure 3.17-6 shows a contour plot oI the gasket pressure distribution at the end oI the third loadcase when the joint
has been Iastened, the temperature has been increased to 180C and the interior pressure has been applied. In
SELECT
SELECT SET
gasket
MAKE VISIBLE
RETURN
PLOT
DRAW
switch off NODES
RETURN
MORE
ISOLATE ELEMENTS
SET
gasket_ring
OK
RETURN
SCALAR PLOT SETTINGS
RANGE
MANUAL
LIMITS
-2
56
# LEVELS
29
LEGEND
INTEGER
RETURN (twice)
SCALAR
Gasket Pressure
CONTOUR BANDS
MONITOR
1119
CHAPTER 3.17
Gaskets
Figure 3.17-7, the plastic gasket closure distribution at the end oI the analysis is depicted. It can be observed Irom both
pictures that, due to the asymmetric Iastening sequence oI the bolts, the plastic deIormation oI the gasket is also
slightly asymmetric.
Figure 3.17-6 Contour Plot of the Gasket Pressure at the end of the Third Loadcase
Figure 3.17-7 Contour Plot of the Plastic Gasket Closure at the end of the Analysis
1120 Marc Users Guide: Part 2
CHAPTER 3.17
In order to compare the response oI the gasket ring with the loading and unloading paths that were read in Irom the
data Iiles, select the nodes where the plastic gasket closure assumes its maximum, create a history plot oI the gasket
pressure versus the gasket closure oI those nodes and copy the plot to the generalized XY plotter Ior comparison.
The resulting plot is displayed in Figure 3.17-8. It shows that when the gasket is being loaded, the response oI the ring
closely Iollows the loading path and that upon unloading, the unloading path is interpolated between the loading path
and the supplied unloading path.
Finally, in Figure 3.17-9, the Iorces on the bolts are depicted and in Figure 3.17-10, the deIormed shape oI the cover
at the end oI the thermo-mechanical loading cycle is shown. In the latter picture, the displacements are enlarge by a
Iactor oI 25.
RESULTS
TOOLS
SELECT BY EXTREMES
NODES MAXIMUM
RETURN
HISTORY PLOT
SET NODES
ALL: SELECT.
COLLECT DATA
0 40 1
NODES/VARIABLES
ADD VARIABLE
Gasket Closure
Gasket Pressure
RETURN
> XY
UTILS
GENERALIZED XY PLOT
FIT
1121
CHAPTER 3.17
Gaskets
Figure 3.17-8 Pressure-closure History of the Gasket Ring at the Nodes where the Plastic
Closure assumes its Maximum
Figure 3.17-9 History Plot of the Bolt Forces
1122 Marc Users Guide: Part 2
CHAPTER 3.17
Figure 3.17-10 Deformation of the Cover (enlarged 25 times) at the end of the Thermo-mechanical
Loading Cycle
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
gasket.proc Mentat procedure Iile to run the above example
gasket_csect.proc Mentat procedure Iile to run the above example
body_loading.raw Mentat procedure Iile to run the above example
body_unloading.raw Mentat procedure Iile to run the above example
ring_loading.raw Mentat procedure Iile to run the above example
ring_unloading.raw Mentat procedure Iile to run the above example
Chapter 3.18: Cantilever Beam
3.18
Cantilever Beam

Summary 1124

Detailed Session Description of Cantilever Beam 1126

Add Plasticity to Cantilever Beam 1130

Run Job and View Results 1133

Input Files 1135

1124 Marc Users Guide: Part 2
CHAPTER 3.18
Summary
Title Cantilever Beam
Problem Ieatures Elastic-plastic solution oI a cantilever beam with a tip load
Geometry
Material properties E 10x10
6
Psi, 0.3, 0.283/386 lbI s
2
/in
4
and workhardening
Analysis type Static with elastic-plastic material behavior
Boundary conditions LeIt end Iixed, right end point load
Element type Plane stress element type 3
FE results Tracking oI stress versus plastic strain behavior
10" X 1" X 1"
500 lbf
v
1
2
3
Strength
3
0
3.05 0
Y (x10000)
True Plastic Strain (x.1)
40
30
50
20
10
0
Equivalent Von Mises Stress Node 4
-1.247e-003
1.624e-002
3.372e-002
5.121e-002
6.869e-002
8.618e-002
1.037e-001
1.211e-001
1.386e-001
1.561e-001
1.736e-001
lcase1
Total Equivalent Plastic Strain
Inc: 50
Time: 1.000e+000
X
Y
Z
1
1125
CHAPTER 3.18
Cantilever Beam
This example session describes the simulation oI loading a cantilever beam with a tip load. This model will be used
later in Chapter 3.35 Ior dynamics and will be saved.
The linear elastic solution is Iound. The bending stresses and tip displacements are then compared to theory.
The material properties are changed to include plasticity with workhardening.
The beam is then loaded with a larger load oI 1500 pounds in 50 equal load steps. We will see how every integration
point must track the material`s constitutive relation.
Figure 3.18-1 Cantilever Beam Problem Description
10" X 1" X 1"
500 lbf
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
0
5000
10000
15000
20000
25000
30000
0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035
0
5000
10000
15000
20000
25000
30000
E = 3.0x10
6
1

0 20,000
0.109 25,000
0.305 30,000
Strength (True Stress)
True Plastic Strain =
p
True Total Strain =
p
+ /E
Strength (True Stress)
-1.247e-003
1.624e-002
3.372e-002
5.121e-002
6.869e-002
8.618e-002
1.037e-001
1.211e-001
1.386e-001
1.561e-001
1.736e-001
lcase1
Total Equivalent Plastic Strain
Inc: 50
Time: 1.000e+000
X
Y
Z
1
1
2
3
Strength
3
0
3.05 0
Y (x10000)
True Plastic Strain (x.1)
40
30
50
20
10
0
Equivalent Von Mises Stress Node 4
1126 Marc Users Guide: Part 2
CHAPTER 3.18
Detailed Session Description of Cantilever Beam
Here is an example oI a cantilever beam below. Later, we will look at its dynamic behavior and the model created here
will be used later.
Figure 3.18-2 Cantilever and Beam Descriptions
FILES
NEW
OK
SAVE AS
beam1
MAIN
MESH GENERATION
NODE ADD
0 0 0
10 0 0
10 1 0
0 1 0
FILL
ELEMENT ADD (Pick above nodes in CCW)
SUBDIVIDE
DIVISIONS
10 4 1
ELEMENTS
ALL: EXISTING
RETURN
SWEEP
ALL & RETURN
RENUMBER
ALL & RETURN
MAIN
10" X 1" X 1"
500 lbf
1127
CHAPTER 3.18
Cantilever Beam
Figure 3.18-3 Cantilever Beam Mesh
Figure 3.18-4 Loads and Boundary Conditions
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
ON X DISPLCEMENT
OK
NODES ADD
Select all nodes on left edge
END LIST
NEW
FIXED DISPLACEMENT
ON Y DISPLACEMENT
OK
NODES ADD
Select bottom node on left edge
END LIST
NEW
POINT LOAD
ON Y FORCE -500
OK
NODES ADD
Select top right node
END LIST
RETURN
ID BOUNDARY CONDITIONS
MAIN
apply1
apply2
apply3
X
Y
Z
1128 Marc Users Guide: Part 2
CHAPTER 3.18
Figure 3.18-5 Material Properties: Isotropic Properties
MATERIAL PROPERTIES (twice)
NEW
STANDARD
STRUCTURAL
E = 3E7
= .3, OK
GENERAL
= .283/386, OK
ELEMENTS ADD
ALL: EXISTING
MAIN
GEOMETRIC PROP.
PLANAR
PLANE STRESS
THICKNESS = 1
ASSUMED STRAIN
OK
ELEMENTS ADD
ALL: EXISTING, MAIN
u

1129
CHAPTER 3.18
Cantilever Beam
Figure 3.18-6 Flog Assumed Strain Formulation
Figure 3.18-7 Bending Stress Contours
JOBS
NEW
MECHANICAL
PROPERTIES
PLANE STRESS
JOB RESULTS
TENSORS STRESS
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR
OK
OPEN POST FILE (RESULTS MENU)
SCALAR
COMP 11 OF STRESS
CONTOUR BANDS
-2.867e+004
-2.295e+004
-1.723e+004
-1.151e+004
-5.788e+003
-6.659e+001
5.655e+003
1.138e+004
1.710e+004
2.282e+004
2.854e+004
job1
Comp 11 of Stress
Inc: 0
Time: 0.000e+000
X
Y
Z
1
1130 Marc Users Guide: Part 2
CHAPTER 3.18
Figure 3.18-8 Y-Displacement Contours
Complete Modeling: Check Load
Peak Bending Stress /- 29Ksi, Max Disp 6.7e-2.
How does this compare to beam theory?
What can improve the results?
Add Plasticity to Cantilever Beam
Here is a cantilever beam. Let`s convert it to an elastic-plastic model and increase the load.
Figure 3.18-9 Beam Dimensions
SCALAR
DISPLACEMENT Y
OK
-6.692e-002
-6.022e-002
-5.352e-002
-4.683e-002
-4.013e-002
-3.343e-002
-2.674e-002
-2.004e-002
-1.334e-002
-6.645e-003
5.164e-005
job1
Displacement Y
Inc: 0
Time: 0.000e+000
X
Y
Z
1
apply1
apply2
apply3
X
Y
Z
10" X 1" X 1"
750 lbf
1131
CHAPTER 3.18
Cantilever Beam
Figure 3.18-10 Workhardening Data is True Stress Versus True Plastic Strain
FILES
OPEN
beam1
SAVE AS
beam1p
OK
RETURN
MATERIAL PROPERTIES (twice)
TABLES
NEW (1 Independent Variable)
TABLE TYPE:
eq_plastic_strain
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
0
5000
10000
15000
20000
25000
30000
0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035
0
5000
10000
15000
20000
25000
30000
E = 3.0x10
6
1

0 20,000
0.109 25,000
0.305 30,000
Strength (True Stress)
True Plastic Strain =
p
True Total Strain =
p
+ /E
Strength (True Stress)
1132 Marc Users Guide: Part 2
CHAPTER 3.18
POINT ADD
0.000 20E3
0.109 25E3
0.305 30E3
FIT
COPY TO GENERALIZED XY PLOTTER
MAIN
MATERIAL PROPERTIES (twice)
TABLE
NEW (1 Independent Variable)
TABLE TYPE:
TIME
FORMULA
ENTER
1.5*V1 (will ramp load from 0 to 750# in one second)
FIT
SHOW MODEL
RETURN
STRUCTURAL
PLASTICITY
INITIAL YIELD STRESS = 1.0
TABLE1 = table1
OK (twice)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply3 (point load)
OK
POINT LOAD
Y FORCE (pick table2, time)
OK
MAIN
LOADCASES
MECHANICAL
STATIC
1133
CHAPTER 3.18
Cantilever Beam
Run Job and View Results
OK
RETURN (twice)
JOBS
PROPERTIES
SELECT lcase1
ANALYSIS OPTIONS
LARGE STRAIN
OK
JOB RESULTS
EQUIVALENT VON MISES STRESS
TOTAL EQUIVALENT PLASTIC STRAIN
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR
OK
OPEN POST FILE (RESULTS MENU)
SCALAR Total Equivalent Plastic Strain
CONTOUR BANDS
DEF & ORIG
LAST
CONTOUR BANDS
1134 Marc Users Guide: Part 2
CHAPTER 3.18
Figure 3.18-11 Plastic Strain Contours on Deformed Shape
RESULTS
HISTORY PLOT
SET LOCATIONS (pick top left node)
END LIST
ALL INCS
ADD CURVES
ALL LOCATIONS
Total Equivalent Plastic Strain
Equivalent Von Mises Stress
FIT
RETURN
COPY TO GENERALIZED XY PLOTTER
UTILS
GENERALIED XY PLOT
FIT
-1.247e-003
1.624e-002
3.372e-002
5.121e-002
6.869e-002
8.618e-002
1.037e-001
1.211e-001
1.386e-001
1.561e-001
1.736e-001
lcase1
Total Equivalent Plastic Strain
Inc: 50
Time: 1.000e+000
X
Y
Z
1
1135
CHAPTER 3.18
Cantilever Beam
This overlays the history plot oI the stress strain response oI this node with the stress-strain material behavior.
Remember that continuum mechanics requires that the continuum be in equilibrium and that every point must track
the constitutive relation.
Figure 3.18-12 Stress Strain Response(+) tracking the Constitutive Relation
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
s2.proc Mentat procedure Iile to run the above problem
1
2
3
Strength
3
0
3.05 0
Y (x10000)
True Plastic Strain (x.1)
40
30
50
20
10
0
Equivalent Von Mises Stress Node 4
1136 Marc Users Guide: Part 2
CHAPTER 3.18
Chapter 3.19: Creep of a Tube
3.19
Creep of a Tube

Summary 1138

Detailed Session Description of Oval Tube 1140

Run Job and View Results 1144

Input Files 1154

1138 Marc Users Guide: Part 2
CHAPTER 3.19
Summary
Title Creep oI a Tube
Problem Ieatures Creep material model with and without a closed cavity
Geometry
Material properties E 21.4x10
6
Psi, 0.3 Creep law
Analysis type Static with automatic load increments
Boundary conditions Symmetry with internal pressure
Element type Plane strain element type 11
FE results Large dimensional changes in tube shape as creep continues
apply1
apply2
apply3
X
Y
Z
v c

c
4x10
24
o
4.51

lcase1
Total Equivalent Creep Strain
-4.539e-005
1.688e-002
3.381e-002
5.073e-002
6.766e-002
8.458e-002
1.015e-001
1.184e-001
1.354e-001
1.523e-001
1.692e-001
Inc: 1342
Time: 3.470e+006
X
Y
Z
Node 80
1139
CHAPTER 3.19
Creep of a Tube
A stainless steel oval tube is pressurized at a uniIorm high temperature and over time will creep. Only halI oI the tube
is modeled due to symmetry.
The material constitutive behavior has the creep strain rate dependent upon the stress level (Norton creep). The
material data has been Iitted with a power relation where the creep strain rate becomes: .
The oval tube will bulge and become a completely circular tube over time. The tube Iinally ruptures due to the large
strains.
Plotting the displacement oI the bulge versus time shows a quick growth Iollowed by a slower growth, because the
stresses drop with time.
A more complex constitutive relation may be easily modeled with the CRPLAW user subroutine.
Figure 3.19-1 Creep of a Tube Problem Description
c

c
4x10
24
o
4.51

apply1
apply2
apply3
X
Y
Z
lcase1
Total Equivalent Creep Strain
-4.539e-005
1.688e-002
3.381e-002
5.073e-002
6.766e-002
8.458e-002
1.015e-001
1.184e-001
1.354e-001
1.523e-001
1.692e-001
Inc: 1342
Time: 3.470e+006
X
Y
Z
Node 80
4.459
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Time (x1e6)
Displacement X Node 80 (x.1)
0 3.47
0.054
1100
100
0
1
1.564
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Total Equivalent Creep Strain Node 80 (x.01)
Equivalent Von Mises Stress Node 80 (x10000)
0 8.23
0.097
1100
100
0
1
1140 Marc Users Guide: Part 2
CHAPTER 3.19
Detailed Session Description of Oval Tube
FILES
NEW
OK
SAVE AS
creep
MAIN
MESH GENERATION
COORDINATE SYSTEM:
SET
GRID (ON)
U DOMAIN
0 1
U SPACING
0.065
V DOMAIN
-1 1
V SPACING
0.065
FILL
RETURN
CURVE TYPE ARC
CENTER/PT/PT
RETURN
CURVES:
ADD (arcs shown)
CURVE TYPE
LINE
RETURN
CURVES:
ADD(lines shown)
GRID (OFF)
SURFACE TYPE
RULED
RETURN
SURFACES ADD (pick interior and opposite
exterior arcs and lines)
CONVERT
DIVISONS
15 4
u=1
v=1
X
Y
Z
1141
CHAPTER 3.19
Creep of a Tube
SURF. TO ELEMS (pick largest surface)
DIVISIONS
10 4
SURF. TO ELEMS
(pick smallest surface)
RETURN
SYMMETRY
NORMAL 0 1 0
ELEMENTS
ALL: EXISTING
RETURN
CHECK
UPSIDE DOWN
FLIP ELEMENTS
ALL: SELECTED
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
MAIN
v=1
u=1
X
Y
Z
S
y
m
m
e
t
r
y
1142 Marc Users Guide: Part 2
CHAPTER 3.19
INITIAL CONDITIONS
STATE VARIABLES
NODAL TEMPERATURE
TEMPERATURE
1660, OK
NODES ADD
ALL: EXISTING
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISP. X=0, RETURN
NODES: ADD
all on x=0 axis, END LIST
NEW
FIX Y=0
NODES: ADD
at line of symmetry y=0
RETURN
NEW
EDGE LOAD
PRESSURE
66, OK
SELECT
METHOD PATH, OK
EDGES (pick node path on interior)
RETURN
EDGES: ADD
ALL: SELECTED
RETURN
STATE VARIABLES
NEW
NODAL TEMPERATURE
TEMPERATURE
1600
OK
NODES ADD
ALL: EXISTING
MAIN
MATERIAL PROPERTIES (twice)
NEW STANDARD
apply1
apply2
apply3
X
Y
Z
1143
CHAPTER 3.19
Creep of a Tube
STRUCTURAL
YOUNGS MODULUS = 21.4E6
POISSONS MODULUS = .3
CREEP (twice)
COEFFICIENT (define the creep strain rate)
4E-24
STRESS DEPENDENCE EXPONENT
4.51
OK (twice)
ELEMENTS ADD
ALL: EXISTING
MAIN
GEOMETRIC PROPERTIES
PLANAR
PLANE STRAIN
THICKNESS
1
CONSTANT DILATATION (on)
ASSUMED STRAIN (on)
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
LOADCASES
MECHANICAL
CREEP
TOTAL LOADCASE TIME
3.47E6
CREEP STRAIN/STRESS PARAMETERS
INITIAL TIME STEP
1
MAX. # INCS
2000
STRESS CHANGE TOLERANCE
1
OK (twice)
MAIN
1144 Marc Users Guide: Part 2
CHAPTER 3.19
Run Job and View Results
JOBS
NEW MECHANICAL
PROPERTIES
lcase1
PLANE STRAIN
ANALYSIS OPTIONS
ADVANCED OPTIONS
LARGE ROTATIONS, OK
FOLLOW FORCE
OK
JOB RESULTS
Equivalent Von Mises Stress
Total Equivalent Creep Strain
Temperature (Integration Point)
OK (twice)
SAVE
RUN
SUBMIT(1)
OK
RETURN
RESULTS
OPEN DEFAULT
DEF & ORIG
CONTOUR BANDS
SCALAR
Total Equiv. Creep Strain
OK
LAST
HISTORY PLOT
SET LOCATIONS
n:80
END LIST
ALL INCS
ADD CURVES
lcase1
Total Equivalent Creep Strain
-4.539e-005
1.688e-002
3.381e-002
5.073e-002
6.766e-002
8.458e-002
1.015e-001
1.184e-001
1.354e-001
1.523e-001
1.692e-001
Inc: 1342
Time: 3.470e+006
X
Y
Z
Node 80
Figure 3.19-2 Analysis Set Nodes at 80
1145
CHAPTER 3.19
Creep of a Tube
ALL LOCATIONS
Time
Displacement X
FIT
RETURN
CLEAR CURVES
ADD CURVES
ALL LOCATIONS
Total Equiv. Creep Strain
Equiv. Von Mises Stress
FIT
RETURN
1146 Marc Users Guide: Part 2
CHAPTER 3.19
Figure 3.19-3 History Plots for Time and Total Equivalent Creep Strain
What can improve the results?
Clearly as the tube creeps, the volume inside the tube increases. Assuming a constant mass oI air in the tube, increasing
the volume decreases the pressure the creep deIormation is reduced. To simulate this eIIect, we can model the cavity
oI air inside the tube. This cavity monitors the volume and adjusts the pressure according to the ideal gas law.
FILES
OPEN
4.459
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Time (x1e6)
Displacement X Node 80 (x.1)
0 3.47
0.054
1100
100
0
1
1.564
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Total Equivalent Creep Strain Node 80 (x.01)
Equivalent Von Mises Stress Node 80 (x10000)
0 8.23
0.097
1100
100
0
1
1147
CHAPTER 3.19
Creep of a Tube
creep
OK
SAVE AS
creep2
OK, MAIN
MESH GENERATION
ELEM. CLASS LINE(2)
ELEMS ADD
(pick interior nodes)
N1, N3
RETURN
MODELING TOOLS
CAVITIES
NEW
SELECT
METHOD PATH
RETURN
EDGES
(pick interior nodes)
N1, N2, N3
END LIST
RETURN
EDGES ADD
ALL: SELECTED
REF. PRESSURE
15
REF. TEMPERATURE
1660
REF. DENSITY
1.8E-5
MAIN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply4
MORE
N1
N3
N2
1148 Marc Users Guide: Part 2
CHAPTER 3.19
CAVITY PRESSURE LOAD
PRESSURE
66
OK
CAVITIES ADD
cavity1, OK
NEW
CAVITY MASS LOAD
MASS
CLOSED CAVITY
OK
CAVITIES ADD
cavity1, OK
MAIN
MESH GENERATION
CHECK
FLIP ELEMENTS
(pick element added to close the
cavity properly. Make sure that all
arrows point from inside to outside
of cavity)
MAIN
LOADCASES
MECHANICAL
CREEP
LOADS
apply3 (off)
apply4 (off)
apply5 (on)
OK (twice)
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
ANALYSIS DIMENSION
X
Y
Z
1149
CHAPTER 3.19
Creep of a Tube
Results
PLANAR
MISCELLANEOUS
171
(pick element added)
OK (twice)
PROPERTIES
INITIAL LOADS
apply3 (off)
apply4 (on)
apply5 (off)
OK (twice)
JOB PARAMETERS
CAVITY PARAMETERS
AMBIENT PRESSURE = 0 (the default value)
OK (thrice)
RUN
STYLE TABLE-DRIVEN -> OLD
SUBMIT(1)
MONITOR
OK, MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
CONTOUR BANDS
SCALAR
Total Equiv. Creep Strain
LAST
HISTORY PLOT
SET LOCATIONS
n:80
END LIST
1150 Marc Users Guide: Part 2
CHAPTER 3.19
Figure 3.19-4 Analysis at Node 80
ALL INCS
ADD CURVES
ALL LOCATIONS
Time
Displacement X
FIT, RETURN
-3.845e-006
1.458e-002
2.916e-002
4.375e-002
5.833e-002
7.292e-002
8.750e-002
1.021e-001
1.167e-001
1.313e-001
1.458e-001
lcase1
Total Equivalent Creep Strain
Inc: 1344
Time: 3.470e+006
X
Y
Z
1
1151
CHAPTER 3.19
Creep of a Tube
Figure 3.19-5 History Plot of Displacement x Node 80
COPY TO GENERALIZED XY PLOTTER
SHOW HISTORY
RETURN
CLEAR CURVES
ADD CURVES
ALL LOCATIONS
Total Equiv. Creep Strain
Equiv. Von Mises Stress
FIT
3.938
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Time (x1e6)
Displacement X Node 80 (x.1)
0 3.47
0.054
1100
100
0
1
1152 Marc Users Guide: Part 2
CHAPTER 3.19
Figure 3.19-6 Total Equivalent Creep Strain at Node 80
CLEAR CURVES
ADD CURVES
GLOBAL
Volume Cavity 1
Pressure Cavity 1
FIT
1.564
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Total Equivalent Creep Strain Node 80 (x.01)
Equivalent Von Mises Stress Node 80 (x10000)
0 7.769
0.155
1100
100
0
1
1153
CHAPTER 3.19
Creep of a Tube
Figure 3.19-7 Cavity Pressure verses Cavity Volume at Node 80
Figure 3.19-8 X-Displacement History with and without the Closed Cavity Feature
CLEAR CURVES
ALL LOCATIONS
Time
Displacement X
FIT
COPY TO GENERALIZED XY PLOTTER
FIT
6.6
200
300
400
500
600
700
800
900
1000
1200
1300
lcase1
Volume Cavity 1 (x.1)
Pressure Cavity 1 (x10)
3.438 6.38
3.557
1100
100
0
1
P = 22.659V
-1.003
R = 1
200
300
400
500
600
700
800
1000
1100
1300
Time (x1e6)
Y (x.1)
0 3.47
0.054
4.459
Displacement X Node 80
1200
100
0
900
1300
Displacement X Node 80
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1
Closed Cavity
1154 Marc Users Guide: Part 2
CHAPTER 3.19
Clearly, the reduction in pressure due to the increase in volume reduced the creep deIormation oI the tube, which is a
more realistic simulation.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
s6.proc Mentat procedure Iile to run the above problem
creep.mud Associated Mentat model Iile without cavity
creep2.mud Associated Mentat model Iile with cavity
Chapter 3.20: Tensile Specimen
3.20
Tensile Specimen

Summary 1156

Detailed Description Session 1157

Run Job and View Results 1164

Modeling Tips 1175

Input Files 1179

1156 Marc Users Guide: Part 2
CHAPTER 3.20
Summary
Title Tensile specimen
Problem Ieatures Simple dog bone uniaxial specimen illustrating several meshing techniques (overlay,
advancing Iront, and mapped methods) and the eIIect oI extending the gage section.
Mentat Ieatures include copy to clipboard and report writer.
Geometry
Material properties E 10x10
6
Psi, 0.3 and Orthotropic
Analysis type Static with elastic material behavior
Boundary conditions LeIt end Iixed, right edge load
Element type Plane stress element type 3
FE results UniIormity oI axial stress original and extended gage section
X
Y
Z
1.0 in
0.9 in
R = 21
R = 21
v
2.120e+004
2.473e+004
2.827e+004
3.181e+004
3.534e+004
3.888e+004
4.242e+004
4.595e+004
4.949e+004
5.303e+004
5.657e+004
Inc: 0
Time: 0.000e+000
1157
CHAPTER 3.20
Tensile Specimen
This example session describes the simulation oI the loading oI a dog-bone tensile specimen. This session builds the
geometry, exports an IGES Iile, and demonstrates diIIerent types oI meshing strategies including: overlay, advancing
Iront, and mapped meshing.
Using the mapped mesh, the tensile specimen is subjected to an axial load and submitted to Marc. Then, Mentat post
processes the results oI the tensile specimen.
AIter the Iirst run, the specimen`s gage section is changed and re-run to compare with the original specimen.
Finally, the material is changed Irom an isotropic to an orthotropic material. The material direction does not line up
with the pull direction, and the deIormed shape becomes skewed.
Figure 3.20-1 Examples of Meshing
Detailed Description Session
Tensile Specimen Analysis
Begin this session at the main menu.
MESH GENERATION
COORDINATE SYSTEM: SET
2.120e+004
2.473e+004
2.827e+004
3.181e+004
3.534e+004
3.888e+004
4.242e+004
4.595e+004
4.949e+004
5.303e+004
5.657e+004
job1
Comp 11 of Stress
Inc: 0
Time: 0.000e+000
X
Y
Z
apply1
apply2
apply3
Skewed Orthotropic
ID Boundary Conditions
Overlay Mesh
Mapped Mesh
Advancing Front Mesh
1158 Marc Users Guide: Part 2
CHAPTER 3.20
GRID ON
U DOMAIN
-1.5 1.5
V DOMAIN
-1.5 1.5
FILL
RETURN
CURVE TYPE
ARCS:
CENTER/POINT/ANGLE
RETURN
CURVES ADD
0 1.5 0
0 -1.5 0
-21 (degrees)
MOVE
TRANSLATIONS
0 1.75 0
CURVES (use left mouse to pick curve, right mouse will END LIST)
RETURN
1159
CHAPTER 3.20
Tensile Specimen
Select pairs oI points beginning at the upper leIt oI the top arc and move counter-clockwise to complete the boundary
oI the model.
SYMMETRY
NORMAL
0 1 0
CURVES select the arc
END LIST
NORMAL
1 0 0
CURVES
ALL: EXISTING
RETURN
COORDINATE SYSTEM: GRID OFF
RETURN
DUPLICATE
TRANSLATIONS
0.425 0 0
POINTS (select two right most points)
END LIST
TRANSLATIONS
-0.425 0 0
POINTS (select two left most points)
END LIST
RETURN
CURVE TYPE
LINE
RETURN
CRVS ADD
1160 Marc Users Guide: Part 2
CHAPTER 3.20
Use the Iollowing steps to save this geometry in an IGES Iile.
The next section shows how to mesh the geometry several ways.
Overlay Technique
FILES
EXPORT
IGES
ten.spec.iges
OK
RETURN
SAVE AS
isotropic
OK
MAIN
MESH GENERATION
AUTOMESH
2D PLANAR MESHING
DIVISIONS
20 20
QUADRILATERALS (OVERLAY): QUAD MESH!
ALL: EXISTING
UNDO
DIVISIONS
40 40
QUADRILATERALS (OVERLAY): QUAD MESH!
ALL: EXISTING
RETURN
UNDO (this will undo your last command)
1161
CHAPTER 3.20
Tensile Specimen
Advancing Front Technique
Mapped Meshing Technique
AUTOMESH 2D PLANAR
QUADRILATERALS (ADV FRNT): QUAD MESH!
ALL: EXISTING
UNDO
RETURN
CURVE DIVISIONS
FIXED AVG LENGTH (ON)
FORCE EVEN DIV
APPLY CURVE DIVISIONS
ALL: EXISTING
RETURN
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT): QUAD MESH!
ALL: EXISTING
UNDO
RETURN
CURVE DIVISIONS
CLEAR CURVE DIVISIONS
ALL: EXISTING
RETURN (twice)
SURFACE TYPE
RULED RETURN
SRFS ADD (pick top left/bottom arcs)
UNDO
1162 Marc Users Guide: Part 2
CHAPTER 3.20
CHECK
FLIP CURVES (pick all top lines & curves)
RETURN
SRFS ADD (pick right top/bottom line)
SRFS ADD (pick left top/bottom line)
SRFS ADD (pick left top/bottom curve)
SRFS ADD (pick right top/bottom curve)
CONVERT
SURFACES TO ELEMENTS (pick left and right curved surfaces)
DIVISIONS
5 10
SURFACES TO ELEMENTS (pick left and right rectangular surfaces)
RETURN
SWEEP
ALL & RETURN
RENUMBER
ALL & RETURN
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
ON DISPLACEMENT X = 0
OK
NODES ADD (select all nodes on left edge)
END LIST
NEW
1163
CHAPTER 3.20
Tensile Specimen
FIXED DISPLACEMENT
ON DISPLACEMENT Y = 0
OK
NODES ADD (select center node on left edge)
END LIST
NEW
EDGE LOAD
ON PRESSURE
-30000
OK
EDGES ADD (select all edges on right edge)
END LIST
RETURN
ID BOUNDARY CONDITIONS
ARROW PLOT SETTINGS
SOLID
RETURN
DRAW
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
NEW
STANDARD
STRUCTURAL
YOUNGS MODULUS = 1E7
POISSONS RATIO = .3
OK
ELEMENTS ADD
ALL: EXISTING
apply1
apply2
apply3
1164 Marc Users Guide: Part 2
CHAPTER 3.20
Run Job and View Results
ID MATERIALS
MAIN
GEOMETRIC PROPERTIES
PLANAR
PLANE STRESS
THICKNESS
0.25
ASSUMED STRAIN (This improves the elements behavior in bending.)
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
JOBS
PROPERTIES
PLANE STRESS
ANALYSIS OPTIONS
LARGE STRAIN
OK
MECHANICAL ANALYSIS CLASS JOB
RESULTS
AVAILABLE ELEMENT TENSORS STRESS
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR (some elements upside/down)
OK
material1
1165
CHAPTER 3.20
Tensile Specimen
Go back to RUN and resubmit. See Figure 3.20-2.
Figure 3.20-2 Run Job Menu. Exit Number 3004 is a successful completion.
MAIN
MESH GENERATION
CHECK UPSIDE DOWN
FLIP ELEMENTS
ALL: SELECTED
UPSIDE DOWN
Number of upside/down elements should now be 0
RETURN (twice)
JOBS
SAVE
RUN
SUBMIT1
MONITOR
OK
1166 Marc Users Guide: Part 2
CHAPTER 3.20
Is the job complete? Exit number 3004 is a successIul completion oI the simulation where the constitutive behavior
oI the material was Iollowed and equilibrium was satisIied, click the EXIT MESSAGE button to display exit message.
Did it do what I expect? However, just because equilibrium is satisIied, the simulation may not accomplish its
objective, Ior example we expect that the net section tensile stress should be:
yet in Figure 3.20-3 the maximum stress in the net section is 56, 570 psi and is higher than our average estimate above.
This suggests our net section has a bit oI stress concentration and needs to be extended to make the stresses as uniIorm
as possible in the net section where strain measurements occur. This extension will lower the net section tensile stress
nearer to the average estimated above (see Figure 3.20-8).
Figure 3.20-3 Analysis of Comp 11 of Stress
MAIN
RESULTS
OPEN DEFAULT
SCALAR
COMP 11 OF STRESS
OK
CONTOUR BANDS
PLOT
NODES (turn nodes off)
RETURN
DEFORMED SHAPE SETTINGS
OUTLINE (on)
RETURN
o p
A
end
A
mi d
-----------
\ .
| |
30x10
4
0.898517t
0.5t
------------------------
\ .
| |
53 911psi ,
2.120e+004
2.473e+004
2.827e+004
3.181e+004
3.534e+004
3.888e+004
4.242e+004
4.595e+004
4.949e+004
5.303e+004
5.657e+004
job1
Comp 11 of Stress
Inc: 0
Time: 0.000e+000
X
Y
Z
1
1167
CHAPTER 3.20
Tensile Specimen
Figure 3.20-4 (A) Example of Reaction Force (B) Example of External Force
RESULTS
MORE
VECTOR Pick Reaction Force
OK
VECTOR PLOT ON
VECTOR Pick External Force
OK
2.757e-004
7.017e+001
1.403e+002
2.105e+002
2.807e+002
3.509e+002
4.210e+002
4.912e+002
5.614e+002
6.316e+002
7.017e+002
job1
Reaction Force
Inc: 0
Time: 0.000e+000
X
Y
Z
0.000e+000
6.739e+001
1.348e+002
2.022e+002
2.696e+002
3.369e+002
4.043e+002
4.717e+002
5.391e+002
6.065e+002
6.739e+002
job1
External Force
Inc: 0
Time: 0.000e+000
X
Y
Z
(A)
(B)
1168 Marc Users Guide: Part 2
CHAPTER 3.20
PLOT
NODES (turn nodes on)
RETURN
DEFORMED SHAPE SETTINGS
SURFACE
RETURN
RESULTS
PATH PLOT
NODE PATH N1 N2
path from N1 to N2
END LIST
ADD CURVES
ADD CURVE
ARC LENGTH
COMP 11 OF STRESS
FIT
RETURN
Y MIN 0
N1
N2
1169
CHAPTER 3.20
Tensile Specimen
Figure 3.20-5 (A) Add Curve Analysis (B) Arc Length Analysis
RESULTS
PATH PLOT
TABLES COPY TO
> 1
FIT
V1 STEPS 100
F1 STEPS 100
INTEGRATE
FIT
Inc : 0
Comp 11 of Stress (x10000)
Arc Length (x.1)
job1
1
7
Time : 0
13
25
31
37
43
49
55
19
61
0
5.224
5
5.657
Inc : 0
Comp 11 of Stress (x10000)
Arc Length (x.1)
job1
1
7
Time : 0
13
25 31 37
43
49
55
19
61
0
0
5
5.657
(A)
(B)
N1
N2
W
1170 Marc Users Guide: Part 2
CHAPTER 3.20
Figure 3.20-6 Tables Menu
where: p 30000, W0.898518, t0.25
Tensile Specimen Uniform Gage Section
The previous stress analysis shows that the stress Iield is not uniIorm in the gage section. Redesign the specimen such
that it has a 1" constant gage section at the center.
MAIN
RESULTS
CLOSE, MAIN
FILES
SAVE AS
isotropic_long, OK
RESET PROGRAM
RETURN
MESH GENERATION
ATTACH
DETACH NODES
o
11
t y d
N1
N2
}
26926t 26826.8t ~ pWt 6739
1171
CHAPTER 3.20
Tensile Specimen
ALL: EXISTING
DETACH ELEMENTS
ALL: EXISTING
SELECT ELEMENTS (pick all elements to the right of the net section)
END LIST
ELEMENTS STORE
right
ALL:SELECTED
RETURN (twice)
SUBDIVIDE
DIVISIONS
1 1 1
ELEMENTS,
ALL:SELECTED
RETURN
MOVE
TRANSLATIONS
1 0 0
ELEMENTS
right
RETURN
SWEEP
REMOVE UNUSED NODES
ALL
FILL
RETURN
PLOT
CURVES OFF
SURFACES OFF
POINTS OFF
REGEN
1172 Marc Users Guide: Part 2
CHAPTER 3.20
Figure 3.20-7 Extension of Gage Section
RETURN to mesh generation
ELEMS ADD
N1, N2, N3, N4
SUBDIVIDE, DIVISIONS
10 10 1
ELEMENT (pick element just added)
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
ELEMENTS ADD
ALL: EXISTING
MAIN
GEOMETRIC PROPERTIES
PLANAR
PLANE STRESS
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
JOBS
RUN
SUBMIT(1)
N1 N2
N4 N3
1173
CHAPTER 3.20
Tensile Specimen
Figure 3.20-8 Axial Stress Contours
Tensile Specimen Composite Material
What about composites? Suppose we want to analyze an orthotropic material whose material axis does not line up with
the structure`s geometric axis.
OPEN POST FILE (RESULTS MENU)
SCALAR
Comp 11 Of Stress
OK
CONTOUR BANDS
FILES
OPEN
isotropic
SAVE AS
orthotropic
RETURN
MATERIAL PROPERTIES (twice)
STRUCTURAL
2.119e+004
2.465e+004
2.811e+004
3.157e+004
3.502e+004
3.848e+004
4.194e+004
4.540e+004
4.886e+004
5.231e+004
5.577e+004
job1
Comp 11 of Stress
Inc: 0
Time: 0.000e+000
X
Y
Z
1
1174 Marc Users Guide: Part 2
CHAPTER 3.20
Re-run and check results, deIormed shape is skewed.
TYPE ELASTIC-PLASTIC ORTHOTROPIC
E11 = 3E7, E22=E33=1E6
ALL s = .3
ALL GS = 5E5
OK
RETURN
ORIENTATIONS
NEW
EDGE41 ON
ANGLE 45
ADD ELEMENTS
ALL: EXISTING
ORIENTATION PLOT SETTINGS
ORIENTATION (on)
FIRST DIRECTION (only)
REGEN
SAVE
u
1175
CHAPTER 3.20
Tensile Specimen
Figure 3.20-9 Axial Stress Contours Orthotropic Material
Modeling Tips
It is very common in the testing oI materials to monitor the load - displacement response oI the specimen since the
testing machine typically records this inIormation. For our current models oI the specimen, this can be a bit awkward,
since we only have nodal Iorces and displacements that would need to be summed up appropriately. For example,
should we need to sum up the external Iorces on the right side oI our specimen, we can use a path plot to collect the
proper inIormation, then copy to the clipboard and place the inIormation into MicrosoIt Excel to sum up the loads; oI
course, this should be the applied pressure on the surIace times its area. Let`s do this Ior our isotropic_long model.
RESULTS
OPEN
isotropic_long_job1.t16, OK
PATH PLOT
NODE PATH
238 188 # (pick lower and upper nodes on right end)
ADD CURVES
ADD CURVE
Arc Length
External Force X
FIT
1.975e+004
2.577e+004
3.179e+004
3.780e+004
4.382e+004
4.984e+004
5.585e+004
6.187e+004
6.788e+004
7.390e+004
7.992e+004
job1
Comp 11 of Stress
Inc: 0
Time: 0.000e+000
X
Y
Z
1
1176 Marc Users Guide: Part 2
CHAPTER 3.20
This places the data into the clipboard and upon starting up Excel, one can paste into a worksheet and sum, sum Fx,
as indicated below. Also other calculations are included to check equilibrium by multiplying the surIace load times its
area. The equilibrium check is very good.
We can export results Irom Mentat to external Iiles by using the report writer, namely:
CLIPBOARD COPY TO
RETURN
MORE
REPORT WRITER
SELECT
METHOD PATH
SHORTCUTS
SHOW MODEL, OK
NODES
238 188 #
RETURN
SELECTION SELECTED ENTITIES
DATA NODAL VALUES
CREATE REPORT
OPEN REPORT
!"# $%&'()*+(%"&,- ./"#% 0
1 2234566
14175789 3:24777
14;:5:12 3:24777
14935888 3:24777
1428561: 3:24777
14665985 3:24777
14825;; 3:24777
14397539 3:24777
14:;77;6 3:24777
14717333 3:24777
147578;: 2234566
<=> .+ ? 3:27477 -@A
<("%<< ? 21111 B<C
,"%, ? 1499639598 C&D9
A/"#% ? 3:2747::8 -@A
1177
CHAPTER 3.20
Tensile Specimen
and the text Iile, isotropic_ling_job1.rpt is created and opened (right). Yet, iI we want to simulate the specimen as being
pulled in a test machine, these procedures would be cumbersome; we would want another way to automatically record
the total load. To accomplish this, we shall use the contact option and take advantage that all Iorces acting on a rigid
body are automatically summed up to a generalized Iorce (Iorces and moments) at the reIerence point on the rigid
body.
To model the specimen being pulled by a test machine, we shall replace the pressure surIace with a rigid body that will
pull the specimen using displacement control; Ior saIety reasons, testing laboratories use displacement control oI the
test machine, rather than load control.
FILES
OPEN
isotropic_long, OK
SAVE AS
isotropic_long_contact, OK
MAIN
MESH GENERATION
CRVS ADD
(pick lower and upper nodes on right end)
MOVE
RESET
SCALE FACTORS
1 1.2 1
CURVE
11 # (pick curve just added)
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE, OK
ELEMENTS ADD
ALL EXISTING
NEW
2-D CURVES ADD
11 #
FLIP CURVES
11 #
1178 Marc Users Guide: Part 2
CHAPTER 3.20
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE time, OK
ADD
0 0 1 1
RETURN
RIGID
POSITION PARAMETERS
X = 0.0121254, TABLE table1 (yields total grip load of 6739 lbf)
OK (twice)
MAIN
BOUNDARY CONDITIONS
EDIT apply3
REM
MAIN
LOADCASES
MECHANICAL
STATIC, OK
MAIN
JOBS
PROPERTIES
lcase1
CONTACT CONTROL
ADVANCED CONTACT CONTROL
SEPERATION FORCE = 1E11 (insures no seperaton)
OK (thrice)
RUN
SAVE MODEL
SUBMIT (1)
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
ALL INCS
ADD CURVES
1179
CHAPTER 3.20
Tensile Specimen
Figure 3.20-10 Load - Displacement Response of Specimen
Since our problem had no nonlinear behavior, the load - displacement response is a straight line ending at a total load
oI -6,739 lbI. The Iorce on the specimen is equal and opposite to the Iorce on the rigid body and the total Iorce agrees
with our other calculations.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
GLOBAL
Pos X cbody2
Force X cbody2
FIT
File Description
s1.proc Mentat procedure Iile to run the above problem
44
45
46
47
48
49
50
lcase1
Pos X cbody2 (x.01)
Force X cbody2 (x1000)
0 1.213
-6.739
0
43
42
41
40
20
21
22
23
24
25
26
27
28
29
31
32
33
34
35
36
37
38
39
30
19
18
16
0
1
2
3
4
5
7
8
9
10
11
12
13
14
15
6
17
1
1180 Marc Users Guide: Part 2
CHAPTER 3.20
Chapter 3.21: Rubber Elements and Material Models
3.21
Rubber Elements and
Material Models

Summary 1182

Lower-order Triangular Rubber Elements 1183

Tube with Friction 1198

Cavity Pressure 1200

Buckling of an Elastomeric Arch 1204

Comparison of Curve Fitting of Different Rubber Models 1212

Input Files 1219

1182 Marc Users Guide: Part 2
CHAPTER 3.21
Summary
Title Rubber elements and material models
Problem Ieatures Compression oI a rubber tube and snap through oI an elastic arch
Geometry
Material properties Elastomeric Materials: Ogden and Mooney
Analysis type Static with elastomeric material behavior
Boundary conditions As show above
Element type Plane stress quad and tri element types
FE results Force displacement response
v=1
apply1
X
Y
Z
1
44
45
46
47
48
49
50
Tube
Pos Y bottom (x.1)
Force Y top
0 4
0
3.181
43
42
41
40
20212223242526272829
313233
34
35
36
37
38
39
30
19 18 17 16
012345 789101112131415
6
1
38
39
40
41
42
43
Arch
Displacement Y Node 11 (x.1)
External Force Y Node 11 (x.01)
-6.774 0
-3
0
37
36
34
17 18
19
20
21
22
23
24
25
26
27 28
29
30
31
32
33 35
16
15
14
13
0
1
2
3
4
5
6 7
8
9
10
11
12
Open Air Cavity
Closed Air Cavity
1183
CHAPTER 3.21
Rubber Elements and Material Models
An example oI a compression oI a rubber tube using Ogden material using both quadrilateral and triangular elements
is presented Iirst. Curve Iitting oI Ogden coeIIicients is demonstrated next. The tube is then considered closed and
Iilled with air. A postbuckling simulation oI a rubber arch is then perIormed. Finally, curve Iitting based upon diIIerent
rubber models ins perIormed.
Lower-order Triangular Rubber Elements
Let`s start with the compression oI a rubber tube in two dimensions assuming plane strain. The tube is compressed by
two rigid bodies on the top and bottom and is modeled with one plane oI symmetry. Units used are inches, pounds,
and seconds.
Figure 3.21-1 Compressed Tube using Quadrilaterals and Triangles
Using Quadrilateral Elements
FILES
SAVE AS
elasto
RETURN
MESH GENERATION
COORDINATE SYSTEM SET:
GRID ON
U DOMAIN
-1.1 1
V DOMAIN
0.000e+000
6.500e+001
1.300e+002
1.950e+002
2.600e+002
3.250e+002
3.900e+002
4.550e+002
5.200e+002
5.850e+002
6.500e+002
Equivalent Von Mises Stress
Inc: 50
Time: 5.000e-001
X
Y
Z Equivalent Von Mises Stress
1184 Marc Users Guide: Part 2
CHAPTER 3.21
0 1
FILL
RETURN
CURVES: ADD
point( 1.0, 0.0, 0.0)
point(-1.1, 0.0, 0.0)
point(-1.1, 1.0, 0.0)
point( 1.0, 1.0, 0.0)
CURVE TYPE
CENTER/POINT/POINT
RETURN
CURVES: ADD
-1.0, 0.5, 0.0
-1.0, 0.0, 0.0
-1.0, 1.0, 0.0
-1.0, 0.5, 0.0
-1.0, 0.1, 0.0
-1.0, 0.9, 0.0
SURFACE TYPE
RULED, RETURN
SURFACES ADD:
4
3
1185
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-2 Tube Geometry
Figure 3.21-3 Quadrilateral Mesh
CONVERT
DIVISION
30 3
SURFACES TO ELEMENTS
all: EXIST.
RETURN
u=1
v=1
X
Y
Z
v=1
u=1
X
Y
Z
1186 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-4 Boundary Conditions
SWEEP
ALL
RETURN
CHECK
UPSIDE DOWN
FLIP ELEMENTS
ALL SELECTED
RETURN
RENUMBER
ALL
RETURN
MAIN
BOUNDARY CONDITIONS
MECHANICAL
DISPLACEMENT X
0
OK
ADD NODES (pick nodes along x=0)
MAIN
MATERIAL PROPERTIES (twice)
v=1
apply1
X
Y
Z
1
1187
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-5 Material Stress Strain Curve
NEW
STANDARD
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
experimental_data
ADD
0 0
.9 100
1.6 250
1.9 300
2.2 500
2.4 600
2.6 700
2.9 1000
FIT
FILLED
3
2
1
1
0
2.9 0
F (x1000)
V1
table1
4
5
7
8
6
1
1188 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-6 Curve Fit of Material Stress Strain Curve
NAME
tension
RETURN
EXPERIMENTAL DATA FIT
UNIAXIAL
tension
ELASTOMERS
OGDEN
UNIAXIAL
POSITIVE COEFFICIENTS (on)
MATHEMATICAL CHECKS (on)
COMPUTE
APPLY
OK
SCALE AXES (to scale the curve)
RETURN (twice)
uniaxial/ogden uniaxial/experiment
1
0
2.9
(x1000)
0
biaxial/ogden planar_shear/ogden
simple_shear/ogden
1
1189
CHAPTER 3.21
Rubber Elements and Material Models
This curve Iit oI the raw data has been applied to this material. It is important to have the other deIormation modes
(biaxial and planar shear) be similar to the tension Iit and not vary Iactors oI two or higher. OI course, it would be best
to have the biaxial and planar shear material data to make a combined mode Iit. More on this later.
Figure 3.21-7 Time Table to move Rigid Bodies
ELEMENTS ADD
all: EXIST.
SHOW MODEL
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE, OK
ELEMENTS ADD
all: EXIST.
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
OK (select OK button only if type time was typed in)
ADD
0 0 0.5 1 1 0
END LIST (#)
FILLED
FIT
table2
V1
F
0 1
0
1
2
3
1
1
1190 Marc Users Guide: Part 2
CHAPTER 3.21
SHOW TABLE
SHOW MODEL
RETURN
MAIN
CONTACT
CONTACT BODIES
NEW
DEFORMABLE
OK
ELEMENTS ADD
ALL: EXISTING
NEW
NAME
top
RIGID
POSITION
PARAMETERS
position (center of rotation) Y
-.4
TABLE
table2 (pick the time table)
OK
DISCRETE (on)
OK
2-D CURVES ADD
(pick top curve)
END LIST (#)
NEW
NAME
bottom
RIGID
POSITION
PARAMETERS
position (center of rotation) Y
.4
TABLE
1191
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-8 Identification of Contact Bodies
table2 (pick the time table)
OK
DISCRETE (on)
2-D CURVES ADD (pick bottom curve)
END LIST (#)
ID CONTACT
MAIN
LOADCASES
MECHANICAL
STATIC
SOLUTION CONTROL
MAX # RECYCLES
30
NON-POSITIVE DEFINITE
DEVIATORIC STRESS,
OK
CONVERGENCE TESTING
DISPLACEMENT
OK
TOTAL LOADCASE TIME
0.5
cbody1
top
bottom
none
X
Y
Z
1
1192 Marc Users Guide: Part 2
CHAPTER 3.21
fixed # OF STEPS
50
OK
COPY
MAIN
JOBS
NEW
MECHANICAL
PROPERTIES
lcase1
lcase 2
CONTACT CONTROL
ADVANCED CONTACT CONTROL
DISTANCE TOLERANCE BIAS
.5
OK (twice)
JOB RESULTS
available element scalars
Equivalent Cauchy Stress
OK (twice)
ELEMENT TYPES
MECHANICAL
ANALYSIS DIMENSION
PLANAR
SOLID
80 (Quad 4)
OK
all: EXIST.
RETURN
RUN
SUBMIT 1
MONITOR
OK
SAVE
RETURN
1193
CHAPTER 3.21
Rubber Elements and Material Models
Results
Figure 3.21-9 Equivalent Cauchy Stress Contours Quadrilateral Mesh
RESULTS
OPEN DEFAULT
SKIP TO INC
50
DEF & ORIG
CONTOUR BAND
SCALAR
Equivalent Cauchy Stress
HISTORY PLOT
SET LOCATIONS
n:1
END LIST (#)
INC RANGE
0 50 1
5.943e+000
7.177e+001
1.376e+002
2.034e+002
2.693e+002
3.351e+002
4.009e+002
4.667e+002
5.326e+002
5.984e+002
6.642e+002
lcase1
Equivalent Von Mises Stress
Inc: 50
Time: 5.000e-001
X
Y
Z
1
1194 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-10 Die Force versus Die Displacement
Using Triangular Elements
ADD CURVES
ALL LOCATIONS
contact body variables
Pos Y bottom
Force Y top
FIT
RETURN
GENERALIZED XY PLOT: COPY TO (add to Generalized XY Plotter)
SHOW HISTORY
SHOW MODEL
RETURN
CLOSE
MAIN
FILES
44
45
46
47
48
49
50
lcase1
Pos Y bottom (x.1)
Force Y top
0 4
0
3.181
43
42
41
40
2021222324252627
28
29
31
32
33
34
35
36
37
38
39
30
19 18 17 16
012345 789101112131415
6
1
1195
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-11 Change Class to Triangles
Run Job and View Results
SAVE AS
triangle
OK
RETURN
MESH GENERATION
CHANGE CLASS
TRIA (3)
ELEMENTS
all: EXIST.
MAIN
JOBS
ELEMENT TYPES
PLANE STRAIN SOLID
155 (Tri 3)
OK
all: EXIST.
X
Y
Z
1196 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-12 Equivalent Cauchy Stress Contours Triangular Mesh
RETURN (Twice)
RUN
SUBMIT 1
MONITOR
OK
SAVE
RETURN
RESULTS
OPEN DEFAULT
SKIP TO INC
50
DEF & ORIG
CONTOUR BAND
SCALAR
Equivalent Cauchy Stress
3.790e+000
2.775e+001
5.172e+001
7.568e+001
9.964e+001
1.236e+002
1.476e+002
1.715e+002
1.955e+002
2.195e+002
2.434e+002
lcase1
Equivalent Von Mises Stress
Inc: 50
Time: 5.000e-001
X
Y
Z
1
1197
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-13 Die Force versus Die Displacement
HISTORY PLOT
SET LOCATIONS
1
END LIST (#)
INC RANGE
0 50 1
ADD CURVES
ALL LOCATIONS
contact body variables
Pos Y bottom
Force Y top
FIT
RETURN
GENERALIZED XY PLOT: COPY TO (add to Generalized XY Plotter)
44
45
46
47
48
49
50
lcase1
Pos Y bottom (x.1)
Force Y top
0 4
-0
2.846
43
42
41
40
202122232425
2627
28
29
31
32
33
34
35
36
37
38
39
30
19 18 17 16
012345
789101112131415
6
1
1198 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-14 Generalized XY Plot to compare Quad and Tri Results
Figure 3.21-14 compares the Iorce displacement response oI the two models with nearly the same curve Ior both
quadrilaterals and triangular elements.
Tube with Friction
Friction between the rigid bodies and the tube is added using the stick-slip Iriction option. A Iriction coeIIicient oI 0.2
is used.
FILES
NEW
OK
OPEN
elasto1.mud
SAVE AS
elasto1f
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE
40
39
38
37
41
36
34
33
32
31
35
42
43
44
Y
0 4
-0
3.181
Force Y top : Pos Y bottom
50
49
48
47
45
30
X (x.1)
29
27
11 10 9 8
12
7 5 4 3 2
6
131415
26 25 24 23 22 21 20 19 18 17 16
28
1 0
46
45
49
50
0
35
34
33
32
31
30
29
28
27
36
26
37
39
48
47
46
44
43
42
41
40
38
Force Y top : Pos Y bottom
25
23
9 8 7 6 5 4 3 2 1
10
24
1113
22 21 20 19 18 17 16 15 14 12
1
1199
CHAPTER 3.21
Rubber Elements and Material Models
FRICTION COEFF
0.2
OK
NEXT
FRICTION COEFF
0.2
OK
NEXT
FRICTION COEFF
2
OK
MAIN
JOBS
PROPERTIES
CONTACT CONTROL
STICK-SLIP
OK (twice)
RUN
SAVE
SUBMIT1
MONITOR
POSTPROCESS
HISTORY
SET LOCATIONS
n:1
#END LIST
ALL INCS
ADD CURVES
GLOBAL
Pos Y top, Force Y top
Pos Y top, Force X top
FIT
1200 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-15 Hysteresis Due to Friction
Cavity Pressure
In the last run, the tube had an open air cavity; here, results are compared to Irictionless tube with a closed cavity
simulating the compression oI the air inside the tube, assuming that the tube is closed and the air cannot escape. Let`s
start with the original model and add a closed cavity representing the air inside the tube.
FILES
NEW
OK
OPEN
elasto1.mud
SAVE AS
elasto1c
RETURN
FILL
MESH GENERATION
ELEMENT CLASS
LINE (2)
RETURN
ADD ELEMENT (pick nodes N3 and N1 in order indicated)
RETURN
MODELING TOOLS
68
697071
67
72
74757677
73
78
66
64
54
55
56
57
65
58
60
61
62
63
59
53
7981
969899100
95
0
Pos Y top (x.1)
Y
-4 0
job1
80
94 92
82838485
93
8688899091
87
-0.345
52
50
12 13 14 15
11
16 18 19 20 21
17
22
108
0 1
9
2 4 5 6 7
3
51
23 25
40
41
42
43
39
44
46
47
48
49
45
24
38
36
26 27 28 29
37
30
32
33
34
35
31
3.2
Force Y top
97
9899100 0
72 71 70 69 68 67 66 65 64 63 62 61
52
53
54
55
56
57
58
59
60
73747576
9091929394959697 89
51
88 86 77787980818283848587
Force X top
50
48
212019181716151413121110 1 2 3 4 5 6 7 8 9
22 23 24 25
39 40 41 42 43
44 45
46
47
38
49
3735
26 27 28 29 30 31 32 33 34 36
1
-4
Force X top (x.1)
Pos Y top (x.1)
job1
0
100 99 98 97 96 95 94 93 92 91 90 89
0
88
-3.45
0.348
87 86 85
62
61
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
63
64
65
66 84
83
82
81
80
79 78
77 76 74 73 72
71
70
69
68
67
75
45
44
43
42
19
18
17
161514131211109876543
20
21 22
23
41
40
39
38
37
36
35
34
2
33
31
30
29
28
27
26
2524
32
10
1201
CHAPTER 3.21
Rubber Elements and Material Models
CAVITY
NEW
SELECT
METHOD PATH
RETURN
EDGES (pick N1 N2 N3)
END LIST
RETURN
EDGES ADD
ALL: SELECTED
REF. PRESSURE
1
REF. TEMPERATURE
1
REF. DENSITY
1.8E-5
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
MORE
N1
N3
N2
1202 Marc Users Guide: Part 2
CHAPTER 3.21
CAVITY MASS LOAD
MASS
CLOSED CAVITY, OK
CAVITIES ADD
cavity1
OD
MAIN
LOADCASE
MECHANICAL
STATIC
LOADS
apply2 (on)
OK (twice)
NEXT
STATIC
LOADS
apply2 (on)
OK (twice)
MAIN
JOBS
PROPERTIES
INITIAL LOADS
apply2 (on)
OK
JOB PARAMETERS
CAVITY PARAMETERS
AMBIENT PRESSURE
1
OK (thrice)
ELEMENT TYPES
MECHANICAL
MISCELLANEOUS
171
OK (pick element previously added)
END LIST
RETURN (twice)
1203
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-16 Force and Pressure History
The Iorce to crush the tube is considerably larger(5x) than beIore and nearly equal to the cavity pressure since the area
oI contact is about one inch.
SAVE
RUN
SUBMIT (1)
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
ALL INCS
ADD CURVES
GLOBAL
Pos Y top
Force Y top
Volume Cavity 1
Pressure Cavity 1
FIT
68 69 70 71 72 73 74 75 76
67 66 65
64
54
55
56
57
77
58
60
61
62
63
59
78 79 80
9597 98 99 94 100
X (x.1)
Y (x10)
-4 2.509
job1
53
93 91 81 82 83 84 92 8587 88 89 90 86
52
51
50
12 13 14 15 11 16 18 19 20 21 17 22
108 0 1 9 2 4 5 6 7 3
-0
23 25
40
41
42
43
39
44
46
47
48
49
45
24
38
36
26 27 28 29
37
30 32 33 34 35
31
2.137
Force Y top : Pos Y top
96 97 99 100 0
72 71 70 69 68 67 66
65
64
63
62
61
52
53
54
55
56
57
58
59
60
73 74 75 76
90 91 92 93 94 95 96 98 89
51
88 86
77 78 79 80 81 82 83 84 85 87
Pressure Cavity 1 : Volume Cavity 1
50
48
21 20 19 18 17 16 15 14 13 12 11 10 1 2 3 4 5 6 7 8 9
22 23 24 25
39
40
41
42
43
44
45
46
47
38
49
37
35
26 27 28 29 30 31 32 33 34
36
1
1204 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-17 Affect of Air Cavity
The deIormed shape at increment 50 Ior the closed air cavity shows the inner walls have yet to close, whereas the inner
walls will touch Ior the open air cavity.
Buckling of an Elastomeric Arch
Overview
An elastomeric arch has a center load applied and the objective oI the analysis is to determine the snap though in the
Iorce displacement response.
An adaptive load stepping method called arc-length (modiIied Riks-Ramm) is used.
SHOW MODEL
RETURN
DEF ONLY
SKIP TO INC
50
Open Air Cavity
Closed Air Cavity
Force
1205
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-18 Elastomeric Arch Problem Description
MESH GENERATION
COORDINATE SYSTEM SET
CYLINDRICAL
SET: GRID ON
RETURN
CURVE TYPE ARC CPP
RETURN
CURVES ADD
0 0 0
.7 30 0
.7 150 0
0 0 0
.8 30 0
.8 150 0
SURFACE TYPE RULED
RETURN
F
o
r
c
e
Displacement
1206 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-19 (A) Model (B) Mesh (C) Geometry
SURFACE: ADD
2 1
CONVERT
DIVISONS
20 3
SURFACES TO ELEMENTS
ALL: EXISTING
RETURN
GRID OFF
FILL
MAIN
(A)
(B)
(C)
1207
CHAPTER 3.21
Rubber Elements and Material Models
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISP
X=0
Y=0
NODES ADD (nodes at both ends)
NEW
POINT LOAD
Y FORCE
-0.03
OK
TABLES
NEW (1 INDEPENDENT VARIABLE)
DATA POINTS ADD
0 0 1 1 2 0
TABLE TYPE TIME
SHOW MODEL
RETURN
NODES ADD (top center node)
POINT LOAD
TABLE
table1 (attach table to y force)
MAIN
MATERIAL PROPERTIES (twice)
STRUCTURAL
TYPE MOONEY
C10
1
OK
ELEMENTS ADD
ALL EXISTING
MAIN
1208 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-20 (A) Nodes Added (B) Show Table (C) Show Model (D) POINT LOAD Submenu and (E)
MOONEY PROPERTIES Submenu
LOADCASES
MECHANICAL
STATIC
ARC LENGTH PARAMETERS
INITIAL FRACTION
0.1
OK (twice)
COPY
MAIN
(A)
(B)
(C)
(D)
(E)
1209
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-21 ADAPTIVE STEPPING (ARC LENGTH) Submenu
Run Job and View Results
JOBS
NEW
PROPERTIES
lcase1
lcase2
ANALYSIS DIMENSION PLANE STRAIN
JOB RESULTS
CAUCHY STRESS TENSOR
OK (twice)
ELEMENT TYPES
MECHANICAL
PLANAR
SOLID
80
OK
ALL EXISTING
RETURN
RUN
SUBMIT1
MONITOR
1210 Marc Users Guide: Part 2
CHAPTER 3.21
OK (twice)
SAVE
RESULTS
OPEN DEFAULT
SKIP TO INC
29
OK
DEF ONLY
CONTOUR BAND
SCALAR
EQ. CAUCHY STRESS
HISTORY PLOT
SET LOCATIONS
n:11
# END LIST (pick top center node)
ALL INCS
1211
CHAPTER 3.21
Rubber Elements and Material Models
Figure 3.21-22 Equivalent Cauchy Stress Contours
ADD CURVES
ALL LOCATIONS
Displacement Y
External Force Y
FIT
1212 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-23 Force-Displacement History with Snap-through
Comparison of Curve Fitting of Different Rubber Models
In the previous example, the material data was invented. Now let`s take a look at how actual data is collected and
various material models are Iit to this data. This is a uniaxial test with the recorded data shown as the thin black line.
Figure 3.21-24 Adjusting Measured Data to Analytical Model
38
39
40
41
42
43
lcase2
Displacement Y Node 11 (x.1)
External Force Y Node 11 (x.01)
-6.774 0
-3
0
37
36
34
17 18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
35
16
15
14
13
0
1
2
3
4
5
6
7
8
9
10
11
12
0.0 0.2 0.4 0.6 0.8
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Adjusted 18th Load Cycle
18th Cycle Loading Data
Raw Data
Uniaxial Test Data Adjustment
Engineering Stress [MPa]
Engineering Strain [1]

m

m
, ( )

c

c
, ( )
, ( )

m

c
( ) 1
c
+ ( ) / =

m

c
( ) 1
c
+ ( ) =
1213
CHAPTER 3.21
Rubber Elements and Material Models
The 6288 data points measured during this uniaxial test must be reduced in a logical manner to produce a single stress
strain diagram suitable Ior elastic materials. The 18th loading cycle was chosen to best represent the application, and
the data (gage length and original area) are adjusted Ior the strain oIIset, , the minimum value oI strain Ior the 18
th

load cycle. The 52 data points oI the 18
th
load cycle oI this uniaxial test are adjusted Ior the strain oIIset to
determine the data as shown in Figure 3.21-24 where adjusts Ior the implied change in
gage length, and adjusts Ior the implied change in original cross sectional area. The nonlinear
elastic theory requires that at zero strain the stress is zero. Repeating this procedure Ior biaxial and planar shear will
yield three stress-strain curves Ior the same material. Each curve represents the stress-strain behavior Ior three strain
states: tension, planar shear and equal biaxial behavior as shown in Figure 3.21-25. Ideally, it is best to use all strain
states when determining the constants used in analytic models such as Mooney, Ogden, Boyce, or Gent.
Figure 3.21-25 Three Basic Strain States
Each oI the curves above actually come Irom three independent tests perIormed on the same material. The process oI
using Mentat to determine the Mooney, Ogden, Boyce, or Gent constants is called Experimental Curve Fitting and we
shall now use Mentat to Iit the data shown in Figure 3.21-25.
MATERIAL PROPERTIES
EXPERIMENTAL DATA FIT
TABLES
READ
RAW
FILTER *data (pick uniaxial.data, biaxial, and planar_shear.data)
OK
c
c
o
m
c
m
,
o c , c c
m
c
c
( ) 1 c
c
( )
o o
m
o
c
( ) 1 c
c
( )
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.5
1.0
1.5
2.0
Biaxial
Planar Shear
Uniaxial
Three Basic Strain States
Engineering Stress [MPa]
Engineering Strain [1]
Uniaxial
Volumetric Specimen Cutouts
Planar Shear
Planar Shear
Biaxial
Biaxial
1214 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-26 Reading Tables of Material Properties
Figure 3.21-27 Displaying Material Data
Having the data Ior the three strain states in Mentat Iacilitates perIorming several curve Iits to pick the most
appropriate material model. Let`s examine a Mooney and Arruda-Boyce material models.
TABLES
GENERALIZED XY PLOT: COPY TO
RETURN, NEXT
GENERALIZED XY PLOT: COPY TO
RETURN, NEXT
GENERALIZED XY PLOT: COPY TO
RETURN, NEXT
RETURN (twice)
12
11
10
9
8
7
5
4
3
2
6
13
14
16
0.018
1.547
Uniaxial
56
55
54
53
52
51
50
49
8.54 0
Stress
Strain (x.1)
48
47
46
45
28
27
26
25
24
29
23
21
20
19
18
17
22
15
30
32
44
43
42
41
40
31
39
37
36
35
34
33
38
29
38
43
30
31
32
33
34
35
36
37
39
40
41
42
28
9
26
10
11
6
5
4
3
2
1
Planar Shear
8
13
14
15
16
17
18
19
20
21
22
23
24
25
12
27
7
26
27
28
29
30
31
32
33
34
35
36
37
46
45
44
43
42
41
40
39
38
25
24
23
22
8
7
6
5
4
3
2
1
Biaxial
9
47
10
12
21
20
19
18
17
16
15
14
13
11
48
49
50
52
51
1215
CHAPTER 3.21
Rubber Elements and Material Models
Mooney
Now let`s associate each table read with the proper strain state and do a Iit.
Figure 3.21-28 Two Constant Mooney only using Uniaxial Data A Poor Fit
EXPERIMENTAL DATA FIT
UNIAXIAL
table1
BIAXIAL
table 2
PLANAR SHEAR
table 3
ELASTOMERS
MOONEY(2)
UNIAXIAL
COMPUTE
OK
SCALE AXES
uniaxial/mooney2 uniaxial/experiment
4.453
0
8.894
(x.1)
0
biaxial/mooney2 planar_shear/mooney2
simple_shear/mooney2
1
1216 Marc Users Guide: Part 2
CHAPTER 3.21
Notice that when Mentat Iits just a single curve oI data, it also plots the predicted other strain states using the current
elastomeric model. In this case, only the uniaxial data was used to Iit the two constants to a Mooney material. Notice
that this is a poor Iit because this model is too stiII in biaxial deIormation.
Now let us try using all oI the strain states again to Iit a two constant Mooney material model.
Figure 3.21-29 Two Constant Mooney only Using Uniaxial Data A Good Fit
MOONEY(2)
USE ALL
DATA
COMPUTE
OK
SCALE AXES
uniaxial/mooney2
1.547
0
9.1 0
(x.1)
uniaxial/experiment
biaxial/experiment
planar_shear/experiment
biaxial/mooney2
planar_shear/mooney2
simple_shear/mooney2
1217
CHAPTER 3.21
Rubber Elements and Material Models
Arruda-Boyce
Clearly using all oI the strain states is the best. However, many times you may not have all oI the data and may be
stuck with just the uniaxial test data. In that case, you may wish to use the Arruda-Boyce model.
Figure 3.21-30 Arruda-Boyce Model only using Uniaxial Data A Good Fit
For completeness, here is the Arruda-Boyce model using all strain states.
ARRUDA-BOYCE
UNIAXIAL
COMPUTE
OK
SCALE AXES
ARRUDA-BOYCE
USE ALL DATA
COMPUTE
OK
SCALE AXES
uniaxial/arruda_boyce uniaxial/experiment
1.005
0
8.894
(x.1)
0
biaxial/arruda_boyce planar_shear/arruda_boyce
simple_shear/arruda_boyce
1218 Marc Users Guide: Part 2
CHAPTER 3.21
Figure 3.21-31 Arruda-Boyce Model only using All Data A Good Fit
Let`s suppose that the actual application was the inIlation oI a tube. The Mooney(2) using only uniaxial data would
require a pressure oI over 4 times that oI the Arruda-Boyce model when the inIlation strains are about 90. There are
several elastomeric material models that Iall into three classes: phenomenological, principal stretch, and micro
mechanical models such as the Arruda-Boyce and Gent models.
There are no good models, only good curve Iits. In this case, a one constant Mooney would have worked Iine. The
important Iact to keep in mind when Iitting elastomeric material models to material data, is to watch the response oI
all strain states shown by the curve Iitting in Mentat. II there are large variations in the curves such as in
Figure 3.21-28, don`t use it. Don`t obtain good agreement, only Iorm one strain state, rather seek a balanced response
to the three diIIerent deIormation states because your application will have all strain states present. That balanced
response usually looks just like the data in the three basic strain states shown in Figure 3.21-25.
uniaxial/arruda_boyce
1.547
0
9.1 0
(x.1)
uniaxial/experiment
biaxial/experiment
planar_shear/experiment
biaxial/arruda_boyce
planar_shear/arruda_boyce
simple_shear/arruda_boyce
1
1219
CHAPTER 3.21
Rubber Elements and Material Models
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
Tube Crush
rubber_a.proc Mentat procedure Iile Ior quads
rubber_b.proc Mentat procedure Iile Ior triangles
rubber_c.proc Mentat procedure Iile Ior quads multi mode Iit
uniaxial.data xy data Ior rubberc.proc
biaxial.data xy data Ior rubberc.proc
planar_shear.data xy data Ior rubberc.proc
s3c.proc adds closed cavity to above
elasto1.mud Mentat model Iile read by s3c.proc
Arch
s9.proc Mentat procedure Iile Ior arch
Curve Fitting
AllTestData.xls Test data Ior all modes
uniaxial.data Uniaxial data taken Irom the 18th load cycle
planar_shear.data Planar shear data taken Irom the 18th load cycle
biaxial.data Biaxial data taken Irom the 18th load cycle
NonlinearFEAofElastomers.pdf White paper on FEA oI elastomers
1220 Marc Users Guide: Part 2
CHAPTER 3.21
Chapter 3.22: Modeling of General Rigid Body Links using RBE2/RBE3
3.22
Modeling of General Rigid
Body Links using RBE2/RBE3

Chapter Overview 1222

Cylindrical Shell 1222

Results 1230

Input Files 1231

1222 Marc Users Guide: Part 2
CHAPTER 3.22
Chapter Overview
This chapter describes the use oI RBE2 link in Marc (The term RBE2 signiIies Rigid Body element type 2 as deIined
in Nastran). RBE2 link is a general multi-point constraint that connect a set oI tied nodes with a reIerence node. The
link between the tied nodes and the reIerence node is generally based on a rigid link connection, but some oI the
degrees oI Ireedom oI the tied nodes can be set Iree (unrestrained). II all degrees oI Ireedom are constrained, then it
results in tying type 80 behavior. To show the Ilexibility oI the RBE2 in simulating general rigid body link, a simple
cylindrical shell model is used as a case study.
Cylindrical Shell
The example described here is a cylindrical shell with a coarse mesh. The boundary condition on the one end oI the
cylinder is Iully clamped. On the other end oI the cylinder, an RBE2 link is deIined to connect all nodes at the
cylindrical edge with a retained node. The tied nodes have TRANSFORMATION data in which the axial (third
direction) is parallel with the x-axis (In this case, it is not actually necessary to deIine TRANSFORMATION on the tied
nodes). The tied nodes are constrained only in the axial direction while the other degrees oI Ireedom are Iree. The axial
direction is updated Iollowing the rotation oI the retained node.
Figure 3.22-1 FE Mesh of a Cylindrical Shell
Mesh Generation
The length and the radius oI the cylinder are 90 mm and 15 mm, respectively. The mesh is generated by Iirst deIining
one 1-D element. This element is then divided into three elements. With the cylindrical model, the elements are
loaded end
clamped end
1223
CHAPTER 3.22
Modeling of General Rigid Body Links using RBE2/RBE3
expanded about the x-axis. The Iinal mesh can be seen in Figure 3.22-1. Two extra nodes are also created. They are
needed to deIine TRANSFORMATION and RBE2 Link.
MESH GENERATION
NODES ADD
0 15 0
90 15 0
FILL
ELEMENT CLASS
LINE(2)
RETURN
ELEMS ADD
1
2
SUBDIVIDE
DIVISIONS
3 1 1
ELEMENTS
1 #
RETURN
SWEEP
NODES
ALL EXIST.
RETURN
EXPAND
ROTATION ANGLES
45 0 0
REPETITIONS
8
ELEMENTS
ALL EXIST
FILL
RETURN
SWEEP
NODES
ALL EXIST.
RETURN
1224 Marc Users Guide: Part 2
CHAPTER 3.22
Boundary Conditions
One end oI the cylinder is clamped, meaning all degrees oI Ireedom are Iixed. The other end oI the cylinder is
constrained to move as general rigid body to be deIined later using RBE2 Link. The nodes that belong to this end will
have local coordinate system deIined with TRANSFORMATION. The retained node oI the RBE2 Link are Iixed in
z-direction, about the x-, and y-direction. As loading, this node is given 30 rotation about z-axis. The time history oI
the load is set by using a table that ramps Irom 0 to 30 in one second.
RENUMBER
NODES
ALL EXIST.
ELEMENTS
ALL EXIST.
RETURN (twice)
NODES ADD
0 0 0
90 0 0
RETURN
BOUNDARY CONDITIONS
MECHANICAL
NAME
fixed
FIXED DISPLACEMENT
DISPLACEMENT X
0
DISPLACEMENT Y
0
DISPLACEMENT Z
0
ROTATION X
0
ROTATION Y
0
ROTATION Z
0
OK
1225
CHAPTER 3.22
Modeling of General Rigid Body Links using RBE2/RBE3
NODES ADD
1 11 10 9 8 7 6 5 #
NEW
NAME
load
FIXED DISPLACEMENT
DISPLACEMENT X
0
ROTATION X
0
ROTATION Y
0
ROTATION Z
-30*PI/180
OK
TABLES
NEW TABLE: 1 INDEPENDENT VARIABLE
NAME
time
TYPE
time
ADD
0 0 1 1
RETURN
FIXED DISPLACEMENT
ROTATION Z: TABLE
TABLE
time
OK (twice)
NODES ADD
34 #
MAIN
1226 Marc Users Guide: Part 2
CHAPTER 3.22
Transformation
The local coordinate system oI the tied nodes is deIined using the CYLINDRICAL option in Mentat. In this case, the
axial direction is chosen to be the same as the axial direction oI the cylinder.
Links
RBE2 link is deIined along one edge oI the cylinder. First, a reIerence node located at (90,0,0) is set. Second, the
constrained degree oI Ireedom is set to three according to the local coordinate system deIined Ior the tied nodes. Third,
the list oI the tied nodes are set that have consistent TRANSFORMATION as deIined in the second step. In this case,
the axial displacement is the third degree oI Ireedom.
Material Properties
The material is elastoplastic with von Mises yield criteria and isotropic hardening parameter.
BOUNDARY CONDITIONS
MECHANICAL
TRANSFORMATION
CYLINDRICAL
0 0 0
90 0 0
2 18 17 16 15 14 13 12 #
MAIN
LINKS
RBE2S
NEW
NODE (RETAINED)
34
DOF
3
ADD (TIED NODES)
2 18 17 16 15 14 13 12 #
MAIN
MATERIAL PROPERTIES
NEW
ISOTROPIC
1227
CHAPTER 3.22
Modeling of General Rigid Body Links using RBE2/RBE3
Geometric Properties
The thickness oI the cylinder is 3 mm.
YOUNGS MODULUS
250000
POISSONS_RATIO
0.3
ELASTIC-PLASTIC
INITIAL YIELD STRESS
1
OK
TABLES
NEW
NAME
plas
TYPE
eq_plastic_strain
ADD
0 500 1 3000
FIT
RETURN
ISOTROPIC
ELASTIC-PLASTIC
YIELD STRESS: TABLE
plas
OK
ELEMENTS ADD
ALL EXIST.
MAIN
GEOMETRIC PROPERTIES
NEW
3-D
SHELL
1228 Marc Users Guide: Part 2
CHAPTER 3.22
Loadcases and Job Parameters
A quasi-static analysis will be perIormed. The convergence criteria are based both on residual Iorces and
displacement increment.
THICKNESS
3
OK
ELEMENTS ADD
ALL EXIST.
MAIN
LOADCASES
NEW
MECHANICAL
STATIC
CONVERGENCE TESTING
RESIDUALS AND DISPLACEMENTS
INCLUDE MOMENTS
INCLUDE ROTATIONS
RELATIVE FORCE TOLERANCE
0.005
RELATIVE MOMENT TOLERANCE
0.005
RELATIVE DISPLACEMENT TOLERANCE
0.01
RELATIVE ROTATION TOLERANCE
0.01
OK
# STEPS
10
OK
MAIN
1229
CHAPTER 3.22
Modeling of General Rigid Body Links using RBE2/RBE3
Save Job, and Run the Simulation
AIter saving the model, two jobs are deIined either with or without LARGE DISP parameter. They are then
submitted sequentially.
JOBS
NEW
NAME
linear
MECHANICAL
SELECTED
lcase1
OK
NEW
NAME
nonlinear
MECHANICAL
SELECTED
lcase1
SOLUTION OPTION
LARGE DISPLACEMENT
OK
ELEMENT TYPES
MECHANICAL
3-D MEMBRANE/SHELL ELEMENT TYPES
THIN SHELL: 139
ALL: EXIST
MAIN
JOBS
PREV
RUN
SUBMIT
NEXT
RUN
SUBMIT
1230 Marc Users Guide: Part 2
CHAPTER 3.22
Results
The deIormed shape oI the cylinder without and with LARGE DISP (large rotation) are shown in Figure 3.22-2 and
Figure 3.22-3, respectively. From these Iigures, it is clearly seen the diIIerence oI ovalization (red line) oI the
cylinder (where rotation is applied). The result using 'large rotation option shows ovalization about the z-axis as the
major one while the 'small rotation the y-axis as the major one. This is known as the Bezier eIIect.
Figure 3.22-2 Deformed Shape without LARGE DISP
Figure 3.22-3 Deformed Shape with LARGE DISP
1231
CHAPTER 3.22
Modeling of General Rigid Body Links using RBE2/RBE3
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
rbe2.proc Mentat procedure Iile Ior quads
1232 Marc Users Guide: Part 2
CHAPTER 3.22
Chapter 2.23: Cyclic Symmetry
3.23
Cyclic Symmetry

Chapter Overview 1234

Pure Torsion 1235

Mechanical Analysis of Friction Clutch 1238

Coupled Analysis of Friction Clutch 1241

Input Files 1246

1234 Marc Users Guide: Part 2
CHAPTER 3.23
Chapter Overview
A special set oI tying constraints Ior continuum elements can be automatically generated by the Marc program to
eIIectively analyze structures with a geometry and a loading varying periodically about a symmetry axis. Figure 3.23-1
shows an example where, on the leIt-hand side, the complete structure is given and, on the right-hand side, a sector to
be modeled.
Figure 3.23-1 Cyclic Symmetric Structure: Complete Model (left) and Modeled Sector (right)
Looking at points and on this segment, the displacement vectors should IulIill:
(3.23-1)
which can also be written as:
(3.23-2)
where the transIormation matrix depends on the symmetry axis (which, in the example above, coincides with the
global Z-axis) and the sector angle (see Figure 3.23-1). In Marc, the input Ior the CYCLIC SYMMETRY option
consists oI the direction vector oI the symmetry axis, a point on the symmetry axis and the sector angle . The
Iollowing items should be noted:
1. The meshes do not need to line up on both sides oI a sector (Ior example, see Figure 3.23-1).
X
Y
X
Y
x y
A
B
a
A B
u'
B
u
A

u
B
Ru
A

R
o
o
1235
CHAPTER 3.23
Cyclic Symmetry
Figure 3.23-1 Finite Element Mesh for Cyclic Symmetric Structure with Different Mesh Densities
on the Sector Sides
2. Any shape oI the sector sides is allowed provided that upon rotating the sector times about the
symmetry axis over the sector angle will result in the complete model.
3. The CYCLIC SYMMETRY option can be combined with the CONTACT option.
4. The CYCLIC SYMMETRY option can be combined with global remeshing.
5. In a coupled thermo-mechanical analysis, the temperature is Iorced to be cyclic symmetric ( as in
Figure 3.23-1).
6. A nodal point on the symmetry axis is automatically constrained in the plane perpendicular to the symmetry
axis.
7. The possible rigid body motion about the symmetry axis can be automatically suppressed.
8. Cyclic Symmetry is valid Ior:
a. Only the continuum elements. However, the presence oI beams and shells is allowed, but there is no
connection oI shells to shells, so the shell part can, Ior example, be a turbine blade and the volume part is
the turbine rotor. The blade is connected to the rotor and iI there are 20 blades, 1/20 oI the rotor is modeled
and one complete blade.
b. It can be used Ior static, dynamic, remeshing, and coupled analysis.
c. It cannot be used Ior pure heat transIer.
d. It can be used Ior all analysis involving contact.
The Iollowing cases will demonstrate many oI the items above and show how this Ieature can save computer time by
taking advantage oI the symmetry oI the structure.
Pure Torsion
A solid rubber cylinder will be subjected to a state oI pure torsion by rotating the ends which are attached to rigid
bodies. Figure 3.23-2 shows the solid rubber cylinder (leIt) and its cyclic symmetry counterpart (right). The torsional
stiIIness oI these two models will be compared to each other as well as the theoretical values.
360 o
o
T
A
T
B

1236 Marc Users Guide: Part 2

CHAPTER 3.23
Figure 3.23-2 Model for Case 1 Pure Torsion
Procedure for Case 1: Build and run the cyclic symmetric model
The cylinder is 10 m in length and 1 m in radius. The ends are glued to the square rigid bodies shown in Figure 3.23-2.
The procedure here only Iocuses upon the cyclic symmetry Ieature and how it is implemented on the cyclic symmetry
model called slice.mud.
You can now review the properties oI this model. The mesh is just a 10
o
sector taken Irom the solid mesh
(solid.mud). The material is a one constant Mooney with C 1 |MPa| or a shear modulus G oI 2 |MPa|. The contact
option identiIies the deIormable slice with the two rigid bodies at each end, where the top rigid body will rotate about
the Z-axis one revolution. The contact table option is used to glue the rigid bodies to the end oI the deIormable slice
with a large separation Iorce. The cyclic symmetry option is located in the JOB menu, to go there simply enter:
FILES
OPEN
slice.mud
MAIN
JOBS
MECHANICAL
CYCLIC SYMMETRY
slice.mud solid.mud
1237
CHAPTER 3.23
Cyclic Symmetry
Figure 3.23-3 Cyclic Symmetry Menu
Here we see that the axis oI symmetry is deIined by a direction and a point, with 36 repetitions. This completes the
deIinition oI cyclic symmetry, now let`s run the two models and compare the results.
AIter running both models, the stresses are shown in Table 2.23-1 where both models have nearly the same maximum
equivalent Cauchy stress and are within 4 oI theory.
Clearly, the slice runs Iaster taking advantage oI the cyclic symmetry oI the structure. Figure 3.23-4 plots the torque
versus rotation oI the two cylinders. Since the stresses in the slice model are integrated over a smaller (1/36 times)
area, remember that the external Iorces need to be multiplied by the number oI repetitions, 36, which is plotted in
Figure 3.23-4 with the diamond symbol.
Table 2.23-1 Results for Pure Torsion
Model Stress [MPa] CPU Times [sec.]
Solid 2.054 1462.70
Slice 2.058 10.65
Theory 1.976 NA
1238 Marc Users Guide: Part 2
CHAPTER 3.23
Figure 3.23-4 Torque versus Rotation
Mechanical Analysis of Friction Clutch
Figure 3.23-5 shows an elastomeric Iriction clutch between two rigid surIaces that will compress the clutch then rotate
relative to each other. This causes the clutch to rotate until the Iriction Iorces are overcome by the torsional moment
in the clutch, and the clutch will slip, limiting the torque transmitted to the smaller rigid surIace. The ribs on the clutch
are to help keep the clutch in better contact with the drive.
Figure 3.23-5 Full Model for Case 2 Friction Clutch, clutch_rib.mud
-8.0 -6.0 -4.0 -2.0 0.0
Angle Position top [radians]
0.0
1.0
2.0
3.0
M
o
m
e
n
t

Z

t
o
p
,

M
z

[
M
N
-
m
]
Solid Rubber Cylinder
Solid
Slice
36*Mz Slice
1239
CHAPTER 3.23
Cyclic Symmetry
The material properties are the same as in the previous case and two loadcases are used to compress then rotate the
clutch. The later loadcase uses variable time stepping. Friction coeIIicients oI 0.5 are entered in contact table and
Coulomb Iriction is used. Cyclic symmetry is used as in the previous case; however, two slices are used with Iour
repetitions as shown in Figure 3.23-6 and Figure 3.23-7. Also, the axis oI cyclic symmetry is now the X-axis.
Figure 3.23-6 Cyclic Symmetry for Case 2 Friction Clutch, clutch_rib_slice1.mud
Figure 3.23-7 Cyclic Symmetry for Case 2 Friction Clutch, clutch_rib_slice2.mud
You may view any oI the models by opening either, clutch_rib.mud, clutch_rib_slice1.mud, or
clutch_rib_slice2.mud. The results are shown in Figure 3.23-8.
1240 Marc Users Guide: Part 2
CHAPTER 3.23
Figure 3.23-8 Stress Contours for Case 2 Friction Clutch Models
clutch_rib
clutch_rib_slice1
clutch_rib_slice2
1241
CHAPTER 3.23
Cyclic Symmetry
AIter running the three models, the stresses are shown in Table 2.23-2 where all models have nearly the same
maximum equivalent Cauchy stress. The run times Ior the slices are, oI course, lower that the Iull model, and the run
times Ior the slices are slightly diIIerent because oI slightly diIIerent meshes.
The stresses above are reported at the maximum torque condition, aIter the clutch slips around 1.5 radians oI angular
motion as shown in Figure 3.23-9. Again as in Case 1, the external Iorces in the slice models must be multiplied by
the number oI cyclic repetitions, Iour, as clearly shown in Figure 3.23-9.
Figure 3.23-9 Wall Torque versus Drive Angular Position
Coupled Analysis of Friction Clutch
Figure 3.23-10 shows a coupled Iriction clutch between two rigid surIaces that compress the clutch then rotate relative
to each other. This causes the clutch to rotate until the Iriction Iorces are overcome by the torsional moment in the
clutch, and the clutch slips limiting the torque transmitted to the smaller rigid surIace. In addition, as the larger rigid
surIace rotates, Iriction generates thermal energy that heats up the clutch. Heat Ilows out the sink where the smaller
end oI the clutch is held at a Iixed temperature. As in Case 2, the Iull model will also be modeled using cyclic
symmetry.
Table 2.23-2 Results for Friction Clutch
Model Stress [MPa] CPU Times [sec.]
clutch_rib.mud 1.475 440.84
clutch_rib_slice1.mud 1.471 55.35
clutch_rib_slice2.mud 1.488 64.80
0.0 2.0 4.0 6.0
Angle Position Drive [radians]
0.000
0.010
0.020
0.030
0.040
M
o
m
e
n
t

X

F
i
x
e
d
,

M
x

[
M
N
-
m
]
Friction Clutch
Full Model
Slice1
Slice2
1242 Marc Users Guide: Part 2
CHAPTER 3.23
Figure 3.23-10 Full Model for Case 3 Coupled Friction Clutch, coupled.mud
The material properties are Ior steel and two loadcases are used to compress then rotate the clutch. Both loadcases use
Iixed time stepping. A Iriction coeIIicients oI 0.2 is entered in contact table and Coulomb Iriction is used. Cyclic
symmetry is used as in the previous case; however, two slices are used with Iour repetitions as shown in Figures
3.23-11 and 3.23-12. Also, the axis oI cyclic symmetry is now the X-axis.
Figure 3.23-11 Cyclic Symmetry for Case 3 Friction Clutch, coupled_slice1.mud
1243
CHAPTER 3.23
Cyclic Symmetry
Figure 3.23-12 Cyclic Symmetry for Case 3 Friction Clutch, coupled_slice2.mud
You may view any oI the models by opening either, coupled.mud, coupled_slice1.mud, or
coupled_slice2.mud. The results are shown in Figure 3.23-13.
1244 Marc Users Guide: Part 2
CHAPTER 3.23
Figure 3.23-13 Stress Contours for Case 3 Friction Clutch Models
coupled_rib
coupled_slice1
coupled_slice2
1245
CHAPTER 3.23
Cyclic Symmetry
AIter running the three models, the results are shown in Table 2.23-3 where all models have nearly the same maximum
temperature. The run times Ior the slices are, oI course, lower than the Iull model, and the run times Ior the slices are
slightly diIIerent because oI slightly diIIerent meshes.
The above temperatures are reported at the maximum torque condition, aIter the clutch slips around 2 radians oI
angular motion as shown in Figure 3.23-14. Again as in the other Cases, the external Iorces in the slice models must
be multiplied by the number oI cyclic repetitions as clearly shown in Figure 3.23-14.
Figure 3.23-14 Wall Torque versus Drive Angular Position
As in the other Cases, the external Iorces, as well as the thermal energy in the slice models, must be multiplied by the
number oI cyclic repetitions as clearly shown in Figure 3.23-15 that plots the total thermal energy history oI the three
models.
Table 2.23-3 Results for Coupled Friction Clutch
Model Temperature [F] CPU Times [sec.]
coupled.mud 52.41 152.31
coupled_slice1.mud 56.26 55.25
coupled_slice2.mud 49.27 32.35
0.0 20.0 40.0 60.0
X : Angle Position Drive [radians]
0e+00
1e+05
2e+05
M
o
m
e
n
t

X

F
i
x
e
d
,

M
x

[
l
b
-
i
n
]
Coupled Friction Clutch
Full Model
Slice 1
Slice 2
1246 Marc Users Guide: Part 2
CHAPTER 3.23
Figure 3.23-15 Thermal Energy History
The thermal energy is automatically placed on the post Iile. For more inIormation on this and other energy calculations
see Chapter 6.8.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
clutch_rib.mud Mentat model Iile
clutch_rib_slice1.mud Mentat model Iile
clutch_rib_slice2.mud Mentat model Iile
coupled.mud Mentat model Iile
coupled_slice1.mud Mentat model Iile
coupled_slice2.mud Mentat model Iile
slice.mud Mentat model Iile
solid.mud Mentat model Iile
0.0 10.0 20.0
X : Time [seconds]
0.0
20.0
40.0
60.0
T
h
e
r
m
a
l

E
n
e
r
g
y

[
B
T
U
]
Coupled Friction Clutch
Full Model
Slice 1
Slice 2
Chapter 3.24: Axisymmetric to 3-D Analysis
3.24
Axisymmetric to 3-D Analysis

Chapter Overview 1248

Simulation of a Rubber Bushing 1248

Automobile Tire Modeling with Rebar Elements 1261

Analysis of a Rubber Cylinder using Remeshing 1268

Input Files 1276

1248 Marc Users Guide: Part 2
CHAPTER 3.24
Chapter Overview
In many cases, it is possible to begin the numerical simulations as a two-dimensional axisymmetric problem even
though the Iinal problem is Iully three-dimensional. This is advantageous because oI the large computational savings.
For this to be useIul, the Iirst stage oI the problem should be truly axisymmetric. The second stage oI the problem can
be Iully three-dimensional. This chapter demonstrates the use oI the data transIer capabilities oI Marc Irom an
axisymmetric analysis to a Iully three-dimensional analysis. For this purpose, three problems will be analyzed: the Iirst
one is simulation oI a rubber bushing problem, the second one is an analysis oI an automobile tire using rebar elements,
and the third is an analysis oI a rubber cylinder with remeshing.
The transIer oI data Irom an axisymmetric simulation to a 3-D simulation involves two parts. The Iirst is the generation
oI a new mesh, which may be either equally or unequally distributed along the circumIerence. The second part
involves reading the post Iile Irom the Iirst simulation containing the state variables (displacements, temperatures,
etc.) and stress/strains etc. using the PRE STATE option in Marc. Prior to the 2007 release this was named the AXITO3D
option.
Simulation of a Rubber Bushing
Simulation oI a rubber bushing in this chapter contains two major parts: Axisymmetric analysis and 3-D analysis. The
detailed description oI the two parts will be presented. The substeps at the beginning oI the second part involving mesh
expansion and data transIer Irom axisymmetric to 3-D cases will be highlighted.
Figure 3.24-1 2-D Axisymmetric to 3-D
Description of Problem
A rubber bushing with an outer diameter oI 10 cm and an inner diameter oI 2 cm is considered. The length oI the rubber
bushing is 8 cm. Both outside and inside surIaces are glued to two steel tubes with corresponding diameters so that the
shape oI the surIaces keeps unchanged during deIormation.
1249
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Two load sequences are applied:
In the Iirst load step, a displacement oI 2 cm along the symmetric axis is applied to the outside steel tube
within ten equal increments. During this load step, the deIormation is purely axisymmetric and thereIore an
axisymmetric analysis is perIormed. AIterwards, the outside steel tube moves 1 cm in the radial (Y) direction
within Iive equal increments.
In the second step, the problem becomes Iully three-dimensional, and a 3-D analysis is perIormed.
The 4-node isoparametric quadrilateral axisymmetric element 10 is used in the axisymmetric run. The corresponding
element type in 3-D run is 7 which is the 8-node isoparametric hexahedral element. In the analysis, both element types
are based on mixed Iormulations and Iormulated on the deIormed (updated) conIiguration. This is activated using
ELASTICITY,2 in the parameter options.
The rubber bushing is modeled using Mooney constitutive model. The material parameters are given as C
1
8.0 N/cm
2

and C
2
2.0 N/cm
2
.
Axisymmetric Analysis
This is a standard axisymmetric analysis. Except Ior speciIying proper output to the post Iile, requested in 3-D
analysis, nothing is special. ThereIore, the description in the step is not in great details.
Model Generation
Model generation contains geometry deIinition, mesh generation using advancing Iront mesher, clear geometry, and
clean mesh.
MESH GENERATION
pts ADD
0 2 0
8 2 0
0 10 0
8 10 0
0 9 0
1 9 0
7 9 0
8 9 0
0 3 0
1 3 0
7 3 0
8 3 0
2 8 0
1250 Marc Users Guide: Part 2
CHAPTER 3.24
2 4 0
6 8 0
6 4 0
crvs ADD
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 3 5 4 8 9 1 12 2
CURVE TYPE
ARCS CENTER/POINT/POINT
RETURN
crvs ADD
1 8 0
2 8 0
1 9 0
1 4 0
1 3 0
2 4 0
7 4 0
6 4 0
7 3 0
7 8 0
7 9 0
6 8 0
AUTOMESH
CURVE DIVISIONS
FIXED AVG LENGTH
AVG LENGTH
0.4
APPLY CURVE DIVISIONS
all: EXIST.
RETURN
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT): QUAD MESH!
all: EXIST.
RETURN
RETURN
CLEAR GEOM
SWEEP
1251
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-2 FE-Mesh for Axisymmetric Analysis
Boundary Conditions
DeIining boundary conditions includes deIining node sets, deIining tables, and adding boundary conditions.
ALL
RETURN
RENUMBER
ALL
RETURN
SELECT
METHOD
BOX
RETURN
NODES
-1 11 9.99 11 -1 1
nodes STORE
outer
OK
all: SELECT.
1252 Marc Users Guide: Part 2
CHAPTER 3.24
CLEAR SELECT
NODES
-1 11 1 2.01 -1 1
nodes STORE
inner
OK
all: SELECT.
CLEAR SELECT
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
TABLE
NEW
1 INDEPENDENT VARIABLE
TYPE
time
OK
ADD
0 0 2 2 3 2
NEW
1 INDEPENDENT VARIABLE
TYPE
time
OK
ADD
0 0 2 0 3 1
RETURN
FIXED DISPLACEMENT
DISPLACEMENT X: (on)
DISPLACEMENT Y: (on)
DISPLACEMENT X TABLE
table1
DISPLACEMENT Y TABLE
table2
OK
1253
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Material Properties
Load Steps and Job Parameters
A displacement oI 2 cm along the symmetric axis is applied to the outside steel tube within 10 equal increments;
element type 10 is used; updated Lagrangian Iormulation is used Ior elasticity; stress tensor, strain tensor, and
equivalent von Mises stress are written into the post Iile.
nodes ADD
outer
NEW
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
OK
nodes ADD
inner
MAIN
MATERIAL PROPERTIES
MORE
MOONEY
C10
8
C01
2
OK
elements ADD
all: EXIST.
MAIN
Note: To use Updated Lagrangian Iormulation Ior elasticity, stress and strain tensors must be written into the post Iile. In the
second part involving the 3-D analysis, both stress and strain tensors are needed.
LOADCASE
NEW
MECHANICAL
STATIC
1254 Marc Users Guide: Part 2
CHAPTER 3.24
TOTAL LOADCASE TIME
2
FIXED PARAMETERS
# STEPS
10
OK
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.01
OK (twice)
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
AXISYMMETRIC SOLID
10
OK
all: EXIST.
RETURN
RETURN
NEW
MECHANICAL
LCASE1
ANALYSIS DIMENS: AXISYMMETRIC
ANALYSIS OPTIONS
ELASTICITY PROCEDURE: LARGE STRAIN - TOTAL LAGRANGE
ELASTICITY PROCEDURE: LARGE STRAIN - UPDATED LAGRANGE
OK
JOB RESULTS
available element tensors
Stress
Strain
available element scalars
Equivalent Von Mises Stress
OK (twice)
1255
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Save Model, Run Job, and View Results
Figure 3.24-3 Deformed Mesh and Distribution of Equivalent von Mises Stress at Increment 10
of Axisymmetric Analysis
FILE
SAVE AS
rubberbushing_axi.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
CLOSE
RETURN
1256 Marc Users Guide: Part 2
CHAPTER 3.24
3-D Analysis
AIter the axisymmetric analysis, a Iully 3-D analysis is perIormed based on the results Irom the axisymmetric analysis.
BeIore the 3-D analysis, a corresponding 3-D mesh on the basis oI the axisymmetric mesh and data transIer Irom the
axisymmetric mesh to the 3-D mesh are required. Compared to the other part oI the job, which is more or less standard,
a more detailed description is given Ior the axisymmetric to 3-D mesh expansion and the data transIer.
Mesh Expansion and Data Transfer from Axisymmetric to 3-D
Mesh expansion Irom axisymmetric to 3-D is based on AXISYMMETRIC MODEL TO 3D option under MESH
GENERATION -> EXPAND. Rotation angles and number oI repetitions must be deIined. To shiIt load table curve, the
time at which the analysis will continue in a Iully 3-D manner must be deIined. See Marc Volume A: Theory and User
Information Ior detailed description oI the shiIt oI load table curves.
Data transIer Irom axisymmetric to 3-D is based on option AXISYMMETRIC TO 3D under INITIAL CONDITIONS->
MECHANICAL. Both stress and strain tensors must be transIerred when the updated Lagrangian Iormulation Ior
elasticity is used; displacement is moved by deIault; the name oI post Iile Irom the completed axisymmetric analysis
must be given.
In this example, the 2-D section is uniIormly expanded over 180 in 12 sections. The time is set to 2, which is the time
at the end oI the previous analysis.
Figure 3.24-4 AXISYMMETRIC MODEL TO 3D Submenu
1257
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-5 FE-Mesh for 3-D Analysis
MESH GENERATION
EXPAND
AXISYMMETRIC MODEL TO 3D
ANGLE
15
REPETITIONS
12
TIME SET
2
EXPAND MODEL
MAIN
INITIAL CONDITIONS
MECHANICAL
AXISYMMETRIC TO 3D
POST FILE
rubberbushing_axi_job1.t16
OK
MAIN
1258 Marc Users Guide: Part 2
CHAPTER 3.24
Boundary Conditions
DeIine boundary conditions including deIining node sets, adding boundary conditions as well as symmetric
conditions. Note that axisymmetric to 3-D model expansion automatically generates a set oI local coordinate systems
iI there are any y or z (radial or circumIerential) boundary conditions on any nodes Irom the 2-D model. The set oI
local coordinate systems are not needed in this job and will be removed.
Load Steps and Job Parameters
The outside steel tube moves 1 cm in the radial (Y) direction within Iive equal increments; element type 7 is used;
updated Lagrangian Iormulation is used Ior elasticity; and equivalent von Mises stress is written into the post Iile.
BOUNDARY CONDITIONS
MECHANICAL
SELECT
METHOD
BOX
RETURN
NODES
-20 20 -20 20 -1 0.01
nodes STORE
symm
OK
all: SELECT.
CLEAR SELECT
RETURN
TRANSFORMS
UNTRANSFORM
all: EXIST.
RETURN
NEW
FIXED DISPLACEMENT
DISPLACEMENT Z (on)
OK
nodes ADD
symm
outer
inner
MAIN
1259
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Save Model, Run Job, and View Results
Save the model with a diIIerent name to avoid overwriting the existing axisymmetric mode.
LOADCASE
MECHANICAL
STATIC
LOADS
ON: apply3
OK
TOTAL LOADCASE TIME
1
# STEPS
5
OK
MAIN
JOBS
MECHANICAL
INITIAL LOADS
ON: apply3
ON: icondl
OK (twice)
FILE
SAVE AS
rubberbushing_3d.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
Equivalent Von Mises Stress
OK
MONITOR
1260 Marc Users Guide: Part 2
CHAPTER 3.24
Figure 3.24-6 Deformed Mesh and Distribution of Equivalent von Mises Stress at
Beginning of 3-D Analysis
Figure 3.24-7 Deformed Mesh and Distribution of Equivalent von Mises Stress at
Increment 5 of 3-D Analysis
1261
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Automobile Tire Modeling with Rebar Elements
A 3-D Iinite element analysis oI automobile tires is complicated because oI the complex structure oI tires which are
made oI several types oI rubber reinIorced with cord layers and a steal bead. This problem demonstrates the use oI
Axisymmetric to 3-D data transIer capability Ior rebar elements.
Description of Problem
An automobile tire with a smooth tread, denoted as 195/65R15, is analyzed. The model consists oI Iive diIIerent rebar
layers with diIIerent materials and three types oI rubber. See tire2d.mud Ior detailed inIormation oI the model
including the Iinite element discretization oI the cross-section, the material properties, and the rebar locations.
The analysis includes the numerical simulation oI three stages:
mounting the tire on the wheel,
inIlating the tire up to 2.0 bar, and
pressing it against a road surIace.
During the Iirst two stages, the deIormation is purely axisymmetric and, thereIore, an axisymmetric analysis is
perIormed. The simulation oI tire mounting on the wheel is carried out using ten equal increments. AIterwards, the
inIlation pressure is applied with ten more equal increments. The 2-D model (mesh along with loads and boundary
conditions) is then expanded to 3-D Ior Iurther analysis. In the third stage, the tire contacts with the road surIace. A
total movement oI 25 mm oI the tire against the load surIace is applied using the AUTO STEP option and the analysis
is completed aIter ten increments. The analysis steps are summarized in Figure 3.24-8.
The element types 10 and 144 are used in the axisymmetric run. The corresponding 3D element types are 7 and 146.
ELASTICITY,2 is used to activate updated Lagrangian Iormulation.
Figure 3.24-8 Data Transfer from Axisymmetric to 3-D Analysis
Step 1: Axisymmetric
Step 2: Model Expansion to 3-D
Step 3: 3-D Analysis
with Contact
Analysis
1262 Marc Users Guide: Part 2
CHAPTER 3.24
Axisymmetric Analysis
This is a standard axisymmetric analysis. No details will be given in this step. The analysis will be perIormed based
on a completed tire2d.mud Iile. Examine sets rebar1 and rebar2 under MATERIAL PROPERTIES-> LAYERED
MATERIALS-> NEW REBARS Ior rebar deIinitions.
FILES
OPEN
tire2d.mud
OK
MAIN
JOBS
RUN
RESET
SUBMIT 1
MONITOR
OK
TOP
RESULTS
OPEN DEFAULT
DEF & ORIG
MONITOR
CLOSE
MAIN
1263
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-9 Axisymmetric Finite Element Mesh
Figure 3.24-10 Deformed and Undeformed Meshes after Tire Inflation
3-D Analysis
Based on the results Irom the axisymmetric analysis, a Iully 3-D analysis will be perIormed Ior the third stage tire
contact against the road surIace. BeIore the 3-D analysis, a corresponding 3-D mesh on the basis oI the axisymmetric
1264 Marc Users Guide: Part 2
CHAPTER 3.24
mesh and the data transIer Irom the axisymmetric to 3-D cases are required. The mesh in the area oI contact and its
vicinity should be Iiner.
Mesh Expansion and Data Transfer from Axisymmetric to 3-D
BeIore the mesh expansion and data transIer, part oI boundary conditions, which are no longer useIul in 3-D case,
should be removed. It includes the symmetric condition and the load to mount the tire into the wheel.
Mesh expansion Irom axisymmetric to 3-D is based on AXISYMMETRIC MODEL TO 3D option under MESH
GENERATION-> EXPAND. Rotation angles and number oI repetitions must be deIined. Non-equal spaced mesh
expansion is used in the problem to Iorm a relatively Iiner mesh in the contact area and its vicinity. Load table curve
is shiIted in order to properly include the load applied in axisymmetric analysis on the 3-D model. The shiIt time at
which the analysis will continue in a Iully 3-D manner must be deIined.
The model oI the rigid road surIace is input in a separate tire_rigid.mud Iile.
Data transIer Irom axisymmetric to 3-D is based on option AXISYMMETRIC TO 3D under INITIAL CONDITIONS->
MECHANICAL. Both stress and strain tensors must be transIerred once updated Lagrangian Iormulation Ior elasticity
is used; displacement is transIerred by deIault; the name oI the post Iile Irom the completed axisymmetric analysis
must be deIined.
BOUNDARY CONDITIONS
NEXT
REM
NEXT
REM
NEXT
REM
MAIN
MESH GENERATION
EXPAND
AXISYMMETRIC MODEL TO 3D
1 ANGLE
28
1 REPETITIONS
5
2 ANGLE
5
2 REPETITIONS
16
3 ANGLE
28
Figure 3.24-11 AXISMETRIC MODEL TO 3D
with 3 Angles and Repetitions
1265
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-12 FE-Mesh for 3-D Analysis
3 REPETITIONS
5
TIME SET
2
EXPAND MODEL
MAIN
FILES
MERGE
tire_rigid.mud
OK
MAIN
INITIAL CONDITIONS
MECHANICAL
AXISYMMETRIC TO 3D
POST FILE
tire2d_job1.t16
OK
MAIN
1266 Marc Users Guide: Part 2
CHAPTER 3.24
New Contact Definition
Add the rigid surIace as a new contact body and deIine moving velocity Ior the body.
Loadcases, Job Parameters, and Results
The rigid road surIace moves 25 mm toward the tire. AUTO STEP option is used.
Initial condition icond1 must be set on in deIining job parameters. Advanced contact option to control separation is
used. BeIore submit the job, save the model with a diIIerent name to avoid overwriting the axisymmetric model.
CONTACT
CONTACT BODIES
NEW
surfaces ADD
3
#
RIGID
VELOCITY
VELOCITY Y
1
OK (twice)
MAIN
LOADCASES
MECHANICAL
STATIC
TOTAL LOADCASE TIME
25
MULTI-CRITERIA
OK
MAIN
JOBS
MECHANICAL
INITIAL LOADS
ON: icond1
OK
CONTACT CONTROL
ADVANCED CONTACT CONTROL
1267
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-13 Deformed Tire Model
SEPARATION INCREMENT NEXT
SEPARATION CHATTERING
SUPPRESSED
OK (thrice)
FILES
SAVE AS
tire3d.mud
OK
RETURN
RUN
RESET
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
MONITOR
1268 Marc Users Guide: Part 2
CHAPTER 3.24
Analysis of a Rubber Cylinder using Remeshing
This problem is created to demonstrate the application oI axisymmetric to 3-D data transIer with remeshing. The
procedure to perIorm an analysis involving axisymmetric to 3-D data transIer with remeshing requires the use oI the
post Iile with manual application oI loads and boundary conditions as opposed to jobs without a remeshing which can
use a model Iile to simpliIy the application oI loads and boundary conditions in 3-D. Please note that only the
axisymmetric analysis uses global remeshing techniques Ior the MSC.Marc 2001 release.
Description of Problem
A rubber cylinder with an inner radius oI 0.2 and an outer radius oI 0.5 is considered. The length oI the rubber cylinder
is 0.6. Both ends oI the cylinder are glued to two Ilat rigid surIace. Two load sequences are applied. In the Iirst load
case, a displacement oI 0.2 along the negative symmetric axis direction is applied to the right side rigid surIace within
10 equal increments. During this load case, the deIormation is purely axisymmetric and, thereIore, an axisymmetric
analysis is perIormed. Two global remeshing steps are applied at increments 4 and 8, respectively. AIterwards, the
right side steel surIace moves 0.15 in the radial (Y) direction within 10 increments. In the second step, the problem
becomes Iully three-dimensional, and a 3-D analysis is perIormed.
The element type 10 is used in the axisymmetric run. The corresponding 3-D element type is 7. ELASTICITY,2 is used
to activate updated Lagrangian Iormulation.
The rubber cylinder is modeled using Mooney constitutive model. The material properties are given as C
1
8 and
C
2
2.
Axisymmetric Analysis
This is a standard axisymmetric analysis. No details are given in this step. The analysis is perIormed based on a
completed crubcyl2d.mud Iile.
FILES
OPEN
rubcyl2d.mud
OK
MAIN
JOBS
RUN
RESET
SUBMIT 1
MONITOR
OK
TOP
1269
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-14 Axisymmetric Finite Element Mesh
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Equivalent Von Mises Stress
OK
CONTOUR BAND
MONITOR
1270 Marc Users Guide: Part 2
CHAPTER 3.24
Figure 3.24-15 Deformed Meshes after Pressing the Rubber Cylinder
3-D Analysis
Based on the results Irom the axisymmetric analysis, a Iully 3-D analysis is perIormed Ior the second part the rubber
cylinder subjected to shear deIormation. BeIore the 3-D analysis, a corresponding 3-D mesh on the basis oI the
axisymmetric mesh and the data transIer Irom the axisymmetric to 3-D cases are required.
Because oI the use oI global remeshing techniques in the axisymmetric analysis, the numerical results obtained at the
end oI the axisymmetric analysis are no longer based on the original mesh at the beginning oI the analysis, but the new
mesh at the end oI the analysis. The standard mesh expansion procedure based on Mentat mud Iile, used in the previous
two problems, is not used in problems involving AXITO3D and remeshing. The user must use the post Iile and the same
steps as those oI MSC.Marc 2000. See Marc Volume C: AXITO3D Model Definition option Ior details. Please note
that all data which is not available in the post Iile have to be redeIined manually.
Mesh Expansion and Data Transfer from Axisymmetric to 3-D
BeIore the mesh expansion, a rezoning step is needed, based on the post Iile oI the axisymmetric analysis to obtain the
deIormed axisymmetric mesh. Please also save the new model as rubcyl3d.mud and clean the Mentat database.
Mesh expansion Irom axisymmetric to 3-D is based on AXISYMMETRIC MODEL TO 3D option under MESH
GENERATION-> EXPAND. Rotation angles and number oI repetitions must be deIined. The time at which the analysis
will continue in a Iully 3-D manner must be deIined.
Data transIer Irom axisymmetric to 3-D is based on option AXISYMMETRIC TO 3D under INITIAL CONDITIONS->
MECHANICAL. Both stress and strain tensors must be moved once updated Lagrangian Iormulation Ior elasticity is
1271
CHAPTER 3.24
Axisymmetric to 3-D Analysis
used; displacement should not be moved since the mesh is already in deIormed conIiguration; the name oI the post Iile
Irom the completed axisymmetric analysis must be deIined.
DEFORMED SHAPE: OFF
SCALAR PLOT: OFF
TOOLS
REZONE MESH
FILES
SAVE AS
rubcyl3d.mud
OK
NEW
OK
OPEN
rubcyl3d.mud
OK
MAIN
MESH GENERATION
EXPAND
AXISYMMETRIC MODEL TO 3D
1 ANGLE
15
1 REPETITIONS
24
TIME SET
0.2
EXPAND MODEL
MAIN
INITIAL CONDITIONS
MECHANICAL
AXISYMMETRIC TO 3D
DISPLACEMENT (off)
POST FILE
rubcyl2d_job1.t16
OK
MAIN
1272 Marc Users Guide: Part 2
CHAPTER 3.24
Figure 3.24-16 FE-Mesh for 3-D Analysis
Material Properties
MATERIAL PROPERTIES
MOR
E
MOONEY
C10
8
C01
2
OK
elements ADD
all: EXIST.
MAIN
1273
CHAPTER 3.24
Axisymmetric to 3-D Analysis
New Contact Definition
Add the rigid surIace as new contact bodies and deIine moving velocity Ior the body.
Loadcases, Job Parameters, and Results
The right side rigid surIace moves 0.15 toward -Y direction in ten increments. Initial condition icond1 must be set on
in deIining job parameters.
CONTACT
CONTACT BODIES
NEXT contact body left
surfaces ADD
1
#
NEXT contact body right
surfaces ADD
2
#
RIGID
VELOCITY PARAMETERS
VELOCITY Y
-1
OK (twice)
RETURN
CONTACT TABLES
NEW
PROPERTIES
ALL ENTRIES - CONTACT TYPE: GLUE
OK
MAIN
LOADCASES
MECHANICAL
STATIC
CONTACT
CONTACT TABLE
ctable1
OK
1274 Marc Users Guide: Part 2
CHAPTER 3.24
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
TOTAL LOADCASE TIME
0.15
# STEPS
10
OK
MAIN
JOBS
MECHANICAL
LOADCASES: lcase1
INITIAL LOADS
ON: icond1
OK
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK (twice)
ANALYSIS OPTION
rubber elasticity procedure - LARGE STRAIN - TOTAL LAGRANGE
rubber elasticity procedure - LARGE STRAIN - UPDATED LAGRANGE
OK
JOB RESULTS
available element tensors
Stress
Total Strain
available element scalars
Equivalent Von Mises Stress
OK (twice)
SAVE
1275
CHAPTER 3.24
Axisymmetric to 3-D Analysis
Figure 3.24-17 Deformed Rubber Cylinder after Shear Deformation
RUN
RESET
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
MONITOR
1276 Marc Users Guide: Part 2
CHAPTER 3.24
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
Rubber Bushing Axisymmetric to 3D
axisym_3d_a.proc Mentat procedure Iile
Automotive Tire
tire.proc Mentat procedure Iile
tire_rigid.mud Mentat model Iile
tire2d.mud Mentat model Iile
Rubber Cylinder with Remeshing
rubcyl2d.proc Mentat procedure Iile
rubcyl2d.mud Mentat model Iile
Chapter 3.25: Interference Fit
3.25
Interference Fit

Summary 1278

Run Job and View Results 1283

Input Files 1285

1278 Marc Users Guide: Part 2
CHAPTER 3.25
Summary
Title InterIerence Iit
Problem Ieatures Two concentric cylinders are Iitted together with an interIerence Iit
Geometry
Material properties E 3x10
7
Psi, 0.3
Analysis type Static
Boundary conditions Symmetry with interIerence Iit that causes stress
Element type Axisymmetric element type 116
FE results Stresses generated by interIerence Iit (radial and hoop components shown)
cbody1
cbody2
cbody3
0.004 in. interference
between cylinders
v
369
385
401
417
433
449
481 497
0
Comp 22 of Stress (radial)
3.679
-3.414
2.1 0
Y (x10000)
353
Arc Length (along x = 0)
337
305
33
49
65
81
97
113
129
145
161
177
193 209 225 241 242
273
289
321
17
1 465
433
481 497
1
289
305
321
337
353
273
369
401
417
449
465
385
242
241 225
17
33
49
65
81
97
113
129
145
161
177
193 209
Comp 33 of Stress (hoop)
ode 1
Node 241
Node 497
Node 242
1279
CHAPTER 3.25
Interference Fit
Two concentric cylinders are Iitted together with an interIerence Iit using the contact option and rigid bodies oI
symmetry. Each cylinder is modeled using axisymmetric elements.
Since the inner cylinder is slightly bigger than the hole in the outer cylinder, stresses are generated as the Iit is Iinished.
The hoop stress oI the outer cylinder goes into tension, and the hoop stress oI the inner cylinder goes into compression.
Figure 3.25-1 Two Concentric Cylinders and Contour Plotting Analysis
The contour plot shows the ratio oI equivalent stress to strength and is largest in the outer cylinder where it touches
the inner cylinder. Plotting the radial and hoop components along the radius at the symmetry plane illustrates the
continuity oI radial stress across the interIace while the hoop stress switches Irom compression in the inner cylinder
to tension in the outer cylinder.
cbody1
cbody2
cbody3
X
Y
Z
3.224e-001
3.839e-001
4.455e-001
5.070e-001
5.685e-001
6.301e-001
6.916e-001
7.531e-001
8.147e-001
8.762e-001
9.377e-001
lcase1
Equivalent Stress/Yield Stress
Inc: 1
Time: 1.000e+000
X
Y
Z
0.004 in. interference
between cylinders
369
385
401
417
433
449
481 497
0
Comp 22 of Stress
3.679
-3.414
2.1 0
Time : 1
Inc : 1
Y (x10000)
353
Arc Length (along x = 0)
337
305
33
49
65
81
97
113
129
145
161 177 193 209 225 241 242
273
289
321
17
lcase1
1 465
433
481 497
1
289
305
321
337
353
273
369
401
417
449
465
385
242
241 225
17
33
49
65
81
97
113
129
145 161 177 193 209
Comp 33 of Stress
Node 1
Node 241
Node 497
Node 242
1280 Marc Users Guide: Part 2
CHAPTER 3.25
FILES
NEW
OK
SAVE AS
interf
MAIN
MESH GENERATION
COORDINATE SYS:
SET GRID ON
U DOMAIN
0 1.1
U SPACING
0.1
V DOMAIN
0 3.1
V SPACING
0.1
FILL
RETURN
CRVS: ADD
POINT(0.0,0.0,0.0)
POINT(0.0,3.1,0.0)
ELEMENTS: ADD
NODE(0.0, 1.0,0.0)
NODE(1.1, 1.0,0.0)
NODE(1.1, 2.0,0.0)
NODE(0.0, 2.0,0.0)
NODE(0.0, 2.1,0.0)
NODE(1.0, 2.1,0.0)
NODE(1.0, 3.1,0.0)
NODE(0.0, 3.1,0.0)
v=3.1
u=1.1 X
Y
Z
1281
CHAPTER 3.25
Interference Fit
SUBDIVIDE
DIVISIONS
15 15 1
ELEMENTS
ALL: EXISTING
RETURN
SWEEP
REMOVE UNUSED:
NODES
ALL
RETURN
RENUMBER
NODES DIRECTED
0.0001 1 0
RETURN
MOVE
TRANSLATIONS
0 -0.1 0
ELEMENTS
(pick top cylinder)
END LIST
MAIN
MATERIAL PROPERTIES (twice)
NEW: STANDARD
STRUCTURAL
YOUNGS MODULUS = 3E7
POISSONS RATION = .3
PLASTICITY (twice)
INITIAL YIELD STRESS
5E4
OK (twice)
ELEMENT ADD
ALL: EXISTING
RETURN
CONTACT
CONTACT BODIES
X
Y
Z
Axis of Rotaton
cbody1
cbody2
cbody3
X
Y
Z
0.004 in. interference
between cylinders
1282 Marc Users Guide: Part 2
CHAPTER 3.25
DEFORMABLE,
OK
ELEMENTS: ADD
(pick inner cylinder)
NEW
DEFORMABLE
OK
ELEMENTS ADD
(pick outer cylinder)
NEW
SYMMETRY
DISCRETE
OK
CURVES ADD
(pick symmetry curve)
ID CONTACT
RETURN
CONTACT TABLES
NEW
PROPERTIES
ALL ENTRIES: CONTACT TYPE: TOUCHING
TOUCHING BODIES (pick T in 1,2 position in table)
cbody1
cbody2
INTERFERENCE CLOSURE
4E-3
OK (twice)
MAIN
LOADCASES
MECHANICAL
STATIC
CONTACT
CONTACT TABLE
ctable1
OK
1283
CHAPTER 3.25
Interference Fit
Run Job and View Results
# STEPS
1
OK
MAIN
JOBS
NEW MECHANICAL
PROPERTIES
lcase1
AXISYMMETRIC
JOB RESULTS
AVAILABLE ELEMENT SCALARS: Equivalent Von Mises Stress
AVAILABLE ELEMENT SCALARS: Equivalent Von Mises Stress/yield Stress Ratio
AVAILABLE ELEMENT TENSORS: Stress
OK (twice)
ELEMENT TYPES
MECHANICAL
ANALYSIS DIMENSION: AXISYMMETRIC
SOLID
116
OK
ALL: EXISTING
RETURN
SAVE
RUN
SUBMIT(1)
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
LAST
1284 Marc Users Guide: Part 2
CHAPTER 3.25
Figure 3.25-2 Equivalent Stress to Strength Ratio
SCALAR
EQ. STRESS/YIELD
OK
CONTOUR BANDS
RESULTS
PATH PLOT
NODE PATH
1 241 242 497
END LIST
ADD CURVES
ADD CURVE
ARC LENGTH
Comp 22 Of Stress
ADD CURVE
ARC LENGTH
3.224e-001
3.839e-001
4.455e-001
5.070e-001
5.685e-001
6.301e-001
6.916e-001
7.531e-001
8.147e-001
8.762e-001
9.377e-001
lcase1
Equivalent Stress/Yield Stress
Inc: 1
Time: 1.000e+000
X
Y
Z
1285
CHAPTER 3.25
Interference Fit
Figure 3.25-3 Stresses Plotted Across Interface
Component 22 oI stress is the radial stress. It is in compression and is continuous across the interIace between the two
cylinders. Also, the radial stress vanishes on the Iree surIaces oI the cylinders.
Component 33 oI stress is the hoop stress, with the inner cylinder being compressed and the outer cylinder being
expanded. The Equivalent Stress/Yield Strength ratio in the contour plot show that the outer cylinder at 94 oI yield.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Comp 33 Of Stress
FIT
File Description
s7.proc Mentat procedure Iile
369
385
401
417
433
449
481 497
0
Comp 22 of Stress (radial)
3.679
-3.414
2.1 0
Time : 1
Inc : 1
Y (x10000)
353
Arc Length (along x = 0)
337
305
33
49
65
81
97
113
129
145
161
177
193 209 225 241 242
273
289
321
17
lcase1
1 465
433
481 497
1
289
305
321
337
353
273
369
401
417
449
465
385
242
241 225
17
33
49
65
81
97
113
129
145
161
177
193 209
Comp 33 of Stress (hoop)
Node 1
Node 241
Node 497
Node 242
1286 Marc Users Guide: Part 2
CHAPTER 3.25
Chapter 3.26: 3-D Remeshing with Tetrahedral Elements
3.26
3-D Remeshing with
Tetrahedral Elements

Chapter Overview 1288

Why Remeshing with Tetrahedral Elements? 1288

Tetrahedral Element Type 157 1288

Tetrahedral Remeshing Criteria 1290

Tetrahedral Remeshing Controls and Meshing Parameters 1291

Tetrahedral Remeshing Tests 1293

Input Files 1313

1288 Marc Users Guide: Part 2
CHAPTER 3.26
Chapter Overview
This chapter describes the capability Ior 3-D global remeshing. For analysis using the updated Lagrange Iormulation
based Iinite element method (FEM), one oIten encounters element distortion in applications that involve large
deIormation. When elements become too distorted the analysis Iails. The global remeshing Ieature alleviates this
situation by automatically generating a new mesh, transIerring history data Irom the previous mesh, and resuming the
analysis. Global remeshing also helps improving the analysis by mesh reIinement in the area where small elements are
required due to contact and geometry change, and speeding up the analysis by generating larger elements in the area
that does not require small elements.
Why Remeshing with Tetrahedral Elements?
The tetrahedral element mesh generator has been proved to be the most robust and Iastest method among other types
oI 3-D mesh generators. It is much easier Ior a mesh generator to automatically Iill in an arbitrary geometry with
tetrahedrons than with other elements oI the diIIerent shapes. The meshing technology, such as Delaunay triangulation
and pavement method, has been used successIully in generating triangular and tetrahedral meshes. Marc uses the mesh
generator Irom Patran (or GS-Mesher) to generate the mesh. A mesh-on-mesh technology (MOM mesher) is employed
to mesh the surIace with triangular elements. Subsequently, a tetrahedral mesh generator, using the Delaunay
triangulation and pavement methods, is used to create the Iinal mesh with the tetrahedral elements.
Tetrahedral Element Type 157
The tetrahedral remeshing uses Marc element type 157. The tetrahedral element type 157 is a Herrmann type element,
which typically uses pressure as well as displacement in the FEM analysis. These elements with the mixed unknowns
(or degrees oI Ireedom) allow the element to model incompressible materials undergoing large shear deIormation.
Standard displacement based tetrahedral element cannot perIorm well in this situation because the element locks and
hence lacks Ilexibility. Element type 157 has 5 nodes, with 4 corner nodes and one interior node. There is one pressure
degree oI Ireedom and three displacement degrees oI Ireedom in each corner node while only three displacement
degrees oI Ireedom are in the interior node.
1289
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Example: An Upsetting Compression to Test Incompressibility and Thermal Coupling
Figure 3.26-1 shows the location oI the two interested points used in the comparison. Figure 3.26-2 displays
temperature distribution in the test and Figure 3.26-3 shows nodal temperature changes and displacement change at
these two nodal positions.
Figure 3.26-1 A Corner and Center Nodes
Figure 3.26-2 Temperature Distribution Comparison
1290 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-3 Temperature and Displacement Comparisons
It can be seen that element 157 behaves very well compared with element 7.
Tetrahedral Remeshing Criteria
The remeshing criteria are used to initiate the remeshing process. There are six remeshing criteria that may be used,
either separate or in combination.
1. Increment Irequency: Users can speciIy remeshing intervals so that aIter certain number oI increments,
global remeshing is perIormed.
2. Strain change: An accumulated incremental strain measure is recorded aIter each remeshing.
When this value reaches or exceeds the maximum allowed, remeshing is
initiated. This criterion controls the magnitude oI the deIormation between each
remeshing step.
3. Penetration: Penetration is checked against each contacting body. When penetration reaches or
exceeds the maximum allowed, the remeshing step starts. The penetration distance
is measured between a triangle Iace element and its central point projection to the
other contact bodies. The penetration limit (deIault value set at two times oI the
contact tolerance) can also be speciIied. This criterion is useIul when contacting
with rigid bodies that have sharp corners. It helps remeshing body correct its
geometry to avoid Iurther penetration. The penetration criterion cannot be used in
selI-contact situation.
4. Volume Ratio Distortion: This criterion checks element distortion based on its volume. A ratio oI the height
and the base triangle is used to make sure the element is in a good shape Ior
computation. A ratio oI 1.0 indicates a good element while a ratio oI 0.0 means a
Ilat element, not suitable Ior the analysis. A control value to avoid large element
distortion can also be speciIied.
1291
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Tetrahedral Remeshing Controls and Meshing Parameters
The tetrahedral element remeshing requires REZONING,2 in the parameter section and an ADAPT GLOBAL option in
the model or history section. A standalone mesh generator, afmesh3d, is needed in the bin directory along with the
Marc FEM solver. The GS-mesher library is linked to afmesh3d to perIorm surIace and tetrahedral meshing (see
Figure 3.26-4). This library is normally located in the lib directory. Without all these components ,the global remeshing
does not perIorm properly.
Figure 3.26-4 Global Remeshing Controls
When remeshing is required, the solver writes an input data Iile, jid_bxx.fem Ior afmesh3d and is then called to
generate a mesh. The output Iile Irom afmesh3d, jid_bxx.feb is read into the solver. The two digit number, xx provides
the contact body number. While remeshing, the solver can be either in a waiting state or terminated temporarily to save
memory Ior the mesh generator. AIter the new mesh is created, the solver automatically resumes.
The remeshing parameters control how the new mesh is created. These control parameters are:
5. Immediate Remeshing: This control is used to remesh the body beIore the next analysis step. It is useIul
when you want to switch a model oI a hexahedral mesh to a tetrahedral mesh beIore
the Iinite element analysis starts. It can also be used with restart option to
immediately remesh the body aIter the restart. Immediate remeshing allows the
change element type Irom hexahedral element type 7 to tetrahedral element type
157 but not vise versa.
6. Forced Global
Remeshing:
This control is used internally together with automatic time step cutback Ieature. II
the global remeshing control is used and a bad mesh is encountered during the
iteration cycle, Marc automatically Iorces the job to create a new mesh. II the new
mesh does not help, the time step cutback is then enIorced.
1. Element Edge Length: This element size controls the size oI the new mesh although some reIinement
and coarsening overrule the element size here.
2. Number oI Elements: This controls approximately the number oI the elements in the new mesh. It
gives a guideline Ior Marc to deIine an element size Ior the mesh generation.
II element edge length and the number oI the elements are not given, the
number oI the elements in the previous mesh are used to create the new mesh.
3. Previous Number oI Elements: It uses the number oI elements in the current mesh as a target to create the
new mesh.
1292 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-5 Interior Coarsening of a Tetrahedral Mesh
4. Feature Edge Angle: An edge is preserved aIter remeshing iI any surIace edge angle exceeds this
value. A Ieature edge angle is measured between two connected Iace elements
in such a way that 0 indicates the two Iace elements lie on a planar surIace
while 180 indicates that the elements are touching each other. The deIault
value is 60.
5. Feature Vertex Angle: While the edge angle controls the edges the vertex angle controls points. II a
point on certain edge is smaller than this value, the point is kept aIter
remeshing. The vertex angle measures the Ieature oI two connecting edges. It
is calculated in such a way that the angle is 180 iI the two edges lie in a
straight line. Thus, a 0 means the two edges are touching each other. The
deIault value is 100.
6. Coarsening Factor: This parameter allows creation oI larger tetrahedral elements in the interior. To
capture contact conditions accurately, the mesh usually contain small elements
on the surIace oI a body. By enlarging the element gradually inside a body, we
reduce the number oI elements in the mesh (see Figure 3.26-5). The
coarsening Iactor scales the element size Irom the surIace inwards. Thus, a
coarsening Iactor oI 1.0 means no coarsening. The deIault value oI the
coarsening Iactor is 1.5 times.
Note: Making element too large will aIIect the accuracy oI the analysis results.
1293
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-6 Divisions of a Curvature Circle
Tetrahedral Remeshing Tests
Many tests have been perIormed with good results.
7. Minimum Edge Length: This parameter controls the smallest elements allowed on the surIace mesh. It is
used when local reIinement or adaptive meshing is required. The deIault value is set
at 1/3 oI the element edge length.
8. Maximum Edge Length: This parameter controls the largest elements allowed on the surIace mesh. It is used
when local reIinement or adaptive meshing is required. The deIault value is set at
three times oI the element edge length.
9. Curvature Controls: This parameter controls adaptive meshing on the surIace based on the surIace
curvature. Thus, a surIace that is curved gets smaller elements than the surIace that
is Ilat. The number oI divisions indicates number oI the divisions to subdivide a
curvature circle (see Figure 3.26-6). It shows the sensitivity oI this curvature
control. By deIault, this control is oII. But a number oI ten is considered a good
number Ior the general applications.
10. Change Element Type: This parameter is only required iI element type is to be changed aIter remeshing. This
is oIten used to switch Irom the original hexahedral mesh to the tetrahedral mesh.
Currently, the only available type is 157 in Marc.
1. Rubber Seal Simulation: This example shows remeshing application in rubber seal simulation and is used to
demonstrate the remeshing in the Iorthcoming sections.See Figures 3.26-7 and
3.26-8.
Division = 3 Division = 6
1294 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-7 Initial Setup with One Element
Figure 3.26-8 Deformation at Increment 50
2, Double-sided Contact: This example shows two deIormable body subjected to contact with remeshing. It
shows possibility oI the global remeshing with multiple deIormable bodies. See
Figures 3.26-9 and 3.26-10.
1295
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-9 Initial Setup of Two Contact Bodies
Figure 3.26-10 Deformation At Increment 20
3. Hot Compression oI a Steel Block: This example shows capability oI remeshing in a thermal-mechanical
coupled analysis. Curvature local reIinement can be seen in the Iinal
results. See Figures 3.26-11 and 3.26-12.
1296 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-11 Initial Setup
Figure 3.26-12 Deformation and Temperature at Increment 100
1297
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Many metal Iorming applications are carried out with the help oI the tetrahedral global remeshing. Listed below are
some oI the industrial examples which would not have been possible without remeshing.
Figure 3.26-13 Initial Mesh of a Connecting Rod
Figure 3.26-14 Final Mesh of a Connecting Rod
1. Connecting Rod Forging: This example shows an open die Iorging simulation. The Ilash being extruded out
Irom the die is seen in the Iinal result. The global remeshing permits the solution oI
the large material Ilow. See Figures 3.26-13 and 3.26-14.
1298 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-15 Initial Mesh of a Turbine Blade Preform
Figure 3.26-16 Final Result of a Turbine Blade Forging
2. Turbine Blade Forging: This example shows bending, twisting, and compression oI the turbine blade. The
global remeshing helps the large deIormation in the compression stage. See Figures
3.26-15 and 3.26-16.
1299
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-17 Initial Shape of the Workpiece
Figure 3.26-18 Final Results
3. Flange Forging: This example shows a closed die Iorging simulation. Material Ilows to Iill up the closed die
cavity. The global remeshing permits the large deIormation observed in the simulation
without external intervention due to mesh distortion. See Figures 3.26-17, 3.26-18, and
3.26-19.
1300 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-19 A Closer Look at the Final Mesh
Elastomeric Seal Simulation
A rubber seal with a rectangular cross-section (1.8x1.2 cm
2
) is pressed laterally by rigid tool. Because oI the symmetry,
only a halI oI the seal is considered. With a thickness oI 0.2cm, the model is setup as a 3-D problem. Assuming this is
a long rubber seal in the thickness direction, two symmetry surIaces are used. For the placement oI the rubber seal and
the rigid tools, see Figure 3.26-20. The tool pressure is applied to the top oI the seal and simulated by moving the top
rigid surIace down with a velocity oI 1cm/sec. In the current release, only volumetric loads are automatically reapplied
aIter remeshing. Total load is reached in 50 steps in the analysis with the time step equal to 0.01 second.
Figure 3.26-20 Initial Setup of the Model
1301
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Although the geometry itselI is simple, without remeshing the severely deIormed conIiguration leads to a premature
termination oI the analysis due to excessive distortion in the elements and penetration between contact bodies.
Remeshing/rezoning operation is clearly required Ior a successIul completion oI the analysis.
The analysis starts with one single hexahedral element (to demonstrate that a very crude model can be initially given,
iI the model is remeshed at increment 0 beIore the analysis begins). AIter remeshing, the hexahedral element is
converted into tetrahedral elements (see Figure 3.26-21).
Figure 3.26-21 Tetrahedral Mesh After Immediate Remeshing
In the rest oI the analysis, the remeshing/rezoning is done based on the penetration check to prevent severe penetration
between contact bodies. An adaptive meshing based on the surIace curvature is used to generate smaller elements near
the curved areas. It allows the analysis to capture the geometry changes correctly in those areas without creating
excessive number oI the elements to slow down the analysis (see Figure 3.26-22). Element type 157 is used in the
analysis within the updated Lagrangian Iramework.
1302 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-22 Adaptive Meshing based on Curvature
The rubber seal is modeled using Mooney constitutive model. The material parameters are given as C
1
8N/cm
2
and
C
2
2N/cm
2
. The bulk modulus is 10000N/cm
2
.
Model Generation
We create the model by reading the predeIined model Iiles. This assumes that the users are Iamiliar with model
generation. Two model Iiles are directly read in - element.mfd and rigid_bodies.mfd.
FILE
OPEN
Open file: element.mfd
OK
MERGE
Merge file: rigid_bodies.mfd
OK
1303
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-23 Read in the Predefined Model
Save the model as tet_rubber.
Material Properties
Mooney type oI material is used Ior the rubber seal.
SAVE AS
Save file: tet_rubber.mfd
OK
MAIN (to return to the main menu)
MATERIAL PROPERTIES
Mechanical material types:
MORE
MOONEY
C10
8
C01
2
BULK MODULUS
10000
OK
1304 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-24 Enter Mooney Material Properties
Contact Definitions
Contact body type:
ELEMENTS ADD
ALL EXIST
MAIN
CONTACT
CONTACT BODIES
NEW
NAME
Rubber
DEFORMABLE
OK
Element: ADD
EXISTING
NEW
NAME
1305
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
bot
Contact body type:
RIGID
DISCRETE
OK
Surfaces: ADD
1 (pick surface number 1)
END LIST (#)
NEW
NAME
top
contact body type:
RIGID
DISCRETE
OK
Surfaces: ADD
4 (pick surface number 4)
END LIST (#)
NEW
NAME
sym1
contact body type:
SYMMETRY
DISCRETE
OK
Surfaces: ADD
6 (pick surface number 6)
END LIST (#)
COPY
NAME
sym2
Surfaces: ADD
5 (pick surface number 5)
END LIST (#)
COPY
NAME
1306 Marc Users Guide: Part 2
CHAPTER 3.26
DeIine the pusher:
sym3
Surfaces: ADD
2 (pick surface number 2)
END LIST (#)
NEW
NAME
push
contact surface type:
RIGID
Body control:
VELOCITY
PARAMETERS
VELOCITY
Y: -1
OK
DISCRETE
OK
Surfaces: ADD
3 (pick surface number 3)
END LIST (#)
ID CONTACT (you can see the contact body IDs now)
MAIN
1307
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-25 Define Contact Bodies
Mesh Adaptivity
You can see many new buttons in this section. Two remeshing criteria will be speciIied Ior the deIormable body.
Additional inIormation will be provided to control the meshing process. It is desired that the new elements have edge
dimensions between 0.03cm and 0.1cm.
GLOBAL REMESHING
PATRAN TETRA (select Patran Tetrahedral Mesher)
Remeshing criteria:
IMMEDIATE (immediate remeshing)
ADVANCED
PENETRATION
USER:LIMIT
0.005 (maximum penetration distance)
OK
Remeshing parameters:
ELEMENT EDGE LENGTH: SET
0.1 (element size for the new mesh)
ADVANCED
MINIMUM EDGE LENGTH
0.03 (limit the smallest element size)
1308 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-26 Adaptive Remeshing Controls
Loadcases
The loadcase option is used to deIine the time period and to activate the global adaptive meshing criteria.
CURVATURE CONTROL:#DIV
10 (curvature control number)
CHANGE ELEMENT TYPE: 157
OK
REMESH BODY
Rubber
MAIN
LOADCASES
MECHANICAL
STATIC
GLOBL REMESHING
adapg1
OK
TOTAL LOADCASE TIME
0.5
1309
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Figure 3.26-27 Loadcase Definition
Jobs and Run Analysis
There are a Iew control settings in this section that need special attention. Because the number oI potential contact
nodes increase due to adaptive meshing, it is necessary to give an upper bound. In the model, 2000 is used.
CONSTANT TIME STEP: #STEPS
50
OK
MAIN
JOBS
MECHANICAL
Available:
lcase1
MESH ADAPTIVITY
MAX.# CONTACT NODES/BODY:
2000
OK
ANALYSIS OPTIONS
LARGE DISPLACEMENT
ADVANCED OPTIONS
1310 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-28 Figure 16-9: Job Definition
Use SAVE to save the model. The job can be run using
UPDATE LAGRANGE PROCEDURE
OK
Rubber elasticity procedure:
LARGE STRAIN-UPDATED LAGRANGE
JOB RESULTS
Cauchy stress
OK
Analysis dimension
3-D
OK
RUN
SUBMIT (1)
MONITOR (this can be used to monitor the job status)
MAIN
1311
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Results
Figure 3.26-29 Deformation at Increment 10
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
Equivalent Cauchy Stress
OK
FILL
DYN.MODEL (use this to rotate the model to good viewing position)
MONITOR
1312 Marc Users Guide: Part 2
CHAPTER 3.26
Figure 3.26-30 Deformation at Increment 25
Figure 3.26-31 Deformation at Increment 50
1313
CHAPTER 3.26
3-D Remeshing with Tetrahedral Elements
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
rubber_seal_3d.proc Mentat procedure Iile
rigid_bodies.mfd Mentat model Iile
element.mfd Mentat model Iile
1314 Marc Users Guide: Part 2
CHAPTER 3.26
Chapter 3.27: Rubber Remeshing and radial Expansion of Rigid Surfaces
3.27
Rubber Remeshing and Radial
Expansion of Rigid Surfaces

Chapter Overview 1316

Model Highlights 1316

Results Highlights 1320

Modeling Tips 1321

Input Files 1321

1316 Marc Users Guide: Part 2
CHAPTER 3.27
Chapter Overview
This Ieature demonstrates how to grow rigid bodies via a remeshing example. The example is a rubber bushing with
a cap. The cap is automatically remeshed while utilizing expandable rigid bodies to expand the bushing into the outer
rigid housing. In actual applications, the rubber bushing would be a Iull cylindrical shape, however, a small segment
is used here to keep run times reasonable. The bushing along with the rigid shaIt and housing is shown in Figure 3.27-1.
The deIormable bushing is bounded by two surIaces oI symmetry shown in Figure 3.27-3.
Figure 3.27-1 Rubber Bushing with Housing and Rigid Shaft
The rubber is modeled with a Mooney material Ior both the bushing and cap. The cap is a separate contact body that
will be glued to the rest oI the bushing. The cap is automatically remeshed during the Iirst increment. The global
remeshing Iills the cap volume with low-order tetrahedral elements. Marc uses the meshing technology in the Patran
GS-mesher to create meshes with tetrahedral elements. The Mesh On Mesh (MOM) surIace mesher and tetrahedral
mesher are called separately within the Marc solver to produce new mesh. The loading only consists oI the inner rigid
shaIt will being expanded using the UGROWRIGID user subroutine.
Model Highlights
First, let`s open the model and examine the model highlights.
FILE
OPEN
rubber_remesh.mud
OK
1317
CHAPTER 3.27
Rubber Remeshing and Radial Expansion of Rigid Surfaces
Figure 3.27-2 Elements Marked for Global Remeshing
VIEW
SHOW VIEW 4
FILL
PLOT
NODES (turn drawing off)
POINTS
CURVES
SURFACES
IDENTIFY
GLOBAL REMESHING CRITERION
DRAW
MAIN
1318 Marc Users Guide: Part 2
CHAPTER 3.27
Figure 3.27-3 Contact Bodies and Contact Table
MESH ADAPTIVITY
GLOBAL REMESHING CRITERION
PATRAN TETRA
OK
MAIN
PLOT
SURFACES (Turn on)
IDENTIFY
CONTACT BODIES
DRAW
MAIN
CONTACT
CONTACT TABLE
PROPERTIES
OK
MAIN
Remeshing occurs every 12 increments, with
an element edge length oI 0.05.
1319
CHAPTER 3.27
Rubber Remeshing and Radial Expansion of Rigid Surfaces
The global remeshing criterion is Ior the cap body, it occurs with a Irequency oI 12 increments, immediately
(increment 1) and with an element size oI 0.05. The loadcase used only 10 increments, so there will only be one
remeshing to keep run times short.
In order to expand the inner surIace, two things must be done. First, we add an additional input string 'umotion,2 to
the model deIinition using the new Additional Input File Text Ieature in the JOBS menu.
Secondly, we need to write the Iollowing Iile, say rubber_remesh.f, that contains the user subroutine, ugrowrigid.f
which is listed below. This Iile is selected in the RUN-> JOB menu.
subroutine ugrowrigid(md,relx,rely,relz,time)
implicit real*8 (a-h,o-z)
c user subroutine for definition of relative size of rigids
c
c md : rigid body number
c relx : relative size in x-direction with respect to original
c rely : relative size in y-direction with respect to original
c relz : relative size in z-direction with respect to original
c time : time
c relx,rely and relz should be defined by the user
relx=1.0d0
rely=1.0d0
relz=1.0d0
if(md.eq.3) then
if(time.le.1.0d0) then
rely=1.0d0 + 1.6d0*time
else
rely=2.6d0
end if
relz=rely
write(6,*) md,relx,rely,relz=,md,relx,rely,relz
end if
return
end
JOBS
ADDITIONAL INPUT FILE TEXT
umotion,2
OK
RUN
SUBMIT(1)
OK
MAIN
1320 Marc Users Guide: Part 2
CHAPTER 3.27
Results Highlights
AIter submitting the job with the user subroutine, the results are shown in Figure 3.27-4.
Figure 3.27-4 Rubber Remeshing
Notice how the cap, originally made oI hexahedral elements, is remeshed with tetrahedral elements as shown in
Figure 3.27-4. Also notice that the internal rigid cylinder expands in the radial direction, as controlled by the
UGROWRIGID user subroutine.
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
EQUIVALENT TOTAL STRAIN
OK
LAST
1321
CHAPTER 3.27
Rubber Remeshing and Radial Expansion of Rigid Surfaces
Modeling Tips
The UGROWRIGID user subroutine can be replaced by placing a table (1 1.6*time) to control the y and z growth
Iactors in the deIinition oI body, rigidshaIt.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
rubber_remesh.mud Mentat model Iile
rubber_remesh.f User subroutine
1322 Marc Users Guide: Part 2
CHAPTER 3.27
Chapter 3.28: Automatic Remeshing/Rezoning
3.28
Automatic Remeshing/
Rezoning

Chapter Overview 1324

Elastomeric Seal Simulation 1325

Tape Peeling Simulation 1335

Input Files 1347

1324 Marc Users Guide: Part 2
CHAPTER 3.28
Chapter Overview
In the analysis oI metal or rubber, the materials may be deIormed Irom some initial shape to a Iinal, very oIten complex
shape. During the process, the deIormation can be so large that the mesh, used to model the materials, may become
highly distorted and the analysis cannot go any Iurther without using some special techniques. Remeshing/rezoning in
Marc is a useIul Ieature to overcome the diIIiculties. The global remeshing described here completely regenerates a
mesh over a speciIied body.
In the releases beIore MSC.Marc 2000, the global remeshing/rezoning was done manually. When the mesh becomes
too distorted because oI the large deIormation to continue the analysis, the analysis is stopped. A new mesh is created
based on the deIormed shape oI the contact body to be rezoned. A data mapping is perIormed to transIer necessary
data Irom the old, deIormed mesh to the new mesh. The contact tolerance is recalculated (iI not speciIied by user) and
the contact conditions are redetected. The analysis then continues.
With the release oI MSC.Marc 2000, the above steps can be done automatically. Based on the diIIerent user-speciIied
remeshing criteria, the program determines when the remeshing/rezoning is required. The automatic remeshing
control can be instructed through the ADAPT GLOBAL option or through automatic time stepping control. With
automatic time stepping control, remeshing can be Iorced when the mesh oI the body is distorted during the analysis.
When remeshing/rezoning on a 2-D application, the program Iinds the outline oI the body to be rezoned and repairs
the outline to remove possible penetrations. Then, the program calls the mesher to create a new mesh based on the
cleaned outline. Furthermore, the program perIorms data transIer Irom the old mesh to the new mesh, redetermines
the contact conditions, and continues the analysis.
The 2-D automatic remeshing includes checking the outline curvature and thin region Ior local mesh reIinement to
produce better mesh that captures the changes in the geometry. Another remeshing criterion allows you to control the
new mesh by speciIying the target number oI elements rather than the element edge length. You can also control
number oI elements based on the previous mesh or using the percentage tolerance. When using the penetration
remeshing criterion, you can speciIy the penetration tolerance to control when to remesh.
This chapter demonstrates the capability oI the automatic remeshing/rezoning Ieature available in Marc by Elastomeric
Seal Simulation and Tape Peeling Simulation. Steps on remeshing criterion deIinition are highlighted.
Notes: All loading and boundary conditions on bodies being remeshed must be applied using contact (rigid) bodies.
However, in the analysis with remeshing, you can apply loads and boundary conditions on the bodies not
being remeshed.
Numbering oI Contact Bodies:
DeIault: When deIining contact bodies Ior a deIormable-to-deIormable analysis, it is important to
deIine them in the proper order. As a general rule:
A body with a Iiner mesh should be deIined beIore a body with a coarser mesh. This rule applies both
beIore and aIter remeshing.
In case oI a contact between bodies with large diIIerence in stiIIness, like rubber and steel, the soIter
body should have the lowest number.
Automatic: This is described in this chapter and can be very important Ior remeshing problems.
1325
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Elastomeric Seal Simulation
A rubber seal with a rectangular cross-section (1.8x1.2) is pressed laterally by rigid die. The plane strain condition is
assumed. Because oI the symmetry, only a halI oI the seal is considered. See Figure 3.28-1 below Ior the placement oI
the rubber seal and the rigid dies. The die pressure is added to the top oI the seal and simulated by moving the top rigid
surIace down. In the current release, other than volumetric loads, one cannot put a boundary condition on a mesh that
will automatically be remeshed. Total load is applied in two steps in the analysis. In the Iirst step, the top rigid surIace
moves down 0.2 cm within 5 equal increments. In the second step, the top rigid surIace moves down 0.5 cm within 95
equal increments.
Figure 3.28-1 Simulation of a Rubber Seal: FE-Mesh and Geometry
Although the geometry itselI is simple, the severely deIormed conIiguration at an intermediate stage leads to a
premature termination oI the analysis due to excessive distortion in the elements and penetration between contact
bodies (see Figure 3.28-2). Remeshing/rezoning operation is clearly required Ior a successIul completion oI the
analysis.
The analysis starts with one single element (obviously, one element is not enough to model the rubber seal). A
remeshing is perIormed at increment 0 to demonstrate that a very crude model can be initially given, iI the model is
remeshed beIore the analysis begins. AIterwards, the remeshing/rezoning is done at each Iive 5 increment interval to
prevent Irom highly distorted elements and severe penetration between contact bodies. Element type 11 is used in the
analysis within the updated Lagrangian Iramework.
The rubber seal is modeled using Mooney constitutive model. The material parameters are given as C
1
8N/cm
2
and
C
2
2N/cm
2
. The bulk modulus is 10000N/cm
2
.
1326 Marc Users Guide: Part 2
CHAPTER 3.28
Figure 3.28-2 Analysis without using Remeshing/Rezoning
Analysis
Model Generation
Model generation contains geometry deIinition, element deIinition, clean geometry, and clean mesh.
See Figure 3.28-1.
MESH GENERATION
PTS ADD
-1.0e+0 -9.0e-1 0.0e+0
9.5e-1 -9.0e-1 0.0e+0
-9.0e-1 1.0e+0 0.0e+0
-9.0e-1 -1.0e+0 0.0e+0
-1.0e+0 9.0e-1 0.0e+0
-2.0e-1 9.0e-1 0.0e+0
-3.0e-1 1.0e+0 0.0e+0
-3.0e-1 -5.0e-1 0.0e+0
-2.0e-1 -6.0e-1 0.0e+0
-1.0e-1 -6.0e-1 0.0e+0
0.0e+0 -6.0e-1 0.0e+0
1327
CHAPTER 3.28
Automatic Remeshing/ Rezoning
1.0e-1 -5.0e-1 0.0e+0
2.0e-1 -5.0e-1 0.0e+0
3.0e-1 -6.0e-1 0.0e+0
4.0e-1 -6.0e-1 0.0e+0
5.0e-1 -5.0e-1 0.0e+0
6.0e-1 -5.0e-1 0.0e+0
7.0e-1 -6.0e-1 0.0e+0
8.0e-1 -6.0e-1 0.0e+0
9.5e-1 -6.0e-1 0.0e+0
CRVS ADD
1 2 3 4 5 6 7 8
CURVE TYPE
INTERPOLATE
RETURN
CRVS ADD
9 10 11 12 13 14 15 16 17 18 19 20
END LIST
CURVE TYPE
FILLET
RETURN
CRVS ADD
4 5 0.1
CURVE TYPE
COMPOSITE
RETURN
CRVS ADD
4 6 5
END LIST
NODES ADD
-9.0e-1 -9.0e-1 0.0e+0
-3.0e-1 -9.0e-1 0.0e+0
-3.0e-1 9.0e-1 0.0e+0
-9.0e-1 9.0e-1 0.0e+0
ELEMS ADD
1 2 3 4
SWEEP
1328 Marc Users Guide: Part 2
CHAPTER 3.28
Material Properties
Contact Definition
In an analysis using automatic remeshing/rezoning techniques, all boundary conditions and loads applied to the body
to be remeshed/rezoned are applied via proper deIinition oI contact (rigid) bodies. The die pressure is simulated by the
motion oI top rigid surIace and the symmetric condition is modeled using a speciIically designed rigid body (leIt rigid
surIace).
ALL
REMOVE UNUSED NODES
REMOVE UNUSED POINTS
RETURN
RENUMBER
ALL
RETURN
MAIN
MATERIAL PROPERTIES
MORE
MOONEY
C10
8
C01
2
BULK MODULUS
10000
OK
ELEMENTS ADD
ALL EXIST
MAIN
CONTACT
CONTACT_BODIES
ID CONTACT
NEW
DEFORMABLE
OK
ELEMENTS ADD
1329
CHAPTER 3.28
Automatic Remeshing/ Rezoning
ALL EXIST
NEW
RIGID
DISCRETE
OK
CURVES ADD
1
END LIST
NEW
RIGID
DISCRETE
OK
CURVES ADD
4
END LIST
NEW
RIGID
VELOCITY PARAMETERS
VELOCITY Y
-1
INITIAL VELOCITY Y
-1
OK
DISCRETE
OK
CURVES ADD
3
END LIST
NEW
SYMMETRY
DISCRETE
OK
CURVES ADD
2
1330 Marc Users Guide: Part 2
CHAPTER 3.28
Remeshing/Rezoning Parameters
DeIinition oI remeshing/rezoning parameters is a new and key step in perIorming an analysis with automatic
remeshing/rezoning. These parameters include the type oI mesher to be used, the remeshing criteria and related
parameters, the element target length in the new mesh, and the contact body to be remeshed.
The meshers available in Marc are advancing Iront mesher Ior both quadrilaterals and triangles, overlay quad mesher,
and Delaunay triangle mesher.
Start Irom the MAIN MENU; Click MESH ADAPTIVITY; Click GLOBAL REMESHING; Choose the type oI mesher;
DeIine remeshing criteria and element target length; speciIy the contact body to be remeshed.
END LIST
FLIP CURVES
2
1
END LIST
MAIN
MESH ADAPTIVITY
GLOBAL REMESHING
NEW
ADVANCING FRONT QUAD use advancing front quad mesher
PENETRATION check penetration each increment
IMMEDIATE remesh before analysis begins (1st criterion)
ANGLE DEVIATION (remesh when angle change from
the undeformed angle)
ELEMENT EDGE LENGTH define element edge length
0.07
OK
REMESH BODY define the body to be remeshed
CBODY1
MAIN
1331
CHAPTER 3.28
Automatic Remeshing/ Rezoning

Figure 3.28-3 Define Remeshing/Rezoning Parameters
Load Steps
Total load is applied in two loadcases in the analysis. In the Iirst loadcase, the top rigid surIace moves down 0.2 cm
within 5 equal increments. In the second step, the top rigid surIace moves down 0.534 cm within 95 equal increments.
In the second loadcase, only the deviatoric part oI stresses is included in stiIIness matrix calculation in order to
improve the convergence oI the calculations.
A new click activates the deIined remeshing/rezoning parameters Ior the required loadcases.
Note: The remeshing/rezoning analysis involves the interpolation and extrapolation oI nodal as well as elemental quantities.
This introduces approximations in the nodal and elemental quantities and can make the step aIter remeshing diIIicult to
converge. For this reason, experience shows that care must be taken to:
not remesh the body too Irequently and
keep the element or use the number oI element control target length such that the change in mesh density
or element length aIter remeshing is not too drastic. In this regard, the criterion based on the percent
change oI number oI elements in the Advanced Remeshing Parameters menu can be used.
LOADCASE
NEW
MECHANICAL
STATIC
GLOBAL REMESHING activate global remeshing for 1st loadcase
ADAPG1
1332 Marc Users Guide: Part 2
CHAPTER 3.28
Job Parameters
Element type 11 is used; Both loadcases are activated; Updated Lagrangian elasticity is used; Stress tensor and
equivalent von Mises stress are written into the post Iile; Plane strain condition is assumed.
An important step in remeshing analysis is to deIine the upper bound to the nodes that lie on the periphery oI any
deIormable surIace, since remeshing may considerably change surIace entities and surIace node.
OK
TOTAL LOADCASE TIME
0.2
FIXED PARAMETERS
# STEPS
5
OK (twice)
NEW
STATIC
GLOBAL REMESHING activate global remeshing for 2nd loadcase
ADAPG1
OK
SOLUTION CONTROL
DEVIATORIC
OK
TOTAL LOADCASE TIME
0.534
FIXED PARAMETERS
# STEPS
95
OK (twice)
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
PLANE STRAIN
11
OK
ALL EXIST
RETURN (twice)
1333
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Save Model, Run Job, and View Results
NEW
MECHANICAL
LCASE1
LCASE2
MESH ADAPTIVITY
MAX # CONTACT NODES
2000
OK
ANALYSIS OPTIONS
ELASTICITY PROCEDURE: LARGE STRAIN - UPDATED LAGRANGE
OK
CONTACT CONTROL
ADVANCED CONTACT CONTROL
PENETRATION CHECK: AUTOMATIC
OK (twice)
JOB RESULTS
ELEMENT TENS: CAUCHY
ELEMENT SCAL: VON MISES
OK
PLANE STRAIN
OK
MAIN
JOBS
RUN
SUBMIT 1
MONITOR
OK
SAVE
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
EQUIVALENT VON MISES STRESS
OK
MONITOR
1334 Marc Users Guide: Part 2
CHAPTER 3.28

Figure 3.28-4 FE-Mesh before Analysis Begins (after First Remeshing)

Figure 3.28-5 Deformed Mesh and Distribution of Equivalent von Mises Stress at Increment 100
1335
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Tape Peeling Simulation
Tape peeling simulation with Marc is done next to predict the strength oI the adhesive in the tape. The model consists
oI two types oI materials Iilm and adhesive, which are glued together. The adhesive layer is glued to a desk (see
Figure 3.28-6). The analysis starts by applying a point load to the tip oI the Iilm to simulate peeling operation. As the
peeling takes place, the adhesive layer gets torn oII the desk surIace. The mesh becomes distorted because oI the large
deIormation in the adhesive layer. Without remeshing, this simulation would terminate earlier.
In this example, we again demonstrate how to use global remeshing Ior this type oI applications. Particularly in this
example, we show how the local reIinement is done based upon the outline curvature detection in the peeling area.
Figure 3.28-6 Adhesive Layer glued to a Desk
Analysis
Model Generation
This example uses triangular element type 155 which can be used Ior analysis oI elastomeric materials. The tape has
a geometry oI 0.002 m in length and 0.0001 m in thickness. Plane strain assumption is used. The Iilm layer and the
adhesive layer both take halI oI the thickness. The geometry model is read Irom the predeIined model:
remeshing_rezoning_b_mesh.proc.
FILE
MERGE
MERGE FILE
remeshing_rezoning_b_mesh.proc
OK
1336 Marc Users Guide: Part 2
CHAPTER 3.28
Figure 3.28-7 Geometry Model using Triangular Element Type 155
Save the model as remeshing_rezoning.mfd.
The model should be saved Irom time to time to avoid any lost oI the input during the preprocessing. Click on SAVE
to do this.
Boundary Condition
A point load associated with a time table is deIined and assigned to the tip oI the Iilm layer (see Figure 3.28-8). Note
that this portion oI the mesh is not being remeshed; hence, application oI loads and boundary conditions on this part
oI mesh is allowed.
BOUNDARY CONDITIONS
MECHANICAL
TABLES
TABLE TYPE: time
ADD POINT
0 0
20 20
SHOW MODEL
RETURN
POINT LOAD
Y FORCE:
1.0
1337
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Figure 3.28-8 Mechanical Boundary Conditions Menu adding Point Loads
Contact Definition
Here, we deIine the contact bodies Iirst. The material properties are assigned to the diIIerent contact bodies later. Three
contact bodies are identiIied: adhesive, Iilm, and desk. The GLUE option is used in the contact table.
TABLE
CURRENTLY DEFINED TABLE
table1
OK (twice)
NODES: ADD
86
END LIST
RETURN
ID BOUNDARY CONDS
RETURN
CONTACT
CONTACT BODIES
NEW
NAME
adhesive
1338 Marc Users Guide: Part 2
CHAPTER 3.28
CONTACT BODY TYPE
deformable
ELEMENTS ADD (add all the elements in the lower part of the mesh)
END LIST
NEW
NAME
film
CONTACT BODY TYPE
deformable
ELEMENTS ADD (add all the elements in the upper part of the mesh)
END LIST
NEW
NAME
desk
CONTACT BODY TYPE
rigid
CURVES ADD
1 2
END LIST
ID CONTACT (See Figure 3.28-9)
RETURN
CONTACT TABLES
NEW
PROPERTIES
FIRST adhesive SECOND film: GLUE
FIRST adhesive SECOND desk: GLUE (See Figure 3.28-10)
OK
RETURN (twice)
1339
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Figure 3.28-9 Contact Bodies Menu defining Three Contact Bodies
Figure 3.28-10 CONTACT TABLE PROPERTIES Submenu identifying Adhesive, Film, and Desk
Material Properties
Use elasto-plastic material Ior the Iilm and Mooney type material Ior the adhesive. Select the contact body to assign
the material properties.
MATERIAL PROPERTIES
NEW
1340 Marc Users Guide: Part 2
CHAPTER 3.28
NAME
film
ISOTROPIC
YOUNGS MODULUS
1.0e9
POISSIONS RATIO
0.3
ELASTIC-PLASTIC
INITIAL YIELD STRESS
1.0e7
OK (twice)
SELECT
SELECT CONTACT BODY ENTITIES
film
OK
RETURN
ELEMENTS ADD
ALL: SELECT
NEW
adhesive
MORE
MOONEY
C10
200000
OK
ELEMENTS ADD
ALL: UNSEL.
RETURN
ID MATERIALS (See Figure 3.28-11)
RETURN
1341
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Figure 3.28-11 MATERIAL PROPERTIES Menu using Adhesive
Remeshing/Rezoning Parameters
We will use triangular mesher Ior remeshing on the adhesive body. Number oI elements desired is 600 and the
remeshing is needed when there is a distortion or at every Iive increments. By deIault, the curvature detection is used
Ior the local reIinement. The minimum element size is 1/3 oI the element size computed Ior the remeshing.
MESH ADAPTIVITY
GLOBAL REMESHING
NEW
ADVANCING FRONT TRIA
INCREMENT FREQUENCY
5
ADVANCED
ELEMENT DISTORTION
OK
# ELEMENTS SET
600
OK
REMESH BODY
adhesive (See Figure 3.28-12)
RETURN
1342 Marc Users Guide: Part 2
CHAPTER 3.28
Figure 3.28-12 ADVANCING FRONT TRIA GLOBAL REMESHING Submenu
Load Steps
The load step control uses adaptive time stepping and automatic time step cutback. This makes the loading control
more user Iriendly.
LOADCASES
NEW
MECHANICAL
STATIC
LOADS
apply1
OK
CONTACT
CONTACT TABLE
Cable1
OK (twice)
GLOBAL REMESHING
Adapg1
OK
TOTAL LOADCASE TIME
15
1343
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Figure 3.28-13 MECHANICAL STATIC PARAMETERS Submenu
Job Parameters
It is important to select correct element type and analysis control parameters. Here, we assign the loadcase, select right
analysis control parameters, and the element type.
ADAPTIVE MULTI-CRITERIA
OK (See Figure 3.28-13)
RETURN (twice)
JOBS
NEW
MECHANICAL
AVAILABLE
lcase1
CONTACT CONTROL
ADVANCED CONTACT CONTROL
PENETRATION CHECK: AUTOMATIC
OK (twice)
MESH ADAPTIVITY
MAX #CONTACT NODES/BODY
4000
1344 Marc Users Guide: Part 2
CHAPTER 3.28
Figure 3.28-14 MECHANICAL ANALYSIS OPTIONS Submenu
OK
ANALYSIS OPTIONS
RUBBER ELASTICITY PROCEDURE:
LARGE STRAIN-UPDATED LAGRANGE
PLASTICITY PROCEDURE:
LARGE STRAIN MULTIPLICATIVE (See Figure 3.28-14)
OK
JOB RESULTS
AVAILABLE ELEMENT SCALARS
Equivalent Cauchy Stress
OK
ANALYSIS DIMENSION: PLANE STRAIN (See Figure 3.28-15)
OK
ELEMENT TYPES
MECHANICAL
PLANE STRAIN
TRIA
155
OK
ALL: EXIST
RETURN (twice)
1345
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Figure 3.28-15 MECHANICAL ANALYSIS CLASS Submenu
Save Model, Run Job, and View Results
Here, we show how a job can be submitted and monitored while the analysis is going on. When the results are
generated, you can view the results without waiting Ior the whole analysis is completed.
JOBS
RUN
SUBMIT1
MONITOR (you can check job running status here)
OK (go to see results)
RETURN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
SCALAR
Equivalent Cauchy Stress
OK
FILL
MONITOR
1346 Marc Users Guide: Part 2
CHAPTER 3.28
Here, the result shows the peeling oI tape at time 15 (Figure 3.28-16),. The local mesh reIinement in the peeling area
can be seen in Figure 3.28-17.
Figure 3.28-16 Peeling of Tape Results
Figure 3.28-17 Mesh Refinement of Peeling Area
The Ilow oI material in the adhesive is captured quite realistically.
1347
CHAPTER 3.28
Automatic Remeshing/ Rezoning
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
remeshing_rezoning.proc Mentat procedure Iile
remeshing_rezoning_b.proc Mentat procedure Iile
remeshing_rezoning_b_mesh.mfd Mentat model Iile
1348 Marc Users Guide: Part 2
CHAPTER 3.28
Chapter 3.29: Multibody Contact and Remeshing
3.29
Multibody Contact and
Remeshing

Chapter Overview 1350

Squeezing of a Rubber Body 1350

Input Files 1370

1350 Marc Users Guide: Part 2
CHAPTER 3.29
Chapter Overview
In Marc, contact between deIormable bodies is taken into account via multipoint constraint equations. The constraint
equations used, and thus the quality oI the solution, may depend on the numbering oI the contact bodies. In general,
the most optimal results are obtained iI the numbering oI the bodies is chosen such that:
In case oI contact between bodies with a large diIIerence in stiIIness, like rubber and steel, the soIter body has
the lowest number;
In case oI contact between bodies with a large diIIerence in mesh density, the body with the Iinest mesh has
the lowest number.
In Marc, the CONTACT TABLE option allows you to change the order in which contact is detected between bodies.
With these options:
1. Ior each set oI deIormable contact bodies, you can indicate in which order the search Ior contact is perIormed
by the program. This is especially important Ior models with several deIormable contact bodies, since it makes
the searching order more or less independent Irom the body numbering and
2. Ior each set oI deIormable contact bodies, the optimal search order can also be determined by the program,
based on the smallest element edge length at the outer boundary oI the contact bodies. This can be important
in an analysis with global remeshing, where the mesh density aIter remeshing oI a contact body can be
signiIicantly diIIerent compared to the density beIore remeshing.
In this chapter, the new Iunctionality is illustrated with analysis involving remeshing oI a rubber body. Additionally,
the use oI stress-Iree projection at initial contact is shown. Stress-Iree projection is aimed to correct small geometry
imperIections in a Iinite element model. This is done by adjusting the coordinates oI a node lying within the contact
tolerance zone, according to the projection oI the node on the contacted segment.
Squeezing of a Rubber Body
A circular rubber body is squeezed between two steel legs. During the analysis, various parts oI the rubber body are
remeshed.
Background information
A circular rubber body is positioned between two steel legs and a rigid body as indicated in Figure 3.29-1. The rubber
material is described by a Mooney-Rivlin material with constants and . The steel part is assumed to
be linear elastic with Young`s modulus and Poisson`s ratio . The steel legs are loaded by two
opposite point Iorces, which magnitude as a Iunction oI time is also given in Figure 3.29-1. To illustrate the new
contact Iunctionality and the remeshing capabilities in Marc, six contact bodies are used: two deIormable bodies Ior
the steel part, three deIormable bodies Ior the rubber part, and one rigid body. The two bodies Ior the steel part and the
three bodies Ior the rubber part are glued together. The rubber body is Irequently remeshed, where the element size
and the remesh Irequency Ior each oI the three parts is diIIerent.
C
10
8 C
01
6
E 3
6
10 v 0.3
1351
CHAPTER 3.29
Multibody Contact and Remeshing
Figure 3.29-1 Rubber Body between two Steel Legs and a Rigid Body (left); Loading History (right)
A plane strain analysis is perIormed based on the updated Lagrange procedure. Both the steel and the rubber part are
modeled using 4-node plane strain elements with Iull integration (Marc element type 11).
Model Generation
The Iinite element model is set up in the Iollowing order:
1. The right steel part is meshed by deIining a number oI quadrilateral surIaces and converting those surIaces into
Iinite elements.
2. Three circles are created and intersected, aIter which the unnecessary curves are removed and the three parts
oI the rubber body are meshed using the advancing Iront quad mesher.
3. With the symmetry option, the elements oI the leIt steel part are easily created.
4. Finally, the coordinates oI two nodes oI the steel part are modiIied, one node oI the rubber is moved to simulate
a geometry imperIection and the rigid body is deIined by a straight line.
5. The various parts oI the mesh are stored in element sets. The complete Iinite element model is shown in
Figure 3.29-2.
FILES
NEW
OK
RESET PROGRAM
VIEW
SHOW VIEW 1
Steel
F
F
F
time
25
2
1
Rubber
1352 Marc Users Guide: Part 2
CHAPTER 3.29
MAIN
MESH GENERATION
PTS ADD
0 0 0
0.3 0 0
0.325 1.2 0
0.225 1.2 0
0.2 0.1 0
0 0.1 0
0.2 0 0
0.3 0.1 0
FILL
SRFS ADD
1 7 5 6
7 2 8 5
5 8 3 4
CONVERT
DIVISIONS
2 1
SURFACES TO ELEMENTS
1
END LIST
DIVISIONS
1 1
SURFACES TO ELEMENTS
2
END LIST
DIVISIONS
1 11
SURFACES TO ELEMENTS
3
END LIST
RETURN
SWEEP
NODES
ALL: EXIST.
1353
CHAPTER 3.29
Multibody Contact and Remeshing
RETURN
CURVE TYPE
CENTER/RADIUS
RETURN
CRVS ADD
0 0.8 0
0.2
DUPLICATE
TRANSLATIONS
0.3 0 0
CURVES
1
END LIST
TRANSLATIONS
-0.3 0 0
CURVES
1
END LIST
RETURN
INTERSECT
CURVE/CURVE
1 3 2
END LIST
RETURN
CRVS REM
18 15 12
END LIST
AUTOMESH
CURVE DIVISIONS
FIXED # DIVISIONS
5
APPLY CURVE DIVISIONS
19 16 11 4
END LIST
FIXED # DIVISIONS
10
1354 Marc Users Guide: Part 2
CHAPTER 3.29
APPLY_CURVE_DIVISIONS
8 14 6 10
END LIST
RETURN (twice)
SWEEP
POINTS
ALL: EXIST.
RETURN
AUTOMESH
2D PLANAR MESHING
ADV FRONT QUAD MESH!
8 16 19 4 11 14 6 10 19 16 14
END LIST
RETURN (twice)
SWEEP
REMOVE UNUSED POINTS
RETURN
SYMMETRY
ELEMENTS
9 10 11 12 13 14 3 4 5 6 7 8 1 2
END LIST
RETURN
NODES EDIT
25
0.2 0.8 0.0
268
-0.2 0.8 0.0
MOVE
TRANSLATIONS
-0.0005 0 0
NODES
70
END LIST
RETURN
SELECT
ELEMENTS
1355
CHAPTER 3.29
Multibody Contact and Remeshing
9 10 11 12 13 14 3 4 5 6 7 8 1 2
END LIST
STORE
steel_r
OK
ALL: SELECT.
CLEAR SELECT
(repeat similar steps to create the element sets
steel_l, rubber_l, rubber_m, and rubber_r)
RETURN
POINTS ADD
0.2 0.6 0
-0.2 0.6 0
CURVE TYPE
LINE
RETURN
CRVS ADD
120 121
MAIN
1356 Marc Users Guide: Part 2
CHAPTER 3.29
Figure 3.29-2 Finite Element Model
Boundary Conditions
Boundary conditions are deIined to clamp the lower edge oI the steel part and to load both steel legs. The load is set
up via a table oI type time.
BOUNDARY CONDITIONS
MECHANICAL
NEW
NAME
Clamped
FIXED DISPLACEMENT
ON X DISPLACEMENT
ON Y DISPLACEMENT
OK
NODES ADD
1 2 3 8 248 249 250 254
END LIST
TABLES
NEW
1357
CHAPTER 3.29
Multibody Contact and Remeshing
NAME
Force_time
TYPE
TIME
OK
ADD POINT
0 0
1 25
2 0
FIT
SHOW MODEL
RETURN
NEW
NAME
Load_left
POINT LOAD
X FORCE
1
TABLE
force_time
OK (twice)
NODES ADD
277
END LIST
NEW
NAME
Load_right
POINT LOAD
X FORCE
-1
TABLE
force_time
OK (twice)
NODES ADD
34
END LIST
MAIN
1358 Marc Users Guide: Part 2
CHAPTER 3.29
Material Properties
Two diIIerent materials are deIined: one Ior the steel part, one Ior the rubber part oI the model.
Contact
Five deIormable contact bodies and one rigid contact body are deIined. The Iirst two deIormable bodies correspond
to the steel part oI the model and are called Steel_left and Steel_right. The remaining three deIormable bodies
MATERIAL PROPERTIES
NEW
NAME
steel
ISOTROPIC
YOUNGS MODULUS
300000
POISSONS RATIO
0.3
OK
ELEMENTS ADD
steel_l
steel_r
END LIST
NEW
NAME
rubber
MORE
MOONEY
C10
8
C01
6
OK
ELEMENTS ADD
rubber_l
rubber_m
rubber_r
END LIST
MAIN
1359
CHAPTER 3.29
Multibody Contact and Remeshing
correspond to the rubber part oI the model and are called Rubber_left, Rubber_middle and Rubber_right (see
Figure 3.29-3) and Ior the Contact Table Entry menus (see Figure 3.29-4). Since, in this way, the deIault contact body
numbering is not optimal, a contact table is deIined to inIluence the order in which the search Ior contact is done. The
optimal search order Ior contact between the Iollowing body pairs is determined by the program:
Steel_left and Rubber_left;
Steel_left and Rubber_middle;
Rubber_left and Rubber_middle;
Rubber_middle and Rubber_right;
Rubber_middle and Steel_right;
Rubber_right and Steel_right.
Figure 3.29-3 Contact Bodies
1360 Marc Users Guide: Part 2
CHAPTER 3.29
Figure 3.29-4 Contact Table Entries
The geometric imperIection are removed by activating stress-Iree projection Ior the pair oI contact bodies
Rubber_left and Rubber_middle. Moreover, although this does not inIluence the results, contact between
Steel_left and Steel_right is Iorced to be Irom Steel_right to Steel_left.
A small nonzero separation Iorce is deIined Ior contact between the bodies Rubber_middle and Rigid_holder to
prevent a rigid body motion oI the rubber part.
CONTACT
CONTACT BODIES
NEW
NAME
Steel_left
DEFORMABLE
OK
ELEMENTS ADD
steel_l
NEW
NAME
Steel_right
DEFORMABLE
OK
ELEMENTS ADD
steel_r
1361
CHAPTER 3.29
Multibody Contact and Remeshing
NEW
NAME
Rubber_left
DEFORMABLE
OK
ELEMENTS ADD
rubber_l
NEW
NAME
Rubber_middle
DEFORMABLE
OK
ELEMENTS ADD
rubber_m
NEW
NAME
Rubber_right
DEFORMABLE
OK
ELEMENTS ADD
rubber_r
NEW
NAME
Rigid_holder
RIGID
OK
CURVES ADD
20
END LIST
PLOT
ELEMENTS SOLID
MORE
IDENTIFY
CONTACT
REGEN
RETURN (thrice)
1362 Marc Users Guide: Part 2
CHAPTER 3.29
CONTACT TABLES
NEW
PROPERTIES
1 2
CONTACT TYPE: GLUE
CONTACT DETECTION METHOD: SECOND->FIRST
1 3
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: AUTOMATIC
1 4
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: AUTOMATIC
2 4
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: AUTOMATIC
2 5
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: AUTOMATIC
3 4
CONTACT TYPE: GLUE
CONTACT DETECTION METHOD: AUTOMATIC
PROJECT STRESS-FREE
4 5
CONTACT TYPE: GLUE
CONTACT DETECTION METHOD: AUTOMATIC
4 6
CONTACT TYPE: GLUE
SEPARATION FORCE
0.1
OK (twice)
MAIN
1363
CHAPTER 3.29
Multibody Contact and Remeshing
Mesh Adaptivity
During the analysis, global remeshing is applied to the rubber contact bodies. The advancing Iront quad mesher is
used. The global remeshing parameters are set as Iollows:
Body Rubber_left: increment Irequency 5, element edge length 0.016;
Body Rubber_middle: increment Irequency 7, element edge length 0.024;
Body Rubber_right: increment Irequency 9, element edge length 0.02.
MESH ADAPTIVITY
GLOBAL REMESHING
ADVANCING FRONT QUAD
INCREMENT
FREQUENCY
5
ELEMENT EDGE LENGTH
SET
0.016
OK
REMESH BODY
Rubber_left
OK
ADVANCING FRONT QUAD
INCREMENT
FREQUENCY
7
ELEMENT EDGE LENGTH
SET
0.024
OK
REMESH BODY
Rubber_middle
OK
ADVANCING FRONT QUAD
INCREMENT
FREQUENCY
9
1364 Marc Users Guide: Part 2
CHAPTER 3.29
Loadcases
A mechanical static loadcase is deIined in which the contact table and global remeshing criteria are selected (the
previously deIined boundary conditions are automatically selected). The total loadcase time is set to 2. A Iixed
stepping procedure is chosen with 50 steps. The deIault control settings Ior the Newton-Raphson iteration process
are used.
ELEMENT EDGE LENGTH
SET
0.02
OK
REMESH BODY
Rubber_right
OK
MAIN
LOADCASES
NEW
MECHANICAL
STATIC
CONTACT
CONTACT TABLE
ctable1
OK (twice)
GLOBAL REMESHING
adapg1
adapg2
adapg3
OK
TOTAL LOADCASE TIME
2
# STEPS
50
OK
TITLE
Squeezing of a rubber body
OK
MAIN
NAME
1365
CHAPTER 3.29
Multibody Contact and Remeshing
Jobs
A mechanical job is deIined in which the previously deIined loadcase is selected. The available CONTACT TABLE is
used also Ior initial contact. Since the steel legs are mainly loaded in bending, the assumed strain Iormulation is
activated. The updated Lagrange large strain elasticity procedure is used Ior the rubber part oI the model. Because oI
remeshing, the upper bound to the number oI contact segments and nodes per contact body is set to 500. A check on
penetration should be perIormed every iteration oI the Newton-Raphson process. The element type Ior the steel and
rubber parts is set to 11. The model is saved and the job is submitted.
JOBS
MECHANICAL
lcase1
CONTACT CONTROL
INITIAL CONTACT
ctable1
OK
ADVANCED CONTACT CONTROL
PER ITERATION
OK (twice)
MESH ADAPTIVITY
MAX # CONTACT NODES / BODY
500
OK
ANALYSIS OPTIONS
ADVANCED OPTIONS
ASSUMED STRAIN
OK
RUBBER ELASTICITY PROCEDURE: LARGE STRAIN-UPDATED LAGRANGE
OK
ELEMENT TYPES
MECHANICAL
PLANE STRAIN
11
OK
RETURN (twice)
TITLE
Squeezing of a rubber body
OK
1366 Marc Users Guide: Part 2
CHAPTER 3.29
Results
In Figure 3.29-5, the position oI node 70 at increment 0 is compared with its original position. This clearly illustrates
the use oI the stress-Iree projection; since node 70 was within the contact tolerance, the gap is closed without
introducing stresses.
Figure 3.29-5 Result of Stress-free Projection of Node 70
In Figure 3.29-6, a contour band plot oI the contact status at increment 0 is given showing the eIIect oI the automatic
search order (contact between rubber and steel), and oI enIorcing contact Irom the second to the Iirst body oI the pair
oI deIormable contact bodies (contact between the two steel bodies). Notice that Ior the boundary conditions only one
node oI body Steel_right touches body Steel_left.
RUN
SAVE MODEL
SUBMIT 1
MONITOR
OK
MAIN
70
70
Model
Post file; increment 0
1367
CHAPTER 3.29
Multibody Contact and Remeshing
Figure 3.29-6 Contact Status at Increment 0
In Figure 3.29-7, contact status is shown Ior the bodies Rubber_middle and Rubber_right at increment 10. Due to
remeshing, the edge length at the boundary oI body Rubber_right becomes signiIicantly smaller than that oI body
Rubber_middle. As a result, there is a change in the search order Ior contact: until increment 9, nodes oI body
Rubber_middle are touching body Rubber_right, at increment 10, nodes oI body Rubber_right are touching
body Rubber_middle.
1368 Marc Users Guide: Part 2
CHAPTER 3.29
Figure 3.29-7 Contact Status at Increment 10
In Figure 3.29-8, the deIormed conIiguration is shown at the maximum load level.
Figure 3.29-8 Deformed Configuration at Maximum Load Level
1369
CHAPTER 3.29
Multibody Contact and Remeshing
Finally, Figure 3.29-9 presents the Iorce versus displacement curves Ior the nodes with point loads. Despite irregular
remeshing, the behavior is remarkably symmetric.
Figure 3.29-9 Displacement versus Time for Nodes with Point Loads
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Contact Status
OK
CONTOUR BANDS
MONITOR
HISTORY PLOT
SET NODES
277 34
END LIST
COLLECT DATA
0 1000 1
NODES/VARIABLES
ADD VARIABLE
Time
Displacement X
Fit
1370 Marc Users Guide: Part 2
CHAPTER 3.29
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
multibody_contact.proc Mentat procedure Iile
Chapter 3.30: Container
3.30
Container

Chapter Overview 1372

Background Information 1372

Detailed Session Description 1376

Conclusion 1401

Input Files 1401

1372 Marc Users Guide: Part 2
CHAPTER 3.30
Chapter Overview
This chapter demonstrates the modeling and analysis oI the bottom oI an aluminum container under internal pressure.
The particular conIiguration oI this container bottom leads to a snap-through problem. Accurate modeling oI the
geometry is essential since it dramatically inIluences the snap-through process.
The primary goal oI this chapter is to show you three Mentat Iunctionalities.
Using nonlinear analysis to solve a snap-through analysis problem
Using the TABLES option to speciIy input data that changes with time, plastic strain, etc.
Animating the results oI an analysis
Background Information
Description
The container, a soIt drink can, is assumed to be a circle cylinder with a radius oI 1.3 inches and a total height oI 4.8
inches. The container (see note below) is made out oI aluminum and has a wall thickness oI 0.025 inches
Figure 3.30-1 Aluminum Container
Note: At times, the container may be reIerred to as can in the text.
1373
CHAPTER 3.30
Container
Idealization
The geometry oI this problem is Iairly simple due to two Iactors. The Iirst is that the geometry and the loading oI the
container are axisymmetric and allow you to perIorm an axisymmetric analysis. The second Iactor is that the Iocus oI
the analysis is restricted to the phenomena that occur at the bottom oI the container. In this analysis, the height oI the
container, h, is limited to a length where the edge eIIects are damped out. The theory behind this assumption is
explained below.
II
where
the solution decreases to about 4 oI its value at the bottom edge. In this example, it means you can saIely ignore the
inIluence oI the top edge since the critical height, h, is equal to 0.4519, calculated as Iollows:
An awareness oI this decay distance is very important in numerical calculations. II you wish to correctly capture the
behavior oI the solution in the edge region, the typical Iinite element size must be small in comparison to the decay
distance.
Figure 3.30-2 Section of Container to be Analyzed
h 2.5 rt
r the radius oI the container,
t the wall thickness,
h 2.5 1.307 0.025 0.4519
8
9
4
5
7
6
1
2
3
Center (0.0, 0.0, 0.0)
Arc 1
Radius 2.354
Angle (0, 22)
Arc 2
Center (2.026, 1.0, 0.0)
Radius 0.063
Angle (129, 265)
Arc 3
Center (2.0, 1.2, 0.0)
Radius 0.125
Angle (296, 334)
Arc 4
Center (2.3, 1.127, 0.0)
Radius 0.18
Angle (90, 162)
Arc 5
Tangent to Arc 1
Radius 0.05
Extending angle 50
Center line
1374 Marc Users Guide: Part 2
CHAPTER 3.30
Requirements for a Successful Analysis
Nonlinear problems that involve buckling or snap-through are prime candidates Ior displacement controlled
incremental strategies. UnIortunately, the problem at hand is a load controlled problem. In order to able to traverse the
load versus displacement curve oI a point on the bottom oI the container you must use a loading pattern such that the
load increment is scaled in size and applied in the correct direction. The arc-length method combined with a Newton-
Raphson iterative scheme will guarantee you that the entire load displacement curve can be traversed. Needless to say,
the solution oI this problem consists oI large displacements and Iinite strains.
Full Disclosure
Analysis Type
Nonlinear snap-through.
Element Type
Marc Element Type 89, axisymmetric shell.
Material Properties
Aluminum with workhardening.
Isotropic with Young's Modulus 11.0e6 p.s.i and Poisson's Ratio 0.3.
The stress-strain data used to deIine the workhardening oI the aluminum is listed in Table 2.30-1 and
graphically represented in Figure 3.30-3.
Table 2.30-1 Stress Strain Data
Log Plastic
Strain (x)
Cauchy
Stress (y)
Total
Engineering
Strain
0.0
0.001748
0.003494
0.06766
0.09531
0.1570
0.2070
0.2623
42000.0
44577.0
45157.0
63665.0
70950.0
81315.0
88560.0
95216.0
0.0038
0.0057
0.0075
0.0755
0.1058
0.1763
0.2365
0.3066
1375
CHAPTER 3.30
Container
Figure 3.30-3 Graphical Representation of Cauchy Stress vs. Logarithmic Plastic Strain Data
Overview of Steps
Step 1: Input all arcs according to the measurements specified in Figure 3.30-2.
Step 2: Input straight lines to connect the arcs.
Step 3: Convert the geometric entities to finite elements.
Step 4: Use SWEEP to eliminate all duplicate nodes, then switch the element class to quadratic shell
elements and attach the midside nodes to the curves.
Step 5: Add kinematic boundary conditions to enforce the symmetry and restrain rigid body motion.
Step 6: Specify edge loads.
Step 7: Rectify connectivity to ensure consistent normals.
Step 8: Add material properties.
Step 9: Add geometric properties.
Step 10: Define the loadcase.
Step 11: Submit the job.
Step 12: Postprocess the results by looking at the deformed shape and the load-displacement curve of the
node located on the symmetry axis.
1
2
3
4
5
6
7
0
true_stress (x10e+5)
log_strain (x.1)
workhard
5
0
1
8
1
1376 Marc Users Guide: Part 2
CHAPTER 3.30
Detailed Session Description
Step 1: Input all arcs according to the measurements speciIied in Figure 3.30-2.
A structure that is modeled with axisymmetric elements requires the global x-axis to point into the axial direction oI
that structure. As a result oI this type oI modeling, the container is displayed in a horizontal position.
As in Chapter 3.6: Tube Flaring, this sample session demonstrates the use oI the geometric meshing technique. The
geometric entities used to create the mesh are two types oI curves: arcs and lines. Once you have generated the
geometric model, the arcs and lines are converted to Iinite elements. ReIer to Chapter 1: Introduction in this manual
Ior more inIormation on mesh generation techniques.
Use the Center/Radius/Angle(begin)/Angle(end) arc type (CRAA) to create the Iirst arcs oI the geometry. Use the
Iollowing button sequence to select and add the CRAA arc type. The values Ior the measurements oI the arcs are given
in Figure 3.30-2.
Switch on the labeling oI points.
MAIN
MESH GENERATION
CURVE TYPE
CENTER/RADIUS/ANGLE/ANGLE
RETURN
crvs ADD
0 0 0 (center)
2.345 (radius)
0 22 (angle limits)
2.026 1 0 (center)
0.063 (radius)
129 265 (angle limits)
2.0 1.2 0 (center)
0.125 (radius)
296 334 (angle limits)
2.3 1.127 0 (center)
0.18 (radius)
90 162 (angle limits)
MAIN
FILL
PLOT
draw POINTS
1377
CHAPTER 3.30
Container
Figure 3.30-4 shows the Iour arcs.
Figure 3.30-4 Using CRAA Type Arcs to Create First Four Curves
The next step is to add a new arc (number 5) so that it is tangent to arc 1, the lower arc, using the Iollowing
button sequence.
LABEL (on)
RETURN
REGEN
MAIN
MESH GENERATION
CURVE TYPE
TANGENT/RADIUS/ANGLE
RETURN
crvs ADD
3 (click end point of arc #1)
0.05 (radius)
50.0 (arc angle)
1378 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-5 Using TRA Type Arc to Create Arc 5
Step 2: Input straight lines to connect the arcs.
Finally, add the straight lines to complete the geometric description oI the model. Set the curve type to LINE. Use the
crvs ADD button on the mesh generation panel and click on the existing points that need to be connected. As we noted
in the section on Idealization, it is necessary to extend the wall oI the cylinder to at least 0.4519 inches Irom the edge
to ensure that the edge eIIects are negligible.
MAIN
MESH GENERATION
CURVE TYPE
LINE
RETURN
crvs ADD
11 14 (Click on points to connect)
4 9 (Click on points to connect)
8 17 (Click on points to connect)
1379
CHAPTER 3.30
Container
The origin chosen Ior this problem is 2.354 inches to the leIt oI point 1 oI arc 1. The total extent oI the wall necessary
is thereIore 2.354 0.4519; that is, approximately three inches. ThereIore, we will add a point at 3.0 1.307 0.0.
Figure 3.30-6 Connect the Arcs by Lines
Step 3: Convert the geometric entities to Iinite elements.
Use the CONVERT processor, located on the mesh generation panel, to convert the geometric entities (in this case,
curves) to Iinite elements. You must speciIy the number oI elements Ior each curve. A higher mesh density is required
at sections oI high curvature and large displacements than in regions where the values Ior stress and strain are expected
to be less severe. For this reason, you must speciIy a larger number oI convert divisions Ior those arcs oI high curvature
and large displacements as indicated in the button sequence given below. Figure 3.30-7 shows the result oI converting
the curves to Iinite elements.
MAIN
MESH GENERATION
pts ADD
3.0 1.307 0.0
FILL
crvs ADD
12 18 (Click on points to connect)
MAIN
PLOT
label POINTS (off)
1380 Marc Users Guide: Part 2
CHAPTER 3.30
Step 4: Use SWEEP to eliminate all duplicate nodes, then switch the element class to quadratic shell elements and
attach the midside nodes to the curves.
The previous operations may have leIt duplicate nodes; that is nodes with diIIerent identiIication numbers but
occupying the same space. In Iinite element terms, these nodes are not connected which may introduce undesirable
mechanisms in the structure.
label CURVES (on)
REGEN
RETURN
MESH GENERATION
CONVERT
DIVISIONS
8 1 (Number of subdivisions)
CURVES TO ELEMENTS
1 9
END LIST (#)
DIVISIONS
6 1
CURVES TO ELEMENTS
4 2
END LIST (#)
DIVISIONS
4 1
CURVES TO ELEMENTS
6 3
END LIST (#)
DIVISIONS
3 1
CURVES TO ELEMENTS
7 5 8
END LIST (#)
1381
CHAPTER 3.30
Container
Figure 3.30-7 Model after Converting Curves to LINE(2) Elements
Use the SWEEP processor introduced to you in the sample session oI Introduction to eliminate the duplicate nodes
that occupy the same location. Since this involves a comparison oI real numbers that cannot be done exactly in a
computer, nodes are swept together iI they are within a certain tolerance Irom each other. This tolerance can be
changed Irom its deIault value. Be careIul when adjusting the tolerance as too large a tolerance can collapse the entire
structure into a single point.
In order to describe the curved geometry as precise as possible, the linear (LINE (2) elements) are converted to
elements with a quadratic interpolation Iunction (LINE (3) elements).
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
MAIN
MESH GENERATION
CHANGE CLASS
LINE (3)
ELEMENTS
all: EXIST.
1382 Marc Users Guide: Part 2
CHAPTER 3.30
The result is shown in Figure 3.30-8.
Figure 3.30-8 Curves and Points turned off
Step 5: Add kinematic boundary conditions to enIorce the symmetry and restrain rigid body motion.
In an axisymmetric shell analysis, there are three types oI applicable displacements or degrees oI Ireedom: axial,
radial, and rotational. The axial degree oI Ireedom is represented as a global x, and the radial as a global y.
For this particular model, the boundary conditions are simple. This is due to the symmetry conditions applied to the
center line node through the suppression oI radial displacement and in-plane rotation.
MAIN
PLOT
draw CURVES (off)
draw POINTS (off)
REGEN
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
DISPLACEMENT Z (see note below) (on)
1383
CHAPTER 3.30
Container
Figure 3.30-9 shows you the model with the boundary conditions added.
Figure 3.30-9 Fixed Displacement at Axis of Rotation
Next, suppress two degrees oI Ireedom oI the extreme node at the circumIerence oI the can:
1. suppress the movement in the axial direction, and
2. suppress the rotational degree oI Ireedom.
OK
nodes ADD
1 (Pick the lower left node)
END LIST (#)
FILL
Note: The buttons displacement x, displacement y, displacement z, rotation x, rotation y, and rotation z reIer to the six degrees
oI Ireedom that generally exist Ior a node oI a 3-D shell element. However, this problem uses an axisymmetric shell
element with the Iollowing three degrees oI Ireedom.: displacement in x-direction, displacement in y-direction, and
rotation about the z-axis. In such cases, the button displacement x reIers to the displacements in x-direction, displacement
y reIers to displacements in y-direction and displacement z reIers to the third degree oI Ireedom, the rotation about the
z-axis.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
1384 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-10 Boundary Conditions at the Circumference of Container
Step 6: SpeciIy edge loads.
To speciIy the loading sequence, use the TABLES option to create the load table using the Iollowing button sequence:
NEW
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Z (on)
OK
nodes ADD
18 (pick the top right node)
END LIST (#)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
1385
CHAPTER 3.30
Container
You will reIer to the table name, loading, when you apply the edge loads. The xmin, xmax, ymin, and ymax values
speciIy the table limits. The 0 to 1 range is Ior the x value and 0 to 500 is the range Ior the y value. The x-axis represents
the time (which is to be regarded a dummy variable Ior this analysis) and the y-axis represents the pressure load. The
loading pattern is speciIied as 0 edge load at time 0 and an edge load oI 500 at time 1. Since the total loadcase time,
used Ior quasi static analysis (LOADCASE menu), is set to 1.0, this table results in a total load oI 500 to be reached at
the end oI the loadcase. The table points can be entered either via the keyboard or by using the mouse to pick the
(0, 0.0) point in the graph Iollowed by the (1, 500.0) point.
time
OK (select OK button only if type time was typed in)
independent variable v1: MIN
0
independent variable v1: MAX
1
function value f: MIN
0
function value f: MAX
500
NAME
loading
ADD
0 0
1 500
MORE
independent variable v1: LABEL
time
function value f: LABEL
pressure
RETURN
FILLED
1386 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-11 Load History Table
As mentioned in Requirements for a Successful Analysis, it is the task oI the analysis program to deIine a load
incrementation that reaches the target load.
Now that the load type and the load path have both been deIined, use the Iollowing button sequence below to speciIy
where to apply this load.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
EDGE LOAD
PRESSURE
1
pressure TABLE
loading
OK
OK
edges ADD
all: EXIST.
FILL
1387
CHAPTER 3.30
Container
The actual load applied to the structure is 1 (the base value entered at the pressure prompt), multiplied by the values
deIined in the table.
Figure 3.30-12 Pressure Load Applied
Step 7: RectiIy connectivity to ensure consistent normals.
Figure 3.30-12 clearly indicates the pressure load has not been applied in the correct direction Ior all elements. This is
caused by the way the curves were created. The outward normal that determines the positive direction oI the load is
directly dependent on which point oI the arc was deIined Iirst. Figure 3.30-13 depicts this dependency on an element
that has a 1-2 connectivity.
Figure 3.30-13 The Outward Normal in Arc Definition
1
y
x
2
n
2
y
x
1
n
1388 Marc Users Guide: Part 2
CHAPTER 3.30
The connectivity oI the elements can easily be corrected using the Iollowing button sequence.
You may want to use the ZOOM option Ior closer view oI the areas where the Ilipped elements are located to make it
easier to pick the elements that have been loaded in the opposite direction. Pick the elements by moving the cursor
over each element and clicking <ML>.
Don't Iorget to speciIy end oI list by clicking <MR> in the graphics area when you have picked all the Ilipped elements.
Click on FILL to rescale the model to Iill the graphics area iI you used the ZOOM option. Figure 3.30-14 shows the
model with corrected loading.
Figure 3.30-14 Correctly Directed Loads for all Elements
MAIN
MESH GENERATION
CHECK
FLIP ELEMENTS
90 88 89 66 67 64 65 63 62 (pick the elements to be flipped)
END LIST (#)
72 73 70 71 69 68 (pick the elements to be flipped)
END LIST (#)
FILL
1389
CHAPTER 3.30
Container
Step 8: Add material properties.
The speciIic values Ior the isotropic material speciIication can be entered using the Iollowing sequence oI buttons:
The stress-strain data oI the material requires the use oI the TABLES option similar to Step 6 when you added the edge
loads. Remember that the table values are multiplied by the base value as was explained beIore in the section on
speciIication oI the pressure load. The table name is speciIied by clicking on the NAME button Iollowed by typing in
workhard as the name oI the table. You will use this table name later. The values Ior the plastic strain and stress are
listed in Table 2.30-1.
Note that based on the solution procedure (i.e., large displacements, updated Lagrange procedure, Iinite strain
plasticity), this data must be oI the Iorm listed in Table 2.30-2.
The Iollowing button sequence deIines the stress-strain data table.
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
11.0e6
POISSON'S RATIO
0.3
OK
Table 2.30-2 Stress Strain Definitions
Procedure Stress Strain
DeIault Engineering Engineering
Large Displacements 2nd Piola-KirchhoII Green-Lagrange
Large Strain Plasticity True (Cauchy) Logarithmic
MAIN
MATERIAL PROPERTIES
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
plastic_strain
OK (select OK button only if type plastic strain was typed in)
independent variable v1: XMIN
0
1390 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-15 Workhardening Slope
independent variable v1: XMAX
0.5
function value f: YMIN
0
function value f: YMAX
10000
NAME
workhard
MORE
independent variable v1: LABEL
log_strain
function value f: LABEL
true_stress
RETURN
ADD (refer to plastic strain stress data in Table 2.30-1)
SHOW TABLE
SHOW MODEL (select SHOW MODEL to switch back to model view)
1391
CHAPTER 3.30
Container
Plasticity may occur due to the extreme loading. Click on the PLASTICITY button which shows the plasticity properties
panel. Choose the deIaults von Mises yield criteria and the isotropic hardening rule. The initial yield stress is set to 1,
which means that the table values are multiplied by a Iactor oI 1.
Step 9: Add geometric properties.
SpeciIy the element thickness Ior all elements using the Iollowing button sequence:
Step 10: DeIine the loadcase.
Now that you have deIined the individual loads and kinematic constraints, deIine a loadcase that combines these
boundary conditions.
A pop-up menu appears over the graphics area containing a list oI available boundary conditions and their status
(selected or not). Combine these individual loads in a loadcase that can be reIerred to by a name. The deIault name Ior
this loadcase is lcase1. Clearly, you want all kinematic constraints and distributed loads (pressures) to be applied so
MAIN
MATERIAL PROPERTIES
ISOTROPIC
PLASTICITY
ELASTIC-PLASTIC
INITIAL YIELD STRESS
1
initial yield stress TABLE 1
workhard
OK (select OK button only if workhard was typed in)
OK (twice)
elements ADD
all: EXIST.
MAIN
GEOMETRIC PROPERTIES
AXISYMMETRIC
SHELL
THICKNESS
0.025
OK
elements ADD
all: EXIST.
1392 Marc Users Guide: Part 2
CHAPTER 3.30
all boundary conditions must be selected. Use the SELECT and DESELECT buttons to activate and deactivate the
boundary conditions respectively. ConIirm the correctness oI this loadcase deIinition by clicking on the OK button.
The Iollowing individual components Ior an analysis have already been speciIied:
1. You deIined the topology and connectivity oI the Iinite element model.
2. You assigned boundary conditions, material properties, and geometric properties.
3. You combined the boundary conditions in a loadcase.
Since snap-through is likely to occur in this problem, the user has to instruct Marc to solve a system oI equations with
a nonpositive deIinite tangent stiIIness matrix.
Furthermore, the deIault settings Ior convergence testing are not well suited Ior this particular problem. Although the
deIault type oI testing, relative testing on residual Iorces, is appropriate, the relative Iorce tolerance needs to be
reduced. The necessity oI this action can be explained by looking at the boundary conditions; the constraint on the
axial displacement degree oI Ireedom is Iound at a large radius (node 18).
Due to the internal pressure, we Iind a large reaction-Iorce at this node. Allowing a certain percentage oI this reaction-
Iorce to be present as residuals anywhere in the structure results in undesired interIerence oI those residuals with the
automatic load stepping process.
Finally, the adaptive load stepping algorithm oI Marc is activated. This algorithm allows Ior the analysis oI snap-
through phenomena in which the load incrementation needs to be scaled depending on the amount oI nonlinearity that
is occurring. Various parameters control this procedure. In this case, we allow Ior a maximum oI 600 increments.
MAIN
LOADCASES
mechanical STATIC
LOADS
OK
MAIN
LOADCASES
mechanical STATIC
SOLUTION CONTROL
NON-POSITIVE DEFINITE (on)
OK
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
1393
CHAPTER 3.30
Container
In the Iirst increment, 0.05 (5) oI the total load is applied. Also, the user needs to speciIy that the arc length never
exceeds the value used in the Iirst increment. The deIault MODIFIED RIKS procedure is used.
Step 11: Submit the job.
It is time to prepare the loadcase Ior a job and to submit it Ior Iinite element analysis.
Prior to deIining the job parameters, the appropriate Marc element type is set.
Next, the analysis class MECHANICAL is activated resulting in a pop-up menu over the graphics area. Click the
SELECT button and pick the lcase1 button Irom the available loadcases list to select the only available loadcase Ior
this job.
SOLUTION CONTROL
ARC LENGTH
arc length PARAMETERS
MAX # INCREMENTS IN LOADCASE
600
INITIAL FRACTION
0.05
MAX RATION ARC LENGTH / INITAL ARC LENGTH
1.0
OK (twice)
MAIN
JOBS
ELEMENT TYPES
mechanical element types AXISYM MEMBRANE/SHELL
89 (LINE 3 / thick shell)
OK
all: EXIST.
RETURN
RETURN
MECHANICAL
loadcases SELECT
lcase1
1394 Marc Users Guide: Part 2
CHAPTER 3.30
As we have indicated beIore, this analysis involves large displacements, Iinite strain plasticity, updated Lagrange
procedure, and Iollower Iorces. The Iinite element program requires directives that indicate this. From the mechanical
analysis pop-up menu, click on the ANALYSIS OPTIONS button and activate the Iollowing options:
The results oI the analysis appear in a results Iile. SpeciIy the results variables you are interested in by clicking on the
JOB RESULTS button Irom the mechanical analysis pop-up menu. Select Equivalent Plastic Strain and Equivalent von
Mises Stress variables. Finally, set the numbers oI layers used Ior integration through the shell thickness to 5.
This analysis may involve a large number oI increments. For this reason, you may want to write the results every 10
increments using the FREQUENCY button. For this sample session, however, write every increment which is the
deIault value oI the FREQUENCY option and conIirm the settings by clicking on the OK button.
MAIN
JOBS
MECHANICAL
ANALYSIS OPTIONS
SMALL STRAIN
(Switch on large strain additive procedure)
NO FOLLOWER FORCE
(to switch to FOLLOWER FORCE)
OK
MAIN
JOBS
MECHANICAL
JOB RESULTS
available element scalars
Equivalent Von Mises Stress
layers: OUT & MID
Total Equivalent Plastic Strain
layers: OUT & MID
OK
JOB PARAMETERS
# SHELL/BEAM LAYERS
5
1395
CHAPTER 3.30
Container
The Iollowing button sequence submits the job. The job can be monitored using the MONITOR option which, in case
oI automatically running the procedure Iile, prevents Mentat Irom proceeding to Step 12 beIore the analysis has run
to completion.
This analysis takes a Iew minutes, depending on the power oI the host to which you are submitting the job.
Step 12: Postprocess the results by looking at the deIormed shape and the load-displacement curve oI the node located
on the symmetry axis.
The results oI the analysis step are stored in a disk Iile. To access the results, it is necessary to open this Iile and
(selectively) extract data Irom it. Use the Iollowing button sequence to open the Iile.
Click on the FILL button located in the static menu area to scale the model to Iill the graphics area. The Iollowing
button sequence removes the node drawing and gives a clearer picture oI the model shown in Figure 3.30-14.
We are interested in the deIormed shape as a Iunction oI the increasing/decreasing load. The best way to obtain a good
overview oI the deIormations is to animate the deIormed shape. Since it is impractical to incorporate the animated
picture in this guide, a detailed description Iollows on how to obtain and play back the animation Irames. Only 3 out
oI a sequence oI 25 Irames are shown in this guide.
BeIore you collect the animation Irames, click on the DEF & ORIG button so that the deIormed and original shapes oI
the model are shown simultaneously.
MAIN
JOBS
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
MAIN
FILL
PLOT
draw NODES (off)
REGEN
RETURN
1396 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-16 Resulting Post File of the Container
The program responds by scanning the results Iile and extracting the appropriate data.
When replaying the sequence oI animation Iiles, you may have to scale the deIormed and original models to Iit in the
graphics area. Use the FILL button located in the static menu to scale the models while still having the last increment
oI the animation sequence displayed. Use the Iollowing button sequence to play the animation sequence.
MAIN
RESULTS
NEXT
DEF & ORIG
MORE
ANIMATION
create INCREMENTS
40 (Number of frames to save)
10 (Increment step size)
MAIN
RESULTS
MORE
ANIMATION
FILL
PLAY
SHOW MODEL
1397
CHAPTER 3.30
Container
The Iollowing two Iigures capture 3 oI the 25 animation Irames.
Figure 3.30-17 and Figure 3.30-18 show you that the bottom ridge Iirst un-rolls and is Iollowed by a snap-through oI
the arch. Ultimately, the shape oI the bottom oI the can becomes spherical as is shown in Figure 3.30-19.
Figure 3.30-17 Original and Deformed Model at Increment 80
The user can use the BEAM CONTOURS or BEAM VALUES plot option to display elemental quantities. Alternatively,
a path plot where the position oI the nodes is plotted on the x-axis and the equivalent plastic strain is plotted on the
y-axis, can be created. The nodes are logically connected though the connectivity oI the elements. The plot shown in
Figure 3.30-20 displays the equivalent plastic strain Ior increment 240 oI this analysis.
MAIN
RESULTS
SKIP TO INC
80
SKIP TO INC
160
SKIP TO INC
240
1398 Marc Users Guide: Part 2
CHAPTER 3.30
Figure 3.30-18 Original and Deformed Model at Increment 160
Figure 3.30-19 Original and Deformed Model at Increment 240
Use the Iollowing button sequence in PLOT to plot the total equivalent plastic strain.
MAIN
PLOT
1399
CHAPTER 3.30
Container
Figure 3.30-20 Path Plot of Equivalent Plastic Strain
It is possible to monitor these diagrams and obtain an overview oI the location and degree oI plastic strain as a Iunction
oI the loading.
Finally, a diagram is given that indicates how the bottom collapses as a Iunction oI the total load that was (adaptively)
put onto the structure. The total load is represented by the reaction Iorce in x-direction oI node 18 which is at the end
oI the line segment. The displacements oI the bottom are represented by the axial displacements oI node 1 which is at
radius 0.0.
draw NODES (on)
REGEN
RETURN
RESULTS
PATH PLOT
NODE PATH
(select nodes from lower left to upper right,
approximately each 10th node will do)
END LIST (#)
VARIABLES
ADD CURVE
Arc Length
Total Equivalent Plastic Strain Layer 3
FIT
1400 Marc Users Guide: Part 2
CHAPTER 3.30
The user can use the Iollowing button sequence to generate this Iigure.
MAIN
RESULTS
HISTORY PLOT
SHOW MODEL
SET NODES
18 (upper right)
1 (lower left)
END LIST (#)
COLLECT DATA
0 1000 1
NODES/VARIABLES
ADD 2-NODE CRV
1
variables at nodes
Displacement X
18
global variables
Dist Load 1
FIT
RETURN
SHOW IDS
10
YMAX
5000.0
1401
CHAPTER 3.30
Container

Figure 3.30-21 Axial Displacement of Bottom versus Applied Pressure Diagram
Conclusion
The loading path versus displacement has been successIully traced Ior the snap-through analysis oI the bottom oI an
aluminum container.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
container.proc Mentat procedure Iile
1402 Marc Users Guide: Part 2
CHAPTER 3.30
Chapter 3.31: Analyses of a Tire
3.31
Analyses of a Tire

Steady State Rolling Analysis 1404

Tire Bead Analysis 1423

Conclusion 1448

Input Files 1448

1404 Marc Users Guide: Part 2
CHAPTER 3.31
Steady State Rolling Analysis
A simpliIied automobile tire model is numerically analyzed. The analysis includes Iive steps:
1. 2-D model generation and tire inIlation,
2. 3-D model generation using 2-D results,
3. Footprint analysis,
4. Steady state rolling analysis using spinning velocity control,
5. Steady state rolling analysis using torque control.
This problem demonstrates Marc capability oI simulating automobile tires subjected to various load conditions. The
speciIic Ieatures used in the analysis contain:
1. The use oI rebar membrane elements along with INSERT option to model cord-reinIorced rubber composites,
2. AXITO3D option to transIer data Irom axisymmetric case to 3-D case, and
3. Steady state rolling,
among others.
Three mfd Iiles are provided in this chapter along with the procedure Iile tire.proc.
Simulation of a Tire
2-D Model Generation and Tire Inflation
The key issue here is to use rebar remeshing Ieature to mesh the given curves and deIine INSERT option.
In the Iirst step oI analysis (i.e., tire inIlation), the deIormation is purely axisymmetric and, thereIore, an axisymmetric
analysis is perIormed. The tire rim is modeled with a set oI Iixed boundary conditions. An inIlation pressure oI 2 bar
is applied to the inner surIace oI the tire within the step.
Open model and merge rebar curves:
tire2d_model.mfd contains the axisymmetric model except Ior rebars. All material properties,
including those Ior rebar elements, as well as boundary and load conditions Ior the
analysis oI tire inIlation, are deIined. The updated Lagrange Iormulation and the
Iollow Iorce option are also speciIied in the Iile.
reb_curves.mfd contains curves indicating rebar layer locations in the rubber matrix.
rigid_road.mfd includes a Ilat, rigid surIace Ior modeling road in 3-D analysis.
FILES
OPEN
tire2d_model.mfd
OK
1405
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-1 2-D Tire Model with Rebar Curves
Mesh rebar curves, generate INSERT and define rebar properties:
INSERT is automatically created iI the CREATE INSERTS under MESH 2D REBAR is ON (deIault).
SAVE AS
tire2d.mfd
OK
FILL
MERGE
reb_curves.mfd
OK
MAIN
MESH GENERATION
AUTO MESH
MESH 2D REBAR
MESH CURVES
all existing
all existing
RETURN (twice)
CLEAR GEOMETRY
1406 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-2 Menu for REBAR Meshing
RENUMBER
ALL
MAIN
MATERIAL PROPERTIES
NEXT (thrice)
LAYERED MATERIAL
ELEMENT ADD
33 to 62
NEXT
ELEMENT ADD
insert6_embed_elements
NEXT
ELEMENT ADD
insert7_embed_elements
SAVE
MAIN
1407
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-3 Tire Model with Rebar Elements and INSERT Option
Run axisymmetric job:
3-D Model Generation
A Iully 3-D model is generated based on the axisymmetric model. AXITO3D option is used to transIer the numerically
obtained results Irom the previous axisymmetric analysis into the 3-D case as initial conditions in the 3-D analysis.
3-D model generation contains the Iollowing steps.
1. Save model as tire3d.mfd
2. Expand model Irom Axisymmetric to 3-D
3. Add rigid body control node
4. Merge rigid road surIace
5. DeIine boundary/load conditions
6. AXITO3D deIinition in initial conditions
7. Contact deIinition
JOBS
RUN
SUBMIT
OK
MAIN
1408 Marc Users Guide: Part 2
CHAPTER 3.31
FILE
SAVE AS
tire3d.mfd
OK
MAIN
MESH GENERATION
EXPAND
AXISYMMETRC MODEL TO 3D
ANGLE
30
REPETITION
5
ANGLE
6
REPETITION
10
ANGLE
30
REPETITION
5
TIME SET
1
EXPAND MODEL
FILL
RETURN (twice)
NODES ADD
0 0 0
MAIN
FILE
MERGE
rigid_road.mfd
OK
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
1409
CHAPTER 3.31
Analyses of a Tire
NAME
fix_xz
FIXED DISPLACEMENT
DISPLACEMENT X: ON
DISPLACEMENT Z: ON
OK
NODES ADD
2181
#
NEW
NAME
disp_y
TABLE
NEW
1 INDEPENDENT VARIABLES
NAME
table_disp
TYPE
time
OK
ADD
0 0 1 1 10 1
FIT
RETURN
FIXED DISPLACEMENT
DISPLACEMENT Y: ON
TABLE
table_disp
OK
DISPLACEMENT Y VALUE: 15
OK
NODES ADD
2181
#
NEW
NAME
1410 Marc Users Guide: Part 2
CHAPTER 3.31
load_y
TABLE
NEW
1 INDEPENDENT VARIABLES
NAME
table_load
TYPE
time
OK
ADD
0 0 1 0 2 1 10 1
FIT
RETURN
POINT LOAD
DISPLACEMENT Y: ON
TABLE
table_load
OK
DISPLACEMENT Y VALUE: 3400
OK
NODES ADD
2181
#
MAIN
INITIAL CONDITIONS
MECHANICAL
AXISYMMETRIC TO 3D
POST FILE
tire2d_job1.t16
OK (twice)
MAIN
CONTACT
CONTACT BODIES
NEW
NAME
tire
1411
CHAPTER 3.31
Analyses of a Tire
DEFORMABLE
OK
ELEMENTS ADD
all_existing
NEW
NAME
road
RIGID
LOAD
DISCRETE
SURFACES ADD
all_existing
CONTROL NODE
2181
RETURN
CONTACT TABLES
NEW
PROPERTIES
1-2
CONTACT TYPE: TOUCHING
OK (twice)
NEW
PROPERTIES
1-2
CONTACT TYPE: TOUCHING
FRICTION COEFFICIENT: 0.5
OK (twice)
MAIN
1412 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-4 3-D Tire Model
Footprint Analysis
In the Iirst loadcase, the rigid road surIace moves up 15 mm against the tire using position control option Ior rigid
contact body. AUTO STEP option is used. The position control is then switched to load control in the second loadcase.
A vertical load oI 3400 N is applied to the road surIace within one increment.
LOADCASES
REM (twice)
NEW
NAME
footprint1
MECHANICAL
STATIC
LOADS
load_y: OFF
OK
CONTACT
CONTACT TABLE
ctable1
OK (twice)
CONVERGENCE TESTING
1413
CHAPTER 3.31
Analyses of a Tire
Steady State Rolling Analysis Spinning Velocity Control
In the Iollowing two loadcases, steady state rolling analysis is perIormed. The tire starts to spin at an angular velocity
oI 11.9 cycle/second and runs at a road velocity oI . Only one increment is required to achieve converged
solutions at the given conditions. AIterwards, the spinning velocity oI the tire gradually increases to 16.4 cycle/second
within 20 equal increments.
RELATIVE FORCE TOLERANCE
0.05
OK
MULTI-CRITERIA
PARAMETERS
INITIAL FRACTION OF LOADCASE TIME
0.1
OK (twice)
NEW
NAME
footprint2
STATIC
LOADS
disp_y: OFF
OK
CONTACT
CONTACT TABLE
ctable1
OK (twice)
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
# STEPS
1
OK
NEW
NAME
100
km
hr
-------
1414 Marc Users Guide: Part 2
CHAPTER 3.31
ss_rolling_1
STEADY STATE ROLLING
LOADS
disp_y: OFF
OK
CONTACT
CONTACT TABLE
ctable2
OK (twice)
STEADY STATE ROLLING
SPINNING BODY
tire
OK
GROUND BODY
road
OK
SPINNING VELOCITY
11.9
GROUND VELOCITY Z
27777.8
GRADUAL FRICTION: ON
OK
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
# STEPS
1
OK
NEW
NAME
ss_rolling_2
STEADY STATE ROLLING
LOADS
disp_y: OFF
1415
CHAPTER 3.31
Analyses of a Tire
OK
CONTACT
CONTACT TABLE
ctable2
OK (twice)
STEADY STATE ROLLING
SPINNING BODY
tire
OK
GROUND BODY
road
OK
SPINNING VELOCITY
16.4
GROUND VELOCITY Z
27777.8
OK
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
# STEPS
20
OK
1416 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-5 Menu to Define Loadcase for Steady State Rolling with Spinning Velocity Control
Steady State Rolling Analysis Torque Control
Free rolling analysis is perIormed using torque control option. The solution is achieved with only one increment at
zero torque.
NEW
NAME
free_rolling
STEADY STATE ROLLING
LOADS
disp_y: OFF
OK
CONTACT
CONTACT TABLE
ctable2
OK (twice)
STEADY STATE ROLLING
SPINNING BODY
tire
OK
GROUND BODY
road
OK
METHOD: TORQUE: ON
1417
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-6 Menu to Define Loadcase for Steady State Rolling with Torque Control
Run Job and View Results
Job definition and run the generated model:
GROUND VELOCITY Z
27777.8
OK
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.05
OK
# STEPS
1
OK
MAIN
JOBS
MECHANICAL
footprint1
foorprint2
ss_rolling_1
ss_rolling_2
free_rolling
1418 Marc Users Guide: Part 2
CHAPTER 3.31
Results deformed tire at the end of footprint:
INITIAL CONDITION
fix_xz: ON
disp_y: ON
icond1: ON
OK
CONTACT
COULOMB FOR ROLLING
INITIAL CONTACT
CONTACT TABLE
ctable1
OK (twice)
OK (twice)
SAVE
RUN
SUBMIT
OK
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG: ON
SKIP TO INC
7
1419
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-7 Deformed Tire at the End of Footprint
Results spinning velocity traction curve:
HISTORY PLOT
COLLECT DATA
8 28 1
NODE/VARIABLES
ADD GLOBAL CURVE
X AXIS: ANGLE VEL TIRE
Y AXIS: FORCE Z ROAD
FIT
1420 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-8 Spinning Velocity Traction Curve
More Results on Contact Friction Stresses
The contact Iriction stresses along the central line oI Iootprint area at the stages oI Iull braking, Iull traction, and Iree
rolling are shown in Figure 3.31-9, Figure 3.31-10, and Figure 3.31-11, respectively.
A comparison oI rolling resistance at diIIerent spinning velocities Ior two diIIerent Iriction coeIIicients, 0.3 and 0.5,
is shown in Figure 3.31-12.
1421
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-9 Contact Friction Stress along the Central Line of Footprint Area Full Braking
Figure 3.31-10 Contact Friction Stress along the Central Line of Footprint Area Full Traction
1422 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-11 Contact Friction Stress along the Central Line of Footprint Area Free Rolling
Figure 3.31-12 Spinning Velocity Rolling Resistance Curves at Difference Frictions
-2.00E+03
-1.50E+03
-1.00E+03
-5.00E+02
0.00E+00
5.00E+02
1.00E+03
1.50E+03
2.00E+03
1.20E+01 1.35E+01 1.50E+01 1.65E+01
Spinning Velocity
R
o
l
l
i
n
g

R
e
s
i
s
t
a
n
c
e
Friction Coefficient: 0.3
Friction Coefficient: 0.5
1423
CHAPTER 3.31
Analyses of a Tire
Tire Bead Analysis
This section describes the analysis oI the cross section oI an automobile tire. The model is loaded by an internal
pressure and the contact between the tire and the rim is to be analyzed.
Overview
The method used in this section to obtain a solution is typical Ior tackling an engineering problem. This section
demonstrates that it is useIul to approach a problem by using simple models Iirst beIore going on to large complicated
structures. This approach not only gives you a better understanding oI your problem, but it also enables you to better
analyze the results.
Background Information
Description
An automobile tire is a complex composite structure consisting oI (nonlinear) materials that comes into contact with
the road.
Figure 3.31-13 identiIies the diIIerent materials and parts oI an automobile tire by part name.
Figure 3.31-13 Cross-Section of Automobile Tire
Tread
Steel belts
Side wall Carcass
Apex
Steel bead
Chafer
Rim
1424 Marc Users Guide: Part 2
CHAPTER 3.31
Idealization
The material properties oI the tread, side wall, chaIer, and apex are isotropic. The carcass is characterized by an
orthotropic material property. The steel belts and beads behave as isotropic materials in the circumIerential direction
oI the tire. In this analysis, both the carcass on the steel belts and beads have been given the same properties as the
rubber and, thus, no special elements are required in modeling these parts. The tire comes into contact at the chaIer
with the wheel rim. The wheel rim is modeled as an inIinitely stiII body.
Figure 3.31-14 Overall Dimensions of the Tire
Level of Analysis Detail
This section describes the diIIerent stages oI idealization that are observed in this analysis. As was noted in the
Overview, the best approach Ior analyzing a complicated structure is to start with simple models. This approach allows
you to gain knowledge and conIidence in problem-solving as you progress through the analysis process. This approach
also helps you to predict behaviors and to identiIy potential problems.
In this sample session, you will only analyze the inIlation process using a crude and easy to generate mesh. The
analysis presents some oI the main components oI detailed analysis.
Analysis
The purpose oI this initial analysis is to describe the inIlation process by means oI an idealization oI the real structure.
Idealization
For purposes oI this simpliIied analysis, assume all materials to be identical and ignore the treads in the tire. An
axisymmetric model is used and because oI symmetry, you only need to analyze halI oI the cross-section.
100 mm
210 mm
1
7
0

m
m
3
0
0

m
m
1425
CHAPTER 3.31
Analyses of a Tire
Requirements for a Successful Analysis
The analysis is considered successIul iI the closing behavior at the rim/chaIer interIace oI this simpliIied model can
be shown.
Full Disclosure
Type of analysis
Contact
Materials
The rubber material Ior this structure is characterized by three Mooney constants Ior which the Iollowing
values are chosen:
Elements
Marc Element Type 82, Iour-noded axisymmetric Herrmann Iormulation.
Overview of Steps
Step 1: Create the boundary using Bezier curves.
Step 2: Use automatic meshing (OVERLAY) to create a mesh.
Step 3: Create the rim as a rigid die and identify the contact bodies.
Step 4: Add boundary conditions.
Step 5: Apply internal pressure.
Step 6: Submit the job.
Step 7: Postprocess the results.
Detailed Session Description
The description oI the tire boundary geometry is well suited Ior the use oI Bezier curves. The deIining polygon oI a
Bezier curve can easily be changed which results in a smooth change in the entire curve. To demonstrate the versatility
oI this curve type, we will generate the boundary oI the tire using Bezier curves exclusively.
Step 1: Create the boundary using Bezier curves.
BeIore entering the Bezier curves, however, Iirst establish an input grid using the Iollowing button sequence.
MAIN
MESH GENERATION
C
10
965kPa
C
20
193kPa
C
30
193kPa
1426 Marc Users Guide: Part 2
CHAPTER 3.31
Observe that the dimension oI the grid size is speciIied in centimeters. The material constants speciIied in Idealization
require a conversion Irom kPa into N/cm
2
, in order to be consistent with the units used here. For a good resolution oI
the Bezier curve drawing, set the plotting oI curves with high accuracy.
Note that when drawing, the number oI subdivisions is merely a drawing resolution. The inIormation on every point on
the curve is preserved. Use the Iollowing button sequence to change the resolution and to set the curve type to Bezier.
The curves are added by clicking on the ADD button oI the crvs panel and entering the points Ior the deIining polygon
vertices oI each curve. The beginning and end points oI the Bezier curve are determined by the Iirst and last point
speciIied. The tangent at either end is deIined by the neighboring points.
SET
U DOMAIN
-17 0
U SPACING
1
V DOMAIN
0 17
V SPACING
1
grid ON (on)
FILL
MAIN
PLOT
curves SETTINGS
predefined settings HIGH
RETURN (twice)
MESH GENERATION
CURVE TYPE
BEZIER
MAIN
MESH GENERATION
crvs ADD (pick the appropriate grid points)
point(0,14,0)
point(-11,13,0)
point(-11,6,0)
1427
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-15 Interior Tire Wall
Increase the resolution oI the grid to 0.5 units:
Create the exterior side oI the wall oI the tire by adding the Iollowing curve, the result oI which is shown in
Figure 3.31-16.
point(-5,4,0)
point(-5,0,0)
END LIST (#)
MAIN
MESH GENERATION
SET
U SPACING
0.5
V SPACING
0.5
MAIN
MESH GENERATION
crvs ADD (pick the appropriate grid points)
1428 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-16 Part of Exterior Tire Wall Added
Switch on the labeling oI points in order to Iacilitate creating the curves as speciIied in the button sequences below.
Add the Iollowing curve to create the lower part oI the exterior wall oI the tire. Even though severe changes in
curvature occur in this part, the overall curve remains smooth. The results are shown in Figure 3.31-17.
point(-9,16,0)
point(-9,14,0)
point(-10.5,13,0)
point(-10.5,8.5,0)
END LIST (#)
MAIN
PLOT
points SETTINGS
LABEL (on)
REGEN
MAIN
MESH GENERATION
crvs ADD
9 (pick point)
1429
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-17 Exterior Tire Wall Completed
Figure 3.31-17 clearly indicates that the shape oI the portion oI the tire that comes into contact with the rim oI the
wheel is incorrect. This is remedied by relocating some oI the support points oI the Bezier curve.
Points 13, 10, and 11 are relocated using the pts EDIT button on the mesh generation panel, the results oI which are
shown in Figure 3.31-18.
point(-10.5,4,0) (pick grid point)
point(-9.5.3.5,0) (pick grid point)
point(-8.5,1.5,0) (pick grid point)
point(-8,0,0) (pick grid point)
point(-7.5,0.5,0) (pick grid point)
5 (pick point)
END LIST (#)
MAIN
MESH GENERATION
pts EDIT
13 (pick point)
-9.5 -0.5 0 (pick grid point)
10 (pick point)
-11.0 -0.5 0 (pick grid point)
1430 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-18 Correction of Tire Geometry
To add the tread, use a Bezier curve made up oI point 6 and three new points. The exact location oI these points are
shown in Figure 3.31-19. The Iollowing button sequence speciIies where these points are located in the local u-v-w
coordinate system. ReIrain Irom entering the points by typing in their coordinates; instead, always use the mouse to
pick the points as it is a much easier method.
11 (pick point)
-7 5 0 (pick grid point)
END LIST (#)
MAIN
MESH GENERATION
crvs ADD
6 (pick point)
point(-7.5,16,0) (pick grid point)
point(-1.5,17,0) (pick grid point)
point(0,17,0) (pick grid point)
END LIST (#)
1431
CHAPTER 3.31
Analyses of a Tire

Figure 3.31-19 Tire Tread
Finally, to complete the boundary, add a straight line between points 1 and 17 which Iorm the symmetry boundary. A
Bezier curve is used here simply to demonstrate how it degenerates into a straight line when only two points are
speciIied. The completed boundary is shown in Figure 3.31-20.
MAIN
MESH GENERATION
crvs ADD
17 (pick point)
1 (pick point)
END LIST (#)
1432 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-20 Completed Boundary
Step 2: Use automatic meshing (OVERLAY) to create a mesh.
Use the automatic overlay meshing option to create the mesh. This automatic mesh generator requires a closed
boundary. The only input needed Irom the user is the number oI subdivisions in the x- and y-direction, respectively,
and the identiIication oI the closed boundary.
MAIN
MESH GENERATION
GRID (off)
FILL
AUTOMESH
2D PLANAR MESHING
DIVISIONS
20 20
quadrilaterals (overlay) QUAD MESH!
all: EXIST.
1433
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-21 Mesh generated by OVERLAY
It should be clear Irom the visual inspection that the mesh is rather coarse in the lower area where the tire comes into
contact with the wheel rim. A local reIinement is necessary and can be accomplished by using the SUBDIVIDE and
REFINE processors.
MAIN
MESH GENERATION
SUBDIVIDE
DIVISIONS
2 1 1
ELEMENTS
51 3 5 7 60 (pick elements)
END LIST (#)
1434 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-22 Step 1 of Mesh Refinement
The REFINE option can be used to eIIectively create a transition between layers or rows oI elements. It requires two
sets oI inIormation:
The node about which the reIinement is made;
The elements that will participate in the reIinement.
Note that only those elements that have the reIined node as part oI the connectivity are eligible.
Complete this action by subdividing two more elements according to Figure 3.31-24.
MAIN
MESH GENERATION
SUBDIVIDE
REFINE
1 (pick refine node)
1 (pick element)
END LIST (#)
MAIN
MESH GENERATION
SUBDIVIDE
ELEMENTS
1435
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-23 Local Refinement around a Node
Remember that some processors such as SUBDIVIDE, EXPAND, SYMMETRY, and DUPLICATE may create
duplicate nodes. Although the nodes are in the same position, they are not connected. The node that is picked as the
reIine node may be part oI one element's connectivity but not oI the neighboring element. The REFINE processor can,
in such an instance, produce unexpected and undesired results. To prevent this, it is usually prudent to activate the
SWEEP processor before a reIine operation is perIormed.
Compression oI all nodes located within a speciIied distance is accomplished by activating the NODES button in the
SWEEP menu Iollowed by a list oI nodes that you want to sweep. Generally, you use the all: EXIST. list button to sweep
all existing nodes. Finally, renumber all items in the database in order to obtain a sequential node and element
numbering.
47 46 (pick elements)
END LIST (#)
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
remove unused NODES
RETURN
RENUMBER
ALL
1436 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-24 Completion of Local Mesh Refinement
Step 3: Create the rim as a rigid die and identiIy the contact bodies.
The rim oI the wheel is considered to be a rigid body and is constructed using a Bezier curve. Use the Iollowing button
sequence to add the rim.
MAIN
MESH GENERATION
FILL
ZOOM BOX
(zoom in on the lower area)
GRID (on)
crvs ADD
5 (pick point)
point(-12,0,0) (pick grid point)
point(-5.5,2.5,0) (pick grid point)
point(-11,3.5,0) (pick grid point)
point(-11,1.5,0) (pick grid point)
END LIST (#)
1437
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-25 Wheel Rim added
The mesh has been conveniently generated so that the origins coincide with the center line and the bottom oI the tire.
Use the Iollowing button sequence to move the entire mesh and rim over a distance that is equivalent to the radius oI
the wheel:
MAIN
MESH GENERATION
GRID (off)
SELECT
NODES
all: OUTLINE
END LIST (#)
RETURN
MOVE
TRANSLATIONS
0 13 0
POINTS
all: EXIST.
NODES
all: UNSEL.
RETURN
FILL
1438 Marc Users Guide: Part 2
CHAPTER 3.31
Not all elements oI the tire come into contact with the rim. You can drastically minimize the analysis time by
identiIying the elements that make up the deIormable body that is expected to come into contact with the rim.
To identiIy the rim (curve) as a rigid contact body, use the Iollowing button sequence:
MAIN
CONTACT
CONTACT BODIES
NEW
DEFORMABLE
OK
elements ADD
(pick the elements that may come
into contact with the rigid body)
END LIST (#)
Note: It is important to use NEW in the Iollowing button sequence. II NEW is not used, the contact body just entered is
written over.
MAIN
CONTACT
CONTACT BODIES
NEW
RIGID
OK
CURVES ADD
6 (pick curve)
END LIST (#)
PLOT
draw POINTS (off)
elements SETTINGS
draw SOLID
RETURN (twice)
ID CONTACT (on)
1439
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-26 Identification of Contact Bodies
The contact bodies are identiIied on the graphics screen by clicking on the ID CONTACT button oI the contact bodies
panel. The curve that represents the rigid body is enhanced by cross-hatching the side where the body is located. II the
display indicates that the body is located on the incorrect side, use the FLIP CURVES option to Ilip the curve. ReIer to
Chapter 3.30: Container, Step 7 Ior a detailed description on using the FLIP CURVES option.
Now, switch oII the identiIication oI contact bodies.
MAIN
CONTACT
CONTACT BODIES
ID CONTACT (off)
PLOT
elements SETTINGS
draw WIREFRAME
RETURN
REGEN
FILL
1440 Marc Users Guide: Part 2
CHAPTER 3.31
Step 4: Add boundary conditions.
Symmetry conditions are applied to the nodes along the symmetry line using the Iollowing button sequence:
The symmetry boundary conditions are displayed in Figure 3.31-27.
Figure 3.31-27 Symmetry Boundary Conditions Applied
Step 5: Apply internal pressure.
The tire is loaded by an internal pressure. Use the Iollowing button sequence to speciIy the loading history through
a table.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
OK
nodes ADD
130 138 142 (pick the three nodes at the right)
END LIST (#)
MAIN
BOUNDARY CONDITIONS
1441
CHAPTER 3.31
Analyses of a Tire
It is important to speciIy the table type because a table will only be applied iI the appropriate type is assigned. For
boundary conditions, only table type time is valid.
Apply this load to the interior oI the tire using the Iollowing button sequence:
MECHANICAL
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
OK (select OK button only if type time was typed in)
NAME
loading
TYPE
time
independent variable v1: MAX
300
function value f: MAX
220
ADD
0 0
300 214
SHOW TABLE
SHOW MODEL (select SHOW MODEL to return to model view)
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
EDGE LOAD
pressure TABLE
loading
OK
SELECT
CLEAR SELECT
...EDGES
1442 Marc Users Guide: Part 2
CHAPTER 3.31
The results oI the applied internal pressure are depicted in Figure 3.31-28.
Figure 3.31-28 Internal Pressure Applied
The material Ior this mesh is assumed to be uniIorm over the entire mesh. SpeciIy the material properties using the
Iollowing button sequence:
42:2
END LIST (#)
SELECT BY
edges by CRVS
1
END LIST (#)
RETURN
RETURN
edges ADD
all SELECT
END LIST (#)
SELECT
CLEAR SELECT
MAIN
MATERIAL PROPERTIES
1443
CHAPTER 3.31
Analyses of a Tire
Step 6: Submit the job.
Use the Iollowing button sequence to prepare a loadcase.
This loadcase is to be used in the job that ultimately is submitted Ior analysis. Use the Iollowing button sequence to
speciIy the job.
MORE
MOONEY
C10
96.5
C20
-19.3
C30
19.3
OK
elems ADD
all EXIST.
MAIN
LOADCASES
mechanical STATIC
LOADS, OK
TOTAL LOADCASE TIME
300
# STEPS
300
OK
MAIN
JOBS
MECHANICAL
loadcases SELECT
lcase1
ANALYSIS OPTIONS
LARGE DISPLACEMENT (on)
NO FOLLOWER FORCE
(to switch to FOLLOWER FORCE)
1444 Marc Users Guide: Part 2
CHAPTER 3.31
Use the Iollowing button sequence to submit the job.
The analysis stops with an exit number 2004, indicating that a rigid body motion is present (the tire separates Irom
the rim).
Step 7: Postprocess the results.
The purpose oI the preliminary analysis is to gain experience in completing a relatively simple analysis. The Iollowing
results are displayed:
1. Animation oI the deIormation. Only the Iirst and last Irame are shown here.
2. Contouring oI the von Mises stress on the tire cross section.
Use the Iollowing button sequence to open the results Iile.
OK
JOB RESULTS
available element tensors
Cauchy Stress
OK
AXISYMMETRIC
OK
ELEMENT TYPES
MECHANICAL
AXISYM SOLID
82 (FULL & HERRMANN FORMULATION/
QUAD (4))
OK
all: EXIST.
RETURN
MAIN
JOBS
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
1445
CHAPTER 3.31
Analyses of a Tire
To Iocus on the geometry, it is necessary to turn the node labeling and Iace identiIication oII as is shown in
Figure 3.31-29.
Use the button sequence given below to turn the node labeling and Iace identiIication oII.
Figure 3.31-29 Mesh without Node Labeling and Face Identification
OPEN DEFAULT
NEXT
MAIN
FILL
PLOT
draw NODES (off)
elements SETTINGS
draw FACES (off)
RETURN
REGEN
1446 Marc Users Guide: Part 2
CHAPTER 3.31
Click on DEF & ORIG to request the original and deIormed structure to be shown. The animation buttons are in the
second part oI the postprocessing results menu and can be reached by clicking on the MORE button. To create the
animation Irames, use the Iollowing button sequence:
The numeral 100 is entered here as a response to the number oI increments that need to be processed.
From the analysis, we know that the results stretch out over approximately 14 increments. Hence, 100 is a saIe upper
limit. The program now prepares the Irames Ior animation. Figure 3.31-30 and Figure 3.31-31 show the second and
last oI the animation Irames.
Figure 3.31-30 Second Animation Frame
MAIN
RESULTS
REWIND
MORE
ANIMATION
INCREMENTS
100
1
1447
CHAPTER 3.31
Analyses of a Tire
Figure 3.31-31 Last Animation Frame
To animate the sequence oI Irames use the Iollowing button sequence:
The equivalent Cauchy stress can be displayed by using the Iollowing button sequence:
Figure 3.31-32 shows the results oI the model with equivalent stress contour bands.
MAIN
RESULTS
FILL
MORE
ANIMATION
FILL
PLAY
MAIN
RESULTS
SCALAR
Equivalent Cauchy Stress
LAST
CONTOUR BANDS
1448 Marc Users Guide: Part 2
CHAPTER 3.31
Figure 3.31-32 Mesh with Equivalent Cauchy Stress Contours
Conclusion
This example demonstrates that it is relatively easy to complete a contact analysis using a simple to generate geometry
and an incompressible material. The tire looses contact with the rim. This is caused by the Iact that the steel belt is not
present in this analysis. Although the results shown in this analysis have little engineering value, the analysis is
valuable in reassuring that the available tools in Marc enable you to solve a more complex problem.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
steady_state.proc Mentat procedure Iile
tire_rim.proc Mentat procedure Iile
tire2d_model.mfd Mentat model Iile used in steady state
rigid_road.mfd Mentat model Iile used in steady state
reb_curves.mfd Mentat model Iile used in steady state
Chapter 3.32: Transmission Tower
3.32
Transmission Tower

Chapter Overview 1450

Background Information 1450

Detailed Session Description 1452

Conclusion 1500

Input Files 1500

1450 Marc Users Guide: Part 2
CHAPTER 3.32
Chapter Overview
This chapter describes a sample session that illustrates the Iunctionality oI the Mentat program through the modeling
and analysis oI a tower structure. The goal oI this chapter is to give you hands-on experience with the Iollowing Mentat
capabilities.
To show you how to create a mesh oI linear beam elements using the Iollowing mesh generation Ieatures:
user-deIined local coordinate systems
node and element creation,
element subdivision and duplication.
To demonstrate a static and a modal analysis oI a model.
To view and examine the results oI an analysis.
Background Information
Tower Description
The tower is 68 Ieet high, 18 Ieet square at the base, 4 Ieet square at the top, and has 6 cable-arms, each 6 Ieet wide.
The tower is made oI steel angles (L3x3x1/4 and L2x2x1/4) and is loaded by member selI-weight, wind, and
cable loads.
The Ieet at the base oI the tower are Iixed. A sketch oI the tower is depicted in Figure 3.32-1.
Figure 3.32-1 Transmission Tower
Arm
Head
Mid-section
Base
1451
CHAPTER 3.32
Transmission Tower
Idealization
By virtue oI the element type used in this model, Marc Element Type 52, the sketch in Figure 3.32-1 and the Iinite
element model itselI are identical. This may not be true in all cases. Members oI the tower are idealized as beam
elements with six degrees oI Ireedom at each node (ux, uy, uz, rxx, ryy, rzz).
The wind loads are applied as distributed loads along the main vertical members oI the tower. Cable loads are applied
as point loads at the end oI the cable-arms. SelI-weight is applied as distributed loads on all members.
Requirements for a Successful Analysis
The analysis is considered successIul iI the displacements oI the structure, as a result oI its external loading can be
determined. The second part oI the analysis (modal analysis) is successIul iI the eigenmodes oI the structure can be
predicted.
Full Disclosure
Analysis Types
Linear static
Modal
Material
Steel, Young`s modulus 4.176e9 psI, Poisson`s ratio 0.3.
Mass density 15.217 slugs/It
3
.
Elements
Marc Type 52, three-dimensional, two-noded beam element.
Element Properties
Those obtained Irom the AISC Steel manual.
L3x3x1/4
Weight 4.9 lbs/It.
Area 0.01 It
2
.
Ixx Iyy 6.0e-5 It
4
.
L2x2x1/4
Weight 3.19 lbs/It.
Area 0.00651 It
2
.
Ixx Iyy 2.0e-5 It
4
.
1452 Marc Users Guide: Part 2
CHAPTER 3.32
Overview of Steps
Step 1: Create the first face of the main tower structure by adding nodes and elements and using user-
defined coordinate systems, subdivision, and symmetry.
Step 2: Duplicate the first face to create the remaining faces of the main tower structure. It is crucial to
sweep nodes and elements after using symmetry and duplicate.
Step 3: Create one cable arm by adding nodes and elements and using subdivide.
Step 4: Use symmetry on the first arm to create the second arm. Then use duplicate on the first two to
create the remaining cable arms.
Step 5: Add boundary conditions.
Step 6: Define the material and apply it to all elements. Define the geometric properties and apply them
to the appropriate elements.
Step 7: Job submission of the static analysis.
Step 8: Static analysis results processing.
Step 9: Job submission of the modal analysis.
Step 10: Modal analysis results processing.
Detailed Session Description
When modeling a structure, it is very important to deIine a coordinate system that can be reIerred to as you create
diIIerent parts oI the structure. The global coordinate system, called the x-y-z system, is the coordinate system
attached to the earth and can be used Ior this purpose. The global coordinate system may not always be the optimal
choice in Mentat because the design oI the program restricts the orientation oI the model, particularly when graphical
input is desired.
A local coordinate system is a set oI three independent directions and an origin, deIined with respect to the global
coordinate system. For easy reIerence, we reIer to the local coordinate system as the u-v-w system. The nature oI this
system may be Cartesian, cylindrical, or spherical according to the commonly accepted deIinitions. Initially, the local
coordinate system coincides with the global x-y-z system. We emphasize once again that the position and orientation
oI the local coordinate system is deIined in terms oI the global coordinate system. Everything else, except viewpoint,
is deIined in terms oI the local coordinate system.
A Note on Grid Space
II you are using the mouse to input entities such as nodes, you need to relate the two-dimensional space oI your screen
to three-dimensional reality. Choose the u-v plane oI the local coordinate system as a plane that is sensitive to mouse
picks. You can orient the input grid anywhere in space simply by translating and rotating the local coordinate system.
By virtue oI the Iact that the local u-v-w coordinate system initially lines up with the global coordinate system, the
input grid also initially lies in the global x-y plane.
1453
CHAPTER 3.32
Transmission Tower
Figure 3.32-2 Local Coordinate System (u-v-w)
A Note on Viewpoint
The ability to orient the local coordinate system anywhere in space does not necessarily mean that it is optimal Ior
graphical input. The best resolution oI the grid is obtained by viewing the grid plane with the eye positioned along the
normal to that plane.
Use VIEWPOINT in the MANIPULATE CAMERA (ABSOLUTE) menu - which can be entered via the VIEW menu to
deIine the appropriate eye position measured with respect to the global coordinate system.
Once again, we emphasize that viewpoint and the position oI the local coordinate system are the only two exceptions
to the rule that everything in Mentat is measured in local coordinates.
Keep the Iollowing three points in mind with respect to viewpoint.
Changing the viewpoint is not the same as rotating the object that you are viewing. Although the end result
may appear to be the same, there is a Iundamental diIIerence. Changing the viewpoint does not change the
position oI the model, while a transIormation oI the object permanently changes the position oI that object.
Changing the viewpoint is not related to changing the local coordinate system. These are two independent
actions.
Changing the position and sense oI the local coordinate system only aIIects the position oI entities that have
yet to be deIined. It does not inIluence or change the position oI entities, such as nodes or points, that have
already been deIined.
Step 1: Create the Iirst Iace oI the main tower structure by adding nodes and elements and using user-deIined
coordinate systems, subdivision, and symmetry.
As described in previous sample sessions, the Iirst step in building a Iinite element mesh is to establish an input grid.
Activate the grid and set the grid spacing to 1 unit and the grid size to 10 units.
x
y
zz
1454 Marc Users Guide: Part 2
CHAPTER 3.32
The best approach to use Ior creating the transmission tower is to align the center line oI the structure with the global
z-axis. Rotate the local coordinate system about the global x-axis over 90, and translate it over nine units in the global
y-direction. Set and activate view 2.
Prior to adding elements, the user has to change the deIault element class Ior newly generated elements Irom QUAD(4)
to LINE (2). Use the ADD button Irom the ELEMS panel to add the Iirst three elements to Iorm a triangle that will
constitute the base oI the tower:
MAIN
MESH GENERATION
SET
U DOMAIN
-10 10
U SPACING
1
V DOMAIN
-10 10
V SPACING
1
grid ON (on)
ROTATE
90 0 0
TRANSLATE
0 9 0
VIEW
activate 2 (on)
show 2
FILL
PLOT
nodes SETTINGS
LABELS (on)
RETURN
elements SETTINGS
LABELS (on)
RETURN
MAIN
MESH GENERATION
1455
CHAPTER 3.32
Transmission Tower
Figure 3.32-3 First Three Elements of Tower Base
Having obtained Figure 3.32-3, continue to subdivide the vertical side and the hypotenuse, and add the two cross
members. Notice how you have started to create to the leIt oI the local v-axis. Although this does not look at all like
the base yet, you will continue to add to the leIt side oI the tower Iace and use symmetry to complete the Iirst Iace.
Figure 3.32-4 shows the results oI this operation.
ELEMENT CLASS
LINE (2)
RETURN
elems ADD
node(-9,0,0) (pick grid point)
node(-9,10,0) (pick grid point)
node(0,10,0) (pick grid point)
1 (pick node)
2 (pick node)
3 (pick node)
MAIN
MESH GENERATION
SUBDIVIDE
ELEMENTS
1 2
1456 Marc Users Guide: Part 2
CHAPTER 3.32

Figure 3.32-4 Left Base of Tower
The next step is to establish a new local coordinate system with a grid spacing oI 2 and a grid size oI 20 to create the
head oI the tower that is parallel to the local coordinate system oI the base but shiIted 50 units in z-direction.
Figure 3.32-5 shows the new coordinate system relative to the part oI the base that you have already deIined. Use the
SHOW ALL VIEWS option to view the model Irom the Iour deIault angles. View 2 shows that the new local coordinate
system is shiIted 7 units in the negative y-direction.
END LIST (#)
RETURN
elems ADD
5 (pick node)
8 (pick node)
8 (pick node)
6 (pick node at upper left position)
MAIN
MESH GENERATION
SET
U DOMAIN
-20 20
U SPACING
2
1457
CHAPTER 3.32
Transmission Tower
Figure 3.32-5 Four Views of New Local Coordinate System
Use the technique oI generating one master element and subdividing it into six equally sized elements to generate the
leIt side oI the head. Keep the bias Iactor equal to zero but change the number oI subdivisions to 6 in the Iirst direction
oI the element.
V DOMAIN
-20 20
V SPACING
2
TRANSLATE
0 -7 50
VIEW
activate 1 (off)
activate 2 (off)
activate 4 (on)
PERSPECTIVE
ACTIVATE ALL
SHOW ALL VIEWS
FILL
show 2
RETURN
1458 Marc Users Guide: Part 2
CHAPTER 3.32
Since line elements are one-dimensional elements, it is not necessary to deIine the number oI elements in the second
and third direction. In general, the direction in which the subdivisions are made is dependent on the connectivity; that
is, the order in which nodes are entered to create an element.
Figure 3.32-6 Creating the Left Side of the Tower Head
The next step is to generate the sloping mid-section. Set the grid spacing to 1 and the grid size to 50 to deIine yet
another local coordinate system to connect the two coordinate systems that were already deIined.
Use ALIGN to line up the local u and v axes so that the plane spanned by these two axes contains the origins oI both
previous coordinate systems. This way you can use the grid to deIine the mid-section elements.
MAIN
MESH GENERATION
elems ADD (pick from the grid)
node(-2,0,0)
node(-2,18,0)
SUBDIVIDE
DIVISIONS
6 2 2
ELEMENTS
10 (the newly generated element)
END LIST (#)
1459
CHAPTER 3.32
Transmission Tower
Figure 3.32-7 Local u and v Axes Aligned
MAIN
MESH GENERATION
SET
ALIGN
0 9 10 (upper right of lower section)
-9 9 10 (upper left of lower section)
-2 2 50 (bottom node of top section)
U DOMAIN
-50 50
U SPACING
1
V DOMAIN
-50 50
V SPACING
1
VIEW
show 3
RETURN
1460 Marc Users Guide: Part 2
CHAPTER 3.32
The subdivisions along the mid-section are not equally spaced. This is where the bias factor can be used to
successIully generate a weighted subdivision. The Iollowing example illustrates the theory behind the bias Iactor.
Suppose you want to subdivide the line element in Figure 3.32-8 so that the length oI the elements are biased towards
the right. The desired number oI subdivisions is assumed to be Iour and a local coordinate is introduced, which
ranges Irom -1 at the Iirst node to 1 at the second node. An unbiased subdivision would produce 4 elements with their
respective nodes located at , , , and . A biased subdivision relocates
these nodes using the Iormula
where is the bias Iactor and is the biased local coordinate.
Using a bias Iactor will result in nodes at , , , and . The
unconditionally valid range oI the bias Iactor is .
Figure 3.32-8 Subdivision of Line Element (Biased to the Right)
The mid-section consists oI a biased and an unbiased part (LO-mid-section and HI-mid-section). Continue to generate
the tower creating two elements and subdividing them, once using a bias Iactor oI 0.2 and once using a bias Iactor oI
0.0.
MAIN
MESH GENERATION
VIEW
show 2
RETURN
nodes ADD
0 13 0 (pick grid point)
elems ADD
t
t 1 t 1 2 t 0 t 1/2 t 1
t
*
t b 1 t
2
( )
b t
*
b 0.5 t
*
1 t
*
1 8 t
*
1/2 t
*
7/8 t
*
1
0.5 b 0.5 s s
-1 +1
-1 0 +1
-1 +1
-1/8 +1/2 +7/8
-1/2 +1/2
t
t
t*
unbiased
subdivision
with bias 0.5
subdivision
original
element
1461
CHAPTER 3.32
Transmission Tower
Figure 3.32-9 Subdivided Elements in the Mid-Section
6 (pick upper left node of lower section)
12 (pick bottom node of top section)
SUBDIVIDE
DIVISIONS
2 1 1
BIAS
-0.2 0 0
ELEMENTS
17
END LIST (#)
DIVISIONS
5 1 1
BIAS
0.2 0 0
ELEMENTS
19
END LIST (#)
DIVISIONS
4 1 1
BIAS
0 0 0
ELEMENTS
18
END LIST (#)
1462 Marc Users Guide: Part 2
CHAPTER 3.32
As a next step, the lower-mid-section is completed and cross members are added in this part.
MAIN
MESH GENERATION
elems ADD
33 (pick middle node of mid-section)
19 (pick auxiliary node at u= 0)
19 (pick auxiliary node at u= 0)
29 (pick upper left node of base)
SUBDIVIDE
DIVISIONS
3 1 1
ELEMENTS
30
END LIST (#)
MAIN
MESH GENERATION
elems ADD
34 (pick auxiliary node at u= 0)
32 (almost horizontal to the left)
32 (work the diagonals)
35
35
31
31
36
36
30
PLOT
label ELEMENTS (off)
REGEN
RETURN
1463
CHAPTER 3.32
Transmission Tower
Figure 3.32-10 Completed Mid-Section
Step 2: Duplicate the Iirst Iace to create the remaining Iaces oI the main tower structure. It is crucial to sweep nodes
and elements aIter using symmetry and duplicate.
The stage is now set Ior a symmetry operation. Please note that this operation is always carried out in the local
coordinate system. The point with coordinates (0,0,0) and the normal vector (1,0,0), both in the U,V,W coordinate
system, deIine the symmetry plane. These are the deIault settings Ior Mentat.
MAIN
MESH GENERATION
SYMMETRY
ELEMENTS
all: EXIST.
1464 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-11 Cross Section of Tower Member
The object oI the symmetry operation is to establish the key points in space so that you can complete one Iace oI the
tower. Use the existing points and click on the nodes in succession to add the cross members. Note that the program
assumes you want to add new elements until you instruct it otherwise. Also, note that it is not necessary to activate the
elems ADD button every time you enter a new element.
MAIN
MESH GENERATION
elems ADD
(first generate the horizontal cross
member between nodes 33 and 64,
then work the diagonals
in the high-mid-section)
PLOT
label NODES (off)
REGEN
RETURN
1465
CHAPTER 3.32
Transmission Tower
Figure 3.32-12 Tower Face Completed
The next step is to duplicate the geometry oI one tower Iace three times to complete the structure. There are some
remarks to be made prior to executing this step:
PerIorm a sweep operation on all nodes in order to remove the duplicate nodes Irom the model.
The current local coordinate system is not ideal to use Ior speciIication oI the point about which to duplicate
nor Ior speciIication oI the rotation vector. Reset the coordinate system so that it lines up with the global
coordinate system: make sure the u-direction is parallel to the x-direction, the v-direction parallel to the
y-direction, and the w-direction parallel to the z-direction.
Now duplicate the Iace you just created. The rotation duplication vector should be 90 in the z-direction and the point
oI duplication 0, 0, 0. Figure 3.32-13 shows the results oI the duplication operation Irom Iour diIIerent viewpoints.
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
RETURN
SET
RESET
grid ON (to switch off grid)
RETURN
DUPLICATE
1466 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-13 Four Views of Tower Face Duplication
The various operations you have just executed may have leIt duplicate nodes, that is nodes with diIIerent identiIication
numbers but occupying the same space. In Iinite element terms, these nodes are not connected which may introduce
undesirable mechanisms in the structure.
Use the SWEEP processor to eliminate the duplicate nodes that occupy the same location into one node with a single
identiIication number. Since this involves a comparison oI real coordinates that cannot be done exactly in a computer,
nodes are swept together iI they are within a certain tolerance Irom each other. This tolerance can be changed Irom its
deIault value. Be careIul when adjusting the tolerance as too large a tolerance can collapse the entire structure into a
single point.
ROTATION ANGLES
0 0 90
REPETITIONS
3
ELEMENTS
all: EXIST.
VIEW
SHOW ALL VIEWS
FILL
RETURN
1467
CHAPTER 3.32
Transmission Tower
Now that you have generated the gross anatomy oI the tower structure, it is useIul to identiIy parts oI the structure by
name. The concept oI set naming is a very powerIul tool; it can be used in any place where a list is required and allows
you to manipulate a group oI entities rather than individual entities.
Figure 3.32-14 Boxes used for Element Set Selection
You have already deIined the base, mid-section, and head oI the tower. Use the Box Pick Method to Ience oII the
diIIerent portions oI the mesh. The STORE ELEMENTS command in the SELECT menu prompts Ior a set name Iirst,
Iollowed by a list oI elements. Position the cursor at one corner oI the portion oI elements to be Ienced oII. Depress
the leIt mouse button. Drag the cursor to the opposite corner oI the box and release.You have just selected every item
that is inside the box, indicated by a change in color. The extent oI the box are and - in the direction
perpendicular to the screen.
MAIN
MESH GENERATION
SWEEP
sweep NODES
all: EXIST.
sweep ELEMENTS
all: EXIST.
MAIN
MESH GENERATION
VIEW

1468 Marc Users Guide: Part 2
CHAPTER 3.32
show 2
RETURN
SELECT
elements STORE
base
(Box Pick the elements in the base of
the tower, see Figure 3.32-14)
END LIST (#)
MAIN
DEVICE
picking PARTIAL
SELECT
elements STORE
mid_sect_lo
(Box Pick the elements, realizing
that all elements partially within
the box will also be included)
END LIST (#)
RETURN
picking COMPLETE
SELECT
elements STORE
mid_sect_hi
(Box Pick the elements)
END LIST (#)
elements STORE
head
(Box Pick the elements)
END LIST (#)
MAIN
PLOT
MORE
IDENTIFY
SETS
1469
CHAPTER 3.32
Transmission Tower
Figure 3.32-15 Element Sets Created
The current structure still contains a number oI mechanisms that need to be eliminated by adding members. For
example, the legs oI the base contain mechanisms that are eliminated by adding cross members. In order to accomplish
this, limit the visible elements to the base. Proceed to the SELECT submenu and click SELECT SET, activate the
desired set and click MAKE VISIBLE. Only the elements contained in the set base` and their nodes remain visible. The
base only occupies a small portion oI the screen. Select view 4 Ior display and FILL the screen.
The current view point is perhaps not optimal. Nodes may overlap, making it impossible to pick nodes at some
locations. Use dynamic viewing to change the position oI the mesh so that you can add the additional cross-members.
Dynamic viewing can be switched on by clicking DYN. VIEW in the static menu. Next, position the cursor in the middle
oI the graphics area and hold down <MM>. Move the mouse to the leIt or right and see how the mesh rotates about the
screen axis.
As an alternative, in the button sequence the model is rotated around its z-axis using one oI the buttons in the
MANIPULATE MODEL menu.
Use the basic element ADD command to add the cross members. Repeat this Ior parts in the lower mid-section.
REGEN
NONE
REGEN
Note: Elements can only be added when dynamic viewing is oII. II you try to add an element with dynamic viewing on, the
result is a null operation.
1470 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-16 Elements to be added in the Base of the Tower
MAIN
MESH GENERATION
PLOT
nodes SETTINGS
LABELS (on)
REGEN
RETURN
SELECT
SELECT SET
base
OK
MAKE VISIBLE
RETURN
VIEW
show 4
FILL
activate 1 (off)
activate 2 (off)
activate 3 (off)
MANIPULATE MODEL
1471
CHAPTER 3.32
Transmission Tower
Figure 3.32-17 Elements to be added in the Lower Mid-Section
rotate in model space Z-
RETURN (twice)
elems ADD
(add elements as indicated in Figure 3.32-16)
SELECT
elements STORE
base
all: VISIBLE
SELECT SET
mid_sect_lo
OK
MAKE VISIBLE
RETURN
FILL
elems ADD
(add elements as indicated in Figure 3.32-17)
SELECT
elements STORE
mid_sect_lo
all: VISIBLE
1472 Marc Users Guide: Part 2
CHAPTER 3.32
Step 3: Create one cable arm by adding nodes and elements and using subdivide.
The next step is to generate the arms oI the tower Irom which the high voltage cables are suspended. Make the head
oI the tower visible so that you can easily pick nodes with the mouse. The position oI the node at the extreme end oI
the arm is known and added through the node ADD command. Use the mouse pick to create the elements between the
node at the extreme end and the diIIerent portions oI the head.
Figure 3.32-18 Elements to be Added
MAIN
MESH GENERATION
SELECT
SELECT SET
head
OK
MAKE VISIBLE
RETURN
FILL
ZOOM BOX
(zoom in on top section)
VIEW
MANIPULATE MODEL
rotate in model space Z- (click 7 times)
RETURN (twice)
1473
CHAPTER 3.32
Transmission Tower
Figure 3.32-19 Elements to be Added
nodes ADD
-8 0 63
elems ADD
(add 4 elements according to Figure 3.32-18)
MAIN
MESH GENERATION
SUBDIVIDE
DIVISIONS
2 1 1
ELEMENTS
(pick the 4 elements just created)
END LIST (#)
RETURN
elems ADD
(add 9 elements according to Figure 3.32-19)
VIEW
ACTIVATE ALL
FILL
show 2
RETURN
1474 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-20 Box Pick of Transmission Tower Arm
Store the elements oI the arm in an element set called arm Ior later reIerence.
Step 4: Use symmetry on the Iirst arm to create the second arm. Then use duplicate on the Iirst two to create the
remaining cable arms.
Use symmetry and duplicate to reproduce the arm on either side oI the tower head.
MAIN
DEVICE
picking PARTIAL
SELECT
elements STORE
arm
(Box Pick the elements in the arm)
END LIST (#)
RETURN
picking COMPLETE
MAIN
MESH GENERATION
DUPLICATE
RESET
1475
CHAPTER 3.32
Transmission Tower
You have now completed Steps 1 through 4 outlined in Overview of Steps. The goal oI these Iour steps was to show
you the importance oI deIining a local coordinate system and using it to add, duplicate, and symmetry parts oI the
tower structure.
ROTATIONS
0 0 180
ELEMENTS
arm
FILL
RESET
TRANSLATIONS
0 0 -6
REPETITIONS
2
ELEMENTS
arm
RETURN
SWEEP
sweep NODES
all: EXIST.
SELECT
ELEMENTS
all: EXIST.
MAKE VISIBLE
FILL
RETURN
PLOT
label NODES (off)
MORE
IDENTIFY
SETS
REGEN
NONE
1476 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-21 Completed Transmission Tower Model
This section outlined the advantage oI using the symmetry oI a structure in two-dimensional beIore creating a three-
dimensional structure using the duplication process.
You have been using the direct meshing technique described in Chapter 1: Introduction to model the transmission
tower. At this point, you have only completed the geometrical part oI the Iinite element mesh. The next step is to
speciIy the correct boundary conditions and loads.
Step 5: Add boundary conditions.
SpeciIy Iixed displacements at nodes to constrain the nodes at the tower base. Apply a gravity load to the elements.
Group the elements in sets so that you can easily reIer to them later. Add distributed wind loads. Apply cable loads as
point loads on the nodes at the ends oI the cable arms.
There are basically two types oI boundary conditions Ior this model:
In Chapter 1: Introduction discussed how the application oI boundary conditions is equal to the process oI Iinding the
answer to the question: 'Apply what, where, and when
First concentrate on what.
The BOUNDARY CONDITIONS button in the main menu reveals a submenu that allows you to speciIy mechanical
boundary conditions.
1. Kinematic: The base oI the structure is attached to the ground.
2. Load: Gravity
Point loads are applied to the cable arms to simulate the weight oI the cables.
Wind loads are applied as distributed loads perpendicular to the tower.
1477
CHAPTER 3.32
Transmission Tower
Kinematic Boundary Conditions
Click on FIXED DISPLACEMENT. A pop-up appears over the graphics area. Constrain the Iirst three degrees oI
Ireedom by clicking the ON button.
Note that while the pop-up is activated, your view oI the graphics area is obstructed and all other buttons oI the regular
menu are inactive. You must conIirm the values entered in the pop-up by clicking on the OK button beIore you can
access the regular menu again.
Now that you have answered what, you can concentrate on where.
II you limit your view to the base, you can easily pick on the nodes that attach the structure to the ground. This
operation completes the where and ties it to the what portion oI the equation. Since the problem is time independent,
the equation is complete because there is no need to answer when. The application oI what is conIirmed by the display
oI arrows in the direction it was applied.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NAME
bolts
SELECT
SELECT SET
base
OK
MAKE VISIBLE
FILL
RETURN
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
DISPLACEMENT Z (on)
OK
nodes ADD
(pick 4 nodes at base)
END LIST (#)
1478 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-22 Boundary Conditions
For Iuture reIerence, it is useIul to group elements and store them in a set that can be reIerenced later by name. We
already mentioned that not all members are oI the same geometry. Use the SELECT option to group all elements that
are L3x3 angles and store in a set called L3x3. All other members are L2x2 angles. Generate a list oI these elements
by inverting the previous list and storing them in a set called L2x2.
Figure 3.32-23 Elements to be Contained in Set L3x3
1479
CHAPTER 3.32
Transmission Tower
MAIN
MESH GENERATION
SELECT
elements STORE
L3x3 (pick elements from base)
END LIST (#)
SELECT SET
mid_sect_lo
OK
MAKE VISIBLE
FILL
elements STORE
L3x3 (pick elements from mid_sect_lo)
END LIST (#)
MAIN
MESH GENERATION
SELECT SET
mid_sect_hi
OK
MAKE VISIBLE
FILL
elements STORE
L3x3 (pick elements from mid_sect_hi)
END LIST (#)
SELECT SET
head
OK
MAKE VISIBLE
FILL
elements STORE
L3x3 (pick elements from head)
END LIST (#)
SELECT SET
arm
OK
1480 Marc Users Guide: Part 2
CHAPTER 3.32
MAKE VISIBLE
FILL
elements STORE
L3x3 (pick elements from arm)
END LIST (#)
ELEMENTS
all: EXIST.
MAKE VISIBLE
FILL
ELEMENTS
all: EXIST.
select mode AND (to switch to EXCEPT)
L3x3
elements STORE
L2x2
all: SELECT.
MAIN
MESH GENERATION
SELECT
CLEAR SELECT
RESET
PLOT
MORE
IDENTIFY
SETS
REGEN
NONE
REGEN
1481
CHAPTER 3.32
Transmission Tower
Figure 3.32-24 Identify Sets
Load (Gravity)
Similar to the FIXED DISPLACEMENT button, when you activate the GRAVITY LOAD button, a pop-up appears over
the graphics area where you can enter appropriate values oI the gravity load. The program expects the magnitude oI
the gravity acceleration in the negative z-direction here. This answers the what portion oI the equation. Use the all:
EXIST. button to answer the where part oI the equation.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
RESET VIEW
FILL
NEW
NAME
gravity
GRAVITY LOAD
ACCELERATION Z
-32.2
OK
elements ADD
all: EXIST.
1482 Marc Users Guide: Part 2
CHAPTER 3.32
Wind Load
Assume that the transmission tower is subjected to a wind load with a stronger load applied to the upper part oI the
tower and a weaker load to the lower part oI the tower. Simulate the wind loads by applying a distributed load in
y-direction. Assume that only one Iace oI the tower is loaded by this wind load.
Prior to applying the loads, element sets oI all elements in the lower and upper Irontal Iace oI the tower are generated.
The upper portion is stored in an element set called hi_front while the lower portion is stored in a set called lo_front.
Figure 3.32-25 Gravity Load for the Structure
1483
CHAPTER 3.32
Transmission Tower
Figure 3.32-26 Box Pick from base for lo_front
Figure 3.32-27 Polygon Pick from mid_sect_lo for lo_front
1484 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-28 Polygon Pick from mid_sect_hi for hi_front
Figure 3.32-29 Box Pick from head for hi_front
1485
CHAPTER 3.32
Transmission Tower
Figure 3.32-30 Box Pick from arm for hi_front
MAIN
BOUNDARY CONDITIONS
MECHANICAL
SELECT
VIEW
show 1
RETURN
SELECT SET
base
OK
MAKE VISIBLE
elements STORE
lo_front (select according to Figure 3.32-26)
END LIST (#)
SELECT SET
mid_sect_lo
OK
MAKE VISIBLE
elements STORE
lo_front (select according to Figure 3.32-27)
1486 Marc Users Guide: Part 2
CHAPTER 3.32
Complete the set hi_front by processing the sets head` and arm` in an identical way.
Now actually apply the loads:
END LIST (#)
SELECT SET
mid_sect_hi
OK
MAKE VISIBLE
elements STORE
hi_front (select according to Figure 3.32-28)
END LIST (#)
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
NAME
hi_wind
GLOBAL LOAD
FORCE Y
-120
OK
elements ADD
hi_front
1487
CHAPTER 3.32
Transmission Tower
Figure 3.32-31 Strong Wind Load
In an identical way, a Y FORCE oI -80 can be applied to all elements contained in lo_front.
Figure 3.32-32 Weak Wind Load
1488 Marc Users Guide: Part 2
CHAPTER 3.32
Point Loads
The cables are suspended Irom the arms oI the tower and are simulated as point loads hanging Irom each tip oI the
arm. A load oI -500 in this direction is applied to each oI the six arm extremities. The boundary conditions menu allows
you to enter these point loads through the POINT LOADS option. The already Iamiliar pop-up appears over the
graphics area. Enter the values in the appropriate Iields.
Use the mouse to pick the nodes that are to receive a load. Enter the end oI list character (#) aIter you have picked the
six nodes.
Figure 3.32-33 Cable Loads
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
NAME
cable_load
POINT LOAD
FORCE Z
-500
OK
nodes ADD (pick the six nodes on the tip of the arms)
END LIST (#)
1489
CHAPTER 3.32
Transmission Tower
Step 6: DeIine the material and apply it to all elements. DeIine the geometric properties and apply them to the
appropriate elements.
Geometric Properties
The Iinite element analysis program requires you to speciIy the properties such as the material type and area oI the
members. Use the GEOMETRIC PROPERTIES processor to enter the area. So Iar, you have already separated the
L3x3 angles Irom the L2x2 angles and can reIer to them by set name. The L3x3 angles have an area oI 0.01 while the
L2x2 angles have an area oI 0.00651. The moments oI inertia about the x and y axis are equal and must be entered
here. The data is listed under Full Disclosure in this chapter. However, beam elements require some additional
geometric data deIining the direction oI the local x and y axis. This data can also be entered through the GEOMETRIC
PROPERTIES processor.
The local x and y axis oI all elements not exactly pointing in z-direction can be deIined using a reIerence vector
(0, 0, 1). For all elements pointing in z-direction we use a reIerence vector (1, 0, 0), although the reality may be more
complex. This requires the deIinition oI a set oI elements called 'upright that contains all elements pointing in
z-direction. As demonstrated previously, the user can make the sets base, mid-sect-lo, mid-sect-hi, head and arms
visible one aIter the other, select the upright elements and add those to the stored list. There is a total oI 44 such
elements to be Iound. (8 in base, 24 in head and 12 in arm).
MAIN
MESH GENERATION
PLOT
draw NODES (off)
RETURN
SELECT
SELECT SET
base
MAKE VISIBLE
FILL
elements STORE
upright (pick the 8 elements pointing in z-direction)
END LIST (#)
SELECT SET
head
(etc., etc.)
1490 Marc Users Guide: Part 2
CHAPTER 3.32
AIter having selected all 44 elements, make all elements visible again.
MAIN
GEOMETRIC PROPERTIES
3-D
NEW
NAME
L3x3_z
ELASTIC BEAM
AREA
0.01
6.0e-05 (for Ixx)
6.0e-05 (for Iyy)
0 (for direction)
0 (for direction)
1 (for direction)
OK
SELECT
SELECT SET
L3x3
OK
select mode AND (to switch to EXCEPT)
SELECT SET
upright
OK
RETURN
elements ADD
all: SELECT.
ID GEOMETRIES (on)
MAIN
GEOMETRIC PROPERTIES
3-D
NEW
NAME
L3x3_x
1491
CHAPTER 3.32
Transmission Tower
ELASTIC BEAM
AREA
0.01
6.0e-05 (for Ixx)
6.0e-05 (for Iyy)
1 (for direction)
0 (for direction)
0 (for direction)
OK
SELECT
CLEAR SELECT
RESET
SELECT SET
upright
OK
select mode AND (to switch to EXCEPT)
SELECT SET
L2x2
OK
RETURN
elements ADD
all: SELECT.
1492 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-34 Geometric Properties Assignment for L3x3 Angles
MAIN
GEOMETRIC PROPERTIES
3-D
NEW
NAME
L2x2_z
ELASTIC BEAM
AREA
0.00651
2.0e-05 (for Ixx)
2.0e-05 (for Iyy)
0 (for direction)
0 (for direction)
1 (for direction)
OK
SELECT
CLEAR SELECT
RESET
SELECT SET
L2x2
OK
1493
CHAPTER 3.32
Transmission Tower
select mode AND (to switch to EXCEPT)
SELECT SET
upright
OK
RETURN
elements ADD
all: SELECT.
MAIN
GEOMETRIC PROPERTIES
3-D
NEW
NAME
L2x2_x
ELASTIC BEAM
AREA
0.00651
2.0e-05 (for Ixx)
2.0e-05 (for Iyy)
1 (for direction)
0 (for direction)
0 (for direction)
OK
SELECT
CLEAR SELECT
RESET
SELECT SET
upright
OK
select mode AND (to switch to EXCEPT)
SELECT SET
L3x3
OK
RETURN
elements ADD
all: SELECT.
1494 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-35 Geometric Properties Assignment for L2x2 Angles
Material Properties
The last step is to assign material properties. The members oI this tower are made out oI steel. For this analysis, you
need to speciIy the Young`s Modulus and Poisson`s Ratio and mass density, all located in the MATERIAL
PROPERTIES menu. Assign this material to all existing elements.
SELECT
CLEAR SELECT
RESET
MAIN
MATERIAL PROPERTIES
NEW
NAME
steel
ISOTROPIC
YOUNGS MODULUS
4.176e9
0.3 (for Poissons Ratio)
15.217 (for mass density)
OK
1495
CHAPTER 3.32
Transmission Tower
Figure 3.32-36 Material Properties Assignment
Step 7: Job submission oI the static analysis.
Job Submission of the Static Analysis
Finally, you submit the job. This is easily done in the JOBS menu. SUBMIT submits the job in the background. The
status oI the job can be checked or monitored continuously. Once you have successIully submitted the job, you must
careIully analyze the results.
elements ADD
all: EXIST.
ID MATERIALS (on)
ID MATERIALS (off)
MAIN
JOBS
NEW
NAME
static
MECHANICAL
INITIAL LOADS (default: all loads selected)
OK (twice)
ELEMENT TYPES
1496 Marc Users Guide: Part 2
CHAPTER 3.32
Step 8: Static analysis results processing.
Static Analysis Results Processing
The static analysis considers the wind load, gravitational load, and point loads. The structure undergoes a bending out
oI the x-z plane as a result oI the wind load. Figure 3.32-37 shows the results. We switched on the AUTOMATIC
DEFORMATION SCALING option resulting in an exaggerated display oI the displacements.
mechanical elements 3-D TRUSS/BEAM
52 (LINE(2), THIN ELASTIC BEAM)
OK
all: EXIST.
RETURN (twice)
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
VIEW
show 3
RETURN
DEF & ORIG
NEXT INC
deformed shape SETTINGS
AUTOMATIC
FILL
1497
CHAPTER 3.32
Transmission Tower
Figure 3.32-37 Deformation of Tower under Combined Load
Step 9: Job submission oI the modal analysis.
Job Submission of the Modal Analysis
To run the modal analysis, it is necessary to restore the model Iile. AIter a new dynamic loadcase has been deIined, a
new job is created and submitted.
MAIN
RESULTS
CLOSE
deformed shape OFF
RETURN
FILES
RESTORE
RETURN
LOADCASE
NEW
NAME
dynamic
DYNAMIC MODAL
# MODES
15
1498 Marc Users Guide: Part 2
CHAPTER 3.32
Step 10: Modal analysis results processing.
Modal Analysis Results Processing
Open the results Iile by clicking the RESULTS button Irom the main menu, Iollowed by the OPEN DEFAULT button.
The modal shapes are stored in subincrements and are accessed through the NEXT INC button. As is demonstrated in
previous chapters, it is useIul to animate the diIIerent modal shapes. Figure 3.32-38 and Figure 3.32-39 display
examples oI mode shapes Iound during this analysis.
The postprocessing is carried out as Iollows:
OK
RETURN
JOBS
NEW
NAME
dynamic
MECHANICAL
loadcases SELECT
dynamic
OK
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
VIEW
show 4
RETURN
DEF & ORIG
NEXT (twice)
deformed shape SETTINGS
AUTOMATIC
FILL
RETURN
DEF ONLY
1499
CHAPTER 3.32
Transmission Tower
Figure 3.32-38 Eigenmode of Tower, f = 5.366 Hz
MORE
animate MODE
15
ANIMATION
FILL
REPEAT
PLAY
STOP
SHOW MODEL
RETURN
PREVIOUS
SKIP TO INC
0:15
DEF & ORIG
FILL
1500 Marc Users Guide: Part 2
CHAPTER 3.32
Figure 3.32-39 Eigenmode of Tower, f = 20.64 Hz
Conclusion
This structure is an example where automatic mesh generators cannot be utilized to create a Iinite element model. It
is demonstrated in this chapter that by using the 'conventional' tools available in Mentat, a Iairly complicated mesh
can be generated without any diIIiculty.
The displacements, as a result oI a load in negative y direction shown in Figure 3.32-37 are as expected. The results
oI the modal analysis can be Iully appreciated by animation oI the diIIerent modes.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
transmission_tower.proc Mentat procedure Iile
Chapter 3.33: Bracket
3.33
Bracket

Chapter Overview 1502

Background Information 1502

Detailed Session Description of the Linear Static Case 1504

Conclusion 1521

Dynamic Modal Shape Analysis 1522

Detailed Session Description of the Modal Shape Analysis 1522

Dynamic Transient Analysis 1526

Detailed Session Description of Dynamic Transient Analysis

1526

Conclusion 1533

Pressure Table 1533

Input Files 1534

1502 Marc Users Guide: Part 2
CHAPTER 3.33
Chapter Overview
The sample session described in this chapter demonstrates a simple linear static and dynamic analysis on a steel
bracket. The bracket restrains a vertical pipe. The bracket also supports some mechanical equipment. First, the bracket
will be subjected to a static load. A dynamic analysis will predict the normal Irequencies and mode shapes oI vibration
to determine iI there is any interaction with the bracket and surrounding excitation Irequencies. Finally, the bracket
will be subjected to a time dependent pressure and the dynamic response will be determined.
Background Information
Description
This problem demonstrates the preparation oI a model using two diIIerent meshing techniques, multiple geometric
properties, three loadcase types, and corresponding boundary conditions and loads. It will also demonstrate the
application oI boundary conditions to geometric entities and the merging oI diIIerent meshes.
The bracket is 15x30x10 with a hole to support a pipe. The bracket must support standard operating loads. It must not
have a Irequency that can be excited by the mechanical equipment which it supports. It must not Iail Ior a given time
dependent pressure loading.
Idealization
The bracket will be represented by Ilat plate elements which have 4 nodes and the thickness is considered a property
oI the element. The vertical support plates will require quadrilateral elements to be degenerated to triangular elements
in the portion where the arc is tangent to the horizontal plate.
The bracket is welded to a column and thereIore will be considered Iully Iixed on that edge. The weight oI the
mechanical equipment will be applied to the cantilevered section oI the horizontal plate as a distributed load oI 1 psi.
Requirements for a Successful Analysis
The analysis will be considered successIul iI none oI the stresses are above 36000 psi during standard operating loads
and there are no modes in the range oI the mechanical equipment. The bracket cannot cause the pipe to break during
a dynamic loading event.
1503
CHAPTER 3.33
Bracket
Figure 3.33-1 Bracket
Full Disclosure
The steel bracket is modeled by Iour-noded plate elements with a Young`s Modulus oI 30e6 psi and a Poisson`s Ratio
oI 0.3. It is assumed that the material will not exceed the yield point oI 36000 psi. The horizontal plate is 15 inches by
30 inches with a hole oI 3.5 inch radius centered in the halI oI the plate where the vertical support plates are attached.
The 2 vertical support plates are 10 inches by 15 inches with a Iilleted edge. The horizontal plate is 0.25 inches thick
and the vertical plates are 0.5 inches thick.
Figure 3.33-2 Dimensions and Loads for the Bracket
X Y
Z
4
X Y
Z
Outside Half of Plate Loaded 1 psi
Length 30 Inch
Width 15 Inch
Height 10 Inch
Arc Center (22.5,0,-27.5)
Hole Center (7.5,7.5)
Radius 3.5 Inch
4
1504 Marc Users Guide: Part 2
CHAPTER 3.33
Overview of Steps
Step 1: Create the boundary of a flat area representing the half of the plate with the hole in it. Use the
overlay mesh generator to create finite elements.
Step 2: Create the cantilevered section of the plate. Convert it to finite elements. Merge the two parts.
Step 3: Fold the vertical sections and modify the elements in the triangular region.
Step 4: Apply boundary conditions.
Step 5: Assign material and geometric properties.
Step 6: Create the loadcases and submit the jobs.
Step 7: Postprocess the results.
Detailed Session Description of the Linear Static Case
Step 1: Create the boundary oI a Ilat area representing the halI oI the plate with the hole in it. Use the overlay mesh
generator to create Iinite elements.
The approach used in this session to generate the model is to use the geometric meshing technique to create two
diIIerent areas and mesh them. The Iirst area is meshed using the overlay mesh generator, and the second is meshed
using the CONVERT processor. The entire model is created as a Ilat piece and, subsequently, the two support pieces
are Iolded.
As in the Sample Session described in Following a Sample Session, the Iirst step in building the mesh is to establish
an input grid. Click on the MESH GENERATION button oI the main menu. Next click on the SET button to access the
coordinate system menu where the grid settings are located. Use the Iollowing button sequence to set the grid spacing
to 5 inches and the grid size oI 30 inches.
MAIN
MESH GENERATION
SET
U DOMAIN
-30 30
U SPACING
5
V DOMAIN
-30 30
V SPACING
5
grid ON (on)
RETURN
1505
CHAPTER 3.33
Bracket
The next step is to create the three vertical lines oI the model. The Iollowing button sequence creates the lines.
Figure 3.33-3 Grid and Straight Line Segments
The next geometric entity to be added is the two Iillets. The curve type must be changed to arc and then the two curves
added. To insure that the arc end points are the end points oI the line the CENTER/POINT/POINT arc type is used. The
Iollowing button sequence adds the two arcs.
MAIN
MESH GENERATION
FILL
crvs ADD (pick points from grid)
point(0,-10,0)
point(0,25,0)
point(15,0,0)
point(15,15,0)
point(30,0,0)
point(30,15,0)
MAIN
MESH GENERATION
CURVE TYPE
CENTER/POINT/POINT
1506 Marc Users Guide: Part 2
CHAPTER 3.33
Figure 3.33-4 Line Segments and Fillets
The next step is the center hole. The coordinate system is moved such that it has an origin that is the center oI the hole.
The hole is added using the grid.
RETURN
crvs ADD
22.5 27.5 0 (center point)
15 0 0 (pick lower end point of the second line)
0 -10 0 (pick lower end point of the first line)
22.5 42.5 0 (center point)
0 25 0 (pick upper end point of the first line)
15 15 0 (pick upper end point of the second line)
MAIN
MESH GENERATION
SET
set origin XYZ
7.5 7.5 0
U DOMAIN
-10 10
U SPACING
1507
CHAPTER 3.33
Bracket
Figure 3.33-5 Generation of the Circular Hole
0.5 0.5
V DOMAIN
-10 10
V SPACING
0.5 0.5
RETURN
ZOOM BOX
(zoom in on the center of the grid)
CURVE TYPE
CENTER/POINT
RETURN
crvs ADD
0 0 0 (pick the center point)
3.5 0 0 (pick a point on the circle)
FILL
1508 Marc Users Guide: Part 2
CHAPTER 3.33
The next step is to create Iinite elements Irom the geometric entities. This is done using the overlay mesh generator.
The Iollowing button sequence generates the mesh.
Figure 3.33-6 The Closed Contour for the OVERLAY Command
MAIN
MESH GENERATION
GRID (off)
AUTOMESH
2-D PLANAR MESHING
quadrilaterals (overlay) DIVISIONS
8 20
quadrilaterals (overlay) QUAD MESH!
1 2 4 5 6 (use the Box Pick Method)
END LIST (#)
RETURN (twice)
1509
CHAPTER 3.33
Bracket
Figure 3.33-7 The Automeshed Part
Step 2: Create the cantilevered section oI the plate. Convert it to Iinite elements. Merge the two parts.
The next step is to mesh the cantilevered portion. This section is modeled as Iour point quadrilateral surIace and then
converted to a 6x6 Iinite element mesh. The Iollowing button sequence creates the mesh.
MAIN
MESH GENERATION
srfs ADD (pick the points in counter-clockwise order)
3
5
6
4
CONVERT
DIVISIONS
8 8
SURFACES TO ELEMENTS
1 (pick surface)
END LIST (#)
RETURN
1510 Marc Users Guide: Part 2
CHAPTER 3.33
The overlay mesh generator may create some unused nodes which must be removed. Furthermore, the nodes on the
interIace oI the two meshes are not coincident. ThereIore, to merge them, the sweep tolerance should be large,
approximately 0.5. and only the nodes along the interIace selected. The Iollowing button sequence merges the nodes.
The sweep tolerance should be changed back to the deIault when the merge operation is Iinished.
Figure 3.33-8 Elements on the Square Surface
MAIN
MESH GENERATION
SWEEP
REMOVE UNUSED NODES
TOLERANCE
0.5
SWEEP NODES
(Box Pick the nodes on the interface)
END LIST (#)
TOLERANCE
0.0001
RETURN
1511
CHAPTER 3.33
Bracket
Figure 3.33-9 Correctly Connected Mesh
Step 3: Fold the vertical sections and modiIy the elements in the triangular region.
The Iinal mesh general operation is to Iold the two sides down. First, the nodes must be along the line oI the Iold. This
is done by creating a line along each edge and attaching the nodes to these lines. Then, the two corner elements are
divided into two triangular elements. These elements must still have the class oI QUAD(4). This is achieved by
generating the triangular elements as degenerated quad elements, double clicking one node in the connectivity list.
The lines are created by using the grid with a spacing oI 5 and a size oI 15. The origin oI the grid must be set to the
global origin. The Iollowing button sequence creates the lines and attachs the nodes to them.
MAIN
MESH GENERATION
SET
RESET
U SPACING
5
V SPACING
5
U DOMAIN
0 15
V DOMAIN
0 15
grid ON (on)
1512 Marc Users Guide: Part 2
CHAPTER 3.33
Figure 3.33-10 Nodes at the Top Half attached to the Line
RETURN
CURVE TYPE
LINE
RETURN
crvs ADD
point(0, 15, 0) (pick grid point)
12 (pick upper left point of the surface)
point(0, 0, 0) (pick grid point)
7 (pick lower left point of the surface)
GRID (off)
MOVE
MOVE TO GEOMETRIC ENTITIES
move nodes CURVE
7 (pick lower line)
(pick nodes near lower line)
END LIST (#)
8 (pick upper line)
(pick nodes near upper line)
END LIST (#)
1513
CHAPTER 3.33
Bracket
The two corner elements at the transition oI the Iillet to the square plate must be removed and two triangular elements
will replace them. To create the triangular elements, the last node should be selected twice. The triangular elements
have to be deIined such that they allow Ior Iolding over the line segment. The Iollowing button sequence creates the
Iirst oI Iour triangular elements.
Add three more triangular elements in the same way.
Figure 3.33-11 Corner Elements replaced by Triangular Elements
MAIN
MESH GENERATION
elems REM
89 59 (pick elements at the triangular corners)
END LIST (#)
elems ADD (pick nodes)
123
64
109 (first click on this node)
109 (second click on this node)
1514 Marc Users Guide: Part 2
CHAPTER 3.33
The next step in the process oI Iolding is to detach the elements. This is done using the ATTACH processor to detach
the elements in the triangular region. The lower edge rotates 90 and the upper edge rotates -90. The Iollowing button
sequence Iolds the two edges.
The Iinal step in the process oI Iolding is to actually move the elements. This is done using the MOVE processor and
rotating the elements. The lower edge rotates 90 and the upper edge rotates -90. The Iollowing button sequence Iolds
the two edges.
MAIN
MESH GENERATION
ATTACH
DETACH
ELEMENTS
1 2 3 4 5 6 7 47 48 49
50 51 52 53 54 55 56 58 170
40 41 42 43 44 45 46 90 93 94 95
96 97 98 99 100 101 102 167
END LIST (#)
MAIN
MESH GENERATION
MOVE
ROTATIONS
90 0 0
POINT
15 0 0 (pick point at end of bottom line
of horizontal plate)
ELEMENTS
(Box Pick the lower elements to be folded)
END LIST (#)
ROTATIONS
-90 0 0
POINT
15 15 0 (pick point at end of top line
of horizontal plate)
ELEMENTS
(Box Pick the upper elements to be folded
END LIST (#)
1515
CHAPTER 3.33
Bracket
The Iollowing button sequence shows all Iour views and turns oII the points and curves. It makes the viewing easier.
Figure 3.33-12 The Complete FE Model
MAIN
MESH GENERATION
SWEEP
remove unused NODES
MAIN
VIEW
SHOW ALL VIEWS
PLOT
draw POINTS (off)
draw CURVES (off)
REGEN
FILL
1516 Marc Users Guide: Part 2
CHAPTER 3.33
Step 4: Apply boundary conditions.
The next step is to apply the boundary conditions. First, the back edge oI the bracket is Iixed in the three translational
degrees oI Ireedom. The Iollowing button sequence Iixes the edge.
The next step is to apply the Iace loads to the cantilevered portion oI the plate. The loads are 1 psi downward to
represent the mechanical equipment. The Iollowing button sequence applies the distributed loads.
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
DISPLACEMENT Z (on)
OK
nodes ADD
(Box Pick the left edge of the plate, preferably in view 1 or 2)
END LIST (#)
NEW
FACE LOAD
PRESSURE
1
OK
surfaces ADD
1 (pick the surface)
END LIST (#)
1517
CHAPTER 3.33
Bracket
Figure 3.33-13 Loads Applied to all Elements attached to the Surface
Step 5: Assign material and geometric properties.
The next step is to assign the material properties. The material is steel and the mass density must be included because
a dynamic analysis will be done. The Iollowing button sequence assigns the material properties.
The next step is to assign the thickness oI the plates. Because the plates have diIIerent thicknesses, two geometric
properties are required. To make the picking oI the elements easier, the user should select the partial picking capability
in the device menu. The deIault is Iull which requires that the item oI the requested type be Iully contained in the
graphical pick. The partial mode selects any item oI the requested type that is partially in the graphical pick. The
Iollowing button sequence changes the picking type and then assigns the geometric properties to the elements.
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
30e6
0.3 (Poisson's ratio)
MASS DENSITY
0.283/386.4 (mass density)
OK
elements ADD
all: EXIST
1518 Marc Users Guide: Part 2
CHAPTER 3.33
To veriIy the geometric property assignment, the Iollowing button sequence changes some plot deIaults and switches
on the identiIication oI geometries.
MAIN
DEVICE
picking PARTIAL
RETURN
GEOMETRIC PROPERTIES
3-D
SHELL
THICKNESS
0.25
OK
elements ADD
(pick the Horizontal Plate Elements)
END LIST (#)
NEW
SHELL
THICKNESS
0.5
OK
elements ADD
(pick the Vertical Plate Elements)
END LIST (#)
MAIN
GEOMETRIC PROPERTIES
3-D
ID GEOMETRIES (on)
PLOT
elements SOLID
REGEN
RETURN
ID GEOMETRIES (off)
1519
CHAPTER 3.33
Bracket
Figure 3.33-14 Graphical Confirmation of Applied Geometries
Step 6: Create the loadcases and submit the jobs.
The next step is to create the static loadcase. It is a linear static with all loads selected and, thereIore, no special options
need to be selected. The Iollowing button sequence creates the loadcase.
The Iollowing button sequence creates the job and executes it.
MAIN
LOADCASE
mechanical analyses STATIC
OK
MAIN
JOBS
MECHANICAL
available (select a loadcase from the list)
lcase1
JOB RESULTS
scalars (select Equivalent von Mises stress for
outer and midplane layers)
vonmises
1520 Marc Users Guide: Part 2
CHAPTER 3.33
Step 7: Postprocess the results.
The Iinal step in any analysis is to postprocess the results. This is done by opening the post Iile and reviewing the
results. The deIormations are drawn using automatic scaling. The Iollowing button sequence does this. However, there
may be other results that the user wishes to review.
OUT
tensors
stress (select stress tensor for outer and midplane layers)
OUT
OK
JOB PARAMETERS
# SHELL/BEAM LAYERS
3 (use 3 layers for thickness integration)
OK (twice)
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
NEXT INC
NEXT INC
SCALAR
Equivalent von Mises Stress Layer 2
OK
CONTOUR BANDS
deformed shape SETTINGS
deformation scaling AUTOMATIC
Note: The results Ior increment 0 and increment 1 are identical. This is caused by the Iact that the pressure loading is still
selected as an initial load. The Marc writer generates Ior increment 1 (i.e. the complete loadcase) the incremental load
between the load vector and the initial load, which is a zero load increment. As a result, the deIormation in increment 0
and increment 1 is identical.
1521
CHAPTER 3.33
Bracket
Figure 3.33-15 Deformed Structure and Contours of von Mises Stress
Conclusion
The bracket sustains the required static loads.
1522 Marc Users Guide: Part 2
CHAPTER 3.33
Dynamic Modal Shape Analysis
Overview of Steps
Step 1: Restore the database from the static analysis.
Step 2: Create a modal dynamic loadcase and submit it.
Step 3: Postprocess the results.
Detailed Session Description of the Modal Shape Analysis
Step 1: Restore the database Irom the static analysis.
For the dynamic analysis, the same geometry is used. The Iirst step is to restore the database. Closing the post Iile
automatically restores the database. In addition, the commands are shown Ior restoring the database and resetting
the program. (These commands are not necessary here). Finally, the plotting oI points, curves, and surIaces are
switched oII.
Step 2: Create a modal dynamic loadcase and submit it.
The next step is to create a modal dynamic loadcase. The deIault is that 10 modes are determined which is enough Ior
this structure. (Note that the determination oI higher-order modes in general required higher mesh densities). The
Iollowing button sequence creates the loadcase.
MAIN
RESULTS
CLOSE
FILES
RESTORE
RESET PROGRAM
RETURN
VIEW
show 4
PLOT
draw POINTS (off)
draw CURVES (off)
draw SURFACES (off)
REGEN
FILL
RETURN
1523
CHAPTER 3.33
Bracket
The next step is to create and execute the modal analysis job. The Iollowing button sequence creates and submits
the job.
Step 3: Postprocess the results.
The next step is to postprocess the results. For modal analyses, not only the values oI the eigenIrequencies but also
the shape oI the deIlections or modal shapes are oI interest. For the deIormed shape, the automatically scaled
deIormations should be viewed. For ease oI understanding, it is best to show all Iour views. The Iollowing button
sequence does the postprocessing.
MAIN
LOADCASE
NEW
DYNAMIC MODAL
OK
MAIN
JOBS
NEW
MECHANICAL
available
lcase2
ANALYSIS OPTIONS
(verify that the LANCZOS method is used)
OK (twice)
RUN
SAVE
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
PLOT
draw NODES (off)
MORE
edges OUTLINE
1524 Marc Users Guide: Part 2
CHAPTER 3.33
Finally, generate an animation sequence oI one modal shape:
RETURN (twice)
SCALAR
Displacement z
OK
CONTOUR BANDS
deformed shape SETTINGS
deformation scaling AUTOMATIC
RETURN
NEXT INC (repeat until all modes have been viewed)
MAIN
RESULTS
MORE
animate MODE
9 (number of frames for animation)
ANIMATION
VIEW
show view 4
FILL
RETURN
PLAY
SHOW MODEL
1525
CHAPTER 3.33
Bracket
Figure 3.33-16 The Second Eigenmode
Observe that the calculated eigenIrequencies and corresponding eigenvectors are stored as so-called subincrements on
the post Iile. The eigenIrequency value (in cycles/time unit) corresponding to a speciIic mode is printed on the top leIt
oI the screen.
Observe that the eigen period oI mode 1 will be 0.036 seconds.
Table 2.33-1 Eigenfrequencies Bracket
increment mode
value
(cycles/time)
0:1 1 27.5
0:2 2 85.3
0:3 3 150.7
0:4 4 240.8
0:5 5 298.6
0:6 6 304.8
0:7 7 413.0
0:8 8 456.9
0:9 9 505.9
0:10 10 569.3
1526 Marc Users Guide: Part 2
CHAPTER 3.33
Dynamic Transient Analysis
Overview of Steps
Step 1: Create an pressure time table for the loading. Then apply it as a face load in a loadcase.
Step 2: Create and submit a transient analysis.
Step 3: Create job and submit it.
Step 4: Postprocess the results.
Detailed Session Description of Dynamic Transient Analysis
Step 1: Create an pressure time table Ior the loading. Then apply it as a Iace load in a loadcase.
Restore the database to continue with the analysis preparation. The next step is to include a time dependent pressure
history using the table option in the boundary condition menu. Here, the table which is included in the Iile bracket.tbl
on the Mentat installation media is used. To load the table, the user needs the Iull path name to the Mentat subdirectory
examples/marc_ug or has to copy this Iile into his own directory. The Iollowing button sequences restores the model
and input the table.
MAIN
RESULTS
CLOSE
FILES
RESTORE
RESET PROGRAM
RETURN
PLOT
draw CURVES (off)
draw POINTS (off)
draw SURFACES (off)
FILL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
TABLE
1527
CHAPTER 3.33
Bracket
The pressure history as Iunction oI the time is shown in Figure 3.33-17.
Figure 3.33-17 Transient Pressure Loading as Function of Time
Observe that the maximum pressure magnitude is one which is equal to the pressure as deIined in the static loadcase.
The maximum pressure is reached within 0.01 seconds and the pressure is equal to zero at 0.04 seconds and kept to
zero until 0.09 seconds. Looking at the results oI the eigenIrequencies in Table 2.33-1, it is expected that the lower
modes are dominating during the transient.
The next step is to assign this table to the pressure to the same elements as in the pressure load used in loadcase 1.
READ
bracket.tbl
FIT
SHOW TABLE
SHOW MODEL (select SHOW MODEL)
RETURN
MAIN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FACE LOAD
pressure TABLE
1528 Marc Users Guide: Part 2
CHAPTER 3.33
Figure 3.33-18 Transient Pressure Loading
Step 2: Create and submit a transient analysis.
The next step is creating the loadcase Ior the transient analysis. The Iollowing button sequence creates the new
loadcase.
table1
OK
faces ADD
(Box Pick the cantilevered elements)
END LIST (#)
MAIN
LOADCASES
NEW
MECHANICAL
DYNAMIC TRANSIENT
LOADS
apply2 (deselect static load)
OK
TOTAL LOADCASE TIME
0.09
1529
CHAPTER 3.33
Bracket
Step 3: Create job and submit it.
The next step is creating the job and submitting it. The Iollowing button sequence creates and submits the job.
# STEPS
90
OK
MAIN
JOBS
NEW
MECHANICAL
available
lcase3 (select loadcase 3)
ANALYSIS OPTIONS
IMPLICIT
SINGLE-STEP HOUBOLT (verify that dynamic transient
operator implicit Single-step Houbolt is used)
OK
INITIAL LOADS
boundary conditions
apply2 (remove static load)
OK
JOB RESULTS
available elemnt scalars
von_mises (select Equivalent von Mises Stress)
OUT & MID (select outer and midplane LAYERS)
OK
JOB PARAMETERS
# SHELL/BEAM LAYERS
3
OK (twice)
SAVE
RUN
SUBMIT 1
MONITOR
1530 Marc Users Guide: Part 2
CHAPTER 3.33
Step 4: Postprocess the results.
The Iinal step in any analysis is postprocessing the results. This is done by opening the post Iile and reviewing the
results. The Iollowing button sequence does this. However, there may be other results that the user wishes to review.
This walks through all 90 increments.
It is also possible to monitor a path plot. The Iollowing button sequence shows this.
MAIN
RESULTS
OPEN DEFAULT
NEXT INC
DEF & ORIG
deformed shape SETTINGS
deformation scaling FACTOR
5.
deformation scaling MANUAL
OK
SCALAR
Equivalent von Mises Stress Layer 2
OK
CONTOUR BANDS
MONITOR
MAIN
RESULTS
REWIND
PATH PLOT
NODE PATH
215
124
123
132
END LIST (#)
SHOW IDS
5
VARIABLES
1531
CHAPTER 3.33
Bracket
The stresses Ior the nodes in the node path do not exceed the yield stress oI 36000 psi.
As a last step, history plots are made. First, a plot oI the stresses versus time is shown in Figure 3.33-19.
ADD CURVE
Arc Length
Equivalent von Mises Stress Layer 1
Arc Length
Equivalent von Mises Stress Layer 2
Arc Length
Equivalent von Mises Stress Layer 3
RETURN
FIT
YMAX
36000
MONITOR
SHOW HISTORY
SHOW MODEL (select SHOW MODEL)
MAIN
RESULTS
REWIND
HISTORY PLOT
SET NODES
208
124
123
END LIST (#)
COLLECT DATA
0 90 1
SHOW IDS
0
NODES/VARIABLES
ADD VARIABLE
global variables
Time
variables at nodes
Equivalent von Mises Stress Layer 3
FIT
1532 Marc Users Guide: Part 2
CHAPTER 3.33
Figure 3.33-19 History Plot of von Mises Stress
Finally, the tip displacement in z-direction are shown and compared with the value oI due to the static load with the
same magnitude (loadcase 1), which was Iound to be 0.28
MAIN
RESULTS
REWIND
HISTORY PLOT
SET NODES
215
END LIST (#)
COLLECT DATA
0 90 1
NODES/VARIABLES
ADD 1-NODE CURVE
215
global variables
Time
variables at nodes
Displacement z
FIT
1533
CHAPTER 3.33
Bracket
Figure 3.33-20 History Plot of Tip Z-displacement
Conclusion
Due to the dynamic loading, a larger value oI the tip displacement is Iound Ior the same value oI the maximum
pressure. The bracket is shown to withstand the dynamic loads. The stresses never exceed yield.
Pressure Table
Pressure as function of time file c14.tbl:
# Title
table1
# X-axis Label
X
# Y-axis Label
Y
# Type
1
# Steps in X and Y
10 20
# X-min, X-max, Y-min, Ymax
0.000000000000e+00 9.000000000000e+00 -1.000000000000e+00
1.000000000000e+00
04
0.000000000000e+00 0.000000000000e+00 1
1.000000000000e+02 1.000000000000e+00 2
4.000000000000e+02 0.000000000000e+00 3
9.000000000000e+02 0.000000000000e+00 4
1534 Marc Users Guide: Part 2
CHAPTER 3.33
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
bracket.proc Mentat procedure Iile
bracket.tbl Table data
Chapter 3.34: Single Step Houbolt Dynamic Operator
3.34
Single Step Houbolt
Dynamic Operator

Chapter Overview 1536

Impact of a Ball on a Plate 1536

Eigenvalue Analysis 1536

Transient Analysis 1546

Input Files 1553

1536 Marc Users Guide: Part 2
CHAPTER 3.34
Chapter Overview
This chapter demonstrates the use oI the Single Step Houbolt method using the numerical simulation oI the impact oI
a ball on a plate.
In a contact analysis involving dynamics, the user oIten encounters a kind oI shock loading iI two moving bodies hit
each other. Depending on the velocity magnitudes, material properties, etc., such a shock loading may trigger
high-Irequency oscillations, which in turn may cause numerical troubles. When use is made oI the well-known
Newmark-beta dynamic operator, the user has to choose speciIic damping properties in order to get rid oI the undesired
oscillations. However, the choice oI damping properties is not trivial, and usually also relevant oscillations may be
damped out.
The Single Step Houbolt method has spectral properties similar to the classical Houbolt method and thus possesses
high-Irequency dissipation. It should be noted that the term 'high-Irequency is always related to the time step
chosen. II is small compared to the time period oI a mode (say ), the mode is properly integrated; iI
is large compared to the time period oI a mode (say ), the mode is damped out rapidly.
Impact of a Ball on a Plate
The simulation consists oI two steps: Iirst an eigenvalue analysis is perIormed to estimate a proper time step, then the
impact simulation itselI is perIormed using a transient dynamic analysis.
Background Information
A circular plate with a thickness oI 0.0025 m and a radius oI 0.05 m is clamped around its circumIerence and hit by a
ball with a radius oI 0.02 m. The initial velocity oI the ball is 2.5 m/s. The material behavior oI both the plate and the
ball is considered to be elastic-plastic. The plate has a Young`s modulus oI 7x10
9
N/m
2
, a Poisson`s ratio oI 0.3, a
density oI 2500 kg/m
3
, an initial yield stress oI 7x10
6
N/m
2
, and a hardening modulus oI 1.4x10
7
N/m
2
. The ball has
a Young`s modulus oI 2x10
11
N/m
2
, a Poisson`s ratio oI 0.3, a density oI 7800 kg/m
3
, an initial yield stress oI 2x10
8

N/m
2
, and a hardening modulus oI 6x10
8
N/m
2
.
The plate and the ball are modeled using 4-node axisymmetric elements with Iull integration (Marc element type 10).
Around the contact area, the mesh oI the plate will be reIined. For the eigenvalue analysis, no contact bodies will be
deIined in order to get the eigenIrequencies oI the plate and the ball independently.
Eigenvalue Analysis
Model Generation
The Iinite element mesh oI the plate will be obtained by subdividing one element and then reIining the mesh near the
center oI the plate by the reIine option. The Iinite element mesh oI the ball will be created by deIining curves and using
the 2-D overlay mesher. The elements oI the plate and the ball will be stored in element sets.
At
At T At T 1 15 <
At T At T 10 >
1537
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
FILES
NEW, OK
RESET PROGRAM
VIEW
SHOW VIEW 1
MAIN
MESH GENERATION
pts ADD
-0.0025 0 0
0 0 0
0 0.05 0
-0.0025 0.05 0
0.02 0 0
0.04 0 0
elem ADD
NODE(-0.0025, 0, 0)
NODE(0, 0, 0)
NODE(0, 0.05, 0)
NODE(-0.0025, 0.05, 0)
SUBDIVIDE
DIVISIONS
4 30 1
ELEMENTS
all: EXIST.
DIVISIONS
2 2 1
ELEMENTS (those shown in Figure 3.34-1a)
2 3 4 5 6 7 8 9 10 32 33 34 35 36 37 38 39 40 62 63
64 65 66 67 68 69 70 92 93 94 95 96 97 98 99 100
ZOOM BOX (zoom transition elements
Figure 3.34-1b)
1538 Marc Users Guide: Part 2
CHAPTER 3.34
Figure 3.34-1 FE-Mesh for Impact Problem (a) Elements (b) Zoom Box
A B
C D
Pick
Pick nodes A, B, C, D
Zoom
(a)
(b)
1539
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
REFINE
45
11
41
END OF LIST (#)
107
71
101
END OF LIST (#)
FILL
RETURN
SWEEP
NODES
all: EXIST.
SELECT
elements STORE
Plate
OK
all: EXIS.
SELECT SET
Plate
OK
MAIN
45
107
11 41 71 101
45
107
Figure 3.34-2 FE-Mesh for Plate
1540 Marc Users Guide: Part 2
CHAPTER 3.34
crvs ADD
2
6
CURVE TYPE
CENTER/RADIUS/ANG/ANG
RETURN
crvs ADD
0.02 0 0 (center)
0.02 (radius)
0 (angle)
180 (angle)
AUTOMESH
2D PLANAR MESHING
QUADRILATERALS
20 20
QUAD MESH!
all: EXIST.
SELECT
elements STORE
Ball
OK
all: UNSEL.
CLEAR SELECT
ID SETS
MAIN
MESH GENERATION
RENUMBER
ALL
MAIN
Figure 3.34-3 FE-Mesh for Ball
1541
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Boundary Conditions
Boundary conditions are deIined to clamp the plate and to introduce symmetry conditions.
Figure 3.34-4 FE-Mesh and Boundary Conditions for Impact Problem
BOUNDARY CONDITIONS
MECHANICAL
NAME
Clamped
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
OK
nodes ADD
3 4 64 95 126
# | End of List
NEW
NAME
Symmetry
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
OK
nodes ADD
1 2 34 65 96 157 193 221 249 420 421 422 423 424 425
426 427 428 429 430 431 432 433 434 435 436 437 438
END OF LIST (#)
Clamped
Symmetry
1542 Marc Users Guide: Part 2
CHAPTER 3.34
Material Properties
For the eigenvalue analysis, only linear elastic material properties are required; the complete material description is
already deIined where the hardening behavior is entered using tables oI type plastic strain.
MATERIAL PROPERTIES
TABLES
NEW
1 INDEPENDENT VAR
NAME
plate_hardening
TYPE
eq_plastic_strain
OK
ADD
0 7e6
1 21e6
function value f
MIN = 0
MAX = 8e8
NEW
1 INDEPENDENT VAR
NAME
ball_hardening
TYPE
eq_plastic_strain
OK
ADD
0 2e8
1 8e8
FILLED
RETURN
NEW
NAME
plate_material
ISOTROPIC
YOUNGS MODULUS = 7e9
Figure 3.34-5 Hardening for Plate and Ball
1543
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Figure 3.34-6 Material Properties of Plate and Ball
POISSONS RATIO = 0.3
DENSITY = 2500.0
ELASTIC-PLASTIC
INITIAL STRESS = 1, TABLE = Plate_Hardening
OK (twice)
elements ADD (select plate elements)
NEW
NAME
ball_material
ISOTROPIC
YOUNGS MODULUS = 2e11
POISSONS RATIO = 0.3
DENSITY = 7800.0
ELASTIC-PLASTIC
INITIAL STRESS = 1, TABLE = Plate_Hardening,
OK (twice)
elements ADD (select ball elements)
MAIN
1544 Marc Users Guide: Part 2
CHAPTER 3.34
Loadcases
A dynamic modal mechanical load case is deIined. The deIault Lanczos method is used. The lowest Irequency is set
to 10 Hz, 20 modes are asked Ior and since the ball still has one rigid body mode, the non-positive deIinite option is
activated.
Jobs
A mechanical job is deIined, in which the previously deIined load case is selected. The lumped mass matrix option is
activated and the element type is set to 10. The model is saved and the job submitted.
LOADCASES
NAME
Modal_analysis
MECHANICAL
DYNAMIC MODAL
LOWEST FREQ.= 10
# MODES = 20
NON-POS. DEFINITE
OK
MAIN
JOBS
NAME
Modal
MECHANICAL
Modal_analysis
AXISYMMETRIC
ANALYSIS OPTIONS
LUMPED
OK (twice)
FILES
SAVE AS
plate_ball_modal
OK
RETURN
RUN
SUBMIT 1
MONITOR
Figure 3.34-7 Modal Loadcase
Figure 3.34-8 Job Submit
1545
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Results
The post Iile is opened. Using the scan option, you can easily get an overview oI the eigenIrequencies. A couple oI
eigenmodes are visualized. Figure 3.34-9 shows the Iirst eigenmode and corresponding eigenIrequency.
Figure 3.34-9 First Eigenmode
RESULTS
OPEN DEFAULT
DEFORMED SHAPE SETTINGS
DEFORMATION SCALING
AUTOMATIC
RETURN
DEF ONLY
NEXT
1546 Marc Users Guide: Part 2
CHAPTER 3.34
Transient Analysis
The results oI the modal analysis clearly show that the eigenIrequencies oI the plate are much lower than the
eigenIrequencies oI the ball. Based on the material properties, it may be assumed that the deIormations mainly occur
in the plate. II it is assumed that modes with Irequencies up to that oI the second eigenmode should be properly
integrated, then the time step Ior the transient analysis can be estimated as
seconds.
Model Generation
The Iinite element model is the same as used Ior the modal analysis.
Boundary Conditions
The boundary conditions are the same as used Ior the modal analysis.
Initial Conditions
All the nodes oI the ball get an initial velocity oI 2.5 m/s in negative x-direction.
FILES
OPEN
plate_ball.mud
OK
SAVE AS
plate_ball_transient
OK
MAIN
INITIAL CONDITIONS
NAME = Initial_velocity
MECHANICAL
VELOCITY
VELOCITY X (on)
-2.5
OK
SELECT
SELECT BALL
OK
MAKE VISIBLE
RETURN
At 1 15f 1 15 3382 ( ) 2 10
5
~
1547
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Contact Bodies
Two deIormable bodies are deIined: one Ior the plate and one Ior the ball. Since in the area oI contact the plate has a
Iiner mesh than the ball, the plate is the Iirst body. For the ball, the analytical description is used in order to get a
smooth, accurate description oI its boundary.
nodes ADD
all: VISIBLE
SELECT
MAKE
INVISIBLE
MAIN
CONTACT
CONTACT BODIES
NAME = Body_Plate
DEFORMABLE,
OK
elements ADD
Plate
NEW
NAME
Body_Ball
DEFORMABLE
OK
elements ADD
Ball
BOUNDARY DECRIPTION ANALYTICAL
nodes ADD
420 438
END OF LIST (#)
ID CONTACT, MAIN
1548 Marc Users Guide: Part 2
CHAPTER 3.34
Figure 3.34-10 Contact Bodies
Material Properties
The material properties are already entered Ior the modal analysis.
Loadcase
The loadcase oI the modal analysis is removed and a new loadcase Ior a transient dynamic analysis is deIined. A total
time oI 0.006 seconds are analyzed in 400 equally sized steps, corresponding to a time step oI second as per
our desire to properly integrate Irequencies in the range oI the second eigenmode.
LOADCASES
MECHANICAL
REM
NAME = Dynamic_transient
DYNAMIC TRANSIENT
SOLUTION CONTROL
MAX # RECYCLES = 20
NON-POSITIVE DEFINITE (on)
PROCEED WHEN NOT CONVERGED (on)
OK
TOTAL LOAD CASE TIME = 0.006
1.5x10
5
1549
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Figure 3.34-11 Dynamic Transient Loadcase
Jobs
The previously deIined load case is used in the mechanical job. The initial conditions are activated. Notice that the
Single Step Houbolt method is the deIault time integration method in Mentat. A distance tolerance bias Iactor oI 0.9
is entered and the iterative increment splitting option is selected. The large strain plasticity method using the additive
decomposition and the constant dilatation option is activated, since relatively large plastic deIormations are to be
expected. As post Iile element variables the equivalent von Mises stress and the total equivalent plastic strain are
selected and as nodal variables the displacements, velocities, accelerations and contact normal Iorces. The model is
saved and the job is submitted.
FIXED # STEPS = 400
OK
MAIN
JOBS
MECHANICAL
NAME
Dynamic_transient
INITIAL LOADS
Initial_velocity (turn on)
OK
CONTACT CONTROL
DISTANCE TOLERANCE BIAS = 0.9
INCREMENT SPLLITTING ITERATIV
OK
1550 Marc Users Guide: Part 2
CHAPTER 3.34
Results
The post Iile is opened and the equivalent plastic strain in the Iinal deIormed conIiguration is plotted. This is shown
in Figure 3.34-13. History plots oI the velocity oI the center oI ball and the displacement oI the center oI the plate are
indicated in Figure 3.34-14 and Figure 3.34-15. Figure 3.34-15 shows a low-amplitude vibration with a time period oI
approximately seconds which can accurately be captured with the chosen time step.
ANALYSIS OPTIONS
CONSTANT DILATATION (on)
plasticity procedure
LARGE STRAIN ADDITIVE (on)
OK
JOB RESULTS
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
NODAL QUANTITIES
CUSTOM
Displacement
Velocity
Acceleration
Cont_Nor_Force
OK (twice)
SAVE
RUN
SUBMIT 1
MONITOR
Note: The deIault transient dynamic operator set in Mentat is the new Single Step Houbolt operator.
RESULTS
OPEN DEFAULT
DEFORMED SHAPE SETTINGS
DEFORMATION SCALING
MANUAL
RETURN
DEF ONLY
SCALAR
Figure 3.34-12 Default Houbolt Operator
8.5x10
4
1551
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Figure 3.34-13 Equivalent Plastic Strain in Deformed Configuration at Increment 400
Total Equivalent Plastic Strain
OK
CONTOUR BANDS
MONITOR
OK
HISTORY PLOT
SET NODES
429 85
# END LIST
COLLECT DATA
0 400 1
NODES/VARIABLES
Add 1-NODE CURVE
429
Time
Velocity X
OK
1552 Marc Users Guide: Part 2
CHAPTER 3.34
Figure 3.34-14 Velocity of Center of Ball as a Function of Time
FIT
YMIN = -3
YMAX = 1
SHOW IDS 0 (none)
CLEAR CURVES
NODES/VARIABLES
Add 1-NODE CURVE
85
Time
Displacement X
OK
FIT
YMIN = -5.5
YMAX = 0
1553
CHAPTER 3.34
Single Step Houbolt Dynamic Operator
Figure 3.34-15 Displacement of Center of Plate as a Function of Time
You may wish to run Mentat procedure Iiles that are in the examples/marc_ug subdirectory under Mentat. The
procedure Iiles modal.proc and transient.proc builds, runs, and postprocesses the modal and transient jobs,
respectively.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
modal.proc Mentat procedure Iile
transient.proc Mentat procedure Iile
plate_ball.mud Mentat model Iile
1554 Marc Users Guide: Part 2
CHAPTER 3.34
Chapter 3.35: Dynamic Analyses of a Cantilever Beam
3.35
Dynamic Analyses of a
Cantilever Beam

Summary 1556

Modal Analysis 1557

Harmonic Analysis 1560

Transient Analysis 1563

Input Files 1571

1556 Marc Users Guide: Part 2
CHAPTER 3.35
Summary
Title Dynamic analysis oI a cantilever beam
Problem Ieatures Static, modal, harmonic and transient with contact
Geometry
Material properties E 30.0x10
6
Psi, 0.3, 7.33161x10
-4
lbI s
2
/in
4
Analysis type Static, modal, harmonic and transient with contact
Boundary conditions Cantilever with tip load
Element type Plane stress element type 3
FE results Characterize static and dynamic response
10" X 1" X 1"
500 lbf
v
146
147
148
149
150
job1
Time (x.001)
Displacement Y Node 55 (x.1)
0 9.231
-1.328
-0.664
145
0
144
142
131
132
133
134
135
136
137
138
139
140
141
143
130
129
127
85
86
87
88
89
90
91
92
93
94
95
84
96
83
81
70
71
72
73
74 75
76
77
78
79
80
82
97
98
99
116
117
118
119
120
121
122
123
124 125
126
115
114
113
112
100
101
102
103
104
128
105
107
108
109
110
111
106
69
68
66
24 25
26
27
28
29
30
31
32
33
34
23
35
22
20
9
10
11
12
13
14
15
16
17
18
19
21
36
37
38
55
56
57
58
59
60
61
62
63
64
65
54
53
52
51
39
40
41
42
43
8
44
46
47
48
49
50
45
7
6
5
67
0
1
2
3
4
S
t
a
t
i
c

S
o
l
u
t
i
o
n
Period
= 1s/325
= 0.0031
1557
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Modal Analysis
Overview
A modal analysis oI a cantilever beam (Figure 3.35-1) determines its natural Irequencies.
Figure 3.35-1 Beam Dimensions and Load
The end load is turned oII prior to the modal analysis. The eIIects oI pre-stress, iI present, change the natural
Irequencies, like the tension in a guitar string. However, it is not modeled here.
The ten lowest natural Irequencies and corresponding mode shapes were requested. Here, the mode shape oI the lowest
natural Irequency oI 325 Hz is shown with displacements automatically magniIied (Figure 3.35-2). As expected, it
shows 'easy wise bending.
Figure 3.35-2 1st Bending Mode of Vibration
10" X 1" X 1"
500 lbf
-1.744e-002
2.312e+000
4.641e+000
6.971e+000
9.300e+000
1.163e+001
1.396e+001
1.629e+001
1.862e+001
2.095e+001
2.328e+001
lcase1
Displacement Y
Inc: 0:1
Time: 0.000e+000
Freq: 3.251e+002
X
Y
Z
1
1558 Marc Users Guide: Part 2
CHAPTER 3.35
Modal Analysis
The cantilever beam used in the static analysis oI Chapter 3.18: Cantilever Beam starts the modeling; using the same
geometry and material, the procedures will be modiIied to perIorm a modal analysis.
FILES
OPEN
d1.mud
SAVE AS
d11
OK
LOADCASES
MECHANICAL
DYNAMIC MODAL
OK
MAIN
JOBS
TYPE: MECHANICAL
PROPERTIES
SELECT lcase1
INITIAL LOADS
turn off point load
OK
OK
RUN
SUBMIT1
OK
SAVE
OPEN POST FILE (RESULTS MENU)
NEXT
DEFORMED SHAPE SETTINGS
10" X 1" X 1"
500 lbf
1559
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Figure 3.35-3 1st Bending Mode of Vibration at 325 Hz
DEFORMATION SCALING AUTOMATIC
OK
DEF & ORIG
CONTOUR BAND
SCALAR
DISPLACEMENT Y, OK
SCAN
-1.744e-002
2.312e+000
4.641e+000
6.971e+000
9.300e+000
1.163e+001
1.396e+001
1.629e+001
1.862e+001
2.095e+001
2.328e+001
lcase1
Displacement Y
Inc: 0:1
Time: 0.000e+000
Freq: 3.251e+002
X
Y
Z
1st Natural frequency at
3.251E+02 Hz.
1560 Marc Users Guide: Part 2
CHAPTER 3.35
Harmonic Analysis
Overview
A harmonic analysis oI a cantilever beam determines the dynamic response oI the cantilever beam to an oscillating tip
load with a 500 pound magnitude. The end load is turned on in the harmonic loadcase, and the range oI excitation
Irequencies is 0 to 400 Hz, in 40 steps oI 10 Hz.
The Iigure below (Figure 3.35-4) plots the tip displacement magnitude along the Irequency range. It contains the
static solution at 0 Hz, the resonance around the Iirst natural Irequency oI 325 Hz, and ends with a phase reversal
above 325 Hz.
Figure 3.35-4 Harmonic Response Summary
Harmonic Analysis and Results
FILES
OPEN
d12.mud
SAVE AS
d13
OK
3.353
4
1 0
lcase1
Frequency (x100)
Displacement Y Node 55
0
-1.5
1
Static Solution
Resonance @ 325 Hz
1561
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
MAIN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply3, OK
HARMONIC BC
POINT LOAD (view load magnitude)
OK
MAIN
LOADCASES
MECHANICAL
DYNAMIC HARMONIC
LOADS (set harmonic point load on)
OK
LOWEST FREQ
0
HIGHEST FREQ
400
# OF FREQS
40
OK
MAIN
JOBS
TYPE MECHANICAL
PROPERTIES
SELECT lcase1, OK
RUN
SUBMIT1
OK
SAVE
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
SET LOCATIONS
n:55 # (pick the one with point load)
INC RANGE
0:0 0:40 1 (colon separates increment and sub increment)
ADD CURVES
ALL LOCATIONS
Frequency
Displacement Y
FIT
1562 Marc Users Guide: Part 2
CHAPTER 3.35
Figure 3.35-5 Harmonic Response 0 to 400 Hz
3.353
4
1 0
lcase1
Frequency (x100)
Displacement Y Node 55
0
-1.5
Static Solution
Resonance @ 325 Hz
1563
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Transient Analysis
Overview
The dynamic response continues Ior a cantilever beam initially at rest subjected to a constant tip load oI 500 pounds
aIter t0. We shall Iind that the kinetic energy will be bounded by the total strain energy oI the static solution discussed
in Chapter 3.18: Cantilever Beam. The dynamic transient loadcase time period is set to 3/(325 Hz) to get three cycles
oI response. Plotting the tip displacement along the time axis shows the tip oscillating about the static solution.
Figure 3.35-6 Transient Response Summary
The second run includes damping; the tip displacement along the time axis plot shows the tip oscillating about the
static solution with the oscillations diminishing with time.
Analysis and Results
The cantilever beam used in the static analysis oI Chapter 3.18: Cantilever Beam starts the modeling; using the same
geometry and material, the procedures will be modiIied to perIorm a transient analysis.
Figure 3.35-7 Transient Response with Contact: Problem Description
146
147
148 149
150
job1
Time (x.001)
Displacement Y Node 55 (x.1)
0 9.231
-1.328
-0.664
145
0
144
142
131
132
133
134
135
136
137
138
139
140
141
143
130
129
127
85
86
87
88
89
90
91
92
93
94
95
84
96
83
81
70
71
72
73 74 75
76
77
78
79
80
82
97
98
99
116
117
118
119
120
121
122
123
124 125
126
115
114
113
112
100 101
102
103
104
128
105
107
108
109
110
111
106
69
68
66
24 25 26
27
28
29
30
31
32
33
34
23
35
22
20
9
10
11
12
13
14
15
16
17
18
19
21
36
37
38
55
56
57
58
59
60
61
62
63
64
65
54
53
52
51
39
40
41
42
43
8
44
46
47
48
49
50
45
7
6
5
67
0
1
2
3
4
S
t
a
t
i
c

S
o
l
u
t
i
o
n
Period
= 1s/325
= 0.0031
146 147 148 149 150
job1
Time (x.001)
Displacement Y Node 55 (x.1)
0 9.231
-1.14
145
0
144
142
131
132
133
134
135
136
137
138
139
140
141
143
130 129
127
85
86
87
88
89
90
91
92
93
94
95
84
96
83
81 70
71
72
73
74 75 76 77
78
79
80
82
97
98 99
116
117
118
119
120
121
122 123 124 125 126
115
114
113
112
100 101 102 103
104
128
105
107
108
109
110
111
106
69
68
66
24 25 26
27
28
29
30
31
32
33
34
23
35
22
20
9
10
11
12
13
14
15
16
17
18
19
21
36
37
38
55
56
57
58
59
60
61
62
63
64
65
54
53
52 51
39
40
41
42
43
8
44
46
47
48
49 50
45
7
6
5
67
0
1
2
3
4
Period
= 1s/325
= 0.0031
10" X 1" X 1"
500 lbf
1564 Marc Users Guide: Part 2
CHAPTER 3.35
FILES
OPEN
d1.mud
SAVE AS
d14
OK
MAIN
LOADCASE
MECHANICAL
DYNAMIC TRANSIENT
TIME
3/325 (remember 1st natural frequency)
# STEPS
150
OK
MAIN
JOBS
TYPE: MECHANICAL
PROPERTIES
SELECT lcase1
OK
SAVE
RUN
STYLE: OLD (use old style table input)
SUBMIT1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET LOCATIONS
n:55 # (pick the one with point load)
ALL INCS
ADD CURVES
ALL LOCATIONS
1565
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Figure 3.35-8 Transient Response at Cantilever Beam
Damping Analysis
What about damping? Physically, we know it is present. Let`s see how to model with damping.
Time
Displacement Y
FIT
FILES
OPEN
d14.mud
SAVE AS
d15
OK
MAIN
MATERIAL PROPERTIES (twice)
STRUCTURAL
DAMPING (twice)
146
147
148
149
150
job1
Time (x.001)
Displacement Y Node 55 (x.1)
0 9.231
-1.328
-0.664
145
0
144
142
131
132
133
134
135
136
137
138
139
140
141
143
130
129
127
85
86
87
88
89
90
91
92
93
94
95
84
96
83
81
70
71
72
73
74 75
76
77
78
79
80
82
97
98
99
116
117
118
119
120
121
122
123
124 125
126
115
114
113
112
100
101
102
103
104
128
105
107
108
109
110
111
106
69
68
66
24 25
26
27
28
29
30
31
32
33
34
23
35
22
20
9
10
11
12
13
14
15
16
17
18
19
21
36
37
38
55
56
57
58
59
60
61
62
63
64
65
54
53
52
51
39
40
41
42
43
8
44
46
47
48
49
50
45
7
6
5
67
0
1
2
3
4
1
Static Solution
Period
= 1s/325
= 0.0031
1566 Marc Users Guide: Part 2
CHAPTER 3.35
Figure 3.35-9 Transient Response with Damping
STIFFNESS MATRIX MULTIPLIER
1E-4
OK (twice)
SAVE
MAIN
JOBS
RUN
SUBMIT1
MONITOR
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET LOCATIONS (pick the one with point load)
ALL INCS
ADD CURVES
ALL LOCATIONS
Time
Displacement Y
FIT
146
147 148 149 150
job1
Time (x.001)
Displacement Y Node 55 (x.1)
0 9.231
-1.14
145
0
144
142
131
132
133
134
135
136
137
138
139
140
141
143
130
129
127
85
86
87
88
89
90
91
92
93
94
95
84
96
83
81 70
71
72
73
74 75 76 77
78
79
80
82
97
98
99
116
117
118
119
120
121
122
123 124 125 126
115
114
113
112
100 101 102 103
104
128
105
107
108
109
110
111
106
69
68
66
24 25 26
27
28
29
30
31
32
33
34
23
35
22
20
9
10
11
12
13
14
15
16
17
18
19
21
36
37
38
55
56
57
58
59
60
61
62
63
64
65
54
53
52
51
39
40
41
42
43
8
44
46
47
48
49 50
45
7
6
5
67
0
1
2
3
4
1
Static Solution
Period
= 1s/325
= 0.0031
-0.664
1567
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Over Hanging Beam Analysis
Now we assume the beam overhangs a rigid bumper; the beam contacts the bumper only on the way down and
separates Irom the bumper when displacing upward. The beam is rest and the load is placed on the end at time t0.
Again, the kinetic energy is bounded by the total strain energy oI the static solution discussed in Chapter 3.18:
Cantilever Beam. An analogous loading scheme would be to displace the tip upward with a 500 pound load then release
the beam, the initial potential energy exchanges with the kinetic energy keeping the total energy constant since the
beam to bumper contact is elastic.
Figure 3.35-10 Beam over Rigid Bumper
FILES
OPEN
d1.mud (start with modal model)
SAVE AS
d16
OK
MAIN
MESH GENERATION
CURVE TYPE, CIRCLE:CENTER,RADIUS
RETURN
CRVS: ADD
5 0 0
.2
MOVE
TRANS.
0 -.23 0
CURVES
ALL: EXISTING
MAIN
10" X 1" X 1"
500 lbf
0.03 gap
Bumper
1568 Marc Users Guide: Part 2
CHAPTER 3.35
CONTACT
CONTACT BODIES
DEFORMABLE,
OK
ELEMENTS: ADD,
ALL: EXISTING
NEW
RIGID, OK
2-D CURVES ADD
ALL: EXISTING
MAIN
LOADCASES
MECHANICAL
DYNAMIC TRANSIENT
CONVERGENCE TESTING
RELATIVE FORCE TOLERANCE
0.001
OK
TOTAL LOADCASE TIME = 3/325
MULTI-CRITERIA: PARAMETERS
INITIAL FRACTION OF LOADCASE TIME = 1E-6
MINIMUM FRACTION OF LOADCASE TIME = 1E-7
DESIRED # RECYCLES/INCREMENT: SET = 3
AUTOMATIC CRITERIA (on)
OK (twice)
MAIN
JOBS
PROPERTIES
lcase1
CONTACT CONTROL
ADVANCED CONTACT CONTROL
DISTANCE TOLERANCE
.003
BIAS
0.9
OK (thrice)
1569
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Figure 3.35-11 Transient Response: Overhang Beam with Contact
Vertical motion oI node 26 (over bumper) is limited by the rigid body and contact oI the beam with the rigid body
increases the Irequency content oI the response. The dominant periods oI the tip and mid span displacements shorten
showing a higher Irequency. From the modal analysis, the second mode has a Irequency oI 1995.23 (see Figure 3.35-3)
approximately six times higher than the Iirst mode. The time period in Figure 3.35-11 would have only three cycles as
shown in Figure 3.35-6 without contact. There are about Iour repetitions above indicating a higher Irequency.
SAVE
RUN
SUBMIT(1)
MONITOR
OK (twice)
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET LOCATIONS
n:26 n:55 #
ALL INCS
ADD CURVES
ALL LOCATIONS
Time
Displacement Y
FIT
SHOW ID: 100
1300
1400
1500
1600
1200
1700
1900
2000
2100
2200
1800
1100
1000
900
0100 200 300
400
500
600
700
800
2300
2400
2600
0.008
Node 26 (over bumper)
-1.144
9.23 0
Displacement Y (x.1)
2700
2800
2900
3000
3100
3300
3400
3500
3600
job1
Time (x.001)
2500 3200
3600
Node 55
0
100
300
400
500
600
200
700
800
900
1000
2600
2700
2800
2900
2500
3000
3200
3300
3400
3500
3100
2400
2300
2200
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
-0.03
B
u
m
p
e
r
S
t
a
t
i
c

S
o
l
u
t
i
o
n
-0.664
10" X 1" X 1"
500 lbf
at node 55
0.03 gap below node 26
Bumper

1570 Marc Users Guide: Part 2
CHAPTER 3.35
Figure 3.35-12 Kinetic Energy History for Overhang Beam with Contact
The kinetic energy history oI the transient response in shown in Figure 3.35-12, it has nearly eight cycles during the
total time period which is close to the nine cycles we have seen in the second mode shape without contact. Clearly,
mid span beam contact with the bumper increases the Irequency with the second mode becoming more dominant.
Furthermore, we can estimate the amplitude oI the kinetic energy Irom the static analysis in Chapter 3.18: Cantilever
Beam. The total strain energy is simply halI the product oI the Iorce (500 lbI) times the tip displacement (6.692 x10
-2

in) or some 16.7 lbI-in; this is very close to the amplitude oI the kinetic energy history.
CLEAR CURVES
ADD CURVES
GLOBAL
Time
Kinetic Energy
FIT
2700
2800
2900
3000
2600
3100
3300
3400
3500
3600
3200
job1
2500
2300
1300
1400
1500
1600
2400
1700
1900
2000
2100
2200
1800
Time (x.001)
Kinetic Energy (x10)
0 9.23
0
1.697
1200 1100
900
1000
0
100
200
300
500
600
700
800
400
1571
CHAPTER 3.35
Dynamic Analyses of a Cantilever Beam
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
d1.proc Mentat procedure Iile
d2.proc Mentat procedure Iile
d3.proc Mentat procedure Iile
d1.mud Mentat model Iile: static model
d12.mud Mentat model Iile: modal
d13.mud Mentat model Iile: harmonic
d14.mud Mentat model Iile: transient
d15.mud Mentat model Iile: transient with damping
d16.mud Mentat model Iile: transient with contact
1572 Marc Users Guide: Part 2
CHAPTER 3.35
Chapter 3.36: Plastic Spur Gear Pair Failure
3.36
Plastic Spur Gear Pair Failure

Summary 1574

Gear Geometry 1575

Material Modeling 1576

Contact 1576

Failure Criteria 1577

Experimental Test Machine 1580

Results & Conclusions 1582

Modeling Tips 1583

Input Files 1583

References 1583

Animation 1584
1574 Marc Users Guide: Part 2
CHAPTER 3.36
Summary
Title Plastic Spur Gear Pair Failure
Problem Ieatures Acetal copolymer gears in contact
UACTIVE user subroutine deactivates Iailed elements
Geometry
Material Properties Elastic-plastic
Analysis type Quasi-static analysis
Boundary conditions Rigid bodies inside shaIt holes hold one gear Iixed and rotate the other.
Element type 4-node plane strain element type 11 with variable thickness
FE results Predicted torque versus twist compared to experimental values.
2.026 in
0.00 0.02 0.04 0.06 0.08 0.10
0
20
40
60
80
100
120
Twist (Radians)
T
o
r
q
u
e

(
i
n
-
l
b
f
)
-53.49
401.86
857.21
1312.56
1767.91
2223.26
2678.60
3133.95
3589.30
4044.65
4500.00
Equivalent Von Mises Stress
Inc: 57
Inc: 79
Inc: 112
Inc: 128
Inc: 140
Experimental
Prediction
1
0
1
1
1
1
1575
CHAPTER 3.36
Plastic Spur Gear Pair Failure
An elastic-plastic Iinite element analysis oI the quasi-static loading oI two acetal copolymer gears in contact is
preIormed. Torque verses twist oI the gear set is compared to actual experimental results. The gear geometry is
modeled by plane strain elements with variable thickness between the rim and web. Gear tooth Iailure is modeled by
deactivating elements when the plastic strain oI 0.15 is exceeded in the tensile regions.
Gear Geometry
Two acetal copolymer spur gears were selected as test specimens. The geometry oI the gear teeth was based on the
American Gear ManuIacturers Association (AGMA) standard: Tooth proportions Ior Plastic Gears (Reference 1). The
entire gear pair is modeled to capture the correct torsional stiIIness oI the gear pair. The speciIications Ior the test gears
used are provided in the table below.
The test gears were assembled at a center distance oI 2.0620 inches. This gave a nominal backlash oI 0.0320 inches.
This relative large backlash permitted the test gears to reach relatively high torque levels without having the gear teeth
roll back on each other, thereby making contact on the backside oI the adjacent tooth. An illustration oI the gear model
(mesh lines included) assembly is shown in Figure 3.36-1. The rim oI the gear teeth is 0.25 inch (geom1) in thickness
and the web thickness (geom2 and geom3) is 0.123.
Basic Specification Data
Number oI Teeth
40
Diametric pitch 20
Standard pressure angle (degrees) 20
Tooth Iorm AGMA PT1
Standard addendum (inch) .0500
Standard whole depth (inch) .1120
Circular thickness on standard pitch circle (inch) .250
Basic Rack Data
Flank angle (degrees)
20
Tip to reIerence line (inch) .0665
Tooth thickness at reIerence line (inch) .250
Tip radius (inch) .0214
1576 Marc Users Guide: Part 2
CHAPTER 3.36
Figure 3.36-1 Geometry and Mesh
Material Modeling
The material is modeled as elastic-plastic with Young`s modulus oI 3.0x10
5
psi with an initial yield strength oI 2500
psi. The Cauchy stress versus true plastic strain curve is shown in Figure 3.36-2.
Figure 3.36-2 Material Behavior
Contact
The contact bodies are shown in Figure 3.36-3 and two circular rigid bodies, drive1 and drive2, are glued to each gear,
gear1 and gear2, respectively. Contact body drive1 rotates about the center oI the gear while drive2 remains stationary.
Two other rigid bodies (drive1out and drive2out) move just like drive1 and drive2, but are noncontacting rigid bodies
via contact table. They appear on the post Iile to visualize where the teeth would be iI they were rigid. Since kinematics
Ior the design oI a gear set assumes the gears to be rigid; it is convenient to see where the teeth would be iI the gear
material was rigid.
geom1
geom2
geom3
X
Y
Z
0.00 0.05 0.10 0.15 0.20
2000
3000
4000
5000
True Plastic Strain
Cauchy Stress (psi)
1577
CHAPTER 3.36
Plastic Spur Gear Pair Failure
Figure 3.36-3 Contact bodies
Failure Criteria
Two user routines are used. The PLOTV user subroutine captures the total equivalent plastic strain and the mean stress
and determines the elements to be deactivated when the mean stress is tensile (~ 1000) and the plastic strain exceeds
15. The ACTIVE user subroutine uses the inIormation Irom PLOTV to actually deactivate the elements selected. The
deactivated elements no longer participate in the analysis. The routines are listed below.
subroutine plotv(v,s,sp,etot,eplas,ecreep,t,m,nn,layer,ndi,
* nshear,jpltcd)
c* * * * * *
c define a variable for contour plotting (user subroutine).
c v variable to be put onto the post file
c s (idss) stress array
c sp stresses in preferred direction
c etot total strain (generalized)
c eplas total plastic strain
c ecreep total creep strain
c t array of state variable (temperature first)
c m(1) user element number
c m(2) internal element number
c m(3) material id
c m(4) internal material id
c nn integration point number
c layer(1) layer number
c layer(2) internal layer number
c ndi number of direct stress components
c nshear number of shear stress components
c jpltcd the absolute value of the user's entered post code
c* * * * * *
gear1
gear1
gear2
gear2
drive1
drive1
drive2
drive2
drive1out
drive1out
drive2out
drive2out
X
Y
Z
1578 Marc Users Guide: Part 2
CHAPTER 3.36
implicit real*8 (a-h,o-z)
common /mydata/ ielem(30000)
dimension s(*),etot(*),eplas(*),ecreep(*),sp(*)
dimension m(2),layer(2),t(2)
kc=1
call elmvar(18,m(1),nn,kc,v)
call elmvar( 7,m(1),nn,kc,ve)
if(nn.eq.1.and.ielem(m(1)).ne.1) ielem(m(1)) = 0
if(v.ge.1.0d3.and.ve.ge.0.15d0 ) ielem(m(1)) = 1
return
end
subroutine uactive(m,n,mode,irststr,irststn,inc,time,timinc)
c* * * * * *
c user routine to activate or deactivate an element
c
c m(1) - user element number
c m(2) - master element number for local adaptivity
c n - internal elsto number
c mode(1)=-1 - deactivate element, remove element from post file
c mode(1)=-11 - deactivate element, keep element on post file
c mode(1)=2 - leave in current status
c mode(1)=1 - activate element and add element to post file
c mode(1)=11 - activate element and keep status on post file
c mode(2)=1 - only activate/deactivate mechanical of coupled
c mode(2)=2 - only activate/deactivate thermal part of coupled
c mode(3)=0 - activation/deactivation at the end of increment
c mode(3)=1 - activation/deactivation at the beg. of increment
c irststr - reset stresses to zero
c irststn - reset strains to zero
c inc - increment number
c time - time at beginning of increment
c timinc - incremental time
c* * * * * *
implicit real*8 (a-h,o-z)
common /mydata/ ielem(30000)
dimension m(2),mode(3)
ie=m(1)
if(ielem(ie).eq.1.and.mode(1).ne.-1) then
mode(1)=-1
write(96,*) 'deactivating element ', ie, ' increment ', inc
else
mode(1)=2
end if
return
end
1579
CHAPTER 3.36
Plastic Spur Gear Pair Failure
Model Review
The model is complete and ready to run; however, we shall review the CONTACT TABLE option used to glue the rigid
bodies drive1 and drive2 onto gear1 and gear2, respectively, while making rigid bodies drive1out and drive2out
noncontacting. Then we shall submit the results and check the results as they are generated.
FILES
OPEN
gearpair.mud
OK
MAIN
CONTACT
CONTACT TABLES
PROPERTIES
MAIN
JOBS
RUN
SUBMIT
OPEN POST FILE (RESULTS MENU)
DEF ONLY
SKIP TO INC 57
SCALAR (Equivalent von Mises Stress)
CONTOUR BANDS
1580 Marc Users Guide: Part 2
CHAPTER 3.36
As expected, the gears become engaged and deIorm as shown in Figure 3.36-4. The noncontacting rigid bodies,
drive1out and drive2out, are shown as green lines representing rigid gear motion making tooth deIormation easy
to visualize.
Figure 3.36-4 Contour Equivalent von Mises Stress at Increment 57
Another important plot is the torque versus twist which is generated by using the history plot Ieature as:
The Iirst load case brings the gears into contact at the end oI increment 1 and this is seen here. Using the copy to
clipboard, the history data can be exported to Excel and the data manipulated and compared to experimental results as
see in Figure 3.36-7.
Experimental Test Machine
A parallel axis gear-testing machine developed by Ticona (www.ticona.com) was used to load and record the load-
displacement response oI the gears (Figure 3.36-5).
HISTORY PLOT
COLLECT GLOBAL DATA
NODES/VARABLES
ADD GLOBAL CURVE
Angle Pos drive1
Moment Z drive1
1
-53.49
401.86
857.21
1312.56
1767.91
2223.26
2678.60
3133.95
3589.30
4044.65
4500.00
lcase2
Equivalent Von Mises Stress
Inc: 57
Time: 2.467e+000
X
Y
Z
0
Moment Z drive1 (x100)
Angle Pos drive1 (x.01)
job1
130
120
110
8.333
100
80
70
60
50
40
30
20
90
-1.048
0
10
0
1581
CHAPTER 3.36
Plastic Spur Gear Pair Failure
Figure 3.36-5 Parallel axis gear-testing machine
The test gears are lubricated with oil prior to loading to eliminate any shearing Iorces acting on the tooth Ilanks that
are in contact. Torque is measured on the stationary side and load is applied on the motor side. Two high precision
encoders are used to measure the angular displacement oI both gears. These encoders have a positional accuracy oI
57600 counts per revolution. The rate oI loading is set by the time Ior encoder position on the motor side. The
stationary is not totally rigid. It requires some angular displacement Ior the torque meter to record data. To obtain the
true angular displacement, the relative displacement between both gears is recorded. This gives a rate Ior the relative
angular displacement between the motor gear and stationary gear to be about 0.002 radians per minute. Five tests are
made per gear set at ambient conditions. A plot oI applied torques verses relative displacement is recorded. The results
are shown in Figure 3.36-6. Test 2 and Test 4 did not reach tooth Iailure. This is due to that Test 4 was not taken up to
the breaking torque and Test 2 reached the set limited encoder position beIore breaking.
Figure 3.36-6 Plot of Experimental Results
0
20
40
60
80
100
120
140
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
T
o
r
q
u
e

(
in
-
lb
)
Radians
Static Mesh Bending on Acetal Copolymer Gears
Test 1
Test 2
Test 3
Test 4
Test 5
1582 Marc Users Guide: Part 2
CHAPTER 3.36
Results & Conclusions
A plot oI applied torque verses twist is made and gives excellent representation oI the experimental results
(Figure 3.36-7). At the beginning, a two teeth pair (on each gear) come into contact, then as these teeth bend, the tooth
leading this pair begins to come into contact (Figure 3.36-7 Inc: 79). Later (Figure 3.36-7 Inc: 112) there are Iour teeth
on each gear in contact with their counterparts. At increment 112, the Iirst element is deactivated (leading tooth on top
moving gear) Iollowed by several more shown in increment 128. AIter increment 128, elements begin to Iail in the
stationary gear and the torque drops oII dramatically. Based on the results oI this analysis, the mechanical behavior
and prediction oI copolymer acetal gears is very complex. The results indicate that to optimize a gear set, a nonlinear
analysis is required to be perIormed. Only under low loads and deIormation can a linear elastic approach be suitable.
Clearly, combining computer simulations with material and component testing has led to a Iar better understanding oI
copolymer acetal gear design; this understanding could not be achieved by either simulation or testing alone. It is
envisioned that with a Iew more material tests, the torque-displacement response oI the gear pair can be simulated with
conIidence thus advancing the technology oI copolymer acetal gear applications.
Figure 3.36-7 Predictions versus Experimental Results of Torque Versus Twist of the Gear Pair
0.00 0.02 0.04 0.06 0.08 0.10
0
20
40
60
80
100
120
Twist (Radians)
T
o
r
q
u
e

(
i
n
-
l
b
f
)
-53.49
401.86
857.21
1312.56
1767.91
2223.26
2678.60
3133.95
3589.30
4044.65
4500.00
Equivalent Von Mises Stress
Inc: 57
Inc: 79
Inc: 112
Inc: 128
Inc: 140
Experimental
Prediction
1
0
1
1
1
1
1583
CHAPTER 3.36
Plastic Spur Gear Pair Failure
Modeling Tips
The material used was Celcon grade M90 (Toughened; Impact ModiIied) which is the red curve taken Irom
Reference 2, Figure 3.1 duplicated below.
It was assumed that this stress strain data was in engineering measures oI stress and strain (s, e) and they needed to be
converted to true values, ( ) where the Cauchy stress becomes, and the true strain becomes,
. The work hardening plot (Figure 3.36-2) then becomes the Cauchy stress versus the total plastic
strain, .
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
References
1. American National Standard/AGMA Standard, Tooth Proportions Ior Plastic Gears, ANSI/AGMA 1006-A97,
1997.
2. Designing with Celcon http://www.kmsbearings.com/pdI/CelconDesign20Guide3.9.07.pdI
File Description
gearpair.mud Mentat model Iile
gearpair_job1.dat Marc input Iile
gearpair.f User subroutine to deIine invoke Iailure criterion
20
40
60
80
100
120
Strain, %
S
t
r
e
s
s
,

M
P
a
0
0 2 4 6 8 10 12 14
25% Glass Coupled
Unflled, 9.0 Melt Flow
Toughened; Impact Modifed
Fig 3.1 Celcon acetal copolymer stress-strain
properties (ISO 527)
o c , o s 1 e ( )
c 1 e ( ) ln
c
p
c o E
1584 Marc Users Guide: Part 2
CHAPTER 3.36
Animation
Click on the Iigure below to play the animation.
Chapter 3.37: Girkmann Verification Problem
3.37
Girkmann Verification Problem

Summary 1586

Detailed Description 1587

Results 1589

Modeling Tips 1591

Input Files 1596

1586 Marc Users Guide: Part 2
CHAPTER 3.37
Summary
Title Girkmann VeriIication Problem
Problem Ieatures The Girkmann problem is a numerical veriIication exercise in solid mechanics proposed
by Juhani Pitkranta, Ivo Babuka, and Barna Szab in 2008. The importance oI
veriIication is selI-evident.
Geometry
Material properties E 2.059x10
10
N/m
2
, 0.0, 3335.71 kg/m
3

Analysis type Static with elastic material behavior
Boundary conditions Axial displacement at centerline 0, pressure and gravity loads indicated
Element type Axisymmetric shell and solid elements are connected with displacement and slope
continuity.
FE results Bending moment and shear Iorce at
shell-ring interIace; meridional angle at
maximum bending moment. The
authors invited their readers to solve this
problem and veriIy the results are
accurate to within 5 percent. The results
herein are within 0.50 percent.
Z
X
Y
p = 27,283.1 N/m
2
C
L
sh
e
ll
=
3
3
3
5
.7
1
k
g
/
m3

ring

R
c
M
M

R
m
= R
c
/sin()
h
A B a
b
R
c
h = 0.06 m
R
c
= 15.00 m
= 2/9
a = 0.60 m
b = 0.50 m
v
p = 27,283.1 N/m
2
C
L
= 3335.71
kg
/m3
10 20 30 40
-50
0
50
100
150
200
250
300
Bending Moment (Nm/m)
Meridional Angle (degrees)
38.137
o
1587
CHAPTER 3.37
Girkmann Verification Problem
The Girkmann problem consists oI a spherical shell connected to a stiIIening ring at the crown radius. The objective
oI the analysis is to accurately estimate: a) the shear Iorce and bending moment acting at the junction between the
spherical shell and the stiIIening ring; b) determine the location (meridional angle) and the magnitude oI the maximum
bending moment in the shell. The model problem was Iirst discussed by Girkmann in 1956, subsequently by
Timoshenko and Woinowski-Krieger in 1959. The results are compared to the solutions by the classical methods to
demonstrate the accuracy.
Detailed Description
Element type 1, an axisymmetric, straight, thick-shell element is used Ior modeling the spherical shell and element
type 10, an axisymmetric, Iour-node, quadrilateral element is used to model the stiIIening ring. The geometry Ior the
Girkmann problem is shown in Figure 3.37-1. The x axis is the axis oI rotational symmetry. A spherical shell oI
thickness h and midsurIace radius Rm, is connected to a stiIIening ring at the meridional angle and a crown radius
oI Rc. The dimensions oI the ring are a and b.
Figure 3.37-1 The Girkmann Problem Geometry
A close-up oI the shell-ring intersection Ior the Girkmann problem is shown in Figure 3.37-2. The mesh consists oI
2208 elements and 2270 nodes.
Figure 3.37-2 The Girkmann Shell - Ring Close-up
The axisymmetric solid elements Ior the stiIIening ring are generated by *add_elements and re-meshed with
*subdivide_elements (Figure 3.37-3). The axisymmetric shell elements Ior the spherical shell are generated using
*expand_nodes.
Z
X
Y
p = 27,283.1 N/m
2
C
L
ring
sh
e
ll
=
3
3
3
5
.7
1
k
g
/m3

R
c
M
M

R
m
= R
c
/sin()
h
A B a
b
R
c
h = 0.06 m
R
c
= 15.00 m
= 2/9
a = 0.60 m
b = 0.50 m
1588 Marc Users Guide: Part 2
CHAPTER 3.37
Figure 3.37-3 Building the Ring using Subdivide
A local Cartesian coordinate system (*new_coord_system) is created with the shell solid intersection node as origin,
the local X axis along the 40
0
inclined edge oI the ring and the local Y axis normal to that. All the nodes on the 40
0

inclined edge oI the ring and the end node oI the shell at the intersection are transIormed into this co-ordinate system
(Figure 3.37-4).
Figure 3.37-4 Coordinate Transformation (colored arrows) for Joining the Shell and Ring
Servo links constrain the translation and rotation displacements oI the end node oI the shell joining the inclined edge
oI the ring. The local Y displacement oI the nodes on the ring edge is the sum (with appropriate sign) oI the local Y
displacement oI the end node oI the shell at the intersection and Z rotation times the distance oI that node Irom the end
node oI the shell (see Figure 3.37-8). The local X and local Y displacement oI the coincident nodes oI solid and shell
at the intersection are constrained to be equal.
The material Ior all elements is linear elastic, isotropic with Young`s modulus oI 2.059e10 N/m
2
and density oI
3335.71 Kg/m
3
. Pressure oI 27283.14706 N/m
2
is applied is applied to the bottom Iace (leIt) oI the ring as an edge
load (Figure 3.37-5). An acceleration oI -9.81 m/s
2
is applied in the X direction (although not necessary the Y and Z
components oI acceleration is set to zero as well) only to the shell elements, whose mass times this acceleration
Z
X
Y
L
o
c
a
l

Y

D
i
r
e
c
t
i
o
n

Z
X
Y
L
o
c
a
l

X

D
i
r
e
c
t
i
o
n
1589
CHAPTER 3.37
Girkmann Verification Problem
determines the weight or gravity load oI the shell structure. The stiIIening ring is assumed to be weightless. The
displacement oI the node on the axis oI symmetry is constrained in the axial direction.
Figure 3.37-5 Loads and Boundary Conditions
By design, the axial (vertical in Figure 3.37-5) Iorce on the ring equilibrates the weight oI the shell.
Results
The internal Iorces Irom the Iorce balance Iile (girkmann_job1.grd) are listed below Ior the node (2270) on shell at the
intersection.
The bending moment at the shell-ring interIace becomes, .
The axial Iorce at the shell-ring interIace becomes, .
The radial Iorce at the shell-ring interIace becomes, .
The shear stress at the shell-ring interIace (Figure 3.37-6) is -15658.2 N/m
2
, and when multiplied by the thickness
gives a shear Iorce oI -939.492 N/m.
p = 27,283.1 N/m
2
C
L
shell
=
3
3
3
5
.7
1
k
g
/m3

R
c
apply1 -> Axial Disp = 0
apply2 -> Pressure Load
apply3 -> Gravity Load
Z
X
Y
node 2270 internal force from element 2208 -0.1571E+07 0.1735E+07 0.0000E+00 -0.3475E+04 0.0000E+00 0.0000E+00
node 2270 externally applied forces -0.4710E+03 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
node 2270 tying/mpc forces 0.1572E+07 -0.1735E+07 0.0000E+00 0.3475E+04 0.0000E+00 0.0000E+00
node 2270 reaction - residual forces 0.6319E-05 -0.7750E-05 0.0000E+00 0.1676E-07 0.0000E+00 0.0000E+00
M
o
3575
tD
----------- -
3575
t 30 ( )
-------------- 36.871 Nm/m
Q
o
1571000
tD
------------------------ 16668.828 N/m
Q
r
1735000
tD
--------------------- 18408.922 N/m
1590 Marc Users Guide: Part 2
CHAPTER 3.37
Figure 3.37-6 Shear Stress (Component 12) at the Shell-Ring Interface
The bending moment is estimated Irom the shell stresses as Iollows:
Using the above, the bending moment can be plotted versus the meridional angle as shown in. The maximum bending
moment is 255.103 Nm/m at a meridional angle oI 38.137
o
.
Figure 3.37-7 Bending Moment versus Meridional Angle
80888.5
71077
62333.3
63572.1
61789.1
43299.7
-32766.4
31703.8
13389.1
29697.9
42517.6
83265.8
84066.9
77680
93494.7
73937.8
66222.7
70742.5
140071
-15236.7
-15404.7
83583.3
-15573.5
-15658.2
-3.277e+004
7.668e+005
job1
Comp 12 of Stress Layer 1
Inc: 0
Time: 0.000e+000
Z
X
Y
o
B
Comp 11 oI Stress at Layer 1 - Comp 11 oI Stress at Layer 5 ( ) 2
M
B
o
B
bd
2
6 ( ) o
B
1 0.06 ( )
2
6 ( )
10 20 30 40
-50
0
50
100
150
200
250
300
Bending Moment (Nm/m)
Meridional Angle (degrees)
38.137
o
1591
CHAPTER 3.37
Girkmann Verification Problem
The results are summarized below and compared to the reIerence values.
(1) The Problem oI VeriIication with ReIerence to the Girkmann Problem by Barna Szab, Ivo Babuka, Juhani
Pitkranta, and Sebastian Nervi. The Institute Ior Computational Engineering and Sciences Report 09-17, 2009. See
www.ices.utexas.edu/research/reports/2009/0917.pdf.
Editorial Comment: The above reIerence is well worth reading; the authors received 15 solutions and among their
comments the Iollowing is worth repeating, namely: 'Another respondent wrote: 'Regarding verification tasks for
structural analysis software that has adequate quality for use in our safety critical profession of structural
engineering, the solution of problems such as the Girkmann problem represents a minuscule fraction of what is
necessary to assure quality. We |the authors| agree with this statement. That is why we Iind it very surprising that the
answers received had such a large dispersion. For example, the reported values oI the moment at the shell-ring
interIace ranged between -205 and 17977 Nm/m. Solution oI the Girkmann problem should be a very short exercise
to persons having expertise in FEA, yet many oI the answers were wildly oII.
Modeling Tips
To review the model, read in the Marc input Iile girkmann.dat into Mentat. All oI the modeling inIormation will be
present. Axisymmetric models in Marc use the global x axis as the axis oI rotation. The meshing is relatively straight
Iorward and is not repeated here. However, an important Ieature in this model are the transIormations and constraints
between the end shell node (n:2770) where it joins the inclined plane oI the ring (n:11). To review the transIormations
and constraints let`s read in the input Iile and go to modeling tools.
Result Marc Reference (1) % Error
Bending Moment (Nm/m) 36.871 36.81 0.17
Axial Force (N/m) -16668.828 -16700 -0.19
Radial Force (N/m) 18408.922 18400 0.05
Shear Force (N/m) -939.492 -943.6 -0.44
Max. Bending Moment in the shell (Nm/m) 255.103 253.97 0.45
Meridional Angle oI Max. BM (degree) 38.137 38.08 0.15
FILES
MARC INPUT FILE
READ
girkmann.dat, OK
SAVE AS
girkmann, OK
FILL
MAIN
1592 Marc Users Guide: Part 2
CHAPTER 3.37
We see that servo link 1 in Figure 3.37-8, constrains the ring node 110 to have its second degree oI Ireedom related to
the second (translation normal to incline) and third (rotation) oI the end shell node 2770 by the moment arm oI length
0.03m. This is repeated 15 more times Ior all nodes along the ring incline edge. Servo link 17 and 18, simply equate
the Iirst and second degrees oI Ireedom to the coincident nodes 2270 oI the shell and 11 oI the ring. These servo links
are automatically generated with the N to 1 SERVOS button, where you need only select the proper nodes and all oI
the coeIIicients (aka moment arms) are computed automatically by Mentat.
Furthermore, since the shell elements have three degrees oI Ireedom per node, while the ring elements have only two
degrees oI Ireedom per node, node 11 and node 2270 should never be the same node number, but constrained together
as shown here. Also since the nodes are separate, the results are not nodal averaged across the shell and solid
axisymmetric elements. Finally, getting this step wrong gives incorrect results that may not be obvious.
Figure 3.37-8 Servo Link 1
MODELING TOOLS
TRANSFORMATONS
TRANSFORMATION PLOT SETTINGS
TRANSFORMATIONS (turn on)
DRAW (should look like Figure 3.37-4)
TRANSFORMATIONS (turn off)
DRAW
MAIN
LINKS
SERVO LINKS (see Figure 3.37-8)
MAIN
98
88
77
66
55
44
33
22
11
2269
99
2268
108 110
2265
2264
2263
2262
187
176
165
154
143
132
121
109
2267
2266
2265
77
66
55
44
33
22
11
88
98
99
108
2264
2263
2262
187
176
2266
165
143
132
121
110 109
154
2267
2269
2270
2268
Z
X
Y
Servo 1: dof 2 n:110 =1*dof 2 n: 2270 + 0.03*dof 3 n:2270
Z, dof 3
L
o
c
a
l
X
,

d
o
f

1 L
o
c
a
l
Y
,

d
o
f

2
0.03
1593
CHAPTER 3.37
Girkmann Verification Problem
You may wish to run the model; to do so simply go to Jobs, run and submit the simulation, Ior example
AIter the simulation completes, let`s examine how we can produce check the validity oI the servo links, the bending
moment versus meridional angle shown in Figure 3.37-7, and some other ways to help visualize the results.
The servo links must keep the angle (a right angle in this case) between the shell and ring edge the same beIore and
aIter deIormation. Since the deIormations are very small, the scaling oI the deIormed shape was set to automatic and
the magniIication Iactor is over 400 in Figure 3.37-9.
Figure 3.37-9 Shell-Ring Edge Originally Perpendicular must remain Perpendicular - Displacements
Automatically Magnified over 400 times.
The strategy to computing the bending moment in the shell is simple; we just collect bending stress along a path Irom
the centerline to the end shell node.
JOBS
RUN
SUBMIT
OPEN POST FILE (RESULTS MENU) (opens results and jumps to results menu)
DEFORMED SHAPE SETTINGS
AUTOMATIC (turn on)
RETURN
DEF & ORIG (should look like Figure 3.37-9)
PATH PLOT
NODE PATH
n:803 n:2270 #
ADD CURVES
ADD CURVE
Arc Length
Inc: 0
Time: 0.000e+000
Z
X
Y
1594 Marc Users Guide: Part 2
CHAPTER 3.37
The xy data is now in the clipboard and can be exported to MicrosoIt Excel Ior additional processing to compute the
bending moment Irom the bending stresses at the top and bottom layers oI the shell element that was used to produce
the plot in Figure 3.37-7.
Figure 3.37-10 Bending Stress of Top and Bottom Shell Layers along Arc Length of Shell Elements from
Center Line to Shell-ring Intersection
Also we can use the expand Ieature to expand (rotated 40
o
) our results about the axis oI rotation.
Comp 11 of Stress Layer 1
Arc Length
Comp 11 of Stress Layer 5
SHOW ID 100
FIT (should look like Figure 3.37-10)
RETURN
CLIPBOARD COPY TO
job1
Y (x1e5)
1030 1130 12301330
1430
1530 1630
1730
1830
1930
2130
2230
Comp 11 of Stress Layer 1
-0.025
-8.529
1.629 0
Time : 0
Inc : 0
Arc Length (x10)
930 830
2030
2230
Comp 11 of Stress Layer 5
830 930 1030 1130 123013301430
1530 1630
1730
1830
1930
2030
2130
1
1595
CHAPTER 3.37
Girkmann Verification Problem
Figure 3.37-11 Axisymmetric Shell Element Results Expanded about the Rotational Symmetry Axis
Finally we can visualize the shell-ring intersection by closing the post Iile and adjusting the plot settings as Iollows:
Figure 3.37-12 Expanded Shell Plot Showing the Shells Thickness at the Shell-Ring Intersection
Hence, we can see that the thickness oI the shell is identical to the length oI the inclined ring edge that has all oI the
servo links illustrated in Figure 3.37-8.
MAIN
RESULTS
CLOSE
MAIN
GEOMETRIC PROPERTIES
PLOT SETTINGS SHELL
PLOT EXPANDED
DEFAULT THICKNESS = 0.06
DRAW (should look like Figure 3.37-12)
-4.712e+005
-4.243e+005
-3.775e+005
-3.306e+005
-2.837e+005
-2.369e+005
-1.900e+005
-1.431e+005
-9.625e+004
-4.939e+004
-2.520e+003
job1
Comp 11 of Stress Layer 1
Inc: 0
Time: 0.000e+000
Z
X
Y
4
Z
X
Y
Ring
Servo Links
Shell
0
.0
6
1596 Marc Users Guide: Part 2
CHAPTER 3.37
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
girkmann.dat Marc input Iile to run the above problem
Section 4: Heat TransIer Analysis
Section 4: Heat Transfer
Analysis
Marc Users Guide -1598
Chapter 4.1: Thermal Contact Analysis of a Pipe
4.1
Thermal Contact Analysis of
a Pipe

Chapter Overview 1600

Pipe in a House 1600

Input Files 1612

1600 Marc Users Guide: Part 3
CHAPTER 4.1
Chapter Overview
This chapter describes the use oI thermal contact in Marc. In previous versions oI Marc, using contact in a thermal
analysis was only possible by doing a coupled mechanical thermal analysis, which led to extra overhead, both in
solution time and memory use. The introduction oI thermal contact provides the ability to take thermal conduction and
small gaps into account. Bodies which are almost touching each other are considered to be in near contact. DiIIerent
physical heat transIer processes can be simulated Ior this type oI contact such as convection, natural convection,
radiation, or distance dependent heat transIer. Near contact can be used in thermal or coupled analyses.
Pipe in a House
The example described here is a pipe which is heated Irom the inside. The pipe runs through a house, where the
dimensions are chosen so that pipe and house initially do not touch. Figure 4.1-1 illustrates this pipe and the house.
When the pipe heats up, it expands and, at a certain moment, it comes in contact with the house. To analyze this, a
coupled thermal mechanical analysis is perIormed. When the pipe is almost touching the house, a distance dependent
heat transIer is considered. Due to symmetry, an axisymmetric analysis is perIormed.
Figure 4.1-1 3-D Model of the Example
Mesh Generation
The mesh is generated by Iirst deIining two elements representing the pipe and the house, where a gap oI 0.5 mm exists
between these two elements. Then these two elements are reIined. The outer diameter oI the pipe is 0.409 m with a
thickness oI 0.01 m. The house has an outer diameter oI 0.63 m and a thickness oI 0.11 m.
MESH GENERATION
NODES ADD
0.1 0.1945 0
0.5 0.1945 0
0.5 0.2045 0
0.1 0.2045 0
0.25 0.205 0
pipe
house
1601
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
Boundary Conditions
The x-displacement is set to zero Ior a node Irom the pipe and a node Irom the house to prevent the rigid body mode.
The temperature is prescribed as a Iunction oI time on the inside oI the pipe, where it is Iirst increased Irom 25C to
225C, and then held constant.
0.35 0.205 0
0.35 0.315 0
0.25 0.315 0
ELEMENTS ADD
1 2 3 4
5 6 7 8
SUBDIVIDE
DIVISIONS
16 3 1
ELEMENTS
1 #
DIVISIONS
6 6 1
ELEMENTS
2 #
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
OK
NODES ADD
1 8 #
1602 Marc Users Guide: Part 3
CHAPTER 4.1
Initial Conditions
Initially, the pipe and house are at room temperature (25C).
RETURN
NEW
THERMAL
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
ADD
0 298
1000 498
5000 498
FIT
RETURN
FIXED TEMPERATURE
TEMPERATURE (TOP)
TABLE
table1
OK
OK
NODES ADD
1 2 11 15 19 23 27 31 35 39 43 47 51 55 59 63 67 #
RETURN (twice)
INITIAL CONDITIONS
THERMAL
TEMPERATURE
TEMPERATURE (TOP)
298
OK
NODES ADD
ALL EXIST
RETURN (twice)
1603
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
Material Properties
The pipe is isotropic and made oI aluminium, the house is also isotropic and made oI copper. The material properties
are listed in Table 4.1-1.

MATERIAL PROPERTIES
NAME
aluminium
ISOTROPIC
YOUNGS MODULUS
7.1e10
Table 4.1-1 Material Properties
Aluminium Copper
Young`s modulus (GPa) 71 124
Poisson`s ratio 0.3 0.3
Thermal Expansion CoeIIicient (K
-1
) 23 x 10
-6
16.8 x 10
-6
Conductivity (W/m/K) 237 390
SpeciIic Heat (J/kg/K) 880 387
Mass Density (kg/m
3
)
0
2700
0
8960
POISSONS RATIO
0.3
MASS DENSIY
2700
THERMAL EXP
THERMAL EXP COEF
2.3e-5
OK (twice)
HEAT TRANSFER
CONDUCTIVITY
237
SPECIFIC HEAT
880
MASS DENSITY
2700
1604 Marc Users Guide: Part 3
CHAPTER 4.1
The material properties Ior the house are added in a similar way.
Contact
The pipe and the house are deIined as separate contact bodies. The contact properties are set in the CONTACT TABLE
option, where the near contact distance is 0.3 mm, the contact heat transIer coeIIicient is 100 W/m
2
, and the distance
dependent heat transIer coeIIicient is 1 W/m
2
. The menu Ior entering the contact table entry properties including the
near contact distance is shown in Figure 4.1-2.
Figure 4.1-2 Menu for entering Contact Table Entry Properties
OK
ELEMENTS
49 to 84 #
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE
OK
ELEMENTS
49 to 84 #
1605
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
Loadcases and Job Parameters
A quasi-static analysis is perIormed, where two loadcases are deIined. In the Iirst loadcase, the temperature inside the
pipe is increased 200 K in 100 increments during 1000 seconds. In the second loadcase, the temperature is Iixed Ior
250 increments during 10000 seconds. The contact bias Iactor is set to 0.9 to provide a more accurate contact
description.
NEW
DEFORMABLE
OK
ELEMENTS
1 to 48 #
RETURN
CONTACT TABLES
NEW
PROPERTIES
1 2
CONTACT TYPE: TOUCHING
CONTACT DETECTION METHOD: FIRST->SECOND
NEAR CONTACT
DISTANCE
0.0003
THERMAL PROPERTIES
CONTACT HEAT TRANSFER COEFFICIENT
LOADCASES
COUPLED
NAME
ramped_temp_nc
QUASI-STATIC
CONTACT
CONTACT TABLE
ctable
OK
CONVERGENCE TESTING
DISPLACEMENTS
OK
1606 Marc Users Guide: Part 3
CHAPTER 4.1
TOTAL LOADCASE TIME
1000
PARAMETERS
#STEPS
100
OK (twice)
COPY
NAME
fixed_temp_nc
QUASI-STATIC
TOTAL LOADCASE TIME
10000
PARAMETERS
#STEPS
250
OK (twice)
RETURN (twice)
JOBS
ELEMENT TYPES
COUPLED
AXISYM SOLID
10
OK
ALL EXIST
RETURN (twice)
COUPLED
ramped_temp_nc
fixed_temp_nc
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
OK
ADVANCED CONTACT CONTROL
DISTANCE TOLERANCE BIAS
1607
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened. A node on the pipe and a node on the
house are selected to generate plots oI the temperature as a Iunction oI time. These are shown as the (black)
and (red) curves in Figure 4.1-3. Plots oI the y-displacement as a Iunction oI time Ior these nodes are shown
as the (black) and (red) curves in Figure 4.1-4.
0.9
OK (thrice)
FILE
SAVE AS
heatpipe.mud
OK
RETURN
RUN
SUBMIT(1)
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
SET NODES
5 34 #
COLLECT GLOBAL DATA
1608 Marc Users Guide: Part 3
CHAPTER 4.1
Figure 4.1-3 Temperature as a Function of Time for two Nodes in Contact,
one from the Pipe and one from the House
Note: Results are with the Near Contact option switched on and oII.
NODES/VARIABLES
ADD VARIABLE
Time
Temperature
RETURN
SHOW IDS
O
FIT
COPY TO (GENERALIZED XY PLOT)
GENERALIZED XY PLOT
SHOW IDS
0
FIT
Time (x10000)
node on house near contact
node on house contact
node on pipe near contact
node on pipe contact
Temperature (x100)
1609
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
Figure 4.1-4 Displacement in the y-Direction as a Function of Time for two Nodes
in Contact, one from the Pipe and one from the House
Next, the near contact option is switched oII to illustrate what the results look like without this option.
Note: Results are with the Near Contact option switched on and oII.
RETURN
CLOSE
RETURN (twice)
CONTACT
CONTACT TABLES
COPY
PROPERTIES
1 2
NEAR CONTACT
OK (twice)
RETURN (twice)
LOADCASES
EDIT
ramped_temp_nc
Time (x10000)
node on house near contact
node on house contact
node on pipe near contact
node on pipe contact
Displacement Y (0.0001)
1610 Marc Users Guide: Part 3
CHAPTER 4.1
OK
COPY
NAME
ramped_temp
COUPLED
QUASI-STATIC
CONTACT
CONTACT TABLE
ctable2
OK (twice)
EDIT
fixed_temp_nc
COPY
NAME
fixed_temp
QUASI-STATIC
CONTACT
CONTACT TABLE
ctable2
OK (twice)
RETURN (twice)
JOBS
NEW
COUPLED
ramped_temp
fixed_temp
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable2
OK
ADVANCED CONTACT CONTROL
DISTANCE TOLERANCE BIAS
0.9
OK (thrice)
1611
CHAPTER 4.1
Thermal Contact Analysis of a Pipe
The (green) and (blue) curves in Figure 4.1-3 show the temperature as a Iunction oI time Ior a node on
the house and a node on the pipe, respectively, Ior the analysis where the near contact option is switched oII. The
(green) and (blue) curves in Figure 4.1-4 shows the y-displacement as a Iunction oI time Ior the two
nodes. When the results where near contact is used and not used are compared, the diIIerence is clear. The pipe is
behaving similar in both cases, but the house has a much smoother temperature response Ior the case where near
contact is used. A similar eIIect is observed Ior the y-displacement, which is smoother Ior the case using near contact.
Note that Ior both cases, contact is only temporarily, once heat transIer develops between the pipe and the house, the
house will expand more due to a larger diameter. This can be seen in Figure 4.1-5, where the contact status oI a node
in contact is plotted as a Iunction oI time Ior the two jobs.
Figure 4.1-5 Contact Status with and without the Near Contact Option Activated
RUN
SAVE MODEL
SUBMIT
contact status
without near contact with near contact
time (x10000)
1612 Marc Users Guide: Part 3
CHAPTER 4.1
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
pipe.proc Mentat procedure Iile to run the above example
Chapter 4.2: Dynamics with Friction Heating
4.2
Dynamics with Friction Heating

Summary 1614

Friction Heat Analysis 1616

Run Jobs and View Results 1623

Input Files 1626

1614 Marc Users Guide: Part 3
CHAPTER 4.2
Summary
Title Dynamics with Iriction heating
Problem Ieatures dynamic coupled mechanical thermal analysis.
Geometry
Material properties E 2.1x10
11
Pa, 0.3, 0.5
k 60.5 W/m/K, C
p
434 J/kg/K, 7854 kg/m
3
Analysis type Coupled Mechanical Heat TransIer
Boundary conditions Y component oI disp. Iixed all nodes. Gravity load a
z
-9.81 m/s
2
.
Initial conditions V
x
4.905 m/s, T 0 K all nodes
Element type Brick element type 7
FE results 1. Block velocity and acceleration history
2. Rise in temperature due to Iriction
X Y
Z
2x2x1 Block
Initial V
x
= 4.905 m/s
8x4 Rigid
Units: m, kg, s, K
v

0
10
20
30
40
50
60
70
80
90
100
-6
-4
-2
0
2
4
6
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Vx (m/s)
Vx (m/s)
Ax (m/s
2
) Ax (m/s
2
)
T (K)
T (K)
1615
CHAPTER 4.2
Dynamics with Friction Heating
A dynamic coupled analysis is perIormed to simulate the behavior oI a block with an initial velocity sliding over a
rigid table. Due to the weight oI the block and Iriction between the block and the table, the block will slow down and
heat up because oI Iriction.
Figure 4.2-1 Problem Description
Mechanical boundary conditions keep the block moving in a straight line. Initial conditions set the initial velocity and
temperature.
The coupled loadcase selected is a dynamic transient with a time period long enough to allow the block to come to rest.
The temperature contours show how the leading edge oI the block touching the table heat up Iaster than other portions
oI the block.
A history plot oI the velocity and acceleration oI the node (Figure 4.2-2) show how the block comes to a stop with the
velocity and acceleration becoming zero at 1.4 seconds.
Figure 4.2-2 Velocity and Acceleration History
icond1
icond2
X Y
Z
4
6.840e-007
8.826e+000
1.765e+001
2.648e+001
3.530e+001
4.413e+001
5.296e+001
6.178e+001
7.061e+001
7.943e+001
8.826e+001
lcase1
Temperature
Inc: 50
Time: 2.000e+000
X Y
Z
4
39 38 37 36 35 34 33 32 31 30 29 40414243
2 0
Y
Time
job1
28
0
49 48 46 45 44 50
-4.815
27
25
8
7
6
5
4
3
2
1
0
9
10
11
12
24
23
22
21
20
26
19
17
16
15
14
13
18
4.905
Velocity X Node 9
47484950 0 35 34 33 32 31
30
29
28
27
36
26
3739 47 46 45 44 43 42 41 40 38
Acceleration X Node 9
25
23
9 8 7 6 5 4
3
2
1
10
24
1113
22 21 20 19 18 17 16 15 14 12
1616 Marc Users Guide: Part 3
CHAPTER 4.2
Friction Heat Analysis
This is a problem oI a block subjected to its own weight that is sliding on a table with an initial velocity. Friction
between the block and table generate heat and reduce the speed. The steel block has an area oI 4 m
2
and a height oI
1 m. The coeIIicient oI Iriction is 0.5.
Figure 4.2-3 Initial Conditions
FILES
NEW
OK
SAVE AS
block
RETURN
MESH GENERATION
VIEW SHOW VIEW 4
OK
ADD ELEMENTS
node(-1.0, -1.0, 0.0)
node( 1.0, -1.0, 0.0)
node( 1.0, 1.0, 0.0)
node(-1.0, 1.0, 0.0)
ADD SURFACES
point( 1.0, -1.0, 0.0)
point(-1.0, -1.0, 0.0)
point(-1.0, 1.0, 0.0)
point( 1.0, 1.0, 0.0)
icond1
icond2
X Y
Z
1617
CHAPTER 4.2
Dynamics with Friction Heating
Figure 4.2-4 Footprint of Block on Table
MOVE
SCALE
4 2 1
SURFACES
ALL: EXISTING
MOVE
RESET
TRANSLATIONS
1.8 0 0
SURFACES
ALL: EXISTING
RETURN
SUBDIVIDE
ELEMENTS
ALL: EXISTING
RETURN
EXPAND
TRANSLATIONS
0 0 1/2
1618 Marc Users Guide: Part 3
CHAPTER 4.2
Figure 4.2-5 Block Mesh
REPETITIONS
2
ELEMENTS
ALL: EXISTING
RETURN
FILL
SWEEP
REMOVE UNUSED
NODES
ALL
RETURN
RENUMBER
ALL
RETURN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISP Y
0
1619
CHAPTER 4.2
Dynamics with Friction Heating
Figure 4.2-6 Boundary Conditions
OK
NODES ADD
ALL: EXISTING
NEW
GRAVITY LOAD
ON Z ACCEL
-9.81(m/s
2
)
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
1620 Marc Users Guide: Part 3
CHAPTER 4.2
Figure 4.2-7 Initial Conditions
INITIAL CONDITIONS
THERMAL
TEMP.
0 (K)
OK
NODES ADD
ALL: EXISTING
RETURN
NEW
MECHANICAL
VELOCITY
VEL X
4.905 (m/s)
OK
NODES ADD
ALL: EXISTING
MAIN
MATERIAL PROP. (twice)
ANAYSIS CLASS
COUPLED
1621
CHAPTER 4.2
Dynamics with Friction Heating
Figure 4.2-8 Isotropic Properties Submenu using Damping
NEW
STANDARD
STRUCTURAL
E = 210E9 (N/m2)
= .3 (return)
GENERAL
= 7854 (Kg/m3)(return)
THERMAL
CONDUCTIVITY
60.5 (W/m/K)
SPECIFIC HEAT
434 (J/Kg/K)
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE
= .5
OK
ELEMENTS ADD
ALL: EXISTING
u

1622 Marc Users Guide: Part 3

CHAPTER 4.2
Figure 4.2-9 Contact Bodies: Block and Table
CONTACT
CONTACT BODIES
NEW
RIGID
= .5
OK
SURFACES ADD
ALL: EXISTING
MAIN
LOADCASES
COUPLED
CONVERGENCE TESTING
DISPLACEMENTS (OK)
DYNAMIC TRANSIENT
TOTAL LOADCASE TIME
2
FIXED # STEPS
50
OK
MAIN

cbody1
cbody2
X Y
Z
4
1623
CHAPTER 4.2
Dynamics with Friction Heating
Run Jobs and View Results
JOBS
NEW
COUPLED
PROPERTIES
lcase1
ANALYSIS OPTIONS
LARGE STRAIN
LUMPED MASS
OK
CONTACT CONTROL
ARCTANGENT
RELATIVE SLIDING VEL
0.1
ADVANCED CONTACT CONTROL
SEP. FORCE
1E11 (keep block on surface)
OK (twice)
JOB RESULTS
EQUIVALENT VM STRESS
TEMPERATURE (Integration Point)
OK
JOB PARAMETERS
HEAT GENERATION (FRICTIONAL)
1E3 (should be 1, but want larger temps for show)
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR
OK
RETURN
1624 Marc Users Guide: Part 3
CHAPTER 4.2
Figure 4.2-10 Temperature Contours
RESULTS
OPEN DEFAULT
CONTOUR BAND
DEF ON
SCALAR Temp.
SKIP TO 50
RESULTS
HISTORY PLOT
SET LOCATION (pick leading node shown)
ALL INCS
ADD CURVES
ALL LOCATIONS
Time
Velocity x
ALL LOCATIONS
Time
Acceleration x
6.840e-007
8.826e+000
1.765e+001
2.648e+001
3.530e+001
4.413e+001
5.296e+001
6.178e+001
7.061e+001
7.943e+001
8.826e+001
lcase1
Temperature
Inc: 50
Time: 2.000e+000
X Y
Z
4
Pick Leading Node
1625
CHAPTER 4.2
Dynamics with Friction Heating
Figure 4.2-11 Velocity, Acceleration and Temperature History of Leading Node
Notice that the eIIect oI Iriction was not 100 since the block should come to a stop at 1 sec. This was due to the ever
slipping Iriction model. Rigid body dynamics gives:
where the initial velocity was selected as . Where is the stopping time or 1 second.
Also Irom the Iriction heating, the Iriction Iorce moves through a distance and this mechanical energy is converted to
thermal energy. This thermal energy is input to the heat transIer portion oI the solution. Equating the conversion Iactor
times the kinetic energy and accounting that Ior rigid contact only halI oI the Irictional heating is added to the block,
the average rise in temperature Ior a block that comes to rest Irom an initial velocity oI , becomes:
In this case, the rise in temperature is 13.86 K. Why is the block hotter at the leading bottom edge? What would you
do to improve the results?
How does this compare with the Marc predictions? To answer this, we can close the post Iile and add another load case
that is 1e6 seconds long allowing the block to come to thermal equilibrium, namely uniIorm temperature.
ALL LOCATIONS
Time
Temperature
FIT
0
10
20
30
40
50
60
70
80
90
100
-6
-4
-2
0
2
4
6
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Vx (m/s)
Vx (m/s)
Ax (m/s
2
) Ax (m/s
2
)
T (K)
T (K)
u

g u

gt u

0
; u g
t
2
2
---- u

0
t u
0
;
u

0
gt
s
t
s
u

0
AT
1
4
-- - conv
f act or
u

0
2
c
p
--------
\ .
|
| |

1626 Marc Users Guide: Part 3

CHAPTER 4.2
This second load case allows the heat to diIIuse into the block leaving a uniIorm temperature oI 14.34 K throughout
the block; this is in good agreement with our estimate oI 13.86K.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
CLOSE
LOADCASES
COUPLED
NEW
CONVERGENCE TESTING
DISPLACEMENTS (OK)
DYNAMIC TRANSIENT
TOTAL LOADCASE TIME
1e6
STEPPING PROCEDURE ADAPTIVE
TEMPERATURE
PARAMETERS
INITIAL TIME STEP
1.0
OK (twice)
MAIN
JOBS
PROPERTIES
lcase2 (OK)
SAVE
RUN
SUBMIT1
MONITOR
OK
RETURN
File Description
h4.proc Mentat procedure Iile to run the above example
Chapter 4.3: Radiation with Viewfactors
4.3
Radiation with Viewfactors

Summary 1628

Detailed Session Description 1630

Run Job and View Results 1633

Input Files 1635

1628 Marc Users Guide: Part 3
CHAPTER 4.3
Summary
Title Radiation with viewIactors
Problem Ieatures Thermal analysis with radiation via viewIactors
Geometry
Material properties k 0.0001 W mm
-1
K
-1
, 0.4, 5.67x10
-14
W mm
-2
K
-4
Analysis type Steady State Heat TransIer
Boundary conditions Radiation using viewIactors
Initial conditions Inner surIace oI inner sphere 400
o
C, outer surIace oI outer sphere 500
o
C
Element type Axisymmetric quadrilateral
FE results Temperature rise between concentric spheres
apply1
apply2
apply3
C
L
r
8 10 12 r=14
c o
Temperature (x100)
Arc Length
72
189
59
158
46
33
127
20
96
7
0
4
5
6
a
b
c
d
a
c
d
b
1629
CHAPTER 4.3
Radiation with Viewfactors
Two concentric spheres have their inner and outer most surIaces held at a Iixed temperature. They exchange heat Ilow
via radiation.
Figure 4.2-1 Thermal Boundary Conditions
Thermal boundary conditions keep the inner and outer most surIaces Iixed at 400 and 500C. Another thermal
boundary conditions identiIies that the outer surIace oI the inner sphere and the inner surIace oI the outer sphere can
radiate.
Figure 4.2-2 Thermal Contours
The heat transIer loadcase selected is a steady state that allows the sphere to exchange heat Ilow via radiation.
C
L
r
8 10 12 r=14
4.000e+002
4.100e+002
4.200e+002
4.300e+002
4.400e+002
4.500e+002
4.600e+002
4.700e+002
4.800e+002
4.900e+002
5.000e+002
lcase1
Temperature
Inc: 1
Time: 1.000e+000
X
Y
Z
a
b
c
d
1630 Marc Users Guide: Part 3
CHAPTER 4.3
The temperature contours shows this Ilow and the path plot shows the radial change in temperature.
Figure 4.2-3 Temperature Versus Radius
Detailed Session Description
This is an axisymmetric model and we can use cylindrical coordinates to deIine the spheres.
MESH GENERATION
COORDINATE SYSTEM
CYLINDRICAL (on)
CURVE TYPE
CENTER/POINT/POINT
RETURN
CURVES ADD
0,0,0, 8,0,0, 8,180,0
0,0,0, 10,0,0, 10,180,0
0,0,0, 12,0,0, 12,180,0
0,0,0, 14,0,0, 14,180,0
SURFACE TYPE
RULED
OK
SURFACE ADD
1, 2
3, 4
Inc : 1
Temperature (x100)
Arc Length
lcase1
72
189
59
158
46
33
127
20
96
Time : 1
7
0
4
5
6
a
b
c
d
a
c
d
b
1631
CHAPTER 4.3
Radiation with Viewfactors
CONVERT
DIVISIONS
12 2
SURFACES TO ELEMENTS
ALL: EXISTING
RETURN
SWEEP
ALL
RETURN
CHECK ELEMENTS
UPSIDE DOWN
FLIP ELEMENTS
ALL SELECTED
UPSIDE DOWN
RETURN
RENUMBER
ALL
MAIN
BOUNDARY CONDITIONS
THERMAL
FIXED TEMP
400
OK (add all nodes for r = 8)
NEW
FIXED TEMP
500
OK (add all nodes for r = 14)
NEW
EDGE RADIATION
SINK TEMPERATURE ON
OK (add all edges r = 10 & 12)
1632 Marc Users Guide: Part 3
CHAPTER 4.3
Figure 4.2-4 Applying Radiation Boundary Conditions
Figure 4.2-5 Viewfactor Control Menu
Note: Try using the path select option to pick the nodes on r8, 14 and the edges on r10, 12. You only need to pick a
beginning middle and ending node Ior path select.
COMPUTE RADIATION VIEWFACTORS
TYPE AX
VIEWFACTOR FILE
model1.vfs
OK
START
OK
MAIN
MATERIAL PROPERTIES (twice)
ANALYSIS CLASS: HEAT TRANSFER
NEW: STANDARD
THERMAL
K = 1E-4
EMISSIVITY
apply1
apply2
apply3
C
L
r
8 10 12 r=14
Note:
Here 1000 rays are randomly cast Irom each oI the 24 edges
to compute the view Iactors. The view-Iactors are stored in
the Iile model1.vIs. II the geometry changes this would need
to be done again.
1633
CHAPTER 4.3
Radiation with Viewfactors
Run Job and View Results
0.4
OK
ELEMENTS ADD
ALL EXISTING
MAIN
LOADCASES
HEAT TRANSFER
STEADY STATE
CONVERGENCE TESTING
MAX ERROR IN TEMPERATURE ESTIMATE
0.05
OK (twice)
MAIN
JOBS
NEW: HEAT TRANSFER
PROPERTIES
lcase1
AXISYMMETRIC
ANALYSIS OPTIONS
RADIATION
VIEWFACTOR FILE
model1.vfs
OK
OK
JOB PARAMETERS
UNITS AND CONSTANTS
TEMPERATURE IN CELSIUS (on)
STEFAN-BOLTZMANN
5.67E-14
OK, (thrice)
RUN
SUBMIT1
1634 Marc Users Guide: Part 3
CHAPTER 4.3
Figure 4.2-6 Temperature Contours
MONITOR
OK
SAVE
MAIN
RESULTS
OPEN DEFAULT
LAST
CONTOUR BAND
PATH PLOT
SET NODES (pick node shown)
(a,b,c,d)
#END LIST
ADD CURVES
ADD CURVE
Arc Length
Temperature
FIT
4.000e+002
4.100e+002
4.200e+002
4.300e+002
4.400e+002
4.500e+002
4.600e+002
4.700e+002
4.800e+002
4.900e+002
5.000e+002
lcase1
Temperature
Inc: 1
Time: 1.000e+000
X
Y
Z
a
b
c
d
1635
CHAPTER 4.3
Radiation with Viewfactors
Figure 4.2-7 Temperature Versus Radius
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
h5.proc Mentat procedure Iile to run the above example
Inc : 1
Temperature (x100)
Arc Length
lcase1
72
189
59
158
46
33
127
20
96
Time : 1
7
0
4
5
6
a
b
c
d
a
c
d
b
1636 Marc Users Guide: Part 3
CHAPTER 4.3
Chapter 4.4: Cooling Fin Analyses
4.4
Cooling Fin Analyses

Summary 1638

Steady State 1639

Transient 1649

Input Files 1655

1638 Marc Users Guide: Part 3
CHAPTER 4.4
Summary
Title Thermal cooling Iin
Problem Ieatures Thermal analysis with convection boundary conditions
Geometry
Material properties k 1.157x10
-4
BTU in
-1
s
-1

o
F
-1
,
C
p
0.146 BTU lbm
-1

o
F
-1
, 0.283 lbm in
-3
Analysis type Steady state and transient heat transIer
Boundary conditions Hot side H 750 BTU hr
-1
It
-2

o
F
-1
, 2500
o
F
Cool side H 500 BTU hr
-1
It
-2

o
F
-1
, 1000
o
F
Initial conditions T 0
o
F
Element type Planar quadrilateral
FE results Cooling eIIectiveness oI the Iin
(0,0)
(0.1,0.15)
H
o
t

S
i
d
e
C
o
o
l

S
i
d
e
H
o
t

S
i
d
e
C
o
o
l

S
i
d
e
Same Mass

96
97
98
99
100
101
95
102
104
105
106
107
108
109
103
94
93
92
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
110
111
112
113
133
134
136137 138 139 140 141 142
Without Fin
7.695
0
2 0
Y (x.1)
132
76
131
129
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
130
Time (x10)
75
73
21 22 23 24 25 26
20
27
29 30 31
32
33
34
28
19 18 17
2345678910 11 12 13 14 15 16
35
36
37
38
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
57
74
56
54
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
55
1 0
135
135 136
0
87
88
89
90
91
92
93
94
95
96
97
98
99
86
100
85
83
70
71
72
73
74
75
76
77
78
79
80
81
82
84
101
102
103
122
123
124
125
126
127
128
129 130131 132 133 134
121
120
119
118
104
105
106
107
108
109
69
110
112
113
114
115
116
117
111
With Fin
Thermal Energy
68
66
18 19 20 21 22 23 24 25 26
27
28
29 30
17
31
16 14
12345678910 11 12 13 15
32
33
34
53
54
55
56
57
58
59
60
61
62
63
64
65
52
51
50
49
35
36
37
38
39
40
67
41
43
44
45
46
47
48
42
q 8.6
1639
CHAPTER 4.4
Cooling Fin Analyses
Steady State
An eIIective means oI augmenting the cooling eIIectiveness oI a given thermal cooling design, is to increase the area
exposed to the cooling Iluid by means oI adding Iins. In the Iin design, the eIIectiveness is judged by comparing the
temperatures oI the structure Ior conditions with and without Iins. This sample problem determines two sets oI
temperatures reIlecting the structure with and without a Iin.
Background Information
Description
This problem demonstrates the preparation oI a heat transIer model including convection boundary conditions.
Idealization
The model is a 0.15" X 0.05" rectangle with a 0.05" square Iin centered vertically on the right side. The vertical sides
have convection boundary conditions and the top and bottom are adiabatic.
Requirements for a Successful Analysis
The analysis is considered completed iI a steady state analysis is perIormed Ior a structure with Iin and a structure
without Iin.
Figure 4.4-1 Cross-Section of Cooling Fin
Full Disclosure
The model is 0.15" high and 0.05" wide. The Iin is centered vertically and is 0.05" square.
The convection on the leIt side is:
with
(0,0)
(0.1,0.15)
(0.1,0.10)
(0.1,0.05)
k = 1.157x10
-4
BTU in
-1
s
-1

o
F
-1
,
C
p
= 0.146 BTU lbm
-1

o
F
-1
, = 0.283 lbm in
-3

H = 750 BTU hr
-1
ft
-2

o
F
-1
= 2500
o
F T
8
Cool Side Hot Side
H = 500 BTU hr
-1
ft
-2

o
F
-1
= 1000
o
F T 8
Material properties
q h T T

( )
1640 Marc Users Guide: Part 3
CHAPTER 4.4
The right side has a convection:
with
The material has a coeIIicient oI thermal conduction
For a steady-state analysis, it is not required to enter the mass density and the heat capacity, however they will be
entered since a transient analysis will be done later.
Overview of Steps
Step 1: Create two surfaces and convert to finite elements.
Step 2: Add convection boundary conditions.
Step 3: Add material data.
Step 4: Create a steady-state loadcase.
Step 5: Create a thermal job and submit.
Step 6: Postprocess results.
Step 7: Delete fin elements.
Step 8: Modify convection boundary conditions.
Step 9: Create new job and submit.
Step 10: Postprocess results.
Detailed Session Description
Step 1: Create two surIaces and convert to Iinite elements.
The Iirst step creates two surIaces and converts them to Iinite elements. The Iollowing button sequence creates the
surIaces and convert them.
h 750 Bt u HR Ft
2
F
o
750
1
3600 12 ( )
2
-------------------------- Bt u i n
2
F
o

2500 F
o

q h T T

( )
h 500 Bt u HR Ft
2
F
o
500
1
3600 12 ( )
2
-------------------------- Bt u i n
2
F
o

1000 F
o

k 1.157 10
4
Bt u i n
2
F
o

1641
CHAPTER 4.4
Cooling Fin Analyses
FILES
SAVE AS
steady_fin
OK
MESH GENERATION
COORDINATE SYSTEM: SET
U SPACING
0.05
V SPACING
0.05
U DOMAIN
0 0.10
V DOMAIN
0 0.15
GRID ON (on)
RETURN
FILL
ZOOM BOX
(box pick right upper half of grid)
SRFS ADD (pick grid points)
point(0,0,0)
point(0.05,0,0)
point(0.05,0.15,0)
point(0,0.15,0)
point(0.05,0.05,0)
point(0.1,0.05,0)
point(0.1,0.1,0)
point(0.05,0.1,0)
GRID (off)
CONVERT
DIVISIONS
3 6
SURFACES TO ELEMENTS
1 (pick the first surface)
END LIST (#)
DIVISIONS
1642 Marc Users Guide: Part 3
CHAPTER 4.4
The next button sequence merges the duplicate nodes on the interIace oI the two surIaces.
Figure 4.4-2 The Mesh Generated using the Convert Option
Step 2: Add convection boundary conditions.
This step adds the convection boundary conditions. The Iollowing button sequence creates the boundary conditions.
3 2
SURFACES TO ELEMENTS
2 (pick the second surface)
END LIST (#)
MAIN
MESH GENERATION
SWEEP
SWEEP NODES
ALL: EXIST.
PLOT
draw POINTS (off)
draw SURFACES (off)
REGEN
RETURN
FILL
MAIN
BOUNDARY CONDITIONS
THERMAL
NAME
hotside
X
Y
Z
1
1643
CHAPTER 4.4
Cooling Fin Analyses
Note that Ior the adiabatic conditions at top and bottom edges, no boundary conditions have to be applied.
Figure 4.4-3 The Film Conditions on the Hot and Coolant Side
EDGE FILM
FILM
AMBIENT TEMPERATURE
2500 (enter value in text box)
COEFFICIENT
750/(3600*144) (enter value in text box)
OK
EDGES: ADD (box Pick the left edge)
END LIST (#)
NEW
NAME
coolant
EDGE FILM
FILM
SINK TEMPERATURE
1000 (enter value in text box)
COEFFICIENT
500/(3600*144) (enter value in text box)
OK
EDGES: ADD (box Pick the right edge;
several Boxes are required)
END LIST (#)
RETURN
ID BOUNDARY CONDS (on/off)
hotside
coolant
X
Y
Z
1
1644 Marc Users Guide: Part 3
CHAPTER 4.4
Step 3: Add material data.
This step adds the material data. The Iollowing button sequence assigns the material properties.
Step 4: Create a steady-state loadcase.
This step creates a steady-state loadcase. The Iollowing button sequence does this.
Step 5: Create a thermal job and submit.
This step creates a thermal job and submits the job Ior analysis. The Iollowing buttons sequence does this.
MAIN
MATERIAL PROPERTIES (twice)
ANALYSIS CLASS: HEAT TRANSFER
NEW standard
THERMAL
K = 1.157e-4(BTU/s/in/F)
OK
SPECIFIC HEAT = 0.146 (BTU/lbm/F)
MASS DENSITY: THERMAL VALUE = 0.283 (lbm/in^3)
OK
ELEMENTS: ADD
ALL: EXIST.
MAIN
LOADCASES
HEAT TRANSFER
STEADY STATE
LOADS
OK (twice)
MAIN
JOBS
NEW HEAT TRANSFER
PROPERTIES
LOADCASES SELECT
lcase1
ELEMENT TYPES
ANALYSIS DIMENSION PLANAR
1645
CHAPTER 4.4
Cooling Fin Analyses
Step 6: Postprocess results.
The Iinal step Ior the Iirst analysis is to postprocess results. The Iollowing button sequence reviews the results.
Figure 4.4-4 Contours of Temperature for Structure with Fin
Step 7: Delete Iin elements.
First, restore the database with the geometry, delete the Iin elements and add a column oI elements to keep the mass
constant between the two models. The Iollowing button sequence modiIies the model.
PLANAR
SOLID
39
OK
ALL: EXIST.
RETURN
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
CONTOUR BANDS
NEXT
MAIN
FILES
1.300e+003
1.367e+003
1.434e+003
1.501e+003
1.568e+003
1.634e+003
1.701e+003
1.768e+003
1.835e+003
1.902e+003
1.969e+003
lcase1
Temperature
Inc: 1
Time: 1.000e+000
X
Y
Z
1
1646 Marc Users Guide: Part 3
CHAPTER 4.4
Figure 4.4-5 Mesh without Fin
SAVE AS
steady
OK
RESULTS
CLOSE
scalar plot OFF
RETURN
FILES
RESTORE
RESET PROGRAM
RETURN
MESH GENERATION
ELEMS REM (box Pick the fin elements)
END LIST (#)
SWEEP
remove unused NODES
RETURN
DUPLICATE
TRANSLATIONS
.05/3 0 0
ELEMENTS (box Pick right column of elements)
SWEEP
SWEEP NODES
ALL: EXIST.
X
Y
Z
1
1647
CHAPTER 4.4
Cooling Fin Analyses
Step 8: ModiIy convection boundary conditions.
This step modiIies convection boundary conditions on the edge where the Iin was previously. The Iollowing button
sequence modiIies the convection.
Figure 4.4-6 New Thermal Boundary Conditions
Step 9: Create new job and submit.
This step creates a new model and submits it. This prevents overwriting oI the previous post Iile. The Iollowing button
sequence does this.
MAIN
BOUNDARY CONDITIONS
ID BOUNDARY CONDS (on/off)
THERMAL
EDIT coolant (to edit the second boundary condition)
EDGES REM
ALL EXISTING
EDGES ADD
(box Pick the right edges)
END LIST (#)
RETURN
ID BOUNDARY CONDS (on/off)
MAIN
JOBS
NEW: HEAT TRANSFER
PROPERTIES
LOADCASES SELECT
lcase1
hotside
coolant
X
Y
Z
1
1648 Marc Users Guide: Part 3
CHAPTER 4.4
Step 10: Postprocess results.
This Iinal step postprocesses results oI the second analysis. The Iollowing button sequence reviews the results.
Figure 4.4-7 Results for Structure without Fin
The user can observe that when the cooling Iin is included such that there is a greater surIace area exposed to the
convective cooling, the temperature is lower when comparing Figure 4.4-4 and Figure 4.4-7.
ANALYSIS DIMENSION PLANAR
OK
SAVE
RUN
SUBMIT 1
MONITOR
MAIN
RESULTS
OPEN DEFAULT
CONTOUR BANDS
NEXT
1.675e+003
1.712e+003
1.750e+003
1.787e+003
1.825e+003
1.862e+003
1.900e+003
1.938e+003
1.975e+003
2.013e+003
2.050e+003
lcase1
Temperature
Inc: 1
Time: 1.000e+000
X
Y
Z
1
1649
CHAPTER 4.4
Cooling Fin Analyses
Transient
A planar slab oI material is subjected to heat loads and the resulting transient response is determined. The slab has
convection boundary conditions on the leIt and right surIaces as shown. The top and bottom horizontal surIaces are
adiabatic. The slab is at an initial temperature oI 0
o
F.
The leIt surIace is exposed to a hot environment whereas the right surIace is exposed to cooling conditions. The
purpose oI the Iin on the right side is to create more surIace area Ior cooling and improve the cooling eIIectiveness oI
the slab.
Figure 4.4-8 Problem Description
The transient solution requires an estimate oI the time it takes to reach0 steady state. To estimate this time, we assume
that system is treated as a lumped mass where the heat transIerred into the body is equated with the thermal energy
stored, namely:
where h is the diIIerence in the heat transIer coeIIicients between the hot and cold sides. Since is constant,
and .
Integrating the above Irom t 0 where , gives where b is a positive quantity with units oI s
-1

called the time constant. II we assume, . then
So let`s choose a time period oI say 20 seconds Ior our transient solution.
Detailed Session Description with Fin
FILES
OPEN
steady_fin
SAVE AS
96
97
98
99
100
101
95
102
104
105
106
107
108 109
103
94 93
92
77
78
79
80
81
82
83 84 85
86
87
88
89
90
91
110
111
112
113
133 134
136137 138 139 140 141 142
Without Fin
7.695
0
2 0
Y (x.1)
132
76
131
129
114 115
116
117
118 119
120
121 122 123
124 125 126
127 128
130
Time (x10)
75
73
21 22 23 24 25 26
20
27
29 30 31 32 33 34
28
19 18 17
2345678910 11 12 13 14 15 16
35 36 37 38
58
59
60
61
62
63
64
65
66
67
68
69
70 71 72
57
74
56
54
39 40 41 42 43 44 45
46
47
48
49
50
51
52
53
55
1 0
135
135 136
0
87 88
89
90
91
92 93 94 95 96
97
98
99
86
100
85
83
70
71
72
73 74 75 76 77
78
79
80
81
82
84
101 102 103
122 123 124 125 126 127 128 129 130131 132 133 134
121 120
119 118
104
105
106 107 108
109
69
110
112 113
114 115 116
117
111
With Fin
Thermal Energy
68
66
18 19 20 21 22 23 24 25 26 27 28 29 30
17
31
16 14
12345678910 11 12 13 15
32 33 34
53
54
55
56
57
58 59 60 61 62 63 64 65
52
51 50 49
35 36 37 38 39
40
67
41
43 44 45 46 47 48
42
(0,0)
(0.1,0.15)
H
o
t

S
i
d
e
C
o
o
l

S
i
d
e
H
o
t

S
i
d
e
C
o
o
l

S
i
d
e
Same Mass
hA
s
T

T ( )dt VC
p
dT
T

dT d T T

( )
d T T

( )
T T

( )
-------------------------
hA
s
VC
p
-------------- t b t
T T
i

T T

( )
T
i
T

( )
----------------------- e
bt

T T

( )
T
i
T

( )
----------------------- 0.1 t
ss
0.1 ( ) ln ( ) b 0.1 ( ) ln ( ) 0.175 13s
1650 Marc Users Guide: Part 3
CHAPTER 4.4
transient_fin
OK
MAIN
LOADCASES
HEAT TRANSFER
TRANSIENT
TOTAL LOADCASE TIME
20
ADAPTIVE LOADING TEMPERATURE PARAMETERS
MAX # INCREMENTS
200
INITIAL TIME STEP
1
OK (twice)
MAIN
JOBS
PROPERTIES
ANALYSIS OPTIONS
LUMPED CAPACITY
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR
OK
MAIN)
1651
CHAPTER 4.4
Cooling Fin Analyses
RESULTS
OPEN DEFAULT
CONTOUR BANDS
LAST
HISTORY PLOT
SET LOCATIONS
(pick those shown)
END LIST
ALL INCS
ADD CURVES
ALL LOCATIONS
Time
Temperature
FIT
RETURN
RETURN
PATH PLOT
SHOW MODEL
NODE PATH
(pick two nodes shown)
END LIST
VARIABLES
ADD CURVE
Arc Length
Temperature
FIT
RETURN
REWIND
MONITOR
1.300e+003
1.367e+003
1.434e+003
1.501e+003
1.568e+003
1.634e+003
1.701e+003
1.768e+003
1.835e+003
1.902e+003
1.969e+003
lcase1
Temperature
Inc: 136
Time: 2.000e+001
X
Y
Z
Pick Nodes
Figure 4.4-9 Temperature Contours
90
91
92
93
94
95
89
96
98
99
100
101
102
103
97
104
88
86
72
73
74
75
76
77
87
78
80
81
82
83
84
85
79
105
106
107
127
128 129 130131 132
126
134 136
lcase1
Time (x10)
Temperature (x1000)
0 2
135
125
124
123
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
71
70
69
68
17
18
19
20
21
22
16
23
25
26
27
28
29
30
24
15
14
13
0
1
2
3
4
5
6
7
8
9
10
11
12
31
0
32
34
54
55
56
57
58
59
53
60
62
63
64
65
66
67
61
52
51
50
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
33
1.958
Node 17
133
134 135 136
0
87 88
89
90
91
92
93
94
95 96
97
98
99
86
100
85
83
70
71
72 73 74 75 76 77
78
79
80
81
82
84
101
102 103
122
123
124 125 126 127 128 129 130131 132 133
121
120
119 118
104
105
106
107
108
109
69
110
112 113
114
115
116
117
111
Node 40
68
66
18 19 20 21 22 23 24 25 26 27 28 29 30
17
31
16
14
12345678910 11 12 13
15
32 33 34
53 54 55 56 57 58 59 60 61 62 63 64 65
52 51 50 49
35 36 37 38 39 40
67
41
43 44 45 46 47 48
42
Figure 4.4-10 Temperature History
1652 Marc Users Guide: Part 3
CHAPTER 4.4
Our estimate oI 20 seconds to reach steady state was overestimated and about 10 seconds appears acceptable. While
obvious here, iI we are not sure the solutions reaches steady state, another steady state loadcase can be added to this
job and repeating the above post processing steps would insure steady state is reached.
Let`s repeat the above steps Ior the model without the cooling Iin and determine the eIIiciency oI the cooling Iin.
Detailed Session Description without Fin
FILES
OPEN
steady
SAVE AS
transient
OK
MAIN
LOADCASES
HEAT TRANSFER
TRANSIENT
TOTAL LOADCASE TIME
20
ADAPTIVE LOADING TEMPERATURE PARAMETERS
MAX # INCREMENTS
200
INITIAL TIME STEP
1
OK (twice)
MAIN
JOBS
PROPERTIES
ANALYSIS OPTIONS
LUMPED CAPACITY
OK (twice)
SAVE
1653
CHAPTER 4.4
Cooling Fin Analyses
DeIining the cooling eIIiciency as:
where E is the thermal energy stored at the end oI the transient solution. Let`s history plot E with and without the Iin.
RUN
SUBMIT1
MONITOR
OK
MAIN
RESULTS
OPEN
transient_job1.t16
OK
HISTORY PLOT
ALL INCS
ADD CURVES
GLOBAL
Time
Thermal Energy
FIT (E = 0.769516)
OK
COPY TO GENERALIZED XY PLOTTER
MAIN
RESULTS
OPEN
transient_fin_job1.t16
OK
HISTORY PLOT
ALL INCS
ADD CURVES
GLOBAL
Time
Thermal Energy
q 1
T
avg
f i n
T
avgno f i n
-------------------------- 1
E
f i n
E
no f i n
-------------------
1654 Marc Users Guide: Part 3
CHAPTER 4.4
Figure 4.4-11 Thermal Energy History with and without Fin
The thermal eIIiciency oI the Iin becomes; . While the temperature contours in
Figure 4.4-4 and Figure 4.4-7 where extrema values oI temperature Ior the Iin are 1300 and 1969 and 1675 and 2050
without the Iin suggest a higher thermal eIIiciency, the thermal energy (or average temperature) used to compute the
thermal eIIiciency is more appropriate. Other Iin arrays are very possible yet they leave smaller and smaller air
channels to circulate coolant that may adversely impacting thermal eIIiciency. Figure 4.4-12 shows possible cooling
channels (cyan color) improve the thermal eIIiciency (Irom 8.6 to 26) provided collant Ilows maintain the same heat
transIer coeIIicients and sink temperature.
Figure 4.4-12 Possible Cooling Channels
FIT (E = 0.703308)
OK
COPY TO GENERALIZED XY PLOTTER
FIT
96
97
98
99
100
101
95
102
104
105
106
107
108
109
103
94
93
92
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
110
111
112
113
133
134
136137 138 139 140 141 142
Without Fin
7.695
0
2 0
Y (x.1)
132
76
131
129
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
130
Time (x10)
75
73
21 22 23 24 25 26
20
27
29
30 31
32
33
34
28
19 18 17
2345678910 11 12 13 14 15 16
35
36
37
38
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
57
74
56
54
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
55
1 0
135
135 136
0
87
88
89
90
91
92
93
94
95
96
97
98
99
86
100
85
83
70
71
72
73
74
75
76
77
78
79
80
81
82
84
101
102
103
122
123
124
125
126
127
128
129
130131 132 133 134
121
120
119
118
104
105
106
107
108
109
69
110
112
113
114
115
116
117
111
With Fin
Thermal Energy
68
66
18 19 20 21 22 23 24 25 26
27
28
29
30
17
31
16 14
12345678910 11 12 13 15
32
33
34
53
54
55
56
57
58
59
60
61
62
63
64
65
52
51
50
49
35
36
37
38
39
40
67
41
43
44
45
46
47
48
42
q 1 0.703308 0.769516 ( ) 8.6
X
Y
Z
1
1655
CHAPTER 4.4
Cooling Fin Analyses
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
h1.proc Mentat procedure Iile to run the above example
h2.proc Mentat procedure Iile to run the above example
cooling_fin.proc Mentat procedure Iile to run the above example
heat1.mud Mentat model Iile
steady.mud Mentat model Iile
steady_fin.mud Mentat model Iile
transient.mud Mentat model Iile
transient_fin.mud Mentat model Iile
1656 Marc Users Guide: Part 3
CHAPTER 4.4
Section 5: Coupled Analysis
Section 5: Coupled Analysis
Marc Users Guide -1658
Chapter 5.1: Coupled Structural Acoustic Analysis
5.1
Coupled Structural
Acoustic Analysis

Chapter Overview 1660

Two Spherical Rooms Separated by a Membrane 1660

Harmonic Analysis with Stress-free Membrane 1661

Harmonic Analysis with Pre-stressed Membrane 1670

Input Files 1673

1660 Marc Users Guide: Part 3
CHAPTER 5.1
Chapter Overview
In a coupled structural-acoustic analysis, both the structure and the acoustic medium are modeled. The interaction
between the structure and the acoustic medium aIIects the total response oI the system. In Marc, the coupling between
the structure and the acoustic medium is implemented via the CONTACT option. This enables easy modeling, since at
the interIace between the structure and the acoustic medium, the Iinite element mesh does not need to line up and no
interIace elements need to be deIined.
The implementation is currently limited to harmonic analyses, which may be preceded by a static analysis to include
the eIIect oI a pre-stress on the response. II during the pre-stress phase severe distortions oI the Iinite element mesh oI
the acoustic medium occur, remeshing is allowed beIore starting the harmonic analysis.
In this chapter, a coupled structural-acoustic analysis is perIormed on two spherical rooms separated by a membrane.
Two Spherical Rooms Separated by a Membrane
First, an analysis is done using a stress-Iree membrane. Then, a similar harmonic analysis is done aIter having
pre-stressed the membrane.
Background Information
Two spherical rooms with a radius oI 0.5 m are connected by a cylinder with a radius oI 0.5sin(20) m and a length oI
0.01 m (see Figure 5.1-1). The rooms are Iilled with air with a bulk modulus oI 1.5x10
5
N/m
2
and a density oI 1 kg/m
3
.
The cylinder contains a membrane oI elastomeric material which is described by a neo-Hookean material with a
constant C
10
equal to 80x10
5
N/m
2
and a density oI 1000 kg/m
3
. The air in the leIt room is locally excited by a sound
pressure and the response near the membrane in the right room is calculated Ior a Irequency range Irom 60 to 90 Hz,
using 100 intervals. The analysis is Iirst done using an unstressed membrane, then with a pre-stressed membrane,
where the pre-stress is caused by increasing the radius oI the membrane by 0.001 m.
Figure 5.1-1 Structural-Acoustic Problem Schematic
20
o
0
.
5
0.01
Air Air
Membrane
1661
CHAPTER 5.1
Coupled Structural Acoustic Analysis
The rooms and the membrane are modeled using 4-node axisymmetric elements with Iull integration (Marc element
type 40 Ior the air and Marc element type 82 Ior the membrane). The number oI contact bodies used are three: one
acoustic body Ior the leIt room, one acoustic body Ior the right room, and one deIormable body Ior the membrane.
Harmonic Analysis with Stress-free Membrane
Model Generation
As a Iirst step in the generation oI the Iinite element mesh, a number oI curves are deIined describing the boundary oI
the leIt room. Then the straight curve at the cylindrical part is expanded over the length oI the cylinder, so that a surIace
is obtained which, in turn, is converted into Iinite elements. The leIt room is meshed separately using the advancing
Iront quad mesher. By using the symmetry option, the elements oI the right room are easily obtained. A detail oI the
mesh is shown in Figure 5.1-2. The elements oI the membrane, leIt and right room are stored in element sets.
MESH GENERATION
CURVE TYPE
CENTER/RADIUS/ANGLE/ANGLE
RETURN
crvs ADD
0 0 0
0.5
20
180
pts ADD
0.5*cos(20*pi/180) 0 0
CURVE TYPE
LINE
RETURN
crvs ADD
5 6 6 1
EXPAND
TRANSLATIONS
0.01 0 0
SAVE
CURVES
3
#
RETURN
1662 Marc Users Guide: Part 3
CHAPTER 5.1
CONVERT
DIVISIONS
6 3
SURFACES TO ELEMENTS
1
#
RETURN
CHECK
UPSIDE DOWN
FLIP ELEMENTS
all: EXIST.
RETURN
AUTOMESH
CURVE DIVISIONS
FIXED # DIVISIONS
# DIVISIONS
4
APPLY CURVE DIVISIONS
3
#
FIXED # DIVISIONS
# DIVISIONS
16
APPLY CURVE DIVISIONS
1
#
FIXED # DIVISIONS
# DIVISIONS
14
APPLY CURVE DIVISIONS
2
#
RETURN
1663
CHAPTER 5.1
Coupled Structural Acoustic Analysis
2D PLANAR MESHING
QUADRILATERALS (ADV FRNT): QUAD MESH!
1 2 3
#
RETURN (twice)
BETWEEN POINT (click two bottom corner points of
quad surface to create a new point)
4.698463103930e-01 0 0
4.798463103930e-01 0 0
SYMMETRY
POINT (click the point just created)
4.748463103930e-01 0 0
ELEMENTS
19 to 108
RETURN
CLEAR GEOM
SELECT
CLEAR SELECT
ELEMENTS: STORE
membrane
OK
1 to 18
ELEMENTS: STORE
room_left
OK
19 to 108
ELEMENTS: STORE
room_right
OK
109 to 198
MAIN
1664 Marc Users Guide: Part 3
CHAPTER 5.1
Figure 5.1-2 Detail of Finite Element Mesh around Membrane
Boundary Conditions
Boundary conditions are deIined to clamp the membrane around its circumIerence and to enter the pressure at a node
in the leIt room.
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
OK
nodes ADD
7 14 21 28
#
RETURN
NEW
ACOUSTIC
FIXED PRESSURE
PRESSURE (on)
1665
CHAPTER 5.1
Coupled Structural Acoustic Analysis
Material Properties
Two diIIerent materials are deIined: one with the mechanical material properties Ior the membrane, and one with the
acoustic material properties Ior the air in the leIt and right room.
10
OK
nodes ADD
63
#
MAIN
MATERIAL PROPERTIES
NEW
MORE (MECHANICAL MATERIAL TYPES)
MOONEY
C10
80e5
MASS DENSITY
1000
OK
elements ADD
membrane
PREVIOUS
NEW
MORE (NON-MECHANICAL MATERIAL TYPES)
ACOUSTIC
BULK MODULUS
1.2e5
MASS DENSITY
1
OK
elements ADD
room_left
room_right
MAIN
1666 Marc Users Guide: Part 3
CHAPTER 5.1
Contact Bodies
First, the elements oI the leIt and right room are assigned to acoustic contact bodies. An acoustic contact body is
completely deIined by a number oI elements and does not need any Iurther properties. An acoustic contact body cannot
be touched; nodes oI an acoustic contact body may touch a deIormable or a rigid body. Properties which are relevant
Ior the interaction between an acoustic body and a deIormable or a rigid body should be deIined either Ior the
deIormable, the rigid body, or via a contact table. The third body is a deIormable body and consists oI the elements
deIining the membrane. In order to make sure that the nodes oI the leIt and right room only contact edges oI the
membrane with a normal vector parallel to the global x-axis, the EXCLUDE option is used to avoid contact with edges
having a normal vector parallel to the global y-axis.
CONTACT
CONTACT BODIES
NEW
ACOUSTIC
OK
ELEMENTS ADD
room_left
NEW
ACOUSTIC
OK
ELEMENTS ADD
room_right
NEW
DEFORMABLE
OK
ELEMENTS ADD
membrane
RETURN
EXCLUDE SEGMENTS
CONTACT BODY
cbody3
OK
edges ADD (select edges with normal vector parallel to y-axis)
1:3 7:3 13:3 6:1 12:1 18:1
#
MAIN
1667
CHAPTER 5.1
Coupled Structural Acoustic Analysis
Loadcases
An acoustic-solid harmonic load case is deIined, in which the Irequency range Irom 60 Hz to 90 Hz Ior the pressure
is entered. The number oI Irequencies is set to 101.
LOADCASES
NEW
NAME
harmonic_analysis
ACOUSTIC-SOLID
HARMONIC
LOWEST FREQUENCY
60
HIGHEST FREQUENCY
90
# FREQUENCIES
101
OK
MAIN
1668 Marc Users Guide: Part 3
CHAPTER 5.1
Jobs
An acoustic-solid job is deIined and the harmonic load case is selected. The element types Ior the membrane and the
air are set to 40 and 82, respectively. The model is saved and the job is submitted.
JOBS
ACOUSTIC-SOLID
harmonic_analysis (select loadcase)
CONTACT CONTROL
INITIAL CONTACT
exseg1 (exclude segments)
OK (twice)
AXISYMMETRIC
OK
ELEMENT TYPES
ACOUSTIC-SOLID
AXISYMMETRIC (acoustic element types)
40
OK
room_left
AXISYMMETRIC (acoustic element types)
40
OK
room_right
AXISYMMETRIC SOLID (mechanical element types)
82
OK
1669
CHAPTER 5.1
Coupled Structural Acoustic Analysis
Results
A plot oI the sound pressure magnitude at node 168 oI the right room as a Iunction oI the Irequency is given in
Figure 5.1-3 and shows a peak value near an eigenIrequency oI the membrane.
membrane
RETURN (twice)
FILE
SAVE AS
structural_acoustic_1.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
NEXT
SCALAR
Sound Pressure Magnitude
OK
CONTOUR BANDS
MONITOR
HISTORY PLOT
SET NODES
168
END LIST (#)
COLLECT DATA
0:1 0:200 1
NODES/VARIABLES
ADD 1-NODE CURVE
nodes
168
1670 Marc Users Guide: Part 3
CHAPTER 5.1
Figure 5.1-3 Sound Pressure Magnitude as a Function of the Frequency
Harmonic Analysis with Pre-stressed Membrane
Where the previous harmonic analysis was based on a stress-Iree membrane, now the membrane is Iirst pre-stressed,
Iollowed by a similar harmonic analysis. In this way, a shiIt oI the pressure peak to a higher Irequency can be expected.
Model Generation
The Iinite element model is the same as used Ior the previous analysis.
global variables
Frequency
variables at nodes
Sound Pressure Magnitude
FIT
FILES
NEW
OK
RESET PROGRAM
1671
CHAPTER 5.1
Coupled Structural Acoustic Analysis
Boundary Conditions
The boundary conditions oI the previous analysis are modiIied to take into account the radius increase oI the
membrane.
Material Properties and Contact Bodies
The material properties and the contact bodies don`t need any changes.
Loadcase
A new loadcase, an acoustic-solid static one, must be deIined to determine the pre-stress in the membrane.
OPEN
acoustic.mud
OK
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
ON Y-DISPLACEMENT
0.001
OK
MAIN
LOADCASES
NEW
NAME
pre_stress
ACOUSTIC-SOLID
STATIC
CONTACT
exseg1
OK (twice)
MAIN
1672 Marc Users Guide: Part 3
CHAPTER 5.1
Jobs
In the acoustic-solid job, two load cases must be selected: Iirst the one corresponding to the pre-stress oI the membrane
and then the one corresponding to the harmonic analysis. Notice that the displacement boundary conditions may not
occur as initial loads. AIter saving the model, the job is submitted.
Results
A similar history plot as in the previous analysis is made, but now based on the sub-increments oI increment 1, thus
reIlecting the harmonic analysis based on the pre-stressed membrane. The sound pressure magnitude as a Iunction oI
the Irequency is shown in Figure 5.1-4 and clearly shows a shiIt oI the peak value to a higher Irequency.
JOBS
ACOUSTIC-SOLID
CLEAR
pre_stress
Harmonic_analysis
INITIAL LOADS
apply1 fixed_displacement (clear)
OK (twice)
FILES
SAVE AS
structural_acoustic_2.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODES
168
END LIST (#)
COLLECT DATA
1:1 1:200 1
1673
CHAPTER 5.1
Coupled Structural Acoustic Analysis
Figure 5.1-4 Sound Pressure Magnitude as a Function of the Frequency
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
NODES/VARIABLES
ADD 1-NODE CURVE
nodes
168
global variables
Frequency
variables at nodes
Sound Pressure Magnitude
FIT
File Description
acoustic.proc Mentat procedure Iile to run the above example
1674 Marc Users Guide: Part 3
CHAPTER 5.1
Chapter 5.2: Coupled Electrical-Thermal-Mechanical Analysis of a Micro Actuator
5.2
Coupled Electrical-Thermal-
Mechanical Analysis of a
Micro Actuator

Chapter Overview 1676

Simulation of a Microelectrothermal Actuator 1676

Input Files 1683

1676 Marc Users Guide: Part 3
CHAPTER 5.2
Chapter Overview
This chapter demonstrates the simulation oI a micro-actuator using coupled electrical-thermal-mechanical analysis.
In Marc, coupled electrical-thermal-mechanical (a.k.a. Joule mechanical) analysis is handled using a staggered
solution procedure similar to the ones used in coupled electrical-thermal (a.k.a. Joule heating) and in
thermal-mechanical analyses. Using this approach, the electrical problem is solved Iirst Ior the nodal voltages. Next,
the thermal problem is solved to obtain the nodal temperatures. Finally, the mechanical problem is solved Ior the nodal
displacements.
Simulation of a Microelectrothermal Actuator
Problem Description
The microelectrothermal actuator, shown in Figure 5.2-1, is a U` shaped MEMS device Iabricated Irom
polycrystalline silicon. Polycrystalline silicon has a higher electrical resistivity than most metals. The actuator uses
diIIerential thermal expansion between the thin arm (hot arm) and the wide arm (cold arm) to achieve motion. Current
Ilows through the device because oI a potential diIIerence applied across the two electrical pads. Because oI the
diIIerent widths oI the two arms oI the 'U' structure, the current density in the two arms is diIIerent leading to diIIerent
amounts oI thermal expansion and hence, bending. II the lateral deIlection oI the tip oI the device is restricted by an
object, a Iorce is generated on that object. Arrays oI actuators can be connected together at their tips to multiply the
Iorce produced.
The material oI the actuator is polycrystalline silicon with a Young's modulus oI 158.0E3 MPa, a Poisson's ratio oI
0.23, a coeIIicient oI thermal expansion oI 3.0E-6 1/K, a thermal conductivity oI 140.0E6 picowatt/micrometer.K, and
a resistivity oI 2.3E-11 teraohm.micrometer. The hot arm is 240 microns long and 2 microns wide. The cold arm is
200 microns long and 16 microns wide. The Ilexure is 40 microns long and 2 microns wide. The gap between the hot
and cold arms is 2 microns wide. The thickness oI the actuator is 2 microns.
Figure 5.2-1 Actuator Geometry
The initial temperature oI the actuator is set to 300K. A potential diIIerence oI Iive volts is applied across the electrical
pads. The temperature oI the pads is Iixed at 300K. The pads are Iixed in space in all three degrees oI Ireedom.
Electrical Pads Flexure Hot Arm Cold Arm
1677
CHAPTER 5.2
Coupled Electrical-Thermal-Mechanical Analysis of a Micro Actuator
Actuator Model
A 3-D single actuator model is shown in Figure 5.2-2. The model is constructed oI 2174 higher-order tetrahedron
elements (element type 127). The model Iile actuator.mfd contains the 3-D geometry and Iinite element mesh Ior
the problem. In the Iollowing, we deIine the boundary conditions, initial conditions, and material properties pertaining
to Joule-mechanical analysis.
Figure 5.2-2 3-D Microelectrothermal Model
To open the model:
AIter opening the model and examining it, Iollow the steps described below to complete the model deIinition:
FILES
OPEN
actuator.mfd
OK
FILL
ZOOM IN
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X
0
DISPLACEMENT Y
0
DISPLACEMENT Z
0
OK
1678 Marc Users Guide: Part 3
CHAPTER 5.2
NODES
ADD
ALL
SET
fixed_nodes
OK
RETURN
NEW
THERMAL
FIXED TEMPERATURE
TEMPERATURE (TOP)
300
OK
NODES
ADD
ALL
SET
fixed_nodes
OK
RETURN
NEW
JOULE
FIXED VOLTAGE
VOLTAGE
5
TABLE
table1
OK
NODES
ADD
ALL
SET
electrical_pad1_nodes
OK
1679
CHAPTER 5.2
Coupled Electrical-Thermal-Mechanical Analysis of a Micro Actuator
Figures 5.2-3, 5.2-4, and 5.2-5 show the Material Properties, Loadcases, and Jobs menus.
RETURN
NEW
JOULE
FIXED VOLTAGE
VOLTAGE
0
OK
NODES
ADD
ALL
SET
electrical_pad2_nodes
OK
MAIN
INITIAL CONDITIONS
NEW
THERMAL
TEMPERATURE
TEMPERATURE (TOP)
300
OK
NODES
ADD
ALL
EXIST.
1680 Marc Users Guide: Part 3
CHAPTER 5.2
Figure 5.2-3 Material Properties Menu
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
158.0e3
POISSON'S RATIO
0.23
THERMAL EXP.
THERMAL EXP. COEF.
3.0e-6
OK (twice)
JOULE HEATING
CONDUCTIVITY
140.0e6
RESISTIVITY
2.3e-11
OK
ELEMENTS
ADD
ALL
EXIST.
1681
CHAPTER 5.2
Coupled Electrical-Thermal-Mechanical Analysis of a Micro Actuator
Figure 5.2-4 Loadcases Menu
Figure 5.2-5 Jobs Menu
MAIN
LOADCASES
JOULE-MECHANICAL
TRANSIENT
# OF STEPS
10
OK
1682 Marc Users Guide: Part 3
CHAPTER 5.2
Run Job and View Results
To run the job:
MAIN
JOBS
JOULE-MECHANICAL
lcase1
OK
FILES
SAVE AS
etm_actuator1.mud
OK
MAIN
JOBS
RUN
RESET
SUBMIT (1)
MONITOR
OK
PLOT
NODES
ELEMENTS
SOLID
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
SCALAR PLOT
SCALAR
TEMPERATURE
OK
CONTOUR BANDS
MONITOR
CLOSE
1683
CHAPTER 5.2
Coupled Electrical-Thermal-Mechanical Analysis of a Micro Actuator
The Iinal deIormed shape with temperature distribution is shown in Figure 5.2-6. The maximum temperature is
1232K and the maximum y-deIlection is 6.058 microns.
Figure 5.2-6 Final Deformed Shape of the Actuator with Temperature Distribution
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
etm_actuator.proc Mentat procedure Iile to run the above example
actuator.mfd Associated Mentat model Iile
1684 Marc Users Guide: Part 3
CHAPTER 5.2
Chapter 5.3: Coupled Transient Cooling Fin
5.3
Coupled Transient Cooling Fin

Summary 1686

Detailed Session Description 1687

Run Jobs and View Results 1688

Input Files 1690

1686 Marc Users Guide: Part 3
CHAPTER 5.3
Summary
Title Coupled transient cooling Iin
Problem Ieatures Coupled thermal stress analysis
Geometry
Material properties k 1.157x10
-4
BTU in
-1
s
-1

o
F
-1
, E 3x10
7
psi, 0.3
C
p
0.146 BTU lbm
-1

o
F
-1
, 0.283 lbm in
-3
, 1.0x10
-5 o
F
-1
Analysis type Steady state and transient heat transIer
Boundary conditions Hot side H 750 BTU hr
-1
It
-2

o
F
-1
, 2500
o
F,
Cool side H 500 BTU hr
-1
It
-2

o
F
-1
, 1000
o
F
u 0, leIt side; v 0 bottom
Initial conditions T 70
o
F
Element type Planar quadrilateral
FE results Stress versus temperature
(0,0)
(0.1,0.15)
apply1 = hot conv
apply2 = cool conv
apply3 = u = 0
apply4 = v = 0
X
Y
Z
v
o
T

97
98
99
100
101
102
103
104
105
106
107 108
109
110
111
96
95
94
93
77 78 79 80 81 82 83
112
84
86 87 88
89
90
91
92
85
113
114
115
136 137 138 139 140 141 142
135
143
146 147
lcase1
Temperature (x1000)
Equivalent Von Mises Stress (x1e5)
0.07 2.069
145
76
134
132
116
117
118
119
120 121 122
133
123
125 126 127
128
129 130 131
124
0
75
73
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
17
16
15
14
0
1
2
3
4
33
5
7
8
9
10
11
12
13
6
34
35
36
57 58 59 60 61 62 63 56
64 66 67 68 69 70 71 72
65
74
55 53
37
38
39
40
41
42
43
54
44
46
47
48 49 50 51 52
45
1.424
Node8
144
145 146 147
0
95 96 97 98 99 100
94
101 103 104 105 106 107 108
102
93 92 91
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
109
75
110 112
132 133 134 135 136 137
131
138
140 141 142 143 144
139
130 129 128
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
111
74 73 72
20 21 22 23 24 25
19
26
28 29 30
31 32 33
27
18 17 16
12345678910 11 12 13 14 15
34 35
36
37
57
58
59 60 61 62 63 64 65 66 67 68 69 70 71
56
Node 47
55
53
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
54
1
1687
CHAPTER 5.3
Coupled Transient Cooling Fin
Mechanical boundary conditions are added to a previous transient thermal model. Here, the bottom horizontal surIace
is constrained not to displace in the vertical direction and the leIt vertical surIace is constrained not to displace in the
horizontal direction. Mechanical properties are also added to the model including the thermal coeIIicient oI expansion.
The transient loadcase is changed to a quasi-static coupled loadcase. Finally, the element types are changed to plane
stress and the job is submitted.
Stresses are generated in the slab because oI nonuniIorm thermal growth constrained by the mechanical boundary
conditions. By plotting the stress at the points shown, we see that the maximum stress on the hot side occurs well
beIore steady state.
Detailed Session Description
Even though the thermal eIIiciency may be better with the cooling Iin, the structural response may not. Let`s see how
to take the model used in Chapter 4.4: Cooling Fin Analyses and convert it into a coupled thermal stress problem.
FILES
OPEN
heat1.mud
SAVE AS
heat1s
OK
RETURN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FIX X
0
NODES ADD (pick nodes on left edge)
NEW
FIX Y
0
NODES ADD (pick nodes on bottom edge)
RETURN (twice)
MATERIAL PROPERTIES (twice)
ANALYSIS CLASS: USE CURRENT JOB (off)
COUPLED
1688 Marc Users Guide: Part 3
CHAPTER 5.3
Run Jobs and View Results
STRUCTURAL
E = 3E7
= .3
THERMAL EXP (twice)
ALPHA = 10E-6
OK (twice)
MAIN
LOADCASES
TYPE COUPLED
QUASI-STATIC
LOADS (pick new bcs)
CONV. TESTING
DISPLACEMENTS,
OK
TOTAL LOADCASE TIME
60
OK
MAIN
JOBS
TYPE: COUPLED
PROPERTIES
INITIAL LOADS (select new bcs)
OK
JOB RESULTS
EQUIVALENT VON MISES STRESS
OK (twice)
ELEMENT TYPES
ANALYSIS DIMENSION PLANAR
SOLID
PLANE STRESS
3
OK
ALL: EXISTING
MAIN
u
1689
CHAPTER 5.3
Coupled Transient Cooling Fin
Figure 5.3-1 Equivalent von Mises Stress Contours at Steady State
SAVE
RUN
SUBMIT1
MONITOR
RESULTS
OPEN DEFAULT
CONTOUR BANDS
LAST
SCALAR
Equivalent Von Mises Stress
HISTORY PLOT
SET LOCATIONS (pick nodes shown)
END LIST
ALL INCS
ADD CURVES
ALL LOCATIONS
Temperature
Equivalent Von Mises Stress
6.486e+003
1.508e+004
2.368e+004
3.228e+004
4.087e+004
4.947e+004
5.806e+004
6.666e+004
7.526e+004
8.385e+004
9.245e+004
lcase1
Equivalent Von Mises Stress
Inc: 147
Time: 6.000e+001
X
Y
Z
1
Pick Nodes
1690 Marc Users Guide: Part 3
CHAPTER 5.3
Figure 5.3-2 Stress Versus Temperature History
Notice how the stress peaks well beIore steady state (on the hot side - yet the cool side stress slowly grows
monotonically to steady state) because oI nonuniIorm expansion during the transient. This is very common in coupled
thermal stress problems.
Plane stress is used in this example. II plane strain elements (types 11, 27, etc.) are used, the out-oI-plane strain Ior
these elements is zero. This generates a large out-oI-plane stress since Ior plane strain we have:
and the last term in the equation dominates Ior large changes in temperature. II there is no out-oI-plane constraint to
the thermal growth physically, plane stress should be used. II the out-oI-plane thermal growth is restricted, such as
these planes remaining plane, generalized plane strain elements (types 19, 29, etc.) should be used. You may wish to
try these elements and observe what happens.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
h3.proc Mentat procedure Iile to run the above example
heat1.mud Associated Mentat model Iile
97 98
99
100
101
102
103
104
105 106 107 108
109
110
111
96 95 94
93
77 78 79 80 81 82 83
112
84
86 87 88
89
90
91
92
85
113 114 115
136 137 138 139 140 141 142
135
143
146 147
lcase1
Temperature (x1000)
Equivalent Von Mises Stress (x1e5)
0.07 2.069
145
76
134
132
116
117
118
119 120 121 122
133
123
125 126 127
128 129 130 131
124
0
75
73
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
17
16
15
14
0
1
2
3
4
33
5
7
8
9
10
11
12
13
6
34
35
36
57 58 59 60 61 62 63 56
64 66 67 68 69 70 71 72
65
74
55 53
37
38
39
40
41
42
43
54
44
46
47 48 49 50 51 52
45
1.424
Node8
144
145 146 147
0
95 96 97 98 99 100
94
101 103 104 105 106 107 108
102
93 92 91
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
109
75
110 112
132 133 134 135 136 137
131
138
140 141 142 143 144
139
130 129 128
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
111
74 73 72
20 21 22 23 24 25
19
26
28 29 30 31 32 33
27
18 17 16
12345678910 11 12 13 14 15
34 35
36
37
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
56
Node 47
55
53
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
54
1
o
zz
E
1 u ( ) 1 2u ( )
---------------------------------------- c
xx
c
yy
1 u ( )oAT | |
Chapter 5.4: Temperature Dependent Orthotropic Thermal Strains
5.4
Temperature Dependent
Orthotropic Thermal Strains

Chapter Overview 1692

Detailed Session Description 1693

Run Jobs and View Results 1696

Thermal Expansion Data Reduction 1698

References 1701
1692 Marc Users Guide: Part 3
CHAPTER 5.4
Chapter Overview
Temperature dependent thermal expansion behavior oI Iiber-reinIorced composite materials presents some unique
Ieatures with respect to more traditional isotropic materials, primarily the change oI the coeIIicient oI thermal
expansion with spatial direction caused by an isotropy. Here, thermal couples and electrical resistance strain gages are
used to measure the expansion oI the BMS material in the Iiber direction (1) and transverse direction (2) as shown in
Figure 5.4-1.
Figure 5.4-1 Measured Temperatures and Thermal Strains
The measurements have strain gages mounted on two specimens: the test specimen, having unknown expansion and
the reIerence specimen, having a known thermal expansion |1|. This is repeated in a direction transverse to the Iiber
direction along with two thermal couples recording the temperatures oI the unknown and reIerence material.
Ultimately, the thermal strain as a Iunction oI temperature results in the two directions as shown in Figure 5.4-2; the
instantaneous coeIIicients oI thermal expansion used in the analysis are simply the slopes oI the thermal strain versus
temperature curves.
T
Figure 5.4-2 Thermal Strains and Instantaneous Coefficients of Thermal Expansion
BAR
VERT.
BAR
HORZ.
BMS
VERT.
BMS
HORZ.
F F
1 7/12/06 12:49:34 PM 86.6 88.0 0 0 0 0
2 7/12/06 12:50:34 PM 87.1 88.6 -1 -1 -2 1
3 7/12/06 12:51:34 PM 87.4 89.2 -3 -2 -3 1
4 7/12/06 12:52:34 PM 88.0 89.7 -4 -4 -4.0 -2.0
5 7/12/06 12:53:34 PM 89.7 92.3 -9.9 -8.9 -10 -10
6 7/12/06 12:54:34 PM 92.5 95.9 -20 -19 -20 -20
7 7/12/06 12:55:34 PM 95.9 100.0 -32 -30 -29.8 -29.8
8 7/12/06 12:56:34 PM 99.7 104.2 -44.7 -42.7 -39 -42
CTE TEST MEASUREMENTS
T/C#1
TI/SIL
BAR
REF.
T/C#2
BMS
REF.
T/C #1 T/C #2
Time Date Scan ID

1047 7/13/06 6:16:04 AM 81.2 82.1 33 32 44 35
50 100 150 200 250 300 350 400
-200
0
200
400
thermstrn e1 = -73.2595 + 0.6467 * T + 0.0024556 * T
2
- 2.6494e-006 * T
3
thermstrn e2 = -42.0634 + 0.26486 * T + 0.0027178 * T
2
- 2.5128e-006 * T
3
e1
e2
50 100 150 200 250 300 350 400
0.5
1
1.5
alpha1 = 0.6467 + 0.0049112 * T - 7.9482e-006 * T
2
alpha2 = 0.26486 + 0.0054356 * T - 7.5383e-006 * T
2
a1
a2 Temperature ( F)
ppm
ppm/ F
o
o
Temperature ( F)
o
1693
CHAPTER 5.4
Temperature Dependent Orthotropic Thermal Strains
Detailed Session Description
BeIore starting, it is worth mentioning that many times the mean (average) coeIIicient oI linear thermal expansion is
given in handbooks. II the mean coeIIicient oI linear thermal expansion versus temperature is known, one would still
need to construct the thermal strain versus temperature curve and supply the slopes (instantaneous coeIIicients oI
thermal expansion) to Marc as described in Volume A. The average coeIIicient oI linear thermal expansion is not a
thermodynamic material property, whereas the instantaneous coeIIicient oI thermal expansion is a thermodynamic
material property. Finally the terms used here Ior the mean and average coeIIicients are consistent with those
deIinitions in the American Society Ior Testing and Materials (ASTM) test method E 228
(2.)
. The term 'thermal strain
used here is simply the 'linear thermal expansion used in ASTM E 228; the change in length per initial length caused
by a change in temperature.
The thermal strains depicted in Figure 5.4-2, may or may not generate stresses in applications depending on how non
uniIorm the temperature distributions become and how the thermal expansion (or contraction) may be restricted. In
order to determine stresses a Iinite element analysis is usually perIormed with the appropriate material properties and
boundary conditions. In order to make sure that the material`s thermal expansion or contraction is correctly modeled,
the analysis done here is a simple one having uniIorm temperatures changing over time and boundary conditions
allowing Iree thermal growth or contraction. The goal is to see iI the Iinite element analysis reproduces the thermal
strains in Figure 5.4-2. It is always a good practice, particularly Ior complex material behavior, to replicate the material
behavior using a Iew elements in a simple scenario similar to the test method to veriIy the modeling. The example
model here uses only three elements as shown in Figure 5.4-3 with horizontal displacements along A-B, vertical
displacements along C-D Iixed to zero, and temperature ramped Irom 87.153
o
F to 400
o
F.
T
Figure 5.4-3 Model Definition
X
Y
Z
1
Node 3 (1,1,0)
A
C D
B
0.0 0.2 0.4 0.6 0.8 1.0
50
100
150
200
250
300
350
400 Temperature ( F)
Time (sec)
o
1694 Marc Users Guide: Part 3
CHAPTER 5.4
The same geometric model uses several diIIerent element types, namely plane stress, plane strain, generalized plane
strain, and axisymmetric Ior the same boundary conditions and material properties as listed below. The model Iiles are
complete and the steps below are used to highlight portions oI the model.
We will examine how the thermal expansion data appears in each oI the models above by examining each oI the models
above, running them and checking the results starting with the plane strain elements.
File Description
type26uc.mud Model Iile using plane stress elements.
type27uc.mud Model Iile using plane strain elements.
type28uc.mud Model Iile using axisymmetric elements.
type29uc.mud Model Iile using generalized plan strain elements.
FILES
OPEN
type27uc.mud
OK
RETURN
MATERIAL PROPERTIES
ORTHOTROPIC
THERMAL EXPANSION PROPERTIES
ALPHA11
1E-6
TABLE
alpha1 (pick table alpha1)
OK
ALPHA22
1E-6
TABLE
alpha2 (pick table alpha2)
OK
ALPHA33
1E-6
TABLE
alpha2 (pick table alpha2)
OK (thrice)
1695
CHAPTER 5.4
Temperature Dependent Orthotropic Thermal Strains
We have assumed that the Iiber direction (table alpha1) is in the global x direction and that the other two principal
material directions are in the epoxy direction (table alpha2). Let`s examine the tables, alpha1 and alpha2.
T
Figure 5.4-4 Fiber Direction Instantaneous Coefficient of Thermal Expansion versus Temperature
TABLES
SHOW TABLE
EDIT (pick table alpha1)
OK
TYPE
temperature
OK
FORMULA
0.6467 + 0.0049112 * V1 + -7.9482e-006 * V1^2
FIT
MAX
400 (set maximum independent
FIT variable)
EDIT (pick table alpha2)
OK
TYPE
temperature
6
5
4
3
2
1
1.404
0.647
4 0
F
V1 (x100)
alpha1
7
8
9
10
11
1696 Marc Users Guide: Part 3
CHAPTER 5.4
These tables enter the Iormulas Ior the instantaneous coeIIicient oI thermal expansion that is used in the analysis. Since
the model Iile as it exists we can simply run the Iile and examine the results
Run Jobs and View Results
OK
FORMULA
0.26486 + 0.0054356 * V1 + -7.5383e-006 * V1^2
FIT
MAIN
JOBS
MECHANICAL
JOB RESULTS
Stress
Total Strain
Thermal Strain
Temperature (Integration Point)
OK
OK
ELEMENT TYPES
MECHANICAL
PLANE STRAIN SOLID
27
OK
ALL: EXISTING
RETURN (twice)
SAVE
RUN
SUBMIT1
MONITOR
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODES
3
1697
CHAPTER 5.4
Temperature Dependent Orthotropic Thermal Strains
Figure 5.4-5 Thermal Strains and Displacements - Plane Strain Case
The experimentally measured thermal strains (e1, e2) are identical to the thermal strains (Comp 11, Comp 22) in the
analysis. Since the initial coordinate oI node 3 is (1,1) the displacement in the x (1) and y (2) directions are
displacements per unit length (apparent thermal strains) that should equal the thermal strains in the corresponding
directions provided no stresses are generated. This is not the case as shown in Figure 5.4-5 and stresses are being
generated. A large amount oI stress in the z (3) out-oI-plane direction is generated because the total out-oI-plane strain
must be zero (plane strain assumption). The out-oI-plane stress Ior plane strain becomes:
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Temperature (Integration Point)
Comp 11 of Thermal Strain
ADD VARIABLE
Temperature (Integration Point)
Comp 22 of Thermal Strain
ADD VARIABLE
Temperature (Integration Point)
Displacement X
ADD VARIABLE
Temperature (Integration Point)
Displacement Y
FIT
50 100 150 200 250 300 350 400
0.0000
0.0001
0.0002
0.0003
0.0004
0.0005
e2
e1
Displacement Y
Displacement X
Comp 22 of Thermal Strain
Comp 11 of Thermal Strain
0.000339
Temperature ( F)
o
1698 Marc Users Guide: Part 3
CHAPTER 5.4
and the last term in the equation will dominate with changes in temperature. The large compressive stress in the out-
oI-plane direction expands the structure in the in-plane directions, and the displacements per unit length become larger
than the thermal strains in Figure 5.4-5. Plot the out-oI-plane stress, , and see iI it is close to the estimate above.
II there is no out-oI-plane constraint to the thermal growth physically, a plane stress conditions could be used. II the
out-oI-plane thermal growth is restricted, such as plane remaining plane, generalized plane strain elements (types 19,
29, etc.) should be used. For example, transient thermals with large thermal gradients where the out-oI-plane thickness
is large enough to allow out-oI-plane thermal growth can stretch with planes remaining plane should use generalized
plane strain elements. You may wish to try the other elements in the other model Iiles and observe what happens (no
stresses will be generated).
Thermal Expansion Data Reduction
The actual measurements shown in Figure 5.4-6 used two well-matched strain gages, with one bonded to a specimen
oI the reIerence material (Titanium Silicate - thermal expansion assumed to be zero see Vishay |1|), and the second to
a specimen oI the test material (a Iiber reinIorced epoxy composite, Boeing BMS 8-256, along with other materials)
in two directions vertical (Iiber direction) and horizontal. Under stress-Iree conditions, the diIIerential output between
the gages on the two specimens, at any common temperature, is equal to the diIIerential unit expansion usually
reported in micro-strain ( ) or parts per million (ppm). Although the test ran Ior many hours the third recycle is
recorded here and we will only process that data necessary Ior the expansion oI the BMS material that occurred in the
Iirst 70 minutes. The MatLab (Natick, Massachusetts) commands used to reduce the raw data into a Iorm convenient
Ior analysis Iollow below with comments narrating the operations.
Figure 5.4-6 Snippet of Recorded Raw Data
%-- read spread sheet CTETST3 in CTETST3.xls and define local arrays
%-- for first n data points scanned.
n = 70;
o
33
E
3
v
13
E
1
------- o
11
v
23
E
2
------- o
22
o
33
T ( ) T d
T
i
T
}

400 F
o
1.13x10
6
0.000339 ( ) 383.07
o
33
c
BAR
VERT.
BAR
HORZ.
BMS
COMP.
VERT.
BMS
COMP.
HORZ.
Column 2 3 4 5 6 7 8 9 10 11 12
Scan ID F F F F F
1 7/12/06 12:49:34 PM 86.6 88.0 87.4 88.0 88.1 0 0 0 0
2 7/12/06 12:50:34 PM 87.1 88.6 87.6 88.0 88.3 -1 -1 -2 1
3 7/12/06 12:51:34 PM 87.4 89.2 88.0 88.0 88.5 -3 -2 -3 1
4 7/12/06 12:52:34 PM 88.0 89.7 88.3 88.1 88.6 -4 -4 -4.0 -2.0
5 7/12/06 12:53:34 PM 89.7 92.3 89.9 88.6 90.7 -9.9 -8.9 -10 -10
6 7/12/06 12:54:34 PM 92.5 95.9 92.3 89.5 93.1 -20 -19 -20 -20
7 7/12/06 12:55:34 PM 95.9 100.0 95.2 90.4 94.7 -32 -30 -29.8 -29.8
8 7/12/06 12:56:34 PM 99.7 104.2 98.5 91.9 97.1 -44.7 -42.7 -39 -42
9 7/12/06 12:57:34 PM 103.8 108.4 101.8 93.3 100.2 -60 -57 -50 -53
10 7/12/06 12:58:34 PM 108.1 112.6 105.2 95.0 102.3 -75 -72 -60 -63
11 7/12/06 12:59:34 PM 112.6 116.8 109.0 97.1 105.4 -90 -87 -72 -77
--- --- --- --- --- --- --- --- --- --- --- ---
--- --- --- --- --- --- --- --- --- --- --- ---
69 7/12/06 1:58:04 PM 350.0 335.9 339.0 317.2 328.1 -1198 -1186 -798 -845
70 7/12/06 1:59:04 PM 350.0 336.5 340.4 319.5 329.5 -1199 -1188 -800 -848
T/C #1 T/C #2
Date Time
T/C#1
TI/SIL
BAR
REF.
T/C#2
BMS
COUP.
REF.
T/C#3
BMI
TOOL
REF.
T/C#4
ALUM.
TOOL
REF.
T/C#5
INVAR
TOOL
REF.
1699
CHAPTER 5.4
Temperature Dependent Orthotropic Thermal Strains
[ndata, headertext] = xlsread('CTETST3.xls', 'CTETST3');
x1 = ndata(1:n, 4); x1t = 'T/C#1 TI/SIL BAR REF. Temperature F';
y1 = ndata(1:n, 9); y1t = 'T/C#1 BAR VERT. STRN ppm';
y2 = ndata(1:n,10); y2t = 'T/C#1 BAR HORZ. STRN ppm';
x2 = ndata(1:n, 5); x2t = 'T/C#2 COMP. COUP. REF. Temperature F';
y3 = ndata(1:n,11); y3t = 'T/C#2 COMP. VERT. STRN ppm';
y4 = ndata(1:n,12); y4t = 'T/C#2 COMP. HORZ. STRN ppm';
%-- fit gage strain data to polynomials, p and evaluate, f quantify error of polynomial fit order 3.
p1 = polyfit(x1,y1,3); p2 = polyfit(x1,y2,3);
p3 = polyfit(x2,y3,3); p4 = polyfit(x2,y4,3);
f1 =polyval(p1,x1); f2 =polyval(p2,x1);
f3 =polyval(p3,x2); f4 =polyval(p4,x2);
%-- check fit of data
mm1 = (max(f1-y1)-min(f1-y1))/(max(y1)-min(y1));
mm2 = (max(f2-y2)-min(f2-y2))/(max(y2)-min(y2));
mm3 = (max(f3-y3)-min(f3-y3))/(max(y3)-min(y3));
mm4 = (max(f4-y4)-min(f4-y4))/(max(y4)-min(y4));
figure
subplot(2,2,1); plot(x1,f1,x1,y1,'o'); title([y1t,' Error ',num2str(mm1)]);
subplot(2,2,2); plot(x1,f2,x1,y2,'o'); title([y2t,' Error ',num2str(mm2)]);
subplot(2,2,3); plot(x2,f3,x2,y3,'o'); title([y3t,' Error ',num2str(mm3)]);
subplot(2,2,4); plot(x2,f4,x2,y4,'o'); title([y4t,' Error ',num2str(mm4)]);
%-- shift temperature for polynomial roots to agree refit p3,p4 so that
%-- p1=p3=0 at same temperature, same for p2=p4=0.
rp1=roots(p1); rp2=roots(p2); rp3=roots(p3); rp4=roots(p4);
shift42=rp4(3)-rp2(3); shift31=rp3(3)-rp1(3);
p4 = polyfit(x2-shift42,y4,3); p3 = polyfit(x2-shift31,y3,3);
%-- compute thermal strains
g1=p3-p1; g2=p4-p2;
%-- shift so g1 and g2 have same root
rg1=roots(g1); rg2=roots(g2);
shiftg=rg2(3)-rg1(3);
g2s = polyfit(x2-shiftg,polyval(g2,x2),3);
g2 = g2s
%-- plot thermal strain, and alpha
stg1 = ['thermstrn e1 = ' num2str(g1(4)) ' + ' num2str(g1(3)) ' * T + ' num2str(g1(2)) ' * T^2 + ' num2str(g1(1)) ' *
T^3'];
0 100 200 300 400
-1500
-1000
-500
0
500
T/C#1 BAR VERT. STRN ppm Error 0.0048992
0 100 200 300 400
-1500
-1000
-500
0
500
T/C#1 BAR HORZ. STRN ppm Error 0.0060918
0 100 200 300 400
-1000
-500
0
500
T/C#2 COMP. VERT. STRN ppm Error 0.0090716
0 100 200 300 400
-1000
-500
0
500
T/C#2 COMP. HORZ. STRN ppm Error 0.012554
1700 Marc Users Guide: Part 3
CHAPTER 5.4
stg2 = ['thermstrn e2 = ' num2str(g2(4)) ' + ' num2str(g2(3)) ' * T + ' num2str(g2(2)) ' * T^2 + ' num2str(g2(1)) ' *
T^3'];
a1=polyder(g1); a2=polyder(g2);
sta1 = ['alpha1 = ' num2str(a1(3)) ' + ' num2str(a1(2)) ' * T + ' num2str(a1(1)) ' * T^2'];
sta2 = ['alpha2 = ' num2str(a2(3)) ' + ' num2str(a2(2)) ' * T + ' num2str(a2(1)) ' * T^2'];
figure
subplot(2,1,1);plot(x1,polyval(g1,x1),x1,polyval(g2,x1));title({stg1;stg2});legend('e1','e2','Location','Southeast');
subplot(2,1,2);plot(x1,polyval(a1,x1),x1,polyval(a2,x1));title({sta1;sta2});legend('a1','a2','Location','Southeast');
%-- output results into excel worksheet
d = {'Temp F', stg1, stg2, sta1, sta2}
xlswrite('tempdata.xls', d, 'Results', 'A1');
xlswrite('tempdata.xls', x1, 'Results', 'A2');
xlswrite('tempdata.xls', polyval(g1,x1), 'Results', 'B2');
xlswrite('tempdata.xls', polyval(g2,x1), 'Results', 'C2');
xlswrite('tempdata.xls', polyval(a1,x1), 'Results', 'D2');
xlswrite('tempdata.xls', polyval(a2,x1), 'Results', 'E2');
The comments and Iigures generated by the MatLab commands should be suIIicient Ior one to Iollow the calculations.
Because two thermal couples are used Ior the reIerence and unknown material, their temperatures are not exactly the
same at the same time. This required shiIting the data to account Ior this diIIerence. Furthermore, it is quite common
(particularly Ior more precise data measurements Irom optical heterodyne interIerometry |3 and 4|) to use cubic
polynomials Ior the thermal strain data Iits as well as reporting the data. BeIore any two polynomials are added or
subtracted, they are adjusted such that the thermal strain is zero at the same temperature; this minor adjustment
prevents introducing numerical artiIacts into the thermal strain and instantaneous coeIIicients oI thermal expansion.
The output Irom the data reductions is place into an Excel shown in Figure 5.4-7.
50 100 150 200 250 300 350 400
-200
0
200
400
thermstrn e1 = -73.2595 + 0.6467 * T + 0.0024556 * T
2
- 2.6494e-006 * T
3
thermstrn e2 = -42.0634 + 0.26486 * T + 0.0027178 * T
2
- 2.5128e-006 * T
3
e1
e2
50 100 150 200 250 300 350 400
0.5
1
1.5
alpha1 = 0.6467 + 0.0049112 * T - 7.9482e-006 * T
2
alpha2 = 0.26486 + 0.0054356 * T - 7.5383e-006 * T
2
a1
a2 Temperature ( F)
ppm
ppm/ F
o
o
Temperature ( F)
o
1701
CHAPTER 5.4
Temperature Dependent Orthotropic Thermal Strains
Figure 5.4-7 Snippet of Output from Data Reduction
The actual part that was used in these measurement is a large cylinder comprised oI this Iiber reinIorced material
shown in Figure 5.4-8.
Figure 5.4-8 Fiber Reinforced Cylinder with Thermal Couples and Strain Gages
References
1. Vishay Micro-Measurements (2007) Measurement oI Thermal Expansion CoeIIicient Using Strain Gages.
Tech Note TN-513-1. http://www.vishay.com/doc?11063~.
2. ASTM, E 228-06 (2006) Standard Test Method Ior Linear Thermal Expansion oI Solid Materials With a
Push-Rod Dilatometer. Annual Book of ASTM Standards.
3. De Bona E, Som A (1997) Thermal Expansion Measurement oI Composites with Optical Heterodyne
InterIerometry. Experimental Mechanics, 37(1):21-25.
4. Chanchani R, Hall P M (1990) Temperature dependence oI thermal expansion oI ceramics and metals Ior
electronic packages. IEEE Trans Comp, Hybrids, Manufact Technol 13:743750.
Temp F
thermstrn e1 =-
73.2595 +0.6467 * T
+0.0024556 * T^2 +-
2.6494e-006 * T^3
thermstrn e2 =-
42.0634 +0.26486 *
T +0.0027178 * T^2
+-2.5128e-006 * T^3
alpha1 =0.6467 +
0.0049112 * T +-
7.9482e-006 * T^2
alpha2 =0.26486 +
0.0054356 * T +-
7.5383e-006 * T^2
86.572 -0.58859224 -0.39505678 1.012299076 0.678934005
87.091 -0.062733297 -0.042134101 1.014131586 0.681075639
87.437 0.288367224 0.193764733 1.015350881 0.682501139
87.957 0.816825379 0.549221636 1.017179767 0.684640114
89.686 2.58076639 1.73909395 1.023229908 0.691722894
92.45 5.42223188 3.666561037 1.032803028 0.702951893
95.9 9.005774959 6.115695376 1.044581715 0.716806221
99.688 12.98680421 8.859445039 1.05729645 0.731811188
103.81 17.37305449 11.9092042 1.070873133 0.747893404
108.1 21.99689848 15.15308489 1.084716327 0.764359041
112.55 26.85526492 18.59194341 1.098766687 0.781145593
BMS
Invar
Al
Gage 2
Gage 1
Leadwire
Gage 2
Leadwire
1702 Marc Users Guide: Part 3
CHAPTER 5.4
Section 6: Miscellaneous Analysis
Section 6: Miscellaneous
Analysis
Marc Users Guide -1704
Chapter 6.1: Magnetostatics: Analysis of a Transformer
6.1
Magnetostatics: Analysis of
a Transformer

Chapter Overview 1706

3-D Analysis of a Transformer 1706

Input Files 1718

1706 Marc Users Guide: Part 3
CHAPTER 6.1
Chapter Overview
The increasing need Ior analysis oI real world magnetostatic applications has been the motivation to revisit the
magnetostatic vector potential Iormulation used in Marc. Compared to the 2-D Iormulation, in 3-D an additional
constraint has to be taken into account to get a uniquely deIined vector potential, which is done using a penalty
Iormulation. The value oI the penalty Iactor is crucial to get accurate results in 3-D applications. This chapter describes
a 3-D analysis oI a transIormer. A coil is placed around an iron Irame. By Iorcing a current in the coil, a magnetic Iield
is generated inside the iron. The coil is assumed to have the same permeability as air, and is modeled using Iace
currents on selected elements. The transIormer is modeled with a region oI air around the iron. Due to symmetry, only
one eighth oI the transIormer is modeled. The magnetic permeability oI iron is taken constant, which is valid Ior low
currents in the coil. Brick element 109 is used both Ior the air and the iron region.
3-D Analysis of a Transformer
The magnetic Iield in and around a transIormer is computed. Figure 6.1-1 shows an outline oI the transIormer.
Figure 6.1-1 Transformer: Problem Description
A current is Ilowing through a coil around the center oI the iron, thus inducing a magnetic Iield inside the iron. The
shape oI the iron is like a Iigure eight with the coil around the center. In this way, most oI the magnetic Iield generated
by the coil stays inside the iron. Figure 6.1-2 shows the part which is modeled.
iron
0.20
0.20
0.12
coil with current
0.08
1707
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
Figure 6.1-2 Transformer: FE Model of Iron and Air
Mesh Generation
The mesh is generated by Iirst deIining surIaces Ior the iron and the air. Then these surIaces are converted to elements,
where a more reIined mesh is used Ior the iron parts. Then the mesh is expanded in the z-direction considering a
reIinement Ior the iron region. Nodes on the outer surIaces are stored in sets so they can be easily used when deIining
boundary conditions.
MESH GENERATION
srfs ADD
0 0 0
0.02 0 0
0.02 0.12 0
0 0.12 0
0.02 0 0
0.08 0 0
0.08 0.12 0
0.02 0.12 0
0.08 0 0
0.1 0 0
0.1 0.12 0
0.08 0.12 0
0.1 0 0
A
B
C
1708 Marc Users Guide: Part 3
CHAPTER 6.1
0.2 0 0
0.2 0.12 0
0.1 0.12 0
CONVERT
DIVISIONS
3 12
SURFACES TO ELEMENTS
1 3 #
DIVISIONS
5 12
SURFACES TO ELEMENTS
2 4 #
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
EXPAND
SHIFT
SCALE FACTORS
1 1 14/15
TRANSLATIONS
0 0 0.015
REPETITIONS
3
ELEMENTS
all: EXIST.
SCALE FACTORS
1 1 1
TRANSLATIONS
0 0 0.0165
REPETITIONS
2
1709
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
ELEMENTS
1 TO 192 #
TRANSLATIONS
0 0 0.02/3
REPETITIONS
3
ELEMENTS
1 TO 192 #
REMOVE
TRANSLATIONS
0 0 0.105
REPETITIONS
1
ELEMENTS
1 TO 192 #
FILL
RETURN
SUBDIVIDE
DIVISIONS
1 1 6
BIAS FACTORS
0 0 -0.3
ELEMENTS
1729 TO 1920 #
RETURN
SWEEP
ALL
RETURN
SELECT
METHOD
BOX
RETURN
ELEMENTS
-0.0001 0.0201
-0.0001 0.0601
-0.0001 0.0951
1710 Marc Users Guide: Part 3
CHAPTER 6.1
0.0799 0.1001
-0.0001 0.0601
-0.0001 0.0951
0.0199 0.0801
-0.0001 0.0601
0.0750 0.0951
STORE
iron
all: SELECT.
CLEAR SELECT
NODES
-0.001 0.001
-0.001 1
-0.001 1
STORE
fix_yz
all: SELECT.
CLEAR SELECT
NODES
-0.001 1
-0.001 0.001
-0.001 1
STORE
fix_xz
ALL SELECT.
CLEAR SELECT
NODES
-0.001 1
-0.001 1
-0.001 0.001
STORE
fix_z
all: SELECT.
CLEAR SELECT
NODES
-0.199 0.201
1711
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
Boundary Conditions
The Iollowing boundary conditions are applied on components oI the vector potential. The Iaces are indicated in
Figure 6.1-2. On 'A (y 0.0) A
1
A
3
0, on 'B (x 0.0) A
2
A
3
0 and on 'C ( ) . So where
current is Ilowing, the magnetic potential A is Iorced to Iollow its direction. Assuming that the amount oI air is
suIIiciently large, at the outer surIace A
1
A
2
A
3
0. A current oI 5000A/m
2
is applied as a Iace current on the
Iaces oI the elements with air properties which are next to the iron in the center oI the transIormer. (See Figure 6.1-3
Ior the direction oI the current).
-0.001 1
-0.001 1
STORE
fix_surfaceA
all SELECT.
CLEAR SELECT
NODES
-0.001 1
-0.001 1
-0.199 0.201
STORE
fix_surfaceB
all SELECT.
CLEAR SELECT
NODES
-0.001 1
0.119 0.121
-0.001 1
STORE
fix_surfaceC
all: SELECT.
BOUNDARY CONDITIONS
MAGNETOSTATIC
NEW
NAME
current_x
z 0.0 A
3
0
1712 Marc Users Guide: Part 3
CHAPTER 6.1
FACE CURRENT
U TANGENTIAL
-5000
OK
faces ADD
247:0 250:0 253:0 254:0 251:0 248:0 249:0 252:0 255:0 #
NEW
NAME
current_y
FACE CURRENT
U TANGENTIAL
-5000
OK
faces ADD
484:3 469:3 454:3 439:3 424:3 409:3 485:3 470:3 455:3
440:3 425:3 410:3 486:3 471:3 456:3 441:3 426:3 411:3 #
NEW
NAME
fix_yz
FIXED POTENTIAL
POTENTIAL Y (on)
POTENTIAL Z (on)
OK
nodes ADD
fix_yz #
NEW
NAME
fix_xz
FIXED POTENTIAL
POTENTIAL X (on)
POTENTIAL Z (on)
OK
nodes ADD
fix_xz #
1713
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
Figure 6.1-3 The Current which is flowing in the Coil of the Transformer
NEW
NAME
fix_z
FIXED POTENTIAL
POTENTIAL Z (on)
OK
ADD NODES
fix_z #
NEW
NAME
fix_outside
FIXED POTENTIAL
POTENTIAL X (on)
POTENTIAL Y (on)
POTENTIAL Z (on)
OK
nodes ADD
fix_surfaceA fix_surfaceB fix_surfaceC #
1714 Marc Users Guide: Part 3
CHAPTER 6.1
Material Properties
For the air, the magnetic permeability is 1.2566 x 10
-6
Hm
-1
, and Ior the iron, 0.005867 Hm
-1
. So in this
example, a linear dependence between the magnetic induction B and the magnetic Iield intensity H is considered
Ior iron.
Loadcases and Job Parameters
The analysis is a steady state simulation to obtain the magnetic Iield inside and outside the iron. So one loadcase is
deIined. Element type 109 is used both Ior air and iron. The analysis runs with the deIault settings. The penalty, by
deIault is 0.0001 was Iound to be adequate. Components oI the magnetic induction and the magnetic Iield intensity
are selected as element quantities to be written on the post Iile.
MATERIAL PROPERTIES
NEW
NAME
air
MAGNETOSTATIC
PERMEABILITY
1.25664E-6
OK
elements ADD
all: EXIST.
NEW
NAME
iron
MAGNETOSTATIC
PERMEABILITY
0.005867
OK
elements ADD
iron #
LOADCASES
NEW
MAGNETOSTATIC
STEADY STATE
OK
1715
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
Save Model, Run Job, and View Results
The analysis can now start. The resulting plot oI the real magnetic induction is shown in Figure 6.1-4. The Iigure shows
that the magnetic induction is concentrated inside the iron. There is very little leakage to the environment.
MAIN
JOBS
ELEMENT TYPES
MAGNETOSTATIC
3-D
109
OK
all: EXIST.
RETURN (twice)
NEW
MAGNETOSTATIC
LCASE1
JOB RESULTS
Magnetic Potential
1st Real Comp Magnetic Induction
2nd Real Comp Magnetic Induction
3rd Real Comp Magnetic Induction
1st Real Comp Magnetic Field Intensity
2nd Real Comp Magnetic Field Intensity
3rd Real Comp Magnetic Field Intensity
OK (twice)
FILE
SAVE AS
transformer.mud
OK
RETURN
RUN
SUBMIT 1
MONITOR
OK
1716 Marc Users Guide: Part 3
CHAPTER 6.1
Figure 6.1-4 Real Magnetic Induction in the Transformer induced by the Current
MAIN
RESULTS
OPEN DEFAULT
NEXT
MORE
VECTOR PLOT ON
VECTOR
Real Magnetic Induction
OK
RETURN
DEF & ORIG
PLOT
NODES (off)
elements SETTINGS
draw EDGES (off)
RETURN (twice)
1717
CHAPTER 6.1
Magnetostatics: Analysis of a Transformer
Figure 6.1-5 Magnetic Field Intensity along the Path specified in Figure 6.1-2
Figure 6.1-5 shows the magnetic Iield intensity oI the transIormer along the line as indicated in Figure 6.1-2 (Ior
comparison see also Zienkiewicz, O.C., Lyness, J., Owen, D.R.J., IEEE Transactions on Magnetics, VOL 13, No 5,
1649-1656 (1977)). The peak value oI the magnetic Iield intensity and the values Ior the regions inside the iron
correspond well with that Iound by Zienkiewicz et. al.
RESULTS
OPEN DEFAULT
NEXT
scalar plot SETTINGS
EXTRAPOLATION
AVERAGE
RETURN (twice)
PATH PLOT
NODE PATH
224 1344 #
VARIABLES
ADD CURVE
Arc Length
Real Magnetic Field Intensity
FIT
1718 Marc Users Guide: Part 3
CHAPTER 6.1
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
transformer.proc Mentat procedure Iile to run the above example
Chapter 6.2: Fracture Mechanics Analysis with the J-integral
6.2
Fracture Mechanics Analysis
with the J-integral

Chapter Overview 1720

Specimen with an Elliptic Crack 1720

Background Information 1720

Run Job and View Results 1739

Input Files 1741

1720 Marc Users Guide: Part 3
CHAPTER 6.2
Chapter Overview
In Iracture mechanics applications, it is oIten necessary to evaluate the so-called J-integral Ior investigating cracks in
a structure. This chapter demonstrates the modeling and evaluation oI the J-integral in an analysis oI a solid block with
an elliptic surIace crack. The example also illustrates the use oI the new load controlled die options and how to create
a parameterized procedure Iile.
Specimen with an Elliptic Crack
This example is a solid steel block with an elliptic surIace crack, see Figure 6.2-1. It is loaded in bending and tension
and clamped at both ends where the loading is applied. The purpose with the analysis is to evaluate the J-integral along
the crack Iront. This is now Ieasible with the improvements to the J-integral option and some other Ieatures oI Marc
and Mentat as is shown below. It is oI particular interest to obtain a local value oI J along the crack Iront.
Background Information
Figure 6.2-1 Problem Description
The geometry shown in Figure 6.2-1 is studied in this example. A solid block with an elliptic surIace crack is loaded
in tension and bending. The material is linear elastic and no nonlinear eIIects are taken into account. Double symmetry
is taken into account so only a Iourth oI the structure is modeled.
L
h
2b
2a1
a2
F
F, M
M
1721
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
The Iollowing values were used in the analysis:
Material, linear elastic:
Loading:
Modeling Strategies
The basic geometry is created using the solid modeler. The region around the crack is then taken out Irom the solid
model so that a Iocused mapped mesh can be put in there. This way, a crack region can be inserted into a general solid
model in diIIerent places. The region outside the crack region is meshed using the automatic tetrahedral mesher while
a mapped mesh is used in the crack region. These two parts do not Iit together node to node so they are connected
using the GLUE option in the contact Ieature oI Marc. For this example, there is no real contact between diIIerent
bodies, the CONTACT option is only used to automatically create the appropriate tyings between the diIIerent parts oI
the model.
Since we want to make use oI symmetry, we need to be able to apply symmetry boundary conditions to the crack
ligament but not to the Iree crack surIace. To accomplish this, we divide the mesh around the crack Iront into two
separate parts which are glued together. The symmetry boundary condition is applied via a rigid contact body so by
using a contact table, we can speciIy that only the crack ligament part should contact the symmetry plane.
This example is also illustrating the use oI parameterized modeling. The source oI the model is the procedure Iile.
When the model is modiIied or parts oI it are replaced, this is all done in the procedure Iile. One key thing Ior
accomplishing this is to make the diIIerent parts modular. One part oI the procedure Iile must not depend on previous
parts. This means that points, nodes, elements, etc. should not be reIerenced explicitly. In this example, we
systematically make previously deIined parts invisible and supply lists with all_visible instead oI picking
elements Irom the screen. We also regularly delete the current geometry when a part oI the meshing is Iinished so that
the numbering oI new points, etc. starts with one. An alternative way is to renumber so that consecutive numbering is
used and make use oI the predeIined variables npoints(), nnodes(), etc. The second latest created point would
have the number npoints()-1. An obvious way to make the model parameterized is to deIine a number oI variables
at the top oI the procedure Iile and make use oI these throughout the procedure Iile.
L 15
b 8
h 4
a
1
1.5
a
2
1
E 2.0 10
5

v 0.3
F 5.0 10
5

M 1.0 10
5

1722 Marc Users Guide: Part 3
CHAPTER 6.2
Mesh Generation
The mesh Ior the region outside the crack is generated by deIining a solid block and subtracting a part around the crack.
Later, a mapped mesh will be used Ior the crack region.
First, we deIine some parameters to use in the procedure Iile:
Then, we deIine the main block and a cylinder that is subtracted Irom the block. The cylinder is deIined at the origin
and then scaled using the quotient a
2
/a
1
as a scaling Iactor. The ability to scale solids in this way is a new Ieature in
this release.
UTILS
PARAMETERS
h 4
b 8
L 15
a1 1
a2 1.5
force -5e5
moment -1e5
nexpand 22
da 0.5
MESH GENERATION
SOLID TYPE
BLOCK
SOLIDS ADD
0 0 0 b h -L
SOLID TYPE
CYLINDER
SOLIDS ADD
0 0 0 0 0 -1 1.5 1.5
MOVE
RESET
SCALE FACTORS
a2/a1 1 1
SOLIDS
2 #
RESET
1723
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
The model so Iar is shown in Figure 6.2-2. Now this solid model is meshed using the tetrahedral automatic mesher.
First, all Iaces oI the solid are converted into surIaces and then a curve division length oI 1 is applied to all curves.
The curves close to the crack region are given a Iiner division oI 0.3. The solids are removed since they are not
needed anymore.
TRANSLATIONS
b h 0
SOLIDS
2 #
SOLIDS
SUBTRACT
1 2 #
SOLID FACES TO SURFACES
all: EXIST.
SOLIDS REM
all: EXIST.
AUTOMESH
CURVE DIVISIONS
FIXED AVG LENGTH
AVG LENGTH
1
APPLY CURVE DIVISIONS
all: EXIST.
FIXED AVG LENGTH
AVG LENGTH
0.3
APPLY CURVE DIVISIONS
4 36 1 31 3 8 6 9 2 7 5 14 #
1724 Marc Users Guide: Part 3
CHAPTER 6.2
Figure 6.2-2 Initial Solid Model
The surIaces are meshed using the Delaunay surIace mesher to produce a triangular mesh oI the exterior. This mesh
is then used by the tetrahedral mesher to create the 3-D mesh. First, the surIaces are matched together so that a
continuous mesh is obtained. The geometry is deleted when the mesh is created. It is not needed anymore and it also
makes it easier to deIine the geometry Ior other parts later on. A mesh transition oI 1.1 is used in the tetrahedral mesher
to obtain a coarser mesh towards the interior oI the body. This new transition Ieature allows Iewer elements to be
created. The elements created so Iar are store in an element set called tets.
MATCH CURVE DIVISIONS
0.01
all: EXIST.
SURFACE MESHING
TRANSITION
1
SURFACE TRI MESH! (Delaunay)
all: EXIST.
CLEAR GEOM
SWEEP
TOLERANCE
0.01
SWEEP NODES
ALL: EXIST.
REMOVE UNUSED NODES
AUTOMESH
SOLID MESHING
1725
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
The mesh is shown in Figure 6.2-3.
Figure 6.2-3 Finite Element Mesh of the Region Outside the Crack
Note that this part is very general. In this example, a simple geometry is used but the steps to create this mesh are
equally simple Ior a more general solid model.
Now with the crack region, we want a reIined mesh Iocused around the crack Iront. The crack region is divided into
three parts; one part which is part oI the ligament, one part containing the crack, and part oI the Iree surIace and a Iill
part. The Iirst part is kept separate Irom the other parts while coincident nodes are swept between the second and third
parts. The meshes Ior these parts are created by deIining a plane mesh and sweeping it. BeIore they are moved into
position, the parts are scaled using the same scale Iactor a
2
/a
1
as Ior the cylindrical solid part.
ALIGN SHELLS
1
TRANSITION
1.1
TET MESH!
ALL: EXIST.
SELECT
ELEMENTS STORE
tets
ALL: EXIST.
CLEAR SELECT
MAKE VISIBLE
1726 Marc Users Guide: Part 3
CHAPTER 6.2
Note that the previous mesh was made invisible so all_visible can be used in the part below. This makes it easy
to modiIy this part oI the procedure Iile iI needed. The mesh around the crack Iront is Iocused with collapsed elements
at the crack Iront. To create the collapsed mesh, a ruled surIace is created Irom two lines. One oI the lines is collapsed
into a point (point number 1 in the procedure Iile) so that a collapsed surIace is obtained. This surIace is then simply
converted into elements.
MESH GENERATION
PTS ADD
0 -1.0 0
0 -1.5 0
0 -1.0 -1.0
0 -1.5 -1.0
0 -0.5 0
0 -0.5 -1.0
CURVE TYPE
POLYLINE
CRVS ADD
2 4 3 #
1 1 #
3 6 5 #
SURFACE TYPE
RULED
SURFACE ADD
1 2 #
2 3 #
CONVERT
DIVISIONS
10 10
SURFACES TO ELEMENTS
1 #
EXPAND
RESET
ROTATIONS
0 0 -90/nexpand
REPETITIONS
nexpand
ELEMENTS
1727
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
The Iill part is meshed using the quadrilateral planar automatic mesher to create a planar mesh which is then expanded
into 3-D elements.
ALL: VISIBLE
SELECT
ELEMENTS STORE
crack1
all: VISIBLE
SWEEP
TOLERANCE
0.001
SWEEP NODES
all: VISIBLE
SELECT
CLEAR SELECT
MAKE VISIBLE
CONVERT
SURFACES TO ELEMENTS
2 #
EXPAND
ELEMENTS
all: VISIBLE
SELECT
ELEMENTS STORE
crack2
all: VISIBLE
SWEEP
TOLERANCE
0.001
SWEEP NODES
all: VISIBLE
SELECT
CLEAR SELECT
MAKE VISIBLE
MESH GENERATION
CLEAR GEOM
1728 Marc Users Guide: Part 3
CHAPTER 6.2
PTS ADD
0 -0.5 0
0 0 0
CURVE TYPE
CENTER/POINT/POINT
crvs ADD
0 0 0
-0.5 0 0
0 -0.4 0
CURVE TYPE
LINE
crvs ADD
2 3 2 1
AUTOMESH
CURVE DIVISION
FIXED # DIVISIONS
# DIVISIONS
10
APPLY CURVE DIVISIONS
2 3 #
# DIVISIONS
nexpand
APPLY CURVE DIVISIONS
1 #
2D PLANAR MESHING
TRANSITION
1
SELECT
ELEMENTS
all: EXIST.
MAKE INVISIBLE
QUAD MESH!
2 3 1 #
EXPAND
RESET
TRANSLATIONS
0 0 -1/5
REPETITIONS
5
ELEMENTS
all: VISIBLE
1729
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
SELECT
ELEMENTS STORE
fill
all: VISIBLE
SWEEP
REMOVE UNUSED NODES
CLEAR GEOM
SELECT
SELECT SET
tets
MOVE
RESET
SCALE FACTORS
a2/a1 1 1
ELEMENTS
ALL: UNSEL.
RESET
TRANSLATIONS
b h 0
ELEMENTS
all: UNSEL.
SELECT
CLEAR SELECT
SELECT SET
crack2
fill
MAKE VISIBLE
SWEEP
TOLERANCE
0.005
SWEEP NODES
ALL: VISIBLE
SELECT
CLEAR SELECT
MAKE INVISIBLE
1730 Marc Users Guide: Part 3
CHAPTER 6.2
Now it Iits into the other mesh as shown in Figure 6.2-4.
Figure 6.2-4 The Completed Mesh
Crack Definitions
The mesh is now Iinished and the next step is to deIine the crack parameters. In order to automatically identiIy the
crack Iront, an arc at that location is deIined and the nodes at the crack Iront are identiIied using the ATTACH option.
The crack Iront consists oI nodes at the lower part oI the crack mesh as well as the upper part. As mentioned above,
these nodes are glued to each other using the CONTACT option. It is important that the crack nodes belonging to the
lower part are used since they will be part oI the crack ligament and thus have symmetry boundary conditions. The
inIormation about the boundary conditions is used when the shiIt directions Ior the J-integral evaluation are
determined. A small (but nonzero) value is used Ior the Multiple Tip Nodes option. This is to make sure that also
the nodes in the upper crack part are part oI the crack Iront so that the rigid regions Ior the J-evaluation can be
determined properly.
SELECT SET
crack1
MAKE VISIBLE
CURVE TYPE
CENTER/POINT/POINT
CRVS ADD
0 0 0
-1 0 0
0 -1 0
1731
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
MOVE
RESET
SCALE FACTORS
a2/a1 1 1
CURVES
ALL: EXIST.
RESET
TRANSLATIONS
b h 0
CURVES
ALL: EXIST.
ATTACH
MODE CLOSEST
LIMIT ON
DISTANCE
0.01
TOLERANCE
1e-5
EDGES -> CURVE
1
ALL: VISIBLE
SELECT
SELECT BY
NODES BY CURVES
1 #
NODES STORE
crackfront
all: SELECT.
CLEAR SELECT
MAKE INVISIBLE
FRACTURE MECHANICS
3-D CRACKS
NEW
AUTOMATIC (TOPOLOGY SEARCH)
1732 Marc Users Guide: Part 3
CHAPTER 6.2
The new menu Ior deIining these parameters is shown in Figure 6.2-5.
Figure 6.2-5 Menu for Defining Crack Properties
We use the deIault method oI deIining the integration paths Ior the J-integral. For each crack Iront node, there will be
a number oI paths deIined (as speciIied by the RIGID REGIONS option) with increasing size. Thus, Ior each crack Iront
node, there will be this number oI evaluations with diIIerent path radii. Analytically, these values should be identical
since the J-integral is path independent in a linear elastic analysis, but due to the discretization they will in general
vary with the radius. This can be used as an indicator oI the accuracy oI the solution. II the variation is large, the results
are probably inaccurate. With the geometry search method, one speciIies one radius Ior each crack Iront node and the
rigid region is deIined by all nodes inside a cylinder aligned with the crack Iront. The MANUAL option allows the nodes
oI the rigid region to be speciIied explicitly.
RIGID REGIONS
7
CRACK TIP NODE PATH SET
crackfront
DISTANCE
0.001
1733
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
Material Properties
This is a simple part. All elements have the same linear elastic material.
Contact Definitions
The mesh currently consists oI three unconnected regions and they need to be connected to each other. This is
accomplished by deIining them as contact bodies and use the GLUE option to tie them together.
MATERIAL
NEW
ISOTROPIC
YOUNG'S MODULUS
2e5
POISSON'S RATIO
0.3
ELEMENTS ADD
all: EXIST.
CONTACT
CONTACT BODIES
NEW
NAME
crack2
DEFORMABLE
OK
ELEMENTS ADD
crack2
fill
NEW
NAME
crack1
DEFORMABLE
OK
ELEMENTS ADD
crack1
NEW
NAME
1734 Marc Users Guide: Part 3
CHAPTER 6.2
Here, we used the previously deIined set names to select elements Ior the contact bodies and avoid giving element
lists. A very important point here is the order in which the bodies are deIined. As usual when using deIormable contact
in Marc, the Iiner bodies are deIined beIore the coarser. That is why the body with the tetrahedral elements is deIined
last. The other two bodies meet node to node, but it is still important to deIine the top part Iirst. The nodes oI the Iirst
body will be tied to the segments oI the second body. Only nodes oI the lower part oI the crack will contact the
symmetry plane at the crack ligament, so they must not be tied to the nodes oI the top part. II the order oI the deIinition
oI the Iirst two contact bodies above is switched, the crack Iront nodes will not stay in contact with the symmetry plane.
BeIore deIining the contact table with the glue entries, the rigid contact surIaces used Ior applying the loads and
boundary conditions are deIined:
tets
DEFORMABLE
OK
ELEMENTS ADD
tets
MESH GENERATION
CLEAR GEOM
PTS ADD
-10 -10 -L
-10 20 -L
18 -10 -L
18 20 -L
SURFACE TYPE QUAD
SRFS ADD
2 1 3 4
PTS ADD
b -10 10
b -10 -25
b 20 -25
b 20 1
SURFACE ADD
6 7 8 5
DUPLICATE
RESET
TRANSLATE
0 0 L
1735
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
For this body, we are using the new modiIied LOAD CONTROL option. One node each is used Ior the Iorce and
moment. In order to be able to reIer to these nodes without using node numbers (which will change iI the model above
is modiIied), a nodal renumbering is made and we reIer to the new nodes using the predeIined variable nnodes().
SURFACES
1 #
CHECK
FLIP SURFACES
1 #
CONTACT
CONTACT BODIES
NEW
NAME
moving
RIGID
LOAD
SURFACES ADD
1 #
MESH GENERATION
RENUMBER
NODES
NODES ADD
b/2 h/2 -L
b/2 h/2 -L
CONTACT
CONTACT BODIES
CONTROL NODE
nnodes()-1
AUX. NODE
nnodes()
BOUNDARY CONDITIONS
NEW
NAME
force
MECHANICAL
1736 Marc Users Guide: Part 3
CHAPTER 6.2
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
NAME
pointloads
ADD
0 0 1 1
SHOW TABLE
SHOW MODEL
POINT LOAD
ON Z FORCE
Z FORCE
force
TABLE
pointloads
NODES ADD
nnodes()-1 #
NEW
NAME
moment
POINT LOAD
ON X FORCE
X FORCE
moment
TABLE
pointloads
NODES ADD
nnodes() #
CONTACT
CONTACT BODIES
NEW
NAME
symmx
SYMMETRY
1737
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
Again, since the previous geometry was deleted, we can reIer to points 1 through 8 regardless oI how many points that
were previously created. Now to the contact table:
SURFACES ADD
2 #
NEW
NAME
cracksym
SYMMETRY
SURFACES ADD
3 #
CONTACT
CONTACT TABLES
NEW
PROPERTIES
BODIES
crack2
tets
NO CONTACT -> TOUCHING -> GLUED
SEPARATION FORCE
1e32
BODIES
crack2
crack1
NO CONTACT -> TOUCHING -> GLUED
SEPARATION FORCE
1e32
BODIES
tets
crack1
NO CONTACT -> TOUCHING -> GLUED
SEPARATION FORCE
1e32
BODIES
tets
1738 Marc Users Guide: Part 3
CHAPTER 6.2
Now only the load case and job options need to be deIined. Since it is a one-step analysis with contact as the only
source oI nonlinearity, these options are simple. Actually, the analysis is still linear since there will be no change in
contact status. We only use contact to apply boundary conditions and tyings. We make sure to activate the contact table
both in the load case and the job (initial contact). A contact tolerance oI 0.01 ('distance below which a node is assumed
to be in contact) is used to assure that the nodes oI the crack part are properly glued to the tetrahedral elements. For
eIIiciency, we use the iterative solver.
moving
NO CONTACT -> TOUCHING -> GLUED
SEPARATION FORCE
1e32
BODIES
crack1
symmx
NO CONTACT -> TOUCHING
BODIES
crack2
symmx
NO CONTACT -> TOUCHING
BODIES
tets
symmx
NO CONTACT -> TOUCHING
BODIES
crack1
cracksym
NO CONTACT -> TOUCHING
BODIES
tets
cracksym
NO CONTACT -> TOUCHING
LOADCASES
NEW
MECHANICAL
STATIC
CONTACT
1739
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
Run Job and View Results
CONTACT TABLE
ctable1
JOBS
NEW
MECHANICAL
LOADCASES
lcase1
CONTACT CONTROL
INITIAL CONTACT
CONTACT TABLE
ctable1
DISTANCE TOLERANCE
0.01
JOB RESULTS
available element tensors
Stress
available element scalars
Equivalent Von MIses Stress
JOB PARAMETERS
SOLVER
ITERATIVE SPARSE
INCOMPLETE CHOLESKI
JOBS
RUN
SUBMIT 1
MONITOR
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
1740 Marc Users Guide: Part 3
CHAPTER 6.2
Figure 6.2-6 Equivalent von Mises Stress Contours
When the analysis is done we can look at the deIormed shape and the stress Iield. The actual J-integral results are
printed to the output Iile. The EDIT OUTPUT FILE button in the Run menu will bring up the output Iile in the deIault
editor. Locate the string 'J-integral estimations in this Iile. Table 6.2-1 shows the values obtained. The results are
given Ior each crack deIined and are grouped Ior each crack Iront node. Figure 6.2-1 shows the results Ior the Iirst two
crack tip nodes. There is a small path dependency, which indicates that one should use a Iiner mesh should or
higher-order elements around the crack. This would be simple to change since the model is parameterized.
Equivalent Von Mises Stress
CONTOUR BANDS
NEXT
Table 6.2-1 J-integral Estimations
Crack Tip Node Path Radius J-integral Value
1235 1.1180E-01 3.1508E03
1235 2.2361E-01 3.2027E03
1235 3.3541E-01 3.1918E03
1235 4.4721E-01 3.1984E03
1235 5.5902E-01 3.1961E03
1235 6.7082E-01 3.1947E03
1235 7.8262E-01 3.1934E03
1741
CHAPTER 6.2
Fracture Mechanics Analysis with the J-integral
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
3656 1.1187E-01 3.1553E03
3656 2.2375E-01 3.2247E03
3656 3.3562E-01 3.2335E03
3656 4.4750E-01 3.2378E03
3656 5.5937E-01 3.2338E03
3656 6.7125E-01 3.2276E03
3656 7.8312E-01 3.2216E03
File Description
fracture_mech.proc Mentat procedure Iile to run the above example
Table 6.2-1 J-integral Estimations
Crack Tip Node Path Radius J-integral Value
1742 Marc Users Guide: Part 3
CHAPTER 6.2
Chapter 6.3: FEM Simulation of NC Machining Process
6.3
FEM Simulation of NC
Machining Process

Chapter Overview 1744

Input Data 1744

Model Generation 1745

Visualization of Results 1754

Input Files 1758

1744 Marc Users Guide: Part 3
CHAPTER 6.3
Chapter Overview
In the manuIacturing industry, NC machining is a material removal process that is widely used to produce a Iinal part
with desired geometry. AIter removal oI the machined material, re-establishment oI equilibrium within the remaining
part oI the structure causes distortion due to the relieI oI insitu residual stresses. The deIormation caused by this
process usually depends on the residual stress magnitude and its distribution inside the part. It also depends on the Iinal
geometry oI the part aIter machining. For Iinal geometries that include thin walls or large plate structures, the
deIormation can be so large that it causes severe distortions oI the part. The highly distorted part may no longer be
able to serve its designated Iunctionality or may require signiIicant reworking to render it Iunctional. These kinds oI
Iailures result in high scrap rates and increased manuIacturing costs.
The Iinite element procedure (FEM) is a powerIul tool to assess potential distortions caused by the machining process.
With the FEM results, it is then possible Ior engineers to predict the potential Iailures and reduce overall
manuIacture costs.
This chapter describes a capability Ior the simulation oI Numerical Control (NC) machining processes. An automated
interIace between Marc and CAD/NC data that describe the cutter shape and cutter path is provided. The cutter motion
is then analyzed to determine the portion oI the Iinite element mesh to be removed. The cutter path data is stored in
either APT source or CL data Iormat. In the current release, APT source data that is output by CATIA V4 are
supported. CL data is a cutter location data that is provided by APT compilers.
The current example is used to demonstrate the utilization oI the MACHINING (i.e. Metal Cutting) Ieature. A detailed
procedure is shown in this chapter to conduct a realistic machining process simulation using associated Marc models
and NC data in either APT source or CL Iormat.
Input Data
Input data required Ior the simulation oI machining process include CAD data Ior deIining the NC machining process
and Marc data Ior the Iinite element analysis. The required data are as Iollows:
NC data to deIine the cutter geometry and cutter path Ior the machining process. (.apt or .ccl Iiles). For details
on the Iormat oI the apt or ccl Iiles, please reIer to Marc Volume A: Theory and User Information and the
reIerences listed there.
Geometry oI the initial workpiece either imported directly Irom the CAD package (through IGES Iormat) or
built in the GUI. It is recommended that the IGES option be used since the workpiece position and orientation
in the CAD package and in the Iinite element analysis should be identical.
Notes: (1) In the current version, circular motion is required to be transIormed into point-to-point motion type when
the APT Iile is output by CAD NC soItware. In addition, the TRACUT and COPY statements are
necessary to be explicitly interpreted into cutter motion statements. Major statement CYCLE is supported
in combination with DRILL minor statement Ior the deIinition oI drilling motion type.
(2) The Ilipping over oI a part during the machining process is supported by converting the Ilipping over oI
the part into a rotation oI cutter axis. MLTAXS statement is used to deIine the rotation oI cutter axis.
1745
CHAPTER 6.3
FEM Simulation of NC Machining Process
Marc input data that includes the Iinite element model deIinition oI the workpiece and the Iile names Ior
cutter path deIinition. Model deIinition includes the workpiece mesh, boundary conditions, material
properties, insitu residual stresses, and the sequence oI events (loadcases) simulating the machining process.
Initial stress data prior to the cutting process. The initial stresses are obtained through experimental
measurements or Irom analytical results. More details on the estimation oI residual stresses in the current
example are explained below.
Model Generation
Mesh Generation
The geometry oI the part beIore cutting is shown in Figure 6.3-1. The geometry is imported into the GUI through an
IGES Iile. The initial part is a block with length, width and thickness 28 x 14 x 4.5 inches. The block is then meshed
with hexahedral elements. In the current example, these elements are obtained by Iirst creating a 2-D mesh in the X-Z
plane and then extruding it in the Y direction. It is very important to have a Iine mesh in the direction in which the
cutting is taking place.
In the current example, the cutting is predominantly in the Z direction. 18 elements, each oI thickness 0.25 inches, are
used in the Z direction. The Iinite element mesh is shown in Figure 6.3-2. This mesh is available in the Mentat database
ex_r01.mud. In the model, 28224 brick elements and 32205 nodes are deIined.
Figure 6.3-1 The Initial Part Geometry

width
thickne
length
width
thickness
length
1746 Marc Users Guide: Part 3
CHAPTER 6.3
Figure 6.3-2 The Definition of Cutting Processes
Residual Stresses
This section brieIly discusses the residual stresses used Ior the model. A 2-D analysis oI the manuIacturing procedures
to make the aluminium block has been previously undertaken. At the end oI the quenching, stretching and release
loadcases, the stress state in the 2-D Iinite element model is available. The given 2-D mesh is rotated into the XZ plane
and the stress components are assigned to each element, aIter converting them into the proper units. For the 3-D
model, the stress distribution is assumed to be constant in the Y direction. Figure 6.3-3 shows the initial stress
distribution Ior the o
xx
, o
yy
, o
zz
, and o
xz
components.
Cutter and its axis definition 1
Surface for the first cut step
Cutter and its axis definition 2
Surface where the second
Z
Y
X
cut step starts
1747
CHAPTER 6.3
FEM Simulation of NC Machining Process
Figure 6.3-3 The Model with Initial Stresses before Machining (a) o
xx
, (b) o
yy
, (c) o
zz
, (d) o
xz
,
Note that o
zy
= o
xy
= 0
Procedure Files
The most signiIicant step-by-step commands needed to set up the machining process are described below. These
commands describe the procedure to deIine the boundary conditions and loadcases in order to conduct the cutting
process sequentially and automatically. All these commands are also stored in procedure Iiles. The user can execute
these procedure Iiles in a step-by-step manner to obtain greater understanding oI the command sequence.
The Iirst procedure Iile, mc_nfg.proc, reads in the base Iinite element model. Prior to this, it is ensured that the cutter
path Iiles are deIined and previously saved in the current working directory. The Iinite element mesh with the residual
stress state in the ex_r01.mud Iile is also stored in the current working directory. The sequence oI commands to
execute the procedure Iile are as Iollows:
UTILS
PROCEDURES
LOAD
mc_nfg.proc
(a) (b)
(c) (d)
1748 Marc Users Guide: Part 3
CHAPTER 6.3
Once the initial model Iile, ex_r01.mud, has been read in, the next step is to deIine boundary conditions and loadcases
beIore submitting the job.
Isotropic material property parameters are used Ior the aluminium block. These are deIined by: E (Young's Modulus)
10000 ksi, Poisson ratio 0.3.
The procedure to deIine the boundary conditions and loadcases is in procedure Iile: machining_rcd.proc. By
loading this procedure Iile and using START/CONT, Mentat automatically completes all the steps. For better
understanding, one may use the STEP button to conduct the procedure step-by-step.
Machining Process Simulation
The machining process includes two cutting steps:
The Iirst step is to cut two inches oII the upper surIace as shown in Figure 6.3-2. The cutting depth oI each
cutting step is deIined in the cutter path data Iile m2q0090s1.ccl.
The second step is to cut two pockets over the lower surIace oI the part aIter the Iirst cut step is done. The
cutter path Ior this step is deIined by the cutter path data Iile m2q0090s2.ccl. The ccl Iiles are created based on
the apt sources generated Irom CATIA V4.
Between the Iirst and second step, the part is supposed to be Ilipped over, so that the cutter axis is unchanged in the
second cut step. However, Ior the convenience oI FE model deIinition and analysis, the Ilipping over oI the part is
equivalently simulated by the rotation oI the cutter. ThereIore, the second cut is conducted by rotating the cutter into
the opposite direction, as shown in Figure 6.3-2.
There are a total oI Iour loadcases deIined in this model. They are:
1. Cut the top part of the workpiece. The cutter Iile used here is m2q0090s1.ccl.
2. Release the bottom b.c. and apply to the upper face. This loadcase is the one to Ilip over the part by
switching the boundary conditions applied at bottom to the newly generated top surIace.
3. Cut the pocket from the lower face. This loadcase is the one used to cut the pocket on the lower side oI the
part. The cut Iile used here is m2q0090s2.ccl.
4. Final release (springback). This loadcase releases all the boundary conditions, except those required to clear
the rigid body motion oI the part.
The total sets oI boundary conditions deIined by this procedure are:
Fixbottom: This set Iixes the x-y-z displacement oI all the nodes at the bottom surIace. It is used in
loadcase 1.
Fixmiddle: This set Iixes the x-y-z displacement oI all the nodes at the top surIace oI the part aIter the Iirst
cut. It is used in loadcase 2 and 3.
Fixxyz: This set Iixes the x-y-z displacement oI node 2266.
OK
START/CONT
1749
CHAPTER 6.3
FEM Simulation of NC Machining Process
Fixx: This set Iixes the x-displacement oI node 9.
Fixy: This set Iixes the y-displacement oI node 32065.
Fixz: This set Iixes the z-displacement oI node 32058. boundry condition sets 3 to 6 are used in the
loadcase 4.
The Mentat commands to deIine all the loadcases are shown below:
Loadcase1 (cut the top part of the workpiece)
MAIN
LOADCASES
NEW
NAME
cutface1
MECHANICAL
STATIC
LOAD
Click on: fix bottom (to the boundary condition for loadcase)
CONVERGENCE (defining convergence criteria)
RESIDUAL
Click on (AUTO SWITCH)
Enter Relative Force Tolerance: 0.01
OK
Click on CONSTANT TIME STEP
Click on STEPS and Enter a number of 10
OK
Click off AUTO TIME STEP CUT BACK
OK
Click on IMPORT for Inactive Elements
Click on CUT FILE
To select file: m2q0090s1.ccl
(name cutter path definition)
click on TITLE and enter the title for this loadcase:
cut the top part of the workpiece
1750 Marc Users Guide: Part 3
CHAPTER 6.3
Figure 6.3-4 The Definition of First Cutting Loadcase
Now, as shown in Figure 6.3-4, the Iirst loadcase has been deIined. Next step is to deIine the loadcase to Ilip over the
part aIter the Iirst cut step is completed.
Loadcase2 (release the bottom boundary condition and apply to the
top face)
NEW
NAME
release_bot
MECHANICAL
STATIC
LOAD
Click off: fixbottom (to free boundary condition for loadcase 1)
Click on: fixmidface (to apply boundary condition on middle surface)
CONVERGENCE (defining convergence criteria)
RESIDUAL
Click on (AUTO SWITCH)
Enter Relative Force Tolerance: 0.01
OK
Click on CONSTANT TIME STEP
Click on STEPS and Enter a number of 1
OK
1751
CHAPTER 6.3
FEM Simulation of NC Machining Process
Now, the second loadcase has been deIined. Next step is to deIine the loadcase to cut the pockets over the other side
oI the part. The procedure is recorded as Iollowing:
Loadcase3 (cut the pocket from the lower face part)
Click off AUTO TIME STEP CUT BACK
OK
Click on MANUAL for Inactive Elements
Click on TITLE and enter the title for this loadcase:
(release bottom boundary condition; apply to top face)
OK
NEW
NAME
cut pocket
MECHANICAL
STATIC
LOAD
Click on: fixbottom (to apply boundary condition)
CONVERGENCE (defining convergence criteria)
RESIDUAL
Click on (AUTO SWITCH)
Enter Relative Force Tolerance
0.01
OK
Click on CONSTANT TIME STEP
Click on STEPS and Enter a number of 10
OK
Click off AUTO TIME STEP CUT BACK
OK
Click on IMPORT for Inactive Elements
Click on CUT FILE
To select file
m2q0090s2.ccl
click on TITLE and enter the title for this loadcase:
(cut the pocket from the lower face part)
OK
1752 Marc Users Guide: Part 3
CHAPTER 6.3
When the second cutting step is Iinished, the Iinal springback needs to be determined. This process requires that all
the restraints are removed except those that are needed to avoid rigid body motion. So a minimum set oI boundary
conditions are deIined Ior this loadcase (only six degrees oI Ireedom are Iixed Ior the whole model). The procedure is
recorded as Iollowing:
Loadcase4 (final release springback)
NEW
NAME
final_release_bc
MECHANICAL
STATIC
LOAD
Click off: fixmidface to free B.C.
Click on: fix_xyz to fix x, y and z.
Click on: fix_x to fix x.
Click on: fix_y to fix y.
Click on: fix_z to fix z.
CONVERGENCE (defining convergence criteria)
RESIDUAL
Click on (AUTO SWITCH)
Enter Relative Force Tolerance: 0.01
OK
Click on CONSTANT TIME STEP
Click on STEPS and Enter a number of 1
OK
Click off AUTO TIME STEP CUT BACK
OK
Click on MANUAL for Inactive Elements
click on TITLE and enter the title for this loadcase: (final release (springback)
OK
1753
CHAPTER 6.3
FEM Simulation of NC Machining Process
Figure 6.3-5 The Definition of Final Springback Process
Job Definition
The element type Ior this analysis is deIined as Iollows:
Job deIinition is done by the Iollowing procedure.
MAIN
JOB
ELEMENT TYPES
MECHANICAL
3-D Solid
select 7
OK
EXIST (to choose all existing element)
MAIN
JOB
NEW
Enter job name: metal_cut
MECHANICAL
Select loadcases 1, 2, 3, and 4 sequentially (applying loadcases)
1754 Marc Users Guide: Part 3
CHAPTER 6.3
AIter the job is deIined, the model is run through the Iollowing command sequence.
The FE analysis oI the machining (namely, metal cutting) process is started. Mentat instantly shows the progress oI
the calculation by clicking the MONITOR button.
Visualization of Results
The results can be opened using the Iollowing procedure:
INITIAL LOADS
Click off all the b.c
INITIAL CONDITIONS (to check if they are all on)
ANLYSIS OPTIONS (to use defaults for this)
JOB RESULTS (to select the results that the user is interested in)
CENTROID (to reduce the post file size by click this button)
OK
JOB PARAMETER
SOLVER (to choose correct solver)
ITERATIVE SPARSE
IMCOMPLETE CHOLESKI
OPTIMIZATION
OK (iterative solver is used to reduce memory
requirement and total computation time)
OK (twice)
MAIN
JOB
RUN
Click SUBMIT (1)
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BAND
1755
CHAPTER 6.3
FEM Simulation of NC Machining Process
The elements being progressively cut oII Irom the part are not displayed in Mentat; only the remaining part oI the FE
model is displayed Ior postprocessing purposes.
The results are presented here with a two-Iold objective:
To check that the cutter path is Iollowed exactly: As shown in Figure 6.3-6 through Figure 6.3-9, it is seen that
the cutter path is Iollowed exactly during the FE analysis. For areas with small radii corners, a very Iine mesh
is required in order to have better resolution oI the part shape aIter machining. These areas are highlighted in
Figure 6.3-8 and Figure 6.3-9. Rezoning/remeshing can be very powerIul tools to reIine such locations.
To check the deIormation oI the Iinal part. The part displays a very obvious deIormation aIter springback, see
Figure 6.3-10. The maximum displacement oI the part is about 20 times larger aIter springback (increases
Irom 0.000568 to 0.01059 in.). Figure 6.3-11 shows a scaled deIormation pattern that demonstrates how the
Iinal part moves aIter the machining process.
The Iollowing Iigures show the results aIter each loadcases, respectively:
1. Cut Upper Face
Figure 6.3-6 The Machining of the Upper Surface
SCALAR
Total Displacement
OK
FILL
MONITOR
1756 Marc Users Guide: Part 3
CHAPTER 6.3
2. Flipping Over (switch boundary conditions)
Figure 6.3-7 The Flip Over of the part after First Cutting Process
3. Cut Pockets
Figure 6.3-8 The Process of Pocket Cutting
1757
CHAPTER 6.3
FEM Simulation of NC Machining Process
Figure 6.3-9 The Geometry after Pocket Cutting
4. Final Release (springback)
Figure 6.3-10 The Final Geometry and Deformation after Springback
1758 Marc Users Guide: Part 3
CHAPTER 6.3
Figure 6.3-11 The Visualization of Deformation (after scaling)
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
mc_nfg.proc Mentat procedure Iile to run the above example
machining_rcd.proc Mentat procedure Iile to run the above example
m2q0090s2.ccl Cutter geometry and path Iile
m2q0090s1.ccl Cutter geometry and path Iile
ex_r01.mud Associated Mentat model Iile
Chapter 6.4: Piezoelectricity Analysis of an Ultrasonic Motor
6.4
Piezoelectric Analysis of an
Ultrasonic Motor

Chapter Overview 1760

Eigenvalue Analysis of the Stator of an Ultrasonic Motor 1760

Harmonic Analysis of the Stator of an Ultrasonic Motor 1773

Transient Analysis of the Stator of an Ultrasonic Motor 1779

Input Files 1783

Reference 1783
1760 Marc Users Guide: Part 3
CHAPTER 6.4
Chapter Overview
This Ieature shows how to simulate piezoelectricity in Marc. The piezoelectric eIIect is the coupling oI stress and
electric Iield in a material. An electric Iield in the material causes the material to strain and vice versa. A piezoelectric
Marc analysis is Iully coupled, thus simultaneously solving Ior the nodal displacements and electric potential. A
typical application oI piezo-electricity can be Iound in a so-called ultrasonic motor, which can be Iound in camera auto
Iocus lenses or watch motors. The principle oI operation oI an ultrasonic motor is shown in Figure 6.4-1.
Figure 6.4-1 Principle of Operation of an Ultrasonic Motor
A rotor is positioned on a stator and a traveling wave in the stator is driving the rotor. Each point in the stator has an
elliptical motion, as shown in Figure 6.4-1. A point in the top plane oI the stator moves up, liIts the rotor, moves it a
bit backwards, goes down, detaches Irom the rotor, and returns to its original position. The traveling wave in the stator
occurs due to exciting two standing waves with a diIIerent phase. The Irequency oI this excitation is in the ultrasonic
range. The diameter oI these motors is in the order oI centimeters, they are lightweight, produce a high torque, operate
at a low rotational speed, and have a simple design.
First, a dynamic modal analysis is perIormed to obtain the resonant modes oI the stator. Then a harmonic analysis is
perIormed, which shows that a traveling wave occurs iI the stator is excited at the right Irequency with the correct
phase diIIerence. Finally, a transient dynamic analysis is perIormed to show the onset oI the motion oI the stator upon
application oI the potential. Model dimensions and material data are taken Irom Reference 1.
Eigenvalue Analysis of the Stator of an Ultrasonic Motor
In this section, the stator oI the ultrasonic motor is analyzed. A detailed conIiguration oI this stator is shown in
Figure 6.4-2.
The stator consists oI a brass ring plate with a piezoelectric ceramic (PZT) attached to the lower surIace. The
piezoelectric ceramic is polarized in the thickness layer direction, and the polarity is reversed at an interval oI ,
where is the wavelength oI the standing wave. The ninth Ilexural mode is the working Irequency oI the motor.
ThereIore, each polarized piezoelectric segment is 1/18th oI the total ring. Figure 6.4-3 shows a close-up oI the
piezoelectric stator to show this polarization in more detail. The arrows indicate the orientation directions, where blue,
green, and red are the Iirst, second, and third direction, respectively.
rotor
stator
traveling wave
orbital of a point of the stator
2

1761
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Figure 6.4-2 Configuration of the Stator of an Ultrasonic Motor
Figure 6.4-3 Close-up of the Stator showing the Orientation of the Piezoelectric Elements
Two electrodes are placed at the lower surIace oI the piezoelectric ceramic and are separated by the unpolarized
regions. In the model, these electrodes are made by tying the potential degree oI Ireedom oI all the nodes belonging
to an electrode to one node. In this way, the admittance can be easily calculated.
electrode A electrode B
brass
piezo_unpolarized
piezo_up
piezo_down
1762 Marc Users Guide: Part 3
CHAPTER 6.4
Both oI these electrodes are able to generate the ninth Ilexural mode. To generate a traveling wave, the two electrodes
need to be driven simultaneously with a phase diIIerence oI 90. The nodes at the interIace oI the brass and the ceramic
are connected to the common ground to make a closed circuit Ior the potential. Element type 163 is used Ior the
piezoelectric material. This element is mechanically equivalent to element 7, but has Iour degrees oI Ireedom, the Iirst
three are Ior the x-, y-, and z-displacement, and the Iourth is Ior the electric potential.
Mesh Generation
The mesh is generated by Iirst deIining a ruled surIace Irom two circles. This surIace is converted into 144 elements
so that each polarized region consists oI elements. The elements are then expanded twice in the z-direction,
once Ior the piezoelectric layer and once Ior the brass layer.
FILE
NEW
RESET PROGRAM
RETURN
MESH GENERATION
CURVE TYPE
ARCS CENTER/ANGLE/ANGLE
RETURN
CRVS ADD
0 0 0
0.02
0 360
0 0 0
0.03
0 360
FILL
SURFACE TYPE
RULED
RETURN
SRFS ADD
1
2
CONVERT
DIVISIONS
144 1
144 18 8
1763
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Boundary Conditions
The potential oI the nodes where the piezoelectric elements are connected with the brass elements are set to zero. The
two electrodes are applied at the lower surIace. This is done by tying all the nodes oI an electrode to one node and
applying the potential at this node. To remove the rigid body modes in the stator SPRINGS TO GROUND with a small
stiIIness are added to three nodes Ior the x-, y-, and z-direction.
SURFACES TO ELEMENTS
1 #
RETURN
EXPAND
SHIFT
TRANSLATIONS
0 0 0.0005
ELEMENTS
ALL EXIST
TRANSLATIONS
0 0 0.0025
REMOVE
ELEMENTS
1 to 144 #
RETURN
SWEEP
ALL
RETURN
LINKS
NODAL TIES
N TO 1 TIES
TYPE
4
OK
NODE 1
106
ADD TIES
42 to 105
187 to 251 #
1764 Marc Users Guide: Part 3
CHAPTER 6.4
NODE 1
110
ADD TIES
1 to 30
111 to 175
255 to 289 #
RETURN (twice)
SPRINGS/DASHPOTS
NEW
TO GROUND
PROPERTIES
STIFFNESS:SET
10
OK
NODE
1271
DOF: 1
COPY
DOF: 2
COPY
DOF: 3
COPY
NODE
1559
DOF: 1
COPY
DOF: 2
COPY
DOF: 3
COPY
NODE
1415
DOF: 1
COPY
DOF: 2
COPY
1765
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
DOF: 3
RETURN
BOUNDARY CONDITIONS
NAME
ground
ELECTROSTATIC
FIXED POTENTIAL
POTENTIAL(TOP)
0
OK
SELECT
METHOD
BOX
RETURN
NODES
-10 10
-10 10
5e-4-1e-6 5e-4+1e-6
RETURN
NODES ADD
ALL SELECT
SELECT
METHOD
SINGLE
RETURN
CLEAR SELECT
RETURN
NEW
NAME
electrode_A
FIXED POTENTIAL
POTENTIAL(TOP)
1
OK
NODES ADD
106 #
1766 Marc Users Guide: Part 3
CHAPTER 6.4
Material Properties
The brass is isotropic with a Young`s modulus oI 100.6 GPa and a Poisson`s ratio oI 0.35. The elastic properties oI the
piezoelectric material are diIIerent Ior the polarized and nonpolarized part. The nonpolarized part is isotropic and has
a Young`s modulus oI 79.0 GPa and a Poisson`s ratio oI 0.32. The elastic properties oI the polarized part are anisotropic
and are given in the Iollowing matrix:
N/m
2
The piezoelectric coupling matrix is
C/m
2
and the dielectric matrix is
F/m
The density oI the brass is 8560 kg/m
3
and oI the piezoelectric ceramic 7600 kg/m
3
. The polarization oI the elements
is deIined using the ORIENTATIONS menu. BeIore entering the material properties, two element sets are deIined; one
Ior each polarity. The material data Ior the piezoelectric material is entered by giving both the mechanical data (here
ISOTROPIC Ior the nonpolarized and ANISOTROPIC Ior the polarized part), and the nonmechanical data
NEW
NAME
electrode_B
FIXED POTENTIAL
POTENTIAL(TOP)
1
OK
NODES ADD
110 #
RETURN (twice)
13.9 7. 78 7.43 0 0 0
7.78 13.9 7.43 0 0 0
7.43 7.43 11.5 0 0 0
0 0 0 2.56 0 0
0 0 0 0 2.56 0
0 0 0 0 0 3.06
10
10

0 0 5.2
0 0 5.2
0 0 15.1
0 0 0
12.7 0 0
0 0 0
6.464 0 0
0 6.464 0
0 0 5.623
10
9

1767
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
(ELECTROSTATIC Ior the dielectric constants, and PIEZO-ELECTRIC Ior the coupling matrix). The menu Ior entering
the piezo-electric coupling matrix is shown in Figure 6.4-4.
Figure 6.4-4 Piezo-electric Coupling Matrix Menu
MATERIAL PROPERTIES
SELECT
ELEMENTS
145 146 147 148 149 158 159 160 161 162 163 164 165 186
187 188 189 190 191 192 193 202 203 204 205 206 207 208
209 218 219 220 221 222 223 224 225 234 235 236 237 238
239 240 241 254 255 256 257 258 259 260 261 270 271 272
273 274 275 276 277 286 287 288 #
STORE
minus_z
OK
ALL SELECT
CLEAR SELECT
ELEMENTS
150 151 152 153 154 155 156 157 166 167 168 169 170 171
172 173 194 195 196 197 198 199 200 201 210 211 212 213
214 215 216 217 226 227 228 229 230 231 232 233 242 243
244 245 246 247 248 249 262 263 264 265 266 267 268 269
278 279 280 281 282 283 284 285 #
STORE
plus_z
OK
1768 Marc Users Guide: Part 3
CHAPTER 6.4
ALL SELECT
CLEAR SELECT
RETURN
NAME
brass
ISOTROPIC
YOUNGS MODULUS
1.006E11
POISSONS RATIO
0.35
MASS DENSITY
8560
OK
ELEMENTS ADD
289 to 432 #
NEW
NAME
piezo_unpol
ISOTROPIC
YOUNGS MODULUS
7.9E10
POISSONS RATIO
0.32
MASS DENSITY
7600
OK
ELECTROSTATIC
OK
PIEZO-ELECTRIC
OK
ELEMENTS ADD
174 175 176 177 178 179 180 181 182 183 184 185 250 251 252
253 #
NEW
NAME
piezo
1769
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
ANISOTROPIC
MASS DENSITY
7600
C(i,j)
11
1.39E11
12
7.78E10
13
7.43E10
22
1.39E11
23
7.43E10
33
1.15E11
44
2.56E10
55
2.56E10
66
3.06E10
OK
ELECTROSTATIC
ORTHOTROPIC
PERMITTIVITY11
6.46357E-9
PERMITTIVITY22
6.46357E-9
PERMITTIVITY33
5.62242E-9
OK
PIEZO-ELECTRIC
231
12.7
1770 Marc Users Guide: Part 3
CHAPTER 6.4
113
-5.2
223
-5.2
333
15.1
OK
SELECT
SELECT SET
minus_z
plus_z
OK
RETURN
ELEMENTS ADD
ALL SELECT
SELECT
CLEAR SELECT
SELECT SET
minus_z
OK
RETURN
ORIENTATIONS
CYLINDRICAL
0 0 0
0 0 -1
ALL SELECT
SELECT
CLEAR SELECT
SELECT SET
plus_z
OK
RETURN
CYLINDRICAL
0 0 0
0 0 1
ALL SELECT
1771
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Loadcases and Job Parameters
The analysis is a modal shape simulation to obtain the eigenIrequencies oI the stator and, speciIically, to Iind the ninth
Ilexural mode. The Irequency range to search the eigenIrequencies is between 1 kHz and 55 kHz, and the LANCZOS
method is used. Element 163 is used Ior the piezo ceramic material and element 7 Ior the brass material. The
ASSUMED STRAIN Iormulation is selected to improve the bending behavior oI these lower order solid elements.
SELECT
CLEAR SELECT
RETURN (thrice)
LOADCASES
NAME
modal
PIEZO-ELECTRIC
DYNAMIC MODAL
FREQUENCY METHOD:RANGE
LOWEST FREQUENCY
1000
HIGHEST FREQUENCY
55000
OK
RETURN (twice)
JOBS
ELEMENT TYPES
PIEZO-ELECTRIC
MECHANICAL ELEMENT TYPES 3-D SOLID
7
OK
289 to 432 #
PIEZO-ELECTRIC ELEMENT TYPES 3-D
163
OK
145 to 288 #
RETURN (twice)
PIEZO-ELECTRIC
modal
1772 Marc Users Guide: Part 3
CHAPTER 6.4
Save Model, Run Job, and View results
AIter saving the model, the job is submitted and the resulting post Iile is opened. Figure 6.4-5 shows the ninth Ilexural
mode oI the ultrasonic motor at a Irequency oI 46615 Hz, where use is made oI the automatic scaling oI the
displacements.
ANALYSIS OPTIONS
ADVANCED OPTIONS
ASSUMED STRAIN
OK (thrice)
FILE
SAVE AS
piezomotor.mud
OK
RETURN
RUN
SUBMIT(1)
MAIN
RESULTS
OPEN DEFAULT
DEFORMED SHAPE SETTINGS
AUTOMATIC
RETURN
DEF ONLY
CONTOUR BAND
SCALAR
Displacement Z
SCAN
O:39
OK
1773
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Figure 6.4-5 Ninth Flexural Mode of the Ultrasonic Motor
Harmonic Analysis of the Stator of an Ultrasonic Motor
This analysis will show that a traveling wave occurs when the stator is excited at the two electrodes at the right
Irequency and with a phase diIIerence oI 90. The Irequency range is chosen around the 46615 Hz as computed in the
previous section.
The model we built in the previous section can be used as a starting point. The Iollowing changes need to be made.
Boundary Conditions
New boundary conditions Ior the electrodes are generated, since the potentials applied to the two electrodes needs
altering in the harmonic boundary conditions, and on electrode B, a phase shiIt oI 90 is applied.
BOUNDARY CONDITIONS
EDIT
electrode_A
OK
COPY
NAME
electrode_harmonic_A
1774 Marc Users Guide: Part 3
CHAPTER 6.4
Loadcases and Job Parameters
A new loadcase is made Ior this harmonic analysis. The Irequency range is set around the ninth Ilexural mode, Irom
40 to 50 kHz in 51 steps.
ELECTROSTATIC
HARMONIC BCs
FIXED ELECTRIC POTENTIAL
OK
EDIT
electrode_B
OK
COPY
NAME
electrode_harmonic_B
FIXED ELECTRIC POTENTIAL
PHASE
90
OK
RETURN (thrice)
LOADCASES
NEW
NAME
harmonic
PIEZO-ELECTRIC
DYNAMIC HARMONIC
LOADS
CLEAR
ground
electrode_harmonic_A
electrode_harmonic_B
OK
LOWEST FREQUENCY
40000
1775
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Save Model, Run Job, and View Results
A new piezoelectric job is deIined, in which the harmonic loadcase is selected.
Figure 6.4-6 shows the z-displacement at a Irequency oI 46600 Hz, which is close to the resonant Irequency.
Figure 6.4-7 shows the phase in the z-direction at the same Irequency. The amplitude oI the displacement in the
z-direction is more or less uniIorm in the circumIerential direction. The momentary displacement, as shown in the two
Iigures, is completely dependant oI the phase. To get a better understanding oI how this structure responds in the time
domain at this Irequency it is possible to make an animation in Mentat Ior one cycle. Then, we see that a traveling
wave occurs in the structure.
HIGHEST FREQUENCY
50000
# FREQUENCIES
51
OK
RETURN (twice)
JOBS
NEW
PIEZO-ELECTRIC
harmonic
ANALYSIS OPTIONS
ADVANCED OPTIONS
ASSUMED STRAIN
OK (thrice)
RUN
SAVE MODEL
SUBMIT(1)
MAIN
RESULTS
OPEN DEFAULT
SCAN
0:34
OK
RX-
RX-
RX-
1776 Marc Users Guide: Part 3
CHAPTER 6.4
Figure 6.4-6 Z-displacement at a Frequency of 46600 Hz
RX-
RX-
RX-
DEFORMED SHAPE SETTINGS
FACTOR
5000
MANUAL
RETURN
DEF ONLY
SCALAR
Displacement Z Magnitude
CONTOUR BANDS
MORE
ANIMATION
HARMONICS
50
REPEAT
PLAY
Note: In Figure 6.4-6, the z-displacement when the stator is driven at the two electrodes with a Irequency oI 46600 Hz and a
phase diIIerence oI 90; displacement is scaled with a Iactor 5000.
1777
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Figure 6.4-7 A Phase in the Z-direction with a Frequency of 46600 Hz
It is also possible to obtain the admittance Irom this simulation. The admittance is calculated as
, (6.4-1)
where is the current and the applied potential. The current is related to the total charge on the electrode surIace as
, (6.4-2)
where is the operating Irequency, is , and is the sum oI the charge on all the nodes belonging to the
electrode. This summation is already done since all the nodes oI an electrode are tied to one node. Since the
rms value oI the admittance then becomes
. (6.4-3)
Figure 6.4-8 shows this admittance as a Iunction oI the Irequency. This plot is obtained by creating a history plot oI
the reaction charge on node 106 as a Iunction oI the Irequency, converting this history plot in a table and manipulating
the table plot by application oI equation (6.4-3). Figure 6.4-8 clearly shows an increase oI the admittance (or a decrease
oI the resistance) around a resonant Irequency oI the stator.
Note: In Figure 6.4-7, the phase in the z-direction when the stator is driven at the two electrodes with a Irequency oI 46600
Hz and a phase diIIerence oI 90; displacement is scaled with a Iactor 5000.
STOP
RETURN
Y I V
I V I
I i eQ
total

e i 1 Q
total
V 1
Y
1
2
------- eQ
t ot al

1778 Marc Users Guide: Part 3

CHAPTER 6.4
RETURN
HISTORY PLOT
SET NODES
106 #
COLLECT DATA
0:1 0:50 1
NODES/VARIABLES
ADD 1-NODE CURVE
106
Frequency
Reaction Electric Charge
RETURN
FIT
SHOW IDS
0
CONVERT>TABLE
1
TABLES
FORMULA | FUNCTION VALUE
abs(f*v1/sqrt(2))
MIN | FUNCTION VALUE
0
MAX | FUNCTION VALUE
0.005
1779
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Figure 6.4-8 Plot of the Electrical Admittance as a Function of the Frequency
Transient Analysis of the Stator of an Ultrasonic Motor
Finally, a transient analysis is perIormed to show that the traveling wave Iorms in the stator, and that nodes have an
elliptical motion as stated in Figure 6.4-1.
New boundary conditions are generated Ior the electrodes, where the sinusodial potentials are prescribed using tables.
BOUNDARY CONDITIONS
EDIT
electrode_A
COPY
NAME
electrode_transient_A
ELECTROSTATIC
TABLES
NEW
1 INDEPENDENT VARIABLE
FORMULA
ENTER
sin(2*pi*46615*v1)
Frequency (x10000)
Admittance (mmho)
1780 Marc Users Guide: Part 3
CHAPTER 6.4
Loadcases and Job Parameters
A new loadcase is made Ior this transient analysis. The total time is set in such a way that 15 cycles are done at the
resonant Irequency oI 46615 Hz, so ms.
MAX | V1
15/46615
STEPS | V1
500
TYPE
time
REEVALUATE
FIT
COPY
ENTER
sin(2*pi*46615*v1-pi/2)
RETURN
FIXED POTENTIAL
TABLE
table1
OK
EDIT
electrode_B
COPY
NAME
electrode_transient_B
FIXED POTENTIAL
TABLE
table2
OK
RETURN (twice)
LOADCASES
NEW
NAME
transient
PIEZO-ELECTRIC
t 15 46615 0.3218
1781
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Save Model, Run Job, and View Results
A new piezoelectric job is deIined, in which the transient loadcase is selected.
DYNAMIC TRANSIENT
LOADS
CLEAR
ground
electrode_transient_A
electrode_transient_B
OK
TOTAL LOADCASE TIME
15/46615
# STEPS
500
OK
RETURN (twice)
JOBS
NEW
PIEZO-ELECTRIC
transient
INITIAL LOADS
CLEAR
ground
electrode_transient_A
electrode_transient_B
OK
ANALYSIS OPTIONS
ADVANCED OPTIONS
ASSUMED STRAIN
OK (twice)
JOB RESULTS
AVAILABLE ELEMENT TENSORS STRESS
OK (twice)
RUN
1782 Marc Users Guide: Part 3
CHAPTER 6.4
Figure 6.4-9 shows the x- and z-displacement oI a node at the top oI the stator. The Iigure shows that it takes a number
oI cycles beIore the traveling wave starts to emerge, and that the amplitude oI the z-displacement is still increasing.
The node is chosen so that its x-displacement is responsible Ior rotating a rotor. Figure 6.4-10 shows that this node has
an elliptical motion. Here the z-displacement is plotted as a Iunction oI the x-displacement Ior the last 100 increments.
The elliptical motion oI this node is in a counterclockwise motion.
Figure 6.4-9 History Plot of the X- and Z-displacement of a Node at the Top of the Stator
SAVE MODEL
SUBMIT(1)
MAIN
RESULTS
OPEN DEFAULT
DEFORMED SHAPE SETTINGS
MANUAL
250000
RETURN
CONTOUR BANDS
SCALAR
Displacement Z
MONITOR
1783
CHAPTER 6.4
Piezoelectric Analysis of an Ultrasonic Motor
Figure 6.4-10 The Z-displacement as a Function of the X-displacement
for the Node at (0,-0.03,0.003) for Increments 400 to 500
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Reference
1. Kagawa, Y., Tsuchiya, T., and Kataoka, T., 'Finite Element Simulation oI Dynamic Responses oI Piezoelectric
Actuators, Journal of Sound and Vibration, Vol 191(4) (1996), pp. 519-538.
File Description
piezomotor.proc Mentat procedure Iile to run the above example
1784 Marc Users Guide: Part 3
CHAPTER 6.4
Chapter 6.5: Analysis Performance Improvements
6.5
Analysis Performance
Improvements

Chapter Overview 1786

Speed and Memory Improvements 1786

Conclusion 1790

Input Files 1790

1786 Marc Users Guide: Part 3
CHAPTER 6.5
Chapter Overview
There are a Iew major areas where the perIormance has been substantially improved Ior large problems over the years
in Marc. One measure oI perIormance is the speed and memory required to run the same problem Irom version to
version. There are three problems in this chapter that have been run in various versions that span nearly seven years.
These demonstration problems address soItware improvements that require less memory and run in less time based
upon relative perIormance oI the same problem using these versions oI Marc. The speed and memory improvements
herein have been obtained by rewriting the contact database and structure over the intervening years. In addition,
substantial speed and memory improvements, especially Ior examples involving bricks and shells, have been obtained
with the introduction oI a multiIrontal sparse solver.
Speed and Memory Improvements
Case 1: Rigid-Deformable Body Contact
Figure 6.5-1 shows the model where the two rollers are compressing the deIormable ring sector. This particular model
uses the Iterative Sparse solver with Incomplete Choleski preconditioner and runs Ior 50 increments using Iixed load
stepping.
Figure 6.5-1 Model for Case 1 Rigid Contact - Click on right figure to play animation (ESC to stop)
Procedure for Case 1: run with several versions of Marc
../marc200k/tools/run_marc -j rigid -v n (k = 0,1,3 & 5)
The improvements oI memory and speed are shown in Figure 6.5-2.
1787
CHAPTER 6.5
Analysis Performance Improvements
Figure 6.5-2 Memory and Speed for Case 1 Rigid Contact
At some point in the near Iuture, this problem will become obsolete Ior perIormance metrics because oI continuing
soItware improvements and Iaster hardware that will render this problem too small. In the meantime, it is useIul to see
perIormance gains made by Marc over the past six years spanning Iour versions. Clearly this particular problem
demonstrates the beneIit gained in using contact, one oI Marc`s most popular Ieatures.
OI course this is only one model, and other models may have diIIerent perIormance metrics. Contact problems have
two types oI contact bodies, rigid and deIormable. In this case, deIormable and rigid bodies come into contact which
is reIerred to as rigid contact. In the next case, we examine where a deIormable body contacts a deIormable body which
is reIerred to as deIormable contact.
Case 2: Deformable-Deformable Contact
Figure 6.5-3 shows an elastomeric seal that comes into selI contact.This model will be run in several versions oI Marc
using the procedure below.
Figure 6.5-3 Model for Case 2 Deformable Contact
Case 1: Rigid Deformable Body Contact
6.0
6.5
7.0
7.5
8.0
8.5
9.0
Version
M
b
y
t
e
s

o
f

M
e
m
o
r
y
400
500
600
700
800
900
1000
1100
1200
1300
1400
C
P
U

T
i
m
e

[
s
e
c
]
Memory
CPU Time
Memory 7.8 8.4 7.2 6.8
CPU Time 1294 732 601 550
2000 2001 2003 2005
1788 Marc Users Guide: Part 3
CHAPTER 6.5
Procedure for Case 2: run with several versions of Marc
../marc200k/tools/run_marc -j deformable -v n (k = 0,1,3 & 5)
The improvements oI memory and speed are shown in Figure 6.5-4 aIter running these versions. This particular model
uses the MultiIrontal Sparse solver with Optimize 11 and runs Ior 10 increments using Iixed load stepping.
Figure 6.5-4 Memory and Speed for Case 2 Deformable Contact
As with the previous example, there is a steady perIormance gain Irom versions 2000 to 2010 oI Marc. For this
problem, these perIormance improvements have been brought about by changes in the algorithm in contact with
updated Lagrangian elasticity.
Case 3: Model with Solid and Shell Elements
The MultiIrontal Sparse solver yields speed improvements using the Optimize 11 optimizer Ior renumbering and
minimizing bandwidth. This is achieved as shown in Figure 6.5-7 that shows a model oI a magnetic disk drive head
used in the data storage device industry. To run this modal analysis in the several versions oI Marc, use the procedures
below.
Case 2: Deformable-Deformable Body Contact
60
80
100
120
140
160
180
Version
M
b
y
t
e
s

o
f

M
e
m
o
r
y
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
C
P
U

T
i
m
e

[
s
e
c
]
Memory
CPU Time
Memory 159 134 113 81
CPU Time 9028 5713 1836 1413
2000 2001 2003 2005
1789
CHAPTER 6.5
Analysis Performance Improvements
Figure 6.5-5 Model for Demonstrating New Solver Capabilities
Procedure:
.../marc200k/tools/run_marc -j disk_head_drive -v n (k = 0,1,3 & 5)
The improvements oI memory and speed are shown in Figure 6.5-6. The MultiIrontal Sparse solver demonstrates
continued improvements in several version.
Figure 6.5-6 Memory and Speed for Case 3 Modal Analysis
The multiIrontal solver is selected in the JOBS menu as shown in Figure 6.5-7.
Case 3: Modal Analysis with Solid and Shell Elements
200
300
400
500
600
700
800
Version
M
b
y
t
e
s

o
f

M
e
m
o
r
y
150
170
190
210
230
250
270
290
310
330
350
C
P
U

T
i
m
e

[
s
e
c
]
Memory
CPU Time
Memory 720 332 260 256
CPU Time 323 199 181 170
2000 2001 2003 2005
1790 Marc Users Guide: Part 3
CHAPTER 6.5
Figure 6.5-7 Solver Submenu
Conclusion
The perIormance enhancements available in Marc allow you to solve your problems Iaster and with lower memory
requirements than ever beIore. Improvements are continually being made in multiple areas: solver, optimizer, domain
decomposer, contact, improvements in element technology, material modeling as well as general improvements.
In particular, major speed and memory improvements have been obtained by rewriting the contact database and
structure. In addition, substantial speed and memory improvements, especially Ior examples involving bricks and
shells, have been obtained with the introduction oI the multiIrontal sparse solver.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
rigid.dat Marc input Iile
disk_drive_head.dat Marc input Iile
deformable.dat Marc input Iile
Chapter 6.6: Robustness of Automatic Load Stepping Schemes
6.6
Robustness of Automatic Load
Stepping Schemes

Chapter Overview 1792

Usage of the Auto Step Feature 1792

Input Files 1802

1792 Marc Users Guide: Part 3
CHAPTER 6.6
Chapter Overview
For problems where large deIormation and/or contact are involved, it is oIten necessary to use an adaptive time step
strategy to ensure the convergence oI the solution. The robustness control oI nonlinear solution strategy is designed
Ior this purpose. This chapter demonstrates the usage oI the robustness control Ior nonlinear analysis using the AUTO
STEP loadcase.
Usage of the Auto Step Feature
This option controls the automatic time/load stepping procedure. By deIault, the time step is adjusted based upon the
number oI recycles in addition to the user criteria. II the user-speciIied desired number oI recycles is exceeded, the
time step is divided by a Iactor. II the increment converges in less than the desired number oI recycles, the time step
is scaled up using the same Iactor. The increment is repeated iI any oI the Iollowing occurs: maximum number oI
iterations reached, elements going inside out, or a contact node slides oII the end oI a rigid body. In this case, the time
step is divided by two. The enhanced variant is available Ior mechanical, thermal, and thermo-mechanically coupled
analyses.
For more control, the time step can also be adjusted based upon the calculated value oI a parameter (strain increment,
plastic strain increment, creep strain increment, stress increment, strain rate, strain energy increment, temperature
increment, displacement increment, rotation) versus a user-deIined maximum. More than one criteria can be speciIied.
II the criteria is not satisIied within an increment, recycling occurs with a reduced time/load applied. AIter the
increment has converged based upon tolerances speciIied on the CONTROL values, the data given here controls the
next increment.
The Iirst example uses the Auto Step Ieature with deIaults, i.e., without a user speciIied criterion.
Procedure A:
MAIN
FILES
MARC INPUT FILE READ
mesha.dat
OK
PLOT
NODES (turn off drawing of nodes)
DRAW
FILL
MAIN
CONTACT
CONTACT BODIES
ID CONTACT
1793
CHAPTER 6.6
Robustness of Automatic Load Stepping Schemes
Figure 6.6-1 Model for Procedure A
Now this model is completely deIined except Ior the load history. All we need to do is deIine a loadcase using the
AUTO STEP Ieature and run the model. The automatic convergence testing procedure is activated by AUTO SWITCH
which permits switching between displacement and residual control.
EDIT
push
OK
RIGID
VELOCITY PARAMETERS
VELOCITY X
-4
OK (twice)
MAIN
LOADCASES
MECHANICAL
STATIC
CONVERGENCE TESTING
AUTO SWITCH
OK
1794 Marc Users Guide: Part 3
CHAPTER 6.6
The multi-criterion adaptive load stepping procedure (AUTO STEP) is then selected.
Figure 6.6-2 Select Multi-Criteria (Auto Step) Feature
We now submit the analysis to run with only the deIaults Ior Auto Step.
MULTI-CRITERIA
OK
MAIN
JOBS
MECHANICAL
lcase1 (select this loadcase)
OK
FILES
SAVE AS
cauto_load_stepping_a.mud
OK
RETURN
RUN
SUBMIT1
MONITOR
1795
CHAPTER 6.6
Robustness of Automatic Load Stepping Schemes
Figure 6.6-3 Monitor of Job Results
In Figure 6.6-3, we see that the job has completed with a successIul number exit oI 3004. There were 379 Newton-
Raphson iteration cycles, 162 contact separations, 2 cut backs, no remeshes, analysis time oI 1 second and a compute
wall time oI 327 seconds. The statistics are cumulative Ior all the loadcases.
Figure 6.6-4 shows the Iinal position oI the key as it was inserted into the surrounding contact bodies, and Figure 6.6-5
shows the variation in the time step during the analysis.
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BANDS
LAST
1796 Marc Users Guide: Part 3
CHAPTER 6.6
Figure 6.6-4 Final Position
Figure 6.6-5 Variation of Time Step for each Increment
Now let`s try to extract the key Irom the surrounding contact bodies by making a table to control the push(ing)
rigid body.
MAIN
RESULTS
0.0 20.0 40.0 60.0
Increment
0.000
0.010
0.020
0.030
0.040
T
i
m
e

S
t
e
p

[
s
e
c
o
n
d
s
]
Key Insertion Problem
Time Step versus Increment
1797
CHAPTER 6.6
Robustness of Automatic Load Stepping Schemes
Figure 6.6-6 Table for Rigid Body Pushs Velocity
CLOSE
MAIN
CONTACT
CONTACT BODIES
EDIT
push
OK
TABLES
NEW
1 INDEPENDENT VARIABLE
TYPE
time
OK
ADD
0 1 1 1 1 -1 2 -1
FILLED
FIT
SHOW MODEL
RETURN
1798 Marc Users Guide: Part 3
CHAPTER 6.6
Figure 6.6-7 Results from Extraction
RIGID
VELOCITY PARAMETERS
VELOCITY X
choose table1
OK
OK
MAIN
FILES
SAVE AS
auto_load_stepping_b.mud
MAIN
LOADCASES
MECHANICAL
COPY
MAIN
JOBS
MECHANICAL
lcase2
OK
SAVE
RUN
SUBMIT1
MONITOR
1799
CHAPTER 6.6
Robustness of Automatic Load Stepping Schemes
The extraction terminates with an exit number oI 3002, which means that the analysis Iailed to converge. Let`s take a
look at the deIormed shape.
Figure 6.6-8 Deformed Shape at Last from Extraction
We notice that there is not much going on and examining at the convergence criteria, we see that the maximum reaction
used in the denominator oI the convergence ratio is very small. Hence, we need to change our convergence criteria to
account Ior this.
OK
SAVE
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
CONTOUR BANDS
LAST
CLOSE
MAIN
LOADCASES
EDIT
lcase2
OK
1800 Marc Users Guide: Part 3
CHAPTER 6.6
Figure 6.6-9 Results from Extraction with Relative/absolute Testing Switch
MECHANICAL
STATIC
CONVERGENCE TESTING
RELATIVE/ABSOLUTE
MINIMUM REACTION FORCE CUTOFF
8e-4
MAXIMUM ABSOLUTE RESIDUAL FORCE
8e-4
OK (twice)
MAIN
SAVE
JOBS
RUN
SUBMIT
MONITOR
RESULTS
OPEN DEFAULT
DEF ONLY
LAST
1801
CHAPTER 6.6
Robustness of Automatic Load Stepping Schemes
Figure 6.6-10 Deformed Shape at Last from Extraction with Relative/absolute Testing Switch
A cutoII maximum absolute residual Iorce and minimum reaction Iorce oI 8e-4 was chosen because it is 10 oI the
largest reaction that occurs during the 88 increments run in the Iirst extraction attempt.
Finally, looking at the time step history notice how the automatic selection oI the time step varies between the insertion
and extraction steps.
Figure 6.6-11 Variation of Time Step for each Increment
This type oI contact problem would be very hard to run without the Auto Step Ieature since the time step changed 25
times throughout this analysis.
0.0 50.0 100.0
Increment
0.000
0.010
0.020
0.030
0.040
T
i
m
e

S
t
e
p
Key Insertion Problem
Time Step versus Increment
1802 Marc Users Guide: Part 3
CHAPTER 6.6
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
auto_load_stepping.proc Mentat procedure Iile
auto_load_stepping_b.proc Mentat procedure Iile
auto_load_stepping_c.proc Mentat procedure Iile
mesha.dat Mentat procedure Iile
Chapter 6.7: Marc Running in Network Parallel Mode
6.7
Marc Running in Network
Parallel Mode

Run CONTACT WITH DDM 1804

Run CONTACT WITH DDM on a Network 1804

Input Files 1809

1804 Marc Users Guide: Part 3
CHAPTER 6.7
Run CONTACT WITH DDM
In this chapter, the procedure oI running an analysis in parallel on a number oI computers connected over a network
is demonstrated. We start by running a contact demonstration problem, then show how to run the same problem on a
network in parallel. Pick the HELP, RUN A DEMO PROBLEM, and CONTACT WITH DDM buttons as shown in
Figure 6.7-1.
Figure 6.7-1 Run CONTACT WITH DDM Demonstration Problem
This procedure automatically builds, runs, and postprocesses a contact problem using two domains. AIter it Iinishes,
close the post Iile and, Irom the main menu, go to JOBS and pick DOMAIN DECOMPOSITION.
Run CONTACT WITH DDM on a Network
BeIore running a job over the network, make sure that the machines are properly connected. Suppose two machines
with host names host1 and host2, respectively, are to be used in an analysis:
UNIX
From host1, access host2 with
rlogin host2
1805
CHAPTER 6.7
Marc Running in Network Parallel Mode
II a password needs to be provided to do the remote login, this has to be taken care oI. II the rlogin is not possible
without providing a password, a network run is not possible. See the man pages on rlogin (see .rhosts Iile) or
contact your system administrator.
Windows
From host1, access host2 with Network Neighborhood. II this Iails, a network run is not possible. Contact your system
administrator.
Figure 6.7-2 shows the domains that were used in this example.
Figure 6.7-2 Identification of the Domains
For now, generate three domains and run on your network. Pick GENERATE and enter 3 domains. The number oI
domains must equal the number oI processors to be used in the analysis. Marc associates a domain with each processor
and creates a separate input data Iile Ior each domain as well as a root input Iile associated with the job ID.
1806 Marc Users Guide: Part 3
CHAPTER 6.7
Figure 6.7-3 Generating Three Domains
From the JOBS menu, pick RUN, then NETWORK.
Figure 6.7-4 Selecting Hosts to Run
Enter the SETTINGS menu and click on HOST FILE. Create a new Iile called hostfile (any name could be chosen)
by typing in its name in the Iile browser. (Caution Ior Windows NT user: II notepad is used as editor, a Iile called
hostfile.txt is created. Use a name with an extension to avoid this problem.) Now, add the text:
host1 is the host name oI the machine on which Mentat is running and Irom which the job is to be started (the root
host). Assume that two processors are used on host1. host2 is the host name oI the other machine (the remote host)
and on which a single processor is used. The word workdir is replaced with the Iull path to the working directory
where I/O takes place Ior the remote host and installdir is replaced with the Iull path to the Marc installation
directory that the remote host will use (installdir is Ior UNIX only). The path should be given so that it can be
reached Irom host2. The working directory Ior the remote host could be the current working directory on the root
host or, typically, a local directory on the remote host containing the input Iiles Ior the domains to be run on this host.
host1 2
host2 1 workdir installdir
1807
CHAPTER 6.7
Marc Running in Network Parallel Mode
For UNIX, assume that the working directory is called /disk1/testing on host1. Further assume that this disk is
NFS mounted so that the working directory can be reached Irom other computers on the network as
/nfs/host1/disk1/testing (using a hypothetical naming convention Ior shared disks). Now the host Iile
would contain:
Similarly Ior Windows, assume the working directory is D:\users\john\testing, which is shared using the
sharename djohntest. For Windows, please note that the installation directory must be shared and available on all
hosts used in the analysis. The host Iile should contain:
It is more eIIicient to have the working directory on the remote host as a local directory rather than using the working
directory on the root host. In the case oI Windows, assume that the working directory on the remote host is
C:\testing and that Marc is installed on the remote host in C:\MSC.Software\Marc2010, the host Iile
should contain:
In this case, the input data Iile Ior the domain to be processed on host2 must exist in the local directory C:\testing
beIore the analysis starts and the post Iile Ior this domain is available in the same directory at the end oI the analysis.
The necessary input Iiles are automatically copied Irom the working directory oI the root host to the local working
directories oI the remote hosts beIore the job starts. AIter the job is Iinished, the post Iiles Irom the remote hosts are
automatically copied back to the root host. The user has the option to suppress this automatic Iile transIer Irom the
NETWORK SETTINGS under the RUN JOB menu in Mentat. Prior to MSC.Marc 2000, copying oI the input and post
Iiles had to be done manually.
Now submit the job, then check your results. By clicking on OPEN DEFAULT, you view the results Ior the complete
model.
host1 2
host2 1 /nfs/host1/disk1/testing /marcinstall/marc/
host1 2
host2 1 \\host1\djohntest \\host1\marcinstall
host1 2
host2 1 C:\testing
1808 Marc Users Guide: Part 3
CHAPTER 6.7
Figure 6.7-5 Checking Your Results
Marc created a post Iile associated with each domain as well as a root post Iile associated with the job ID. For the
previous model, 1model1_job1.t16, 2model1_job1.t16, and 3model1_job1.t16 are the processor Iiles, while
model1_job1.t16 is the root Iile.
II the model is very large, it can be convenient to view only a portion oI the model by selecting any one oI the processor
post Iiles, such as 2model1_job1.t16 shown in Figure 6.7-6. This Iile contains only data associated with domain 2
as selected in the domain decomposition menu in Figure 6.7-3.
Open Default
Skip to Inc 50
Def Only
Contour Bands
Select Total Equivalent
Plastic Strain"
1809
CHAPTER 6.7
Marc Running in Network Parallel Mode
Figure 6.7-6 Results for Domain 2
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
s4_help.proc Mentat procedure Iile
s4p_help.proc Mentat procedure Iile
1810 Marc Users Guide: Part 3
CHAPTER 6.7
Chapter 6.8: Convergence Automation and Energy Calculations
6.8
Convergence Automation and
Energy Calculations

Chapter Overview 1812

Convergence Automation 1812

Energy Calculation 1819

Input Files 1829

1812 Marc Users Guide: Part 3
CHAPTER 6.8
Chapter Overview
This chapter contains two sections:
Convergence Automation
In nonlinear FEM analysis, it is oIten necessary to check the convergence based on the relative or absolute tolerance
oI solution variables. When the convergence testing is done based on relative values oI solution variables
(displacements/rotations and/or residual forces/moments), the solution is recognized as converged iI the ratio oI
maximum iterative correction to the maximum increment oI solution variable is smaller than the speciIied tolerance.
However, iI maximum displacement increment becomes extremely small, e.g., smaller than the computer's cut-oII
value a likely scenario in the analysis involving springback or constraint thermal expansion (Table 6.8-1), the
convergence testing is less meaningIul. In such cases, Marc automatically switches the convergence testing according
to relative residual forces/moments, iI the AUTO SWITCH option is chosen in the convergence control menu oI Mentat.
Similarly, iI the maximum reaction Iorce/moment becomes extremely small, e.g., smaller the computer's cut-oII value
a likely scenario in the analysis involving stress-Iree-motion (i.e., rigid body motion) or Iree thermal expansion
(Table 6.8-1), the convergence testing is less meaningIul. In such cases, Marc automatically switches the convergence
testing according to relative displacements/rotations, iI the AUTO SWITCH option is chosen in the convergence
control menu oI Mentat.
Also, iI an analysis involves the cases where the deIormable body is totally Iree oI stress and deIormation, then Marc
checks absolute value oI strain energy density and converges iI the absolute strain energy density becomes extremely
small.
.
Convergence Automation introduces a new algorithm Ior automatically choosing convergence criteria
and the new option, AUTO SWITCH. AIter a brieI overview, an example is
given to show how this algorithm can be used in Marc.
Calculation of Energy summarizes the values Ior various types oI mechanical analysis. An example
is given to demonstrate the use oI these energy values to check the balance oI
energy. It is also shown how to plot and view the energy values in Mentat.
Table 6.8-1 Effectiveness of various Relative Tolerance Convergence Testing Criteria
Analysis Type
Convergence Criteria
Displacement/
Rotation
Residual
Force/Torque Strain Energy
Stress-Iree motion Yes No No
Springback No Yes No
Free Thermal Expansion Yes No No
Constraint Thermal Expansion No Yes Yes
Yes relative tolerance testing works.
No relative tolerance testing doesn`t work.
1813
CHAPTER 6.8
Convergence Automation and Energy Calculations
This section shows the use oI AUTO SWITCH option so that the appropriate convergence type oI testing can be
perIormed during the analysis.
AUTO SWITCH Option
This option allows the program to automatically switch on the appropriate convergence testing to address standard as
well as special cases.
This example illustrates the use oI this Ieature Ior a brake-bending oI a two-dimensional workpiece and its springback.
There is one loadcase each Ior the brake-bending and springback. In this analysis, the convergence testing is deIined
to be done on relative displacement.
Let's start Irom an existing model and Iocus on the selection oI the convergence testing scheme Ior analysis oI
this model:
The above procedure creates a model with material/geometry properties and contact/boundary conditions. Now let's
deIine the history data and run this job. There are two loadcases in this job: (1) loadcase1 Ior brake bending, which is
deIined by procedure A. (2) loadcase2 Ior springback and is deIined by procedure B. The commands Ior both
procedure loadcase 1 and 2 are shown below.
Procedure A for Loadcase 1:
MAIN
FILES
NEW
OK
UTILS
PROCEDURES
LOAD
convergence_a.proc
OK
START/CONT
OK
MAIN
LOADCASES
MECHANICAL
STATIC
TOTAL LOADCASE TIME
0.5
1814 Marc Users Guide: Part 3
CHAPTER 6.8
Procedure B for Loadcase 2:
# STEPS
50
CONVERGENCE TESTING
DISPLACEMENTS
RELATIVE DISPLACEMENT TOLERANCE
0.1
OK (twice)
MAIN
MAIN
LOADCASES
MECHANICAL
STATIC
TOTAL LOAD CASE TIME
0.5
# STEPS
20
CONVERGENCE TESTING
DISPLACEMENTS
RELATIVE DISPLACEMENT TOLERANCE
0.1
AUTO SWITCH
OK (twice)
MAIN
1815
CHAPTER 6.8
Convergence Automation and Energy Calculations
Figure 6.8-1 Select AUTO SWITCH Feature
The Mentat commands oI procedure Ior loadcase2 is also shown in Figure 6.8-1. AIter deIining these two loadcases,
we do the analysis with both the draw-bending and springback processes:
MAIN
JOBS
MECHANICAL
lcase1 (selecting loadcase 1)
lcase2 (selecting loadcase 2)
ANALYSIS OPTIONS
LARGE DISPLACEMENT
ADVANCED OPTIONS
CONSTANT DILATATION
OK
under plasticity procedure
click on SMALL STRAIN to see:
LARGE STRAIN ADDITIVE
PLANE STRAIN
JOB RESULTS
Equivalent Von Mises Stress
Total Equivalent Plastic Strain
OK
CONTACT CONTROL
1816 Marc Users Guide: Part 3
CHAPTER 6.8
AIter the job completes two loadcases, the output Iile shows the analysis automatically converged based on small
strain energy density aIter increment 55. We can check the results by the Iollowing Mentat commands:
ADVANCED OPTION
SEPARATION FORCE
0.1
OK (thrice)
ELEMENT TYPE
MECHANICAL
PLANE STRAIN
11
EXIST.
OK
MAIN
JOBS
RUN
SUBMIT 1
OUTPUT FILE
OK
MAIN
VISUALIZATION
COLORS
2 (colormap)
6 (contourmap)
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
NEXT
SCALAR
Total Equivalent Plastic Strain
CONTOUR BANDS
MONITOR
1817
CHAPTER 6.8
Convergence Automation and Energy Calculations
Figure 6.8-2 The Deformed Workpiece and Tools after Springback
Figure 6.8-2 shows the geometry aIter the springback process. We also see the energy balance between the total strain
energy and total work done by external Iorces using the Iollowing Mentat commands:
HISTORY
COLLECT GLOBAL DATA
SHOW IDS
50
NODES/VARIABLES
ADD GLOBAL CRV (choosing within VARIABLES)
Time
Total Strain Energy
Time
Total work
FIT
1818 Marc Users Guide: Part 3
CHAPTER 6.8
Figure 6.8-3 The Energy Balance of the Draw-bending/Springback Process
As shown in Figure 6.8-4, we can see that most oI the elastic strain energy has been released due to springback (Irom
increment 51).
Figure 6.8-4 The Release of Elastic Strain Energy during Springback
1819
CHAPTER 6.8
Convergence Automation and Energy Calculations
Energy Calculation
This Ieature includes the calculation oI energy values Ior various types oI mechanical analysis as listed below:
Note that damping energy and total work done by Iriction Iorces have negative values. In the current implementation,
the damping energy is calculated Ior mass dampers.
For analysis with dynamics, the energy is balanced between the change oI kinetic energy and the work done by external
Iorces, excluding the energy dissipated by plastic/creep strain and dampers.
CONSTANT is the kinetic energy at initial time.
SE + CSE + KE - DE = WE + CONSTANT (6.8-1)
The total work done by external Iorces should be viewed as:
WE = WC + WA + WF (6.8-2)
For static analysis, the energy balance can be calculated as:
WE = SE + CSE + ES + EF (6.8-3)
From equations (6.8-2) and (6.8-3), the energy balance can be calculated by equation (6.8-4):
WC + WA + WF - ES - EF = SE + CSE (6.8-4)
Total strain energy (SE)
Total elastic strain energy (ESE)
Total plastic strain energy (PSE)
Total creep strain energy (CSE)
Thermal energy (ME)
(available Ior heat transIer or coupled stress/thermal analysis)
Total work by all external Iorces (WE)
Within which various contributions are also calculated as:
total work by contact Iorces (WC)
total work by applied Iorces (WA)
total work by Iriction Iorces (WF)
Total kinetic energy (KE)
Total energy dissipated by dampers (DE)
Total energy contributed by springs (ES)
Total energy contributed by Ioundations (EF)
Note: Energy loss is possible Ior dynamic analysis because oI numerical dissipation.
1820 Marc Users Guide: Part 3
CHAPTER 6.8
The energy values mentioned above can be viewed in Mentat. A brieI summary is also given in the output Iile Ior
convenience.
Usage of the Energy Values
We are going to use an example to show the calculation and balance oI various energies. This example models the heat
generated due to Iriction Ior block sliding with an initial velocity and coming to a stop in due time.
As shown in Figure 6.8-5, the block has dimensions oI length x width x height 1.0 m x1.0 m x0.5 m, which is
modeled with 8 brick elements. Element 7 is used Ior this analysis. The material I the block is assumed isotropic Ior
both mechanical and thermal analysis. The Young`s modulus is 210 GPa and the Poisson`s ratio is 0.3. Mass density
is given as 7854 kg/m
3
Ior both dynamic and heat transIer analysis. The conductivity is 60.5 W/(m K) and the speciIic
heat is set as 434 w/(J K). Only proportional mass damping is applied with a ratio oI 0.3. Lumped mass matrix is used
in the example. The conversion rate Ior Iriction work into thermal energy is given as 1.0.
In order to keep the block sliding on the surIace, an acceleration body oI -9.81 m/s
2
is applied to each element along
the z-direction. The initial velocity oI 4.905 m/s is given along the x-direction. The initial temperature is set as 0.0C.
Coulomb model Ior Iriction is chosen with a Iriction coeIIicient oI 0.5 based on nodal Iorces. The relative sliding
velocity Ior Iriction below which a node is assumed to be sticking to a contact surIace is set as 0.1 m/s. The nodal
reaction Iorce required to separate a contacting node Irom its contact surIace is assumed to be 1x10
11
N to keep the
block on the surIace.
The Single Step Houbolt (SSH) method coupled with heat transIer analysis is used Ior the dynamics analysis.
Figure 6.8-5 Initial Geometry and Velocity of the Sliding Block
1821
CHAPTER 6.8
Convergence Automation and Energy Calculations
The procedure convergence_b.proc shows the Mentat commands to create this model:
MAIN
FILES
NEW
OK
RESET PROGRAM
VIEW
4 (under show all view)
RESET VIEW
MAIN
VISUALIZATION
COLOR
1 (colormap)
1 (contourmap)
OK
MAIN
MESH GENERATION
GRID
ADD (elements)
point (-1,-1,0)
point ( 1,-1,0)
point ( 1, 1,0)
point (-1, 1,0)
ADD (SFRS)
point (-1,-1,0)
point ( 1,-1,0)
point ( 1, 1,0)
point (-1, 1,0)
MOVE
SCALE FACTORS
4 2 1
SURFACES
EXIST.
1822 Marc Users Guide: Part 3
CHAPTER 6.8
RESET
TRANSLATIONS
1.8 0 0
SURFACES
EXIST.
RETURN
SUBDIVIDE
ELEMENTS
EXIST.
RETURN
EXPAND
TRANSLATIONS
0 0 0.5
ELEMENTS
EXIST.
RETURN
SWEEP
(remove unused) NODES
EXIST.
RETURN
RENUMBER
ALL
RETURN
FILL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
Y displacement
OK
ADD (nodes)
EXIST.
NEW
GLOBAL LOAD
1823
CHAPTER 6.8
Convergence Automation and Energy Calculations
Z force
Z FORCE
-9.81
OK
ADD (elements)
EXIST.
MAIN
INITIAL CONDITION
THERMAL
TEMPERATURE
temperature (TOP)
OK
ADD (nodes)
EXIST.
RETURN
NEW
VELOCITY
X LINEAR
X LINEAR
4.905
OK
ADD (nodes)
EXIST.
MAIN
MATERIAL PROPERTIES
ISOTROPIC
YOUNG'S MODULUS
210e9
POISSON'S RATIO
0.3
MASS DENSITY
7854
DAMPING
MASS MATRIX MULTIPLIER
0.3
1824 Marc Users Guide: Part 3
CHAPTER 6.8
Until now, the model has been created. Now, we add the loadcase to move the block over the surIace and calculate the
thermal energy generated Irom Iriction:
OK (twice)
HEAT TRANSFER
CONDUCTIVITY
60.5
SPECIFIC HEAT
434
MASS DENSITY
7854
OK
ADD (elements)
EXIST.
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE
FRICTION COEFFICIENT
0.5
OK
ADD (elements)
EXIST.
NEW
RIGID
FRICTION COEFFICIENT
0.5
OK
ADD (surfaces)
EXIST.
ID BACKFACES
ID BACKFACES
MAIN
LOADCASES
MECHANICAL
1825
CHAPTER 6.8
Convergence Automation and Energy Calculations
COUPLED
TITLE
block sliding with friction
OK
DYNAMIC TRANSIENT
CONVERGENCE TESTING
DISPLACEMENT
OK
TOTAL LOADCASE TIME
2.0
# STEPS
50
OK
MAIN
JOBS
COUPLED
"lcase1"
SOLUTION OPTIONS
LARGE DISPLACEMENT
LUMPED MASS & CAPACITY
OK
JOB RESULTS
Equivalent Von Mises Stress
OK
JOB PARAMETERS
CONVERSION FACTOR
1.0
OK
CONTACT CONTROL
COULOMB
RELATIVE SLIDING VELOCITY
0.1
ADVANCED
SEPARATION FORCE
1e11
OK (thrice)
MAIN
JOBS
1826 Marc Users Guide: Part 3
CHAPTER 6.8
Figure 6.8-6 The Dynamic Analysis of a Block Sliding over a Surface with Friction
COUPLED
OK
SUBMIT 1
MONITOR
OK
MAIN
VISUALIZATION
COLORS
2 (colormap)
OK
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Equivalent Von Mises Stress
CONTOUR BANDS
1827
CHAPTER 6.8
Convergence Automation and Energy Calculations
From Figure 6.8-6, we can see the stress generated during the sliding process. The Iollowing Mentat commands show
the energy values and balance during the sliding process.
Figure 6.8-7 The Energy changes during the Sliding Process
MAIN
RESULTS
HISTORY
SHOW ID
50
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD GLOBAL CRV (choose from VARIABLES)
Time
Kinetic Energy
Time
Damping Energy
Time
Total Work
Time
Thermal Energy
Time
Total work by friction force
FIT
1828 Marc Users Guide: Part 3
CHAPTER 6.8
As shown in Figure 6.8-7, the kinetic energy eventually gets dissipated as damping energy and work done by Iriction
Iorces. The energy is nearly conserved as shown by equation (6.8-5).
SE + KE - DE - WE = CONSTANT (6.8-5)
The energy dissipated due to Iriction is converted to thermal energy. In this example, it is halI oI the work done by
Iriction Iorces, because the conversion Iactor is given as 1.0 and only halI oI this is contributed to the deIormable body.
In absence oI plastic strain, the total strain energy value is the same as the total elastic strain energy (Figure 6.8-8).
Figure 6.8-8 The Strain Energy generated in the Sliding Block
ADD GLOBAL CRV (choose from VARIABLES)
Time
Total Strain Energy
Time
Total Elastic Strain Energy
1829
CHAPTER 6.8
Convergence Automation and Energy Calculations
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
convergence_a.proc Mentat procedure Iile
convergence_b.proc Mentat procedure Iile
1830 Marc Users Guide: Part 3
CHAPTER 6.8
Chapter 6.9: Capacitors
6.9
Capacitors

Chapter Overview 1832

Capacitance Computation in Symmetric Multiconductor Systems 1833

Results and Discussion 1849

Input Files 1850

Reference 1850
1832 Marc Users Guide: Part 3
CHAPTER 6.9
Chapter Overview
In electromagnetism and electronics, capacitance is the ability oI a body to hold an electrical charge. Capacitance is
also a measure oI the amount oI electric charge stored (or separated) Ior a given electric potential. It is commonly
Iound in electromagnetic Iields, which exert some sort oI physical Iorce on particles. The Iorce causes the particles to
exhibit motion which results in an electric charge.
Electronics use capacitance as a basic component in nearly every Iacet. Devices known as semiconductors help the
Ilow oI electrons through conductors made oI nonmetallic materials. They work with other electronic devices, most
notably capacitors, to make that Ilow work to power and control a large amount oI components.
Capacitors are the main component that harnesses the electric charge. These are essentially a pair oI conductors which
contain movable electric charge separated by a dielectric or insulator. In order Ior an electric Iield to be present inside
the insulator, a diIIerence between the voltages oI each conductor must be present. This is known as the potential
diIIerence. As the energy is stored, a mechanical Iorce is produced between the conductors. This is most common
between Ilat and narrowly separated conductors.
When two capacitors are placed close together Ior a period oI time, they create an eIIect known as 'stray capacitance.
This means that the electric charge loses some oI its signal and begins to leak within the isolated currents. Stray
capacitance is detrimental Ior the proper Iunction oI high Irequency currents.
A common Iorm oI charge storage device is a two-plate capacitor. Capacitance is directly proportional to the surIace
area oI the conductor plates and inversely proportional to the separation distance between the plates. II the charges on
the plates are and , and give the voltage between the plates, then the capacitance is given by
The capacitance, , above is also known as selI-capacitance. The energy (measured in Joules) stored in a capacitor is
equal to the work done to charge it. Consider a capacitance , holding a charge on one plate and on the other.
Moving a small element oI charge Irom one plate to the other against the potential diIIerence requires
the work :
where is the work measured in Joules, is the charge measured in Coulombs, and is the capacitance, measured
in Farads.
We can Iind the energy stored in a capacitance by integrating this equation. Starting with an uncharged capacitance
( ) and moving charge Irom one plate to the other until the plates have charge and requires the work :
SelI-capacitance must also be used within a number oI electrical devices. This occurs by increasing the electrical
charge by the amount that is needed to raise the potential by one volt. One way to allow selI-capacitance is by placing
Q Q V
C
Q
V
----
C
C q q
dq V q C
dW
dW
q
C
---- dq
W q C
q 0 Q Q W
W
ch i ng arg
q
C
---- q d
0
Q
}
1
2
---
Q
2
C
------
1833
CHAPTER 6.9
Capacitors
a hollow conducting sphere between the conductors. This makes the capacitor regulate itselI in regards to electrical
charge.
Capacitance is generally considered the inverse oI inductance, the concept oI resisting a change in current Ilow. Both
phenomena can be measured by substituting the voltage and current number within each equation with the opposite
measurement. In the same way, an inductor oIIsets the Iunction oI a capacitor.
The holding oI an electric charge is measured in Iarads. This is the amount oI electric charge potential that can change
one volt within a capacitor. It also measures the amount oI electric charge that can be transported in a single second
by a steady current. The SI unit oI capacitance is the Farad; 1 Farad 1 Coulomb per volt.
A problem with more than two conductors is best described by a capacitance matrix that Iully predicts the behavior oI
the problem under diIIerent voltage excitations. The capacitance matrix is essentially the measure oI charge storage
on a singly excited conductor as well as the charge storage on conductor due to excitation oI another conductor.
This chapter introduces the procedure Ior capacitance matrix computation. It also compares the capacitance matrix
values with known reIerence results and brieIly describes how the matrix can be used Ior practical problems. In
practical circuits, a number oI such capacitors are used as part oI the circuit, and the capacitance matrix values are
required to calculate the response oI the circuit to applied loads.
Computation oI capacitance matrix is a special case oI electrostatic analysis and is activated in the load cases. It
requires the deIinition oI capacitors as contact bodies. Boundary conditions are only required at Iar Iield location or
on ground objects. The voltage loads on conductors are automatically applied in Marc.
The procedure Iile to demonstrate this example is called capac.proc under mentat2009 /examples
/marc_ug/s6/c6.9. The required mud Iile is cap_init.mud.
Capacitance Computation in Symmetric Multiconductor
Systems
A capacitor system is made up oI three parallelepipedal conductors on a ground plane, Ior which two symmetry planes
exist: x 0 and y 0. Each conductor or capacitor is a rectangular plate oI uniIorm thickness( t) and their rectangular
dimensions are equal. The three rectangular plates are placed parallel to the XY plane and a distance (h) above it. The
plates extend equally on either side oI the YZ plane and adjacent plates are separated by a distance (s). Figure 6.9-1
shows all geometrical dimensions. They are:
Vacuum electrical permeability is assumed everywhere.
1 20 m
w 5 m
t 0.5 m
h 0.8 m and
s 10 m
c
0
1834 Marc Users Guide: Part 3
CHAPTER 6.9
In this problem, only Iar Iield and ground plane boundary conditions are applied. No applied voltage loads are
required. There is only one loadcase with one increment. The number oI subincrements n, where n is the total number
oI capacitors or conductors. The ground plane or the inIinite boundary is not counted as a conductor.
Figure 6.9-1 Three Parallel Rectangular Conductors on a Ground Plane
Each has the same dimensions , , and . Adjacent conductors are separated by a
distance . The conductors are placed at a distance above the ground plane. All materials in the
problem have the same permittivity, that oI vacuum 8.854E-12 Farads/m
Mesh Generation
In this problem, there is symmetry about the XZ and YZ planes. Only the symmetry about the XZ plane is used Ior
modeling, since the capacitance matrix is desired Ior all three conductors. Since the ground plane is considered inIinite,
it is not required to model below the ground plane. The problem domain above the ground plane extends to inIinity,
and it is required to consider a large domain above this plane. The problem has to be analyzed in 3-D. The domain is
modeled by a mixture oI 8-noded hexahedral and 6-noded pentahedral elements. The problem is Iirst modeled in 2-D
and then expanded in 3-D.
Element and Node Set Selection
Elements and nodes are grouped in sets Ior easy deIinition oI material properties and contact bodies and Ior detailed
viewing in post processing. The sets also help in manipulation oI the model. In this case, there are Iour element sets
and one nodal set. The element sets are:
1. conductor1
2. conductor2
3. conductor3
4. air

h
l
s
w
t
1
2
3
1 20 m w 5 m t 0.5 m
s 10 m h 0.3 m
1835
CHAPTER 6.9
Capacitors
There is one nodal set corresponding to all nodes on the Iar Iield boundary. This set is used to deIine the Iar Iield
boundary condition. This set is named as:
exterior
For the initial model, three element sets are created, conductor1, conductor2, and air. The button sequence Ior
creating element sets is:
SELECT
ELEMENT STORE
STORE ELEMENTS INTO NEW SET:
conductor1
OK
ZOOM
*zoom_box
*zoom_box)(1,0.259868,0.645074,0.429825,0.716599
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
181 182 183 184 185 186 187 188 189 190 191 192
# | End of List
IDENTIFY
SETS
REGENERATE
RETURN
ELEMENT STORE
STORE ELEMENTS INTO NEW SET:
Conductor2
OK
FILL
ZOOM
*zoom_box
*zoom_box(1,0.101974,0.658570,0.206140,0.716599)
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
75 76 77 78 79 80 81 82 83 84
# | End of List
1836 Marc Users Guide: Part 3
CHAPTER 6.9
For the initial model, one nodal set is created, exterior. The button sequence Ior creating the nodal set is:
SELECT SET
SELECT EXISTING SETS:
conductor1
conductor2
OK
ELEMENTS STORE
STORE ELEMENTS INTO NEW SET:
Air
OK
ALL: UNSEL
SELECT ALL UNSELECTED ELEMENTS:
CLEAR SELECT
FILL
RETURN
RETURN
SELECT METHOD
PATH
RETURN
NODES
1
1807
1708
1635
723
# | End of List
NODES STORE
STORE ELEMENTS INTO NEW SET:
Exterior
OK
ALL SELEC
CLEAR SELECT
RETURN
1837
CHAPTER 6.9
Capacitors
The model is then duplicated about the Y axis and the third element set, conductor3, is created. The button
sequence is:
The model is expanded in three dimensions in Iront oI the XY plane. The region behind the XY plane represents
symmetry and is not required to be modeled. The nodes lying in the extreme Iront represent part oI the Iar Iield region.
These nodes lie on the shiIted two dimensional elements and are added to the nodal set (exterior). The button
sequence is given as:
SELECT
RESET
ZOOM
*zoom_box
*zoom_box(1,0.360272,0.381107,0.641239,0.612378)
ELEMENTS
SELECT THE FOLLOWING ELEMENTS:
1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884
1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900
1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916
1917 1918 1919 1920 1921 1922 1923 1924
# | End of List
DEL ENTRIES
SELECT SET
SELECT EXISTING SET:
conductor1
ELEMENTS STORE
STORE ELEMENTS INTO NEW SET:
conductor3
ALL: SELEC
RETURN
RETURN
SELECT
SELECT BY
NODES ELEM
SELECT EXISTING ELEMENT SET:
This is the presently selected element set and consists of the 2-D elements only.
ALL: SELEC
1838 Marc Users Guide: Part 3
CHAPTER 6.9
Material Properties
All objects in this problem have the same value oI electric permittivity as vacuum and is 8.854E-12 Farads/meter.
Two materials are deIined: conductor and air. Material properties are speciIied Ior all objects using these base
materials.
Two materials are deIined with the same value oI permittivity:
The button sequence Ior speciIying the material properties reads:
RETURN
NODES STORE
SELECT EXISTING ELEMENT SET:
Exterior
OK
ALL: SELEC
RETURN
CLEAR SELECT
RETURN
1. Conductor
2. Air
MATERIAL PROPERTIES
MATERIAL PROPERTIES
ANALYSIS CLASS
ELECTROSTATIC
NEW
STANDARD
ELECTRIC PERMITTIVITY
ENTER PERMITTIVITY VALUE
8.854e-12
OK
NAME
ENTER MATERIAL NAME:
air
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
1839
CHAPTER 6.9
Capacitors
Contact
Elements in the three conductors are deIined as three electromagnetic contact bodies:
No properties are required Ior these contact bodies. A contact table is also not required. The button sequence Ior
speciIying the material properties reads:
Air
RETURN
NEW
STANDARD
ELECTRIC PERMITTIVITY
ENTER PERMITTIVITY VALUE
8.854e-12
OK
NAME
ENTER MATERIAL NAME:
conductor
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
SELECT SET
SELECT EXISTING ELEMENT SETS:
Conductor1
Conductor2
Conductor3
OK
RETURN
RETURN
RETURN
RETURN
1. Cond1
2. Cond2
3. Cond3
CONTACT
CONTACT BODIES
1840 Marc Users Guide: Part 3
CHAPTER 6.9
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME:
cond1
ELECTROMAGNETIC
USE ELECTROMAGNETIC PROPERTIES:
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
conductor1
OK
IDENTIFY
IDENTIFY CONTACT BODIES:
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME:
cond2
ELECTROMAGNETIC
USE ELECTROMAGNETIC PROPERTIES:
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
Conductor2
OK
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME:
cond2
ELECTROMAGNETIC
USE ELECTROMAGNETIC PROPERTIES:
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
1841
CHAPTER 6.9
Capacitors
Boundary Conditions
A Iixed electric potential 0 volts is applied on all Iar Iield and ground plane nodes. No other boundary conditions
are required.
The button sequence Ior speciIying the material properties reads:
Loadcase and Job Parameters
There is only one loadcase named capcase. This a basic electrostatic analysis repeated three times Ior computing the
capacitance matrix. The electric potential load on each oI the conductors is applied internally in the Marc code. Only
the Iar Iield and ground plane is constrained to zero potential as described above. The steps to describe the loadcase
are shown in Figure 6.9-2.
Conductor3
OK
RETURN
RETURN
RETURN
BOUNDARY CONDITIONS
ANALYSIS TYPE:
ELECTROSTATIC
NEW
CREATE NEW ELECTROSTATIC BOUNDARY CONDITION:
NAME
ENTER BOUNDARY CONDITION NAME:
fix_pot
FIXED POTENTIAL
APPLY FIXED POTENTIAL (TOP):
POTENTIAL (TOP)
ENTER THE VALUE:
0
NODES ADD
SELECT EXISTING ELEMENT SET:
Exterior
OK
RETURN
RETURN
1842 Marc Users Guide: Part 3
CHAPTER 6.9
Figure 6.9-2 Steps for showing Button Click to define the Loadcase (capcase)
1843
CHAPTER 6.9
Capacitors
The button sequence Ior speciIying the material properties reads:
The job speciIication includes only one loadcase mentioned above. The elemental post processing selected are the X,
Y, and Z components oI the electric Iield intensity and the electric Ilux densities.
The steps to describe the loadcase are shown in Figure 6.9-3.
LOADCASE
CREATE NEW LOADCASE:
NEW
CREATE NEW LOADCASE TYPE:
ELECTROSTATIC
NAME
ENTER LOADCASE NAME:
Capcase
STEADY STATE
SELECT LOAD
fix_pot
OK
CAPACITANCE CALCULATION
SELECT CONTACT BODIES FOR CAPACITANCE CALCULATION:
ADD ALL
ADD ALL BODIES:
OK
OK
RETURN
RETURN
RETURN
1844 Marc Users Guide: Part 3
CHAPTER 6.9
Figure 6.9-3 Steps for showing Button Click to Define the JOB.
1845
CHAPTER 6.9
Capacitors
The button sequence Ior speciIying the material properties reads:
Save Model, Run Job and View Results
The Iinite element model is Iirst saved as a mud Iile, capcase.mud.
JOBS
CREATE NEW JOB:
NEW
ANALYSIS TYPE: ELECTROSTATIC
ELECTROSTATIC
PROPERTIES
SELECT VARIOUS PROPERTIES:
SELECT LOAD CASE:
capcase
JOB ANALYSIS
OK
JOB RESULTS
SELECT THE VARIOUS ELEMNTAL POST QUANTITIES:
1st Component of Electric Field Intensity
2nd Component of Electric Field Intensity
3rd Component of Electric Field Intensity
1st Component of Electric Displacement
2nd Component of Electric Displacement
3rd Component of Electric Displacement
OK
JOB PARAMETERS
OK
OK
RETURN
RETURN
FILE
SAVE AS
SAVE MODEL IN FILE:
Capcase.mud
OK
1846 Marc Users Guide: Part 3
CHAPTER 6.9
It is desired to view the contour plots Ior the electric potential Ior each oI the three subincrements. These contours give
an idea oI how the electric potential load is applied to each conductor. It can also reveal visually detectable errors in
the solution.
To see the electric potential contour plots:
Now rotate and translate the model so as to see the electric potential contours Ior subincrement 3. Then use the PLOT
settings to get a better picture oI the contour plot:
RETURN
JOBS
RUN
RUN THIS JOB:
SUBMIT 1
MONITOR
OK
RETURN
RESULTS
OPEN DEFAULT
OPEN DEFAULT POST FILE:
RETURN
SCALAR
Electric Potential
VIEW ELECTRIC POTENTIAL CONTOURS:
OK
CONTOUR BANDS
PLOT
NODES
SET NODE PLOT OFF
ELEMENTS SETTING
OUTLINE
SET ELEMENT PLOT TO OUTLINE:
REGENERATE
RETURN
RETURN
1847
CHAPTER 6.9
Capacitors
The contour plot Ior subincrement 3 is shown in Figure 6.9-4.
Figure 6.9-4 Electric Potential Distribution for Subincrement 3
The contours are seen wrapped around conductor 3.
To view the contour plots Ior subincrement 2 and then subincrement 3:
PREV
PREV
1848 Marc Users Guide: Part 3
CHAPTER 6.9
Figure 6.9-5 Electric Potential Distribution for Subincrement 2
The contours are seen wrapped around conductor 2.
Figure 6.9-6 Electric potential distribution for Subincrement 1.
The contours are seen wrapped around conductor 1.
1849
CHAPTER 6.9
Capacitors
Results and Discussion
The present problem is reIerred IromReference 1.
In this reIerence, the values oI the direct capacitance matrix are computed using the boundary element method.
The capacitance matrix is computed by Marc and stored in the capcase.out out Iile. These values are shown in
Table 6.9-1.
The above values apply to the problem with halI-symmetry and have to be multiplied by two to get the correct values
Ior the matrix. This is shown in Table 6.9-2.
The direct capacitances are given by the expressions:
and
The direct capacitance values are calculated Irom the above table and are tabulated in Table 6.9-3.
Table 6.9-1 Capacitance Matrix for the Three Conductors in Farads
Column 1 2 3
Row
1 3.8687E-15 -2.0164E-17 -8.6667E-19
2 -2.0164E-17 3.8771E-15 -2.0164E-17
3 -8.6667E-19 -2.0164E-17 3.8687E-15
Table 6.9-2 Capacitance Matrix for the Three Conductors in Farads
Column 1 2 3
Row
1 7.7374E-15 -4.0328E-17 -1.7333E-18
2 -4.0328E-17 7.7542E-15 -4.0328E-17
3 -1.7333E-18 -4.0328E-17 7.7374E-15
Table 6.9-3 Direct Capacitance Matrix for the Three Conductors in Farads
Column 1 2 3
Row
1 7.6953E-15 4.0328E-17 1.7333E-18
2 -4.0328E-17 7.6735E-15 4.0328E-17
3 -1.7333E-18 4.0328E-17 7.6953E-15
C
di i
C
i j
j 1
3

C
di j
C
i j

1850 Marc Users Guide: Part 3
CHAPTER 6.9
The Iirst row oI the above matrix is compared with the results oI the reIerence above in Table 6.9-4.
The direct capacitance value indicates the behavior oI a conductor when connected to an electric circuit and this is the
value to be used to Iind the circuit response. This value is the measure oI the charge induced on the conductor when it
is in the proximity oI other conductors and a ground plane (iI it exists) and when it is excited by an electric potential.
The remaining conductors are grounded.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Reference
1. G. Aiello, S. AlIonzetti and S. Coco: 'Capacitance Computation in Symmetric Multiconductor Systems,
IEEE Transactions ON MAGNETICS, VOL. 30, NO. 5, SEPTEMBER 1994, pp 2952-2955.
Table 6.9-4 Comparison of Direct Capacitance Values for the Three Conductors with the Reference
Values (Farads)
Cd11 Cd12 Cd13
Marc Results 7.69534E-15 4.03280E-17 1.73334E-18
ReIerence 7.34000E-15 4.60000E-17 3.00000E-18
File Description
capac.proc Mentat procedure Iile
cap_init.mud Mentat model Iile
Chapter 6.10: Inductance Between Two Long Conductors
6.10
Inductance Between Two Long
Conductors

Summary 1852

Inductance Computation in Two Infinitely Long

Rectangular Conductors 1854

Results and Discussion 1872

Input Files 1872

Reference 1873
1852 Marc Users Guide: Part 3
CHAPTER 6.10
Summary
Title Inductance Between Two Long Conductors
Problem Ieatures Inductance matrix calculation
Geometry
Material properties Air and conductor have same magnetic permeability oI a vacuum.
Analysis type Magnetostatic
Boundary conditions Fixed magnetic scalar potential Az 0 all Iar Iield nodes. A current oI 100 A applied in
opposite directions in the two conductors
Element type Planar
FE results Magnetic Induction (outer air and conductor elements removed)
Y
X
Z
1
w
s
t
Conductor 1
Conductor 2
Current
Air all around t = 0.0025 m
s = 0.04 m
w = 0.01 m
8.958e-009
5.479e-004
1.096e-003
1.644e-003
2.192e-003
2.740e-003
3.288e-003
3.835e-003
4.383e-003
4.931e-003
5.479e-003
indcase
Magnetic Induction
Inc: 1:2
Time: 2.000e+000
X
Y
Z
1
1853
CHAPTER 6.10
Inductance Between Two Long Conductors
Inductance is the property in an electrical circuit where a change in the current Ilowing through that circuit induces an
electromotive Iorce (EMF) that opposes the change in current. In electrical circuits, any electric current produces a
magnetic Iield and hence generates a total magnetic Ilux acting on the circuit. This magnetic Ilux, due to Lenz's law,
tends to act to oppose changes in the Ilux by generating a voltage (a back EMF) that counters or tends to reduce the
rate oI change in the current.
The quantitative deIinition oI the (selI-) inductance oI a wire loop in SI units (Webers per ampere) is
where denotes the magnetic Ilux through the area spanned by the loop, and is the number oI wire turns. The Ilux
linkage thus is
However, there may be contributions Irom other circuits. Consider Ior example two circuits , , carrying the
currents , . The Ilux linkages oI and are given by
According to the above deIinition, and are the selI-inductances oI and , respectively. It can be shown
(see below) that the other two coeIIicients are equal: , where is called the mutual inductance oI the
pair oI circuits. The number oI turns and occur somewhat asymmetrically in the deIinition above. But actually
always is proportional to the product , and thus, the total currents contribute to the Ilux.
SelI and mutual inductances also occur in the expression
Ior the energy oI the magnetic Iield generated by electrical circuits where is the current in the nth circuit. This
equation is an alternative deIinition oI inductance that also applies when the currents are not conIined to thin wires so
that it is not immediately clear what area is encompassed by the circuit nor how the magnetic Ilux through the circuit
is to be deIined.
The deIinition , in contrast, is more direct and more intuitive. It may be shown that the two deIinitions are
equivalent by equating the time derivative oI and the electric power transIerred to the system.
An inductor is a passive electronic component that stores energy in the Iorm oI a magnetic Iield. Inductors are used
with capacitors in various wireless communications applications. An inductor connected in series or parallel with a
capacitor can provide discrimination against unwanted signals. Large inductors are used in the power supplies oI
electronic equipment oI all types, including computers and their peripherals. In these systems, the inductors help to
smooth out the rectiIied utility AC, providing pure, battery-like DC.
i
u
L
Nu
i
---------
u N
Nu
Nu Li
C
1
C
2
i
1
i
2
C
1
C
2
N
1
u
1
L
11
i
1
L
12
i
2

N
2
u
2
L
21
i
1
L
22
i
2

L
11
L
22
C
1
C
2
L
12
L
21
M M
N
1
N
2
L
mn
N
m
N
n
N
m
i
m
W
1
2
-- - L
m n ,
i
m
i
n
m n , 1
K

K i
n
L Nu i
W
1854 Marc Users Guide: Part 3
CHAPTER 6.10
The procedure Iile to demonstrate this example is called indcase.proc under:
mentat2009 /examples /marc_ug/s6/c6.10
Inductance Computation in Two Infinitely Long
Rectangular Conductors
An inductor system is made up oI two inIinitely long and parallel rectangular conductors. The conductors are parallel
to the Z axis and are symmetrically placed about the XZ and YZ planes. The two conductors carry the same current
I 100 Amperes, but in opposite directions. Figure 6.10-1 shows all geometrical dimensions. They are:
Figure 6.10-1 Two Parallel Infinitely Long Rectangular Conductors
Vacuum magnetic permeability is assumed everywhere.
t 0.0025 meters
w 0.01 meters
s 0.04 meters
Y
X
Z
w
s
t
Conductor 1
Conductor 2
Current
Air all around
t = 0.0025 m
s = 0.04 m
w = 0.01 m

0
1855
CHAPTER 6.10
Inductance Between Two Long Conductors
In this problem, the Iar Iield and coil current boundary conditions are applied. There is only one loadcase with one
increment. The number oI subincrements n * (n1) / 2, where n is the total number oI inductors or conductors. In
this problem, three new Ieatures are introduced Ior magnetostatic analysis:
Electrical winding speciIication
Coil Current boundary condition
Inductance matrix computation.
Electrical windings and Coil Current boundary condition
In many electrical problems a magnetic circuit is excited by electrical windings. A winding is a set oI multiturn coils.
A set oI multiturn coils end in two termination points called terminals. Each coil has the same cross-section, that is,
with respect to shape and size. Usually, a coil has a rectangular or circular shape. Any other shape is rarely used. Each
coil has a thin insulation coating. The multiturns are bound together by some insulation material. The set oI multiturn
itselI deIines a cross-section. This is the winding cross-section. A winding cross-section is usually rectangular,
sometimes circular. A uniIorm electric current is assumed to Ilow through each coil turn. Electrical winding is
speciIied by its cross-section type and cross-section dimensions, number oI coil turns, the coil cross-section type and
cross-section dimensions and the path it Iollows along with its orientation. These speciIications are applied in the
winding sub-menu oI modeling tools. The actual coil current and the elements to which it applies are speciIied in the
coil current boundary condition.
Inductance matrix computation
For this computation, the speciIication oI the Electrical windings and Coil Current boundary condition is required. The
elements deIining the winding in the coil current boundary condition have to be deIined as electromagnetic contact
bodies. These contact bodies identiIy the inductors. The inductance matrix computation is activated in the loadcase
Ieature by selecting the required contact bodies Ior the matrix computation.
As shown in Figure 6.10-1, there are two parallel inIinitely long rectangular conductors carrying the same current but
in opposite direction. Each has the same dimensions t 0.0025 m, w 0.01 m. Adjacent conductors are separated by
a distance s 0.04 m. All materials in the problem have the same permeability, that oI vacuum 1.25664E-06 Henry/m
Mesh Generation
In this problem, there is symmetry about the XZ and YZ planes. The symmetry about the XZ plane can be used Ior
modeling, but is not used in order to get the correct values in the inductance matrix. The whole problem domain
consisting oI the two conductors is modeled. The problem domain extends to inIinity and it is required to consider a
large domain about the conductors. The problem has to be analyzed in 2D, since both conductors are inIinite in the Z
direction. The domain is modeled by 8 noded hexahedral elements. The problem is Iirst modeled in 2D in the Iirst
quarter oI the XY plane. Then symmetry is used to duplicate the model in the remaining three quarters.
1856 Marc Users Guide: Part 3
CHAPTER 6.10
Element Selection as Sets
The materials in the problem and contact bodies are a set oI elements and these sets are selected as a collection oI
elements in the two conductors and air. The material names Ior the conductors are:
Conductor_1
Conductor_2
Three sets are created Ior the collection oI elements in the three materials. These sets are named as:
Cond1
Cond2
air
The creation oI the above sets gives a convenient way oI deIining materials and contact bodies and also in viewing
oI post processing results. The basic model contains only conductor1 and air and elements sets are created Ior
these Iirst.
The button sequence Ior speciIying the material properties reads:
SELECT
ELEMENTS
SELECT NEW ELEMENT SET:
ZOOM
zoom_box(1,0.121711,0.711201,0.214912,0.784076)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
92 93 94 95 96 97 98 99 100
# | End of List
ELEMENTS STORE
STORE IN NEW ELEMENT SET:
cond1
RETURN
IDENTIFY SETS
REGENERATE
ELEMENTS STORE
STORE IN NEW ELEMENT SET:
air
ALL: UNSEL
ELEMENTS: CLR
1857
CHAPTER 6.10
Inductance Between Two Long Conductors
The boundary oI the one-Iourth quadrant deIines the Iar Iield boundary and the nodes on this boundary have a Iixed
potential oI 0 volts. The nodes on this boundary are deIined as a set as Iollows:
The Iinite element model is then duplicated about the X axis and the conductor2 is extracted as a set oI elements
Irom the duplicated elements.
CLEAR ALL ELEMENT SELECTIONS:
FILL
RETURN
PLOT
NODES
SET NODES PLOTTING ON:
REGENERATE
RETURN
SELECT
STORE NODE PATH
ENTER NODE SET NAME:
Exterior
SELECT THE NODES ALONG THE PATH:
529
976
825
815
# | End of List
RETURN
RETURN
SELECT
ELEMENTS
SELECT EXISTING SET:
Cond1
SELECT MODE
SELECT INTERSECT MODE:
INTERSECT
RETURN
SELECT METHOD
1858 Marc Users Guide: Part 3
CHAPTER 6.10
Material Properties
All objects in this problem have the same value oI magnetic permeability as vacuum and is 1.25664E-06
Henry/meter. But it is desirable to deIine diIIerent material identiIication Ior each conductor as well as the surrounding
air. This helps in creation oI element sets, which are used Ior creating contact bodies and Ior viewing results in post
processing.
Three materials are deIined with the same value oI permeability:
Conductor_1
Conductor_2
Air
The button sequence Ior speciIying the material properties reads:
SELECT METHOD TYPE:
BOX
ELEMENTS
Enter Range in X direction: -1 1
Enter Range in Y direction: -1 0
Enter Range in Z direction: -1 1
DEL ENTRIES
REMOVE ABOVE ELEMENTS FROM THE SET cond1:
OK
ELEMENTS STORE
STORE SELECTED ELEMENTS IN NEW ELEMENT SET:
Cond2
OK
ALL: SELEC
RETURN
RETURN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
SET ANALYSIS CLASS:
ANALYSIS CLASS:MAGNETOSTATIC
SET STANDARD MATERIAL:
NEW: STANDARD
NAME
CREATE NEW MATERIAL NAME:
1859
CHAPTER 6.10
Inductance Between Two Long Conductors
Contact
Elements in the three conductors are deIined as three electromagnetic contact bodies:
Cond1
Cond2
Cond3
air
MAGNETIC PERMEABILITY
MU
SET MAGNETIC PERMEABILITY VALUE:
1.25664e-6
OK
ELEMENTS ADD
SELECT EXISTING SET:
air
# End List
ID MATERIALS
SET STANDARD MATERIAL:
NEW: STANDARD
NAME
CREATE NEW MATERIAL NAME:
conductor
MAGNETIC PERMEABILITY
MU
SET MAGNETIC PERMEABILITY VALUE:
1.25664e-6
OK
ELEMENTS ADD
SELECT EXISTING SET:
Cond1
Cond2
# End List
RETURN
RETURN
RETURN
1860 Marc Users Guide: Part 3
CHAPTER 6.10
No properties are required Ior these contact bodies. A contact table is also not required.
The button sequence Ior speciIying the material properties reads:
Modeling Tools
The electric current in the rectangular conductors is speciIied through the winding Ieature. This is required Ior the
inductance matrix computation. Each rectangle has one turn and its cross-section is given by and
. The coil, in this case, is the same as the winding and one can use the same speciIication Ior the coil or
use a circular coil with an appropriate radius. The coil data is not used in the inductance calculation. The winding path
and its orientation are speciIied by a single node each Ior the 2-D case. The corresponding electromagnetic contact
body associated with the winding is selected here. Two windings are deIined Ior the two conductors.
CONTACT
CONTACT BODIES
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME
Cond1
ELECTROMAGNETIC
DEFINE ELECTROMAGNETIC CONTACT BODY:
OK
ELEMENTS: ADD
Cond1
# End List
ID CONTACTS
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME
Cond2
ELECTROMAGNETIC
DEFINE ELECTROMAGNETIC CONTACT BODY:
OK
ELEMENTS: ADD
Cond2
# End List
w 0.01 m
t 0.0025 m
1861
CHAPTER 6.10
Inductance Between Two Long Conductors
The button sequence Ior speciIying the windings in modeling tools reads:
MODELING TOOLS
WINDINGS
NEW
CREATE NEW WINDINGS:
PROPERTIES
DEFINE WINDING PROPERTIES
CROSS-SECTION: RECTANGULAR
LENGTH
ENTER LENGTH VALUE
.01
WIDTH
ENTER WIDTH VALUE
.0025
COIL:CROSSSECTION: CIRCULAR
RADIUS
ENTER RADIUS VALUE
.0001
CONDUCTOR BODY
SELECT CONTACT BODY:
Cond1
OK
SEGMENTS: ADD
ADD WINDING SEGMENTS:
SELECT OPTION: NODE PATH
CENTERLINE: NODES: ADD
SELECT NODE ID:
1
# End List
CROSS-SECTION ORIETNTATION: NODES: ADD
SELECT NODE ID:
8
# End List
FILL
NEW
1862 Marc Users Guide: Part 3
CHAPTER 6.10
Boundary Conditions
A Iixed magnetic scalar potential(Az) 0 volts is applied on all Iar Iield nodes. The coil current boundary condition
is used to apply uniIorm current in both conductors. For this case, the current value and the winding identiIication path
is speciIied on the appropriate conductor elements.
CREATE NEW WINDINGS:
PROPERTIES
DEFINE WINDING PROPERTIES
CROSS-SECTION: RECTANGULAR
LENGTH
ENTER LENGTH VALUE
.01
WIDTH
ENTER WIDTH VALUE
.0025
COIL:CROSSSECTION: CIRCULAR
RADIUS
ENTER RADIUS VALUE
.0001
CONDUCTOR BODY
SELECT CONTACT BODY:
Cond2
OK
SEGMENTS: ADD
ADD WINDING SEGMENTS:
SELECT OPTION: NODE PATH
CENTERLINE: NODES: ADD
SELECT NODE ID:
1137
# End List
CROSS-SECTION ORIETNTATION: NODES: ADD
SELECT NODE ID:
1144
# End List
SEGMENTS: REVERSE
REVERSE OPTION IS SET:ON:
1863
CHAPTER 6.10
Inductance Between Two Long Conductors
The button sequence Ior speciIying the boundary conditions reads:
BOUNDARY CONDITIONS
NEW
SELECT: MAGNETOSTATIC ANALYSIS
NAME
ENTER THE FIXED POTENTIAL BOUNDARY CONDITION NAME:
Fix_pot
FIXED POTENTIAL
POTENTIAL
ENTER POTENTIAL VALUE:
0
OK
NODES: ADD
ENTER NODE SET:
exterior
# | End of List
NEW
NAME
ENTER THE COIL CURRENT BOUNDARY CONDITION NAME:
wind1
COIL CURRENT
COIL CURRENT
ENTER CURRENT VALUE
100
WINDING PATH
SELECT WINDING PATH:
winding1
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
Cond1
# | End of List
FILL
NEW
1864 Marc Users Guide: Part 3
CHAPTER 6.10
Loadcase and Job Parameters
There is only one loadcase named indcase. This a basic magnetostatic analysis repeated three times Ior computing
the inductance matrix. The electric coil current on each oI the conductors is modiIied internally in the Marc code. Only
the Iar Iield is constrained to zero magnetic potential as described above. The steps to describe the loadcase are shown
in Figure 6.10.2.
NAME
ENTER THE COIL CURRENT BOUNDARY CONDITION NAME:
Wind2
COIL CURRENT
COIL CURRENT
ENTER CURRENT VALUE
100
WINDING PATH
SELECT WINDING PATH:
Winding2
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
Cond2
# | End of List
RETURN
RETURN
1865
CHAPTER 6.10
Inductance Between Two Long Conductors
Figure 6.10-2 Steps for Showing Button Click to Define the Loadcase indcase
1866 Marc Users Guide: Part 3
CHAPTER 6.10
The button sequence Ior speciIying the material properties reads:
The job speciIication includes only one loadcase mentioned above. The elemental post processing selected are the X
and Y components oI the magnetic Iield intensity and the magnetic Ilux density.
The steps to describe the loadcase are shown in Figure 6.10-3.
LOADCASE
NEW
CREATE NEW LOADCASE:
NAME
ENTER LOADCASE NAME:
indcase
MAGNETOSTATIC
STEADY STATE
LOADS
SELECT LOADS:
Fix_pot
Wind1
Wind2
INDUCTANCE CALCULATION
INDUCTANCE CALCULATION
SELECT ALL CONDUCTING BODIES:
ADD ALL
OK
OK
RETURN
RETURN
1867
CHAPTER 6.10
Inductance Between Two Long Conductors
Figure 6.10-3 Steps for Showing Button click to Define the JOB.
1868 Marc Users Guide: Part 3
CHAPTER 6.10
The button sequence Ior speciIying the material properties reads:
Save Model, Run Job, and View results
The Iinite element model is Iirst saved as a mud Iile, indcase.mud.
JOBS
NEW
SELECT ANALYSIS: MAGNETOSTATIC
PROPERTIES
ADD LOAD CASE:
indcase
SELECT JOB: PLANAR
JOB RESULTS
SELECT THE FOLLOWING QUANTITIES:
1st Component of Magnetic Induction
2nd Component of Magnetic Induction
1st Component of Magnetic Field Intensity
2nd Component of Magnetic Field Intensity
OK
OK
RETURN
FILE
SAVE AS
Ind_final.mud
OK
RETURN
JOBS
RUN
SUBMIT 1
MONITOR
OK
RETURN
RESULTS
OPEN DEFAULT
*post_open_default
1869
CHAPTER 6.10
Inductance Between Two Long Conductors
It is desired to view the contour plots Ior the external electric current Ior each oI the three subincrements. These
contours give an idea oI how the external electric current load is distributed in each conductor. It can also reveal
visually detectable errors in the solution.
To see the electric potential contour plots:
Observe the external electric current contours Ior subincrement 3. Then use the PLOT settings to get a better picture
oI the contour plot:
The contour plot Ior subincrement 3 is shown in Figure 6.10-4.
Figure 6.10-4 Electric External Current Distribution for Subincrement 3
SCALAR
External Electric Current
OK
CONTOUR BANDS
PLOT
A NODE PLOTTING IS SET OFF:
NODES
ELEMENTS SETTING
OUTLINE
REGENRATE
RETURN
1870 Marc Users Guide: Part 3
CHAPTER 6.10
To view the contour plots Ior subincrement 2 and then subincrement 3:
You will observe that the external electric current is same Ior all three subincrements.
Now, repeat the above steps Ior magnetic induction contour bands. The plots are shown in:
Figure 6.10-5 Magnetic Induction Distribution for Subincrement 3
PREV
PREV
Figure 6.10-5 The contours are seen wrapped around conductor 2.
Figure 6.10-6 The contours are seen wrapped around both conductors.
Figure 6.10-7 The contours are seen wrapped around conductor 1.
1871
CHAPTER 6.10
Inductance Between Two Long Conductors
Figure 6.10-6 Magnetic Induction Distribution for Subincrement 2
Figure 6.10-7 Magnetic Induction Distribution for Subincrement 1
8.958e-009
5.479e-004
1.096e-003
1.644e-003
2.192e-003
2.740e-003
3.288e-003
3.835e-003
4.383e-003
4.931e-003
5.479e-003
indcase
Magnetic Induction
Inc: 1:2
Time: 2.000e+000
X
Y
Z
1
1872 Marc Users Guide: Part 3
CHAPTER 6.10
Results and Discussion
The present problem is reIerred Irom Reference 1.
In this reIerence, the method oI direct integration using Maple is used. The Iollowing is taken Irom the reIerence.
II the rectangular conductor is very thin or and , then the inductance is approximately given by
(6.10-1)
The total inductance oI the two rectangular conductors is given approximately by:
(6.10-2)
The total inductance implies that in the present problem:
(6.10-3)
Using the dimensions chosen Ior this problem and the equations (1), (2) and (3), gives
L1.06509E-06 Henry
The inductance matrix is computed by Marc and stored in the out Iile: indcase.out. These values are shown in
Table 6.10-1.
Using the Marc results Irom the Table 6.10-1, gives
L 1.1105E-06 Henry
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
Table 6.10-1 Inductance Matrix for the Two Conductors in Henries
Column 1 2
Rows
1 6.464300E-07 9.117700E-08
2 9.117700E-08 6.464300E-07
File Description
indcase.proc Mentat procedure Iile
ind_init.mud Mentat model Iile
t 0 s w > L t ( 0 )
L t ( 0 )

0
t
-----
s
w
--- -
\ .
| |
3
2
---
\ .
| |
ln
L L t ( 0 )

0
t
3w
--------

0
24t
--------- 2
w
2
s
2
t
2 w
2
s
2
( )
-----------------------
\ .
| |
25
3
------ ln
t
w
----
\ .
| |
2

L L
11
L
22
2L
12

1873
CHAPTER 6.10
Inductance Between Two Long Conductors
Reference
1. K.F. Goddard, A.A. Roy and J.K. Sykulski: 'Inductance and resistance calculations Ior a pair oI rectangular
conductors, IEE Proc.-Sci. Meas. Technol., Vol. 152, No. 2, March 2005, pp 73-78.
1874 Marc Users Guide: Part 3
CHAPTER 6.10
Chapter 6.11: Lamination Loss in Magnetostatic-Thermal coupling
6.11
Lamination Loss in
Magnetostatic-Thermal
coupling

Summary 1876

Lamination Loss Computation and ohmic Winding Loss in a C

Core Cylindrical Inductor 1877

Results and Discussion 1899

Input Files 1899

1876 Marc Users Guide: Part 3
CHAPTER 6.11
Summary
Title Lamination loss in magnetostatic-thermal coupling
Problem Ieatures 'C core cylindrical inductor
Geometry
Material properties Typical properties oI the air, core, and winding along with lamination loss data are used
herein.
Analysis type Magnetostatic-Thermal
Boundary conditions The outer boundary oI the air has a Iixed magnetic potential oI 0. Each coil, radius oI
1x10
-5
m with a current oI 2 A, is wound 100 times in the circumIerential direction The
windings span the rectangular cross section shown above.
Initial conditions T 30
o
C at all nodes
Element type Planar axisymmetric
FE results Magnetic Induction
r
1
r
2
r
3
l
d
d = 0.02 m
l = 0.05 m
r
1
= 0.02 m
r
2
= 0.04 m
r
3
= 0.06 m
Coil Winding Cross Section
Core
Core-Air Boundry
100 windings
7.845e-005
1.045e-001
2.090e-001
3.134e-001
4.179e-001
5.223e-001
6.268e-001
7.313e-001
8.357e-001
9.402e-001
1.045e+000
lamcase
Core Only (expanded)
Magnetic Induction
Inc: 10
Time: 1.000e+002
Z
Y
X
1877
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Core loss in a magnetic material occurs when the material is subjected to a time varying magnetic Ilux. The actual
physical nature oI this loss is still not completely understood and a simplistic explanation oI this complex mechanism
is as Iollows. Energy is used to eIIect 'magnetic domain wall motion as the domains grow and rotate under the
inIluence oI an externally applied magnetic Iield. When the external Iield is reduced or reversed Irom a given value,
domain wall motion again occurs to realize the necessary alignment oI domains with the new value oI the magnetic
Iield. The energy associated with domain wall motion is irreversible and maniIests itselI as heat within the magnetic
material. The rate at which the external Iield is changed has a strong inIluence upon the magnitude oI the loss, and the
loss is generally proportional to some Iunction oI the Irequency oI variation oI the magnetic Iield. The metallurgical
structure oI the magnetic material, including its electrical conductivity, also has a proIound eIIect upon the magnitude
oI the loss. In electrical machines, this loss is generally termed the core loss.
The core loss contains three components: hysteresis loss, eddy loss, and extra loss. The extra loss clubs together all
losses not accounted by hysteresis and eddy losses and contains higher order harmonic losses.
The statistical loss theory provides the best method to compute the core loss accurately. Please see Marc Volume A:
Theory and User Information Ior details.
Practicing engineers would like to know the core losses as accurately as possible, since they are important Ior analysis
oI large electrical machinery and transIormers. Prediction oI these losses and their location in the device is useIul in
design. Even a small percentage reduction oI loss amounts to huge savings.
The computation oI these losses assumes the availability oI lamination loss curves. These curves are provided by the
manuIacturers oI the magnetic lamination sheets. The manuIacturers obtain the curve data by actual measurements at
diIIerent Irequencies and magnetic induction values. The data is expressed as a series oI curves, one each Ior a single
operational Irequency. Each curve shows the variation oI power loss per unit volume with the magnetic induction B.
To use these curves, the user has to extract suIIicient number oI data points Irom these curves. For each curve, the user
should write down the power loss with respect to the magnetic induction B. This has to be repeated Ior diIIerent
Irequencies. The whole data must be put into table Iormat as it has to be entered as a table in Mentat. Mentat can also
read a preIormatted data Iile that contains the tabulated data points. In this example, a preIormatted data Iile is used.
The procedure Iile to demonstrate this example is called lamcase.proc under
mentat2009/examples/marc_ug/s6/c6.11. The procedure requires an initial mud Iile, lam_init.mud.
Lamination Loss Computation and ohmic Winding Loss in a
C Core Cylindrical Inductor
A cylindrical inductor is made in C` shape. The inductor is like a coaxial cable made oI magnetic lamination material.
The inner core is a solid cylinder and the outer core is a cylindrical shell made Irom laminations. Both sides oI this
structure are capped on both sides by two cylindrical disks and made Iorm the same lamination material. The space
between the inner and outer cylinder is occupied by current carrying winding coils. The winding is made oI 100 turns
and each coil carries a current oI 2.0 Amperes. The initial temperature throughout the domain is 30 Celsius and is valid
just beIore the current application. This is an axisymmetric problem about the cylindrical axis.
1878 Marc Users Guide: Part 3
CHAPTER 6.11
Figure 6.11-1 shows all geometrical dimensions. They are:
Figure 6.11-1 Cylindrical C Core Inductor with a Quarter Cut in the Inductor
The winding completely Iills the gap in the magnetic core. The various dimensions are shown in Figure 6.11-1. The
core is made oI magnetic lamination held in place by a suitable mechanism. The winding consists oI a coil oI 100 turns.
There are three objects/materials in this problem: air, magnetic core, and winding. The winding actually contains
conductor like copper and insulation material. The insulation material is very thin and can be neglected. The magnetic
core is made oI lamination sheets held in place by some mechanism. The lamination sheets are coated with some
insulation, which is neglected as it is very thin. In eIIect, there are three materials: air, core, and winding (conductor).
l 0.05 meters
d 0.02 meters
r1 0.02 meters
r2 0.04 meters
r3 0.06 meters
r
1
r
2
r
3
l
d
d = 0.02 m
l = 0.05 m
r
1
= 0.02 m
r
2
= 0.04 m
r
3
= 0.06 m
Coil Windings
Core
Core-Air Boundry
1879
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
The material properties oI the three materials are given below:
1. Air
a. Thermal
Thermal conductivity: K 0.1Watts/(meter-Kelvin)
SpeciIic Heat: C 900.0 Joules/(kg-Kelvin)
Mass density: 1000.0 kg/(cu. Meter)
b. Electrical
Magnetic permeability 1.25664E-06 Henry/meter
Electric conductivity 0 Siemens/meter
2. Core
a. Thermal
Thermal conductivity: K 10.0Watts/(meter-Kelvin)
SpeciIic Heat: C 100.0 Joules/(kg-Kelvin)
Mass density: 7000.0 kg/(cu. Meter)
b. Electrical
Magnetic permeability 0.01 Henry/meter
Electric conductivity 1.0E08 Siemens/meter
3. Winding
a. Thermal
Thermal conductivity: K 0.2Watts/(meter-Kelvin)
SpeciIic Heat: C 600.0 Joules/(kg-Kelvin)
Mass density: 3000.0 kg/(cu. Meter)
b. Electrical
Magnetic permeability 1.25664E-06 Henry/meter
Electric conductivity 5.0E05 Siemens/meter
In this problem, the Iar Iield and coil current boundary condition is applied. There is only one loadcase with 10
increments corresponding to transient thermal analysis. In this problem, Iour new Ieatures are introduced:
Electrical winding speciIication
Coil Current boundary condition
Lamination loss computation.
Coupled Magnetostatic-thermal analysis
Electrical windings and Coil Current boundary condition
In many electrical problems, a magnetic circuit is excited by electrical windings. A winding is a set oI multiturn coils.
A set oI multiturn coils end in two termination points called terminals. Each coil has the same cross section, that is,
1880 Marc Users Guide: Part 3
CHAPTER 6.11
with respect to shape and size. Usually, a coil has a rectangular or circular shape. Any other shape is rarely used. Each
coil has a thin insulation coating. The multiturns are bound together by some insulation material. The set oI multiturn
itselI deIines a cross section. This is the winding cross-section. A winding cross-section is usually rectangular;
however, sometimes it is circular. A uniIorm electric current is assumed to Ilow through each coil turn. Electrical
winding is speciIied by its cross-section type and cross-section dimensions, number oI coil turns, the coil cross-section
type and cross-section dimensions and the path it Iollows along with its orientation. These speciIications are applied
in the winding submenu oI modeling tools. The actual coil current and the elements to which it applies are speciIied
in the coil current boundary condition.
Lamination loss computation
For this computation, the lamination loss curve is required. The loss curve is, in Iact, a set oI curves showing the
variation oI the lamination power loss with the magnetic induction B. Each curve corresponds to a particular operating
Irequency. The set oI curves used in this problem are shown in Figure 6.11-2, where the each data point has a
multiplying Iactor oI 7.0. The curves give the power loss density Ior diIIerent values oI the magnetic induction B and
Irequency I. SuIIicient data points are extracted Irom the set oI curves and input as a table. This data is then entered
as a Mentat table. The tabular data extracted Irom the curves is shown in Table 6.11-1. It is assumed that the device is
operating at a Irequency I 25 Hertz. The elements deIining the core laminations have to be deIined as an
electromagnetic contact body. This contact body identiIies the laminations. The data in the table is used to extract loss
coeIIicients, which are then used to Iind the elemental lamination loss depending on the value oI the operational
Irequency and the elemental value oI magnetic induction B.
Figure 6.11-2 Specified Lamination Loss Curves
1881
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Coupled magnetostatic-thermal analysis
The current Ilowing in the winding coils produce ohmic heat and the lamination losses also produce heat called
lamination loss. The magnetostatic analysis is Iirst carried out and the heat generated due to these sources is calculated.
Both these generated heats are next used in the coupled thermal analysis to calculate the temperature distribution.
Mesh Generation
In this problem, there is an axisymmetry about the axis oI the cylindrical inductor. This Iact is used to create a two
dimensional axisymmetric Iinite element model. An axial cross section oI the inductor is taken and only one halI oI
the section above the axis is modeled. The problem domain extends to inIinity and it is required to consider a large air
Table 6.11-1 Lamination Loss Data Extracted from the Loss Curves of Figure 6.11-2
Lamination Power Loss in Watts/Cubic Metters Multiplied by 7.0
Magnetic Induction B
(Tesla) Frequency = 10 Hz Frequency = 20 Hz Frequency = 30 Hz
0.0 0.00 0.00 0.00
0.1 288.06 1005.11 2141.04
0.2 1003.41 3688.82 8027.63
0.3 2137.15 8024.96 17610.86
0.4 3686.41 14004.78 30874.15
0.5 5649.68 21623.53 47808.42
0.6 8026.11 30878.35 68408.00
0.7 10815.19 41767.37 92669.13
0.8 14016.68 54289.44 120589.30
0.9 17630.52 68443.90 152166.91
1.0 21656.80 84230.50 187401.09
1.1 26095.76 101649.33 226291.54
1.2 30947.77 120700.84 268838.54
1.3 36213.35 141385.75 315042.88
1.4 41893.14 163705.13 364905.88
1.5 47987.97 187660.40 418429.39
1.6 54498.83 213253.33 475615.86
1.7 61426.89 240486.12 536468.38
1.8 68773.58 269361.44 600990.77
1.9 76540.58 299882.50 669187.66
2.0 84729.86 332053.12 741064.64
1882 Marc Users Guide: Part 3
CHAPTER 6.11
domain about the inductor. The problem has to be analyzed as a 2-D, axisymmetric problem.. The domain is modeled
by 4-noded quadrilateral elements. The element type 40 is used. The problem is Iirst modeled in 2-D in the Iirst quarter
oI the XY plane. Then symmetry is used to duplicate the model about the Y axis.
Element selection as Sets
The materials in the problem and contact bodies are a set oI elements and these sets are selected as a collection oI
elements in the objects air, core and winding. The material names Ior these objects are:
Air
Core
Winding
Three sets are created Ior the collection oI elements in the three materials. These sets are named as:
air
core
windings
The creation oI the above sets gives a convenient way oI deIining materials and contact bodies and also in viewing oI
post processing results.
The button sequence Ior speciIying the element sets reads:
SELECT
ELEMENTS
1 SELECT NEW ELEMENT SET:
ZOOM
zoom_box(1,0.215420,0.807504,0.323129,0.939641)
ELEMENTS STORE
STORE IN NEW ELEMENT SET:
core
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205
206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221
# | End of List
RETURN
1883
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
IDENTIFY SETS
REGENERATE
SELECT SETS
Core
OK
SELECT MODE: INTERSECT
RETURN
ELEMENTS SELECT
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
# | End of List
DEL ENTRIES
SELECT SET
Core
OK
ALL: SELEC
ELEMENTS STORE
STORE IN NEW ELEMENT SET:
windings
ALL: SELEC
ELEMENTS: CLR
CLEAR ALL ELEMENT SELECTIONS:
FILL
SELECT SET
Core
Windings
OK
ELEMENTS STORE
STORE IN NEW ELEMENT SET:
air
ALL: UNSEL
RETURN
1884 Marc Users Guide: Part 3
CHAPTER 6.11
The boundary oI the one-Iourth quadrant deIines the Iar Iield boundary and the nodes on this boundary have a Iixed
magnetic potential oI 0 volts. The nodes on this boundary are deIined as a set as Iollows:
Material Properties
There are three materials in this problem: air, core, and windings. DeIine three materials Ior them and enter the material
properties as detailed in the problem deIinition.
The button sequence Ior speciIying the material properties reads:
PLOT
NODES
SET NODES PLOTTING ON:
REGENERATE
RETURN
SELECT
STORE NODE PATH
ENTER NODE SET NAME:
Far_field
SELECT THE NODES ALONG THE PATH:
980
824
668
330
642
# | End of List
RETURN
RETURN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
SET ANALYSIS CLASS:
ANALYSIS CLASS:MAGNETOSTATIC-THERMAL
SET STANDARD MATERIAL:
NEW: STANDARD
NAME
CREATE NEW MATERIAL NAME:
air
1885
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
THERMAL
K
SET THERMAL CONDUCTIVITY VALUE:
0.1
SPECIFIC HEAT
SET SPECIFIC HEAT VALUE:
900.0
MASS DENSITY: THERMAL
SET MASS DENSITY VALUE:
1000.0
OK
MAGNETIC PERMEABILITY
MU
SET MAGNETIC PERMEABILITY VALUE:
1.25664e-6
OK
ELECTRIC CONDUCTIVITY
SIGMA
SET ELECTRIC CONDUCTIVITY VALUE:
0.0
OK
ELEMENTS ADD
SELECT EXISTING SET:
air
# End List
ID MATERIALS
SET STANDARD MATERIAL:
NEW: STANDARD
NAME
CREATE NEW MATERIAL NAME:
core
THERMAL
K
SET THERMAL CONDUCTIVITY VALUE:
10.0
1886 Marc Users Guide: Part 3
CHAPTER 6.11
SPECIFIC HEAT
SET SPECIFIC HEAT VALUE:
100.0
MASS DENSITY: THERMAL
SET MASS DENSITY VALUE:
7000.0
OK
MAGNETIC PERMEABILITY
MU
SET MAGNETIC PERMEABILITY VALUE:
0.01
OK
ELECTRIC CONDUCTIVITY
SIGMA
SET ELECTRIC CONDUCTIVITY VALUE:
1.0E+08
OK
ELEMENTS ADD
SELECT EXISTING SET:
core
# End List
SET STANDARD MATERIAL:
NEW: STANDARD
NAME
CREATE NEW MATERIAL NAME:
winding
THERMAL
K
SET THERMAL CONDUCTIVITY VALUE:
0.2
SPECIFIC HEAT
SET SPECIFIC HEAT VALUE:
600.0
MASS DENSITY: THERMAL
1887
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Contact
The lamination loss is required Ior the core material only. Hence, only the elements oI the set core are deIined as an
electromagnetic contact body. The loss multiplying Iactor and lamination loss table is selected here. A contact table is
also not required.
The button sequence Ior speciIying the material properties reads:
SET MASS DENSITY VALUE:
3000.0
OK
MAGNETIC PERMEABILITY
MU
SET MAGNETIC PERMEABILITY VALUE:
1.25664e-6
OK
ELECTRIC CONDUCTIVITY
SIGMA
SET ELECTRIC CONDUCTIVITY VALUE:
5.0E+05
OK
ELEMENTS ADD
SELECT EXISTING SET:
winding
# End List
RETURN
RETURN
RETURN
CONTACT
CONTACT BODIES
NEW
CREATE NEW CONTACT BODY:
NAME
ENTER CONTACT BODY NAME
Core
1888 Marc Users Guide: Part 3
CHAPTER 6.11
Modeling Tools
The electric current in the rectangular winding is speciIied through the winding Ieature. The rectangle has a coil oI
100 turns and its cross section is given by l 0.05 m and width 0.02 m. Each coil is a circular coil with an appropriate
radius oI 1.0E-5 meters. The winding path and its orientation are speciIied by a single node each Ior the 2-D
axisymmetric case.
The button sequence Ior speciIying the windings in modeling tools reads:
ELECTROMAGNETIC
DEFINE ELECTROMAGNETIC CONTACT BODY:
LOSS CURVE
ENTER THE LOSS CURVE MULTIPLYING FACTOR
7.0
OK
ELEMENTS: ADD
Cond1
# End List
ID CONTACTS
RETURN
RETURN
MODELING TOOLS
WINDINGS
NEW
CREATE NEW WINDINGS:
PROPERTIES
DEFINE WINDING PROPERTIES
CROSS-SECTION: RECTANGULAR
LENGTH
ENTER LENGTH VALUE
0.1
WIDTH
ENTER WIDTH VALUE
0.02
COIL: # COIL
ENTER NUMBER OF TURNS
1889
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Initial and Boundary Conditions
A Iixed magnetic scalar potential (Az) 0 volts is applied on all Iar Iield nodes. The coil current boundary condition
is used to apply uniIorm current in the winding. For this case, the current value and the winding identiIication path is
speciIied on the appropriate conductor elements.
The button sequence Ior speciIying the boundary conditions reads:
100
COIL:CROSSSECTION: CIRCULAR
RADIUS
ENTER RADIUS VALUE
1.0E-05
CONDUCTOR BODY
SELECT CONTACT BODY:
core
OK
SEGMENTS: ADD
ADD WINDING SEGMENTS:
SELECT OPTION: NODE PATH
CENTERLINE: NODES: ADD
SELECT NODE ID:
105
# End List
CROSS-SECTION ORIETNTATION: NODES: ADD
SELECT NODE ID:
113
# End List
FILL
RETURN
RETURN
BOUNDARY CONDITIONS
NEW
SELECT: MAGNETOSTATIC ANALYSIS
1890 Marc Users Guide: Part 3
CHAPTER 6.11
NAME
ENTER THE FIXED POTENTIAL BOUNDARY CONDITION NAME:
Fix_pot
FIXED POTENTIAL
POTENTIAL
ENTER POTENTIAL VALUE:
0
OK
NODES: ADD
ENTER NODE SET:
Far_field
# | End of List
NEW
NAME
ENTER THE COIL CURRENT BOUNDARY CONDITION NAME:
coilcur
COIL CURRENT
COIL CURRENT
ENTER CURRENT VALUE
2.0
WINDING PATH
SELECT WINDING PATH:
winding1
OK
ELEMENTS ADD
SELECT EXISTING ELEMENT SET:
windings
# | End of List
FILL
RETURN
RETURN
1891
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
The initial temperature at all nodes in the model is deIined as 30 Celsius. This is done using the Iollowing button
sequence
Table
The lamination loss data has to be deIined as a table in Mentat. The data in Table 6.11-1 can be entered as individual
values in Mentat or as an externally created input Iile. Usually, the number oI data points is large and entry oI
individual data points is tedious in Mentat. In this problem, the data is entered as an external input Iile,
loss_curve.dat. The external input is created in a predeIined Iormat which can be read by Mentat.
INITIAL CONDITIONS
NEW
SELECT: THERMAL
NAME
ENTER THE INITIAL TEMPERATURE CONDITION NAME:
Init_temp
TEMPERATURE
TEMPERATURE
ENTER TEMPERATURE VALUE:
30
OK
NODES: ADD
ENTER NODE SET:
ALL: EXISTS
# | End of List
RETURN
RETURN
CONTACT
CONTACT BODIES: THIS SHOWS THE CORE CONTACT BODY
TABLE
READ
READ THE EXTERNAL FILE:
Loss_curve.dat
OK
FIT
PLOT TABLE GRAPH AND FIT IN GIVEN RANGE
Core
RETURN
ELECTROMAGNETIC
TABLE
ASSOCIATE TABLE WITH THIS CONTACT BODY:
1892 Marc Users Guide: Part 3
CHAPTER 6.11
Loadcase and Job Parameters
There is only one loadcase named lamcase. This is a basic coupled magnetostatic thermal analysis in which heat is
generated to ohmic current Ilow in the winding coils and due to lamination loss in the core material. The magnetostatic
analysis is perIormed Iirst and the heat generated as mentioned above is used in the thermal analysis to compute the
nodal temperature distribution. The steps to describe the loadcase are shown in Figure 6.11-3.
The button sequence Ior speciIying the material properties reads:
Lamin_loss
OK
RETURN
RETURN
LOADCASE
NEW
CREATE NEW LOADCASE:
NAME
ENTER LOADCASE NAME:
lamcase
MAGNETOSTATIC-THERMAL
TRANSIENT
LOADS
SELECT LOADS:
Fix_pot
coilcur
LAMINATION CALCULATION
LAMINATION LOSS
SELECT THE CORE CONDUCTING BODY:
Core
TOTAL LOADCASE TIME
ENTER THE TOTAL LOADCASE TIME:
100
PARAMETERS
# STEPS
ENTER NUMBER OF TIME STEPS:
10
OK
OK
OK
RETURN
RETURN
1893
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Figure 6.11-3 Steps for showing Button Click to Define the Loadcase lamcase
1894 Marc Users Guide: Part 3
CHAPTER 6.11
The job speciIication includes only the one loadcase mentioned above. The element post processing selected is the X
and Y components oI the magnetic Iield intensity and magnetic Ilux density, temperature, lamination loss, and
generated heat.
The steps to describe the loadcase are shown in Figure 6.11-4.
Figure 6.11-4 Steps for showing Button Click to Define the JOB
1895
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
The button sequence Ior speciIying the material properties reads:
Save Model, Run Job, and View Results
The Iinite element model is Iirst saved as a mud Iile, lam_final.mud.
JOBS
NEW
SELECT ANALYSIS: MAGNETOSTATIC-THERMAL
PROPERTIES
ADD LOAD CASE:
lamcase
SELECT JOB: PLANAR
JOB RESULTS
SELECT THE FOLLOWING QUANTITIES:
1st Component of Magnetic Induction
2nd Component of Magnetic Induction
1st Component of Magnetic Field Intensity
2nd Component of Magnetic Field Intensity
Lamination Loss
Generated Heat
OK
OK
RETURN
FILE
SAVE AS
lam_final.mud
OK
RETURN
JOBS
RUN
SUBMIT 1
MONITOR
OK
RETURN
RESULTS
OPEN DEFAULT
*post_open_default
1896 Marc Users Guide: Part 3
CHAPTER 6.11
It is desired to view the contour plots Ior the Iollowing:
1. External Electric Current
2. Magnetic Induction
3. Generated Heat
4. Lamination Loss
The plots can reveal visually detectable errors in the solution.
Since a transient thermal analysis is applied with 10 time steps, there are 10 increments. The results are shown below
Ior the 10th increment. It should be noted that the magnetostatic run Ior each increment runs with the same loads,
boundary conditions and material properties and the solutions are the same Ior each increment. The thermal analysis
also runs with the same material properties, but the generated heat and lamination loss adds up in each increment
resulting in progressively increasing temperature.
To see the electric potential contour plots:
Observe the external electric current. Then use the PLOT settings to get a better picture oI the contour plot:
The contour plot Ior increment 10 is shown in Figure 6.11-5.
SCALAR
External Electric Current
OK
CONTOUR BANDS
PLOT
A NODE PLOTTING IS SET OFF:
NODES
ELEMENTS SETTING
OUTLINE
REGENRATE
RETURN
1897
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Figure 6.11-5 Electric External Current Distribution
Now repeat the above steps Ior magnetic induction, temperature, generated heat, and lamination loss contour bands
Ior the tenth increment. The plots are shown in Figures 6.11-6, 6.11-7, 6.11-8, and 6.11-9.
Figure 6.11-6 Magnetic Induction Distribution
7.845e-005
1.045e-001
2.090e-001
3.134e-001
4.179e-001
5.223e-001
6.268e-001
7.313e-001
8.357e-001
9.402e-001
1.045e+000
lamcase
Core Only (expanded)
Magnetic Induction
Inc: 10
Time: 1.000e+002
Z
Y
X
1898 Marc Users Guide: Part 3
CHAPTER 6.11
Figure 6.11-7 Temperature Distribution
Figure 6.11-8 Generated Heat Distribution
1899
CHAPTER 6.11
Lamination Loss in Magnetostatic-Thermal coupling
Figure 6.11-9 Lamination Loss Distribution
Results and Discussion
The total lamination loss is printed in the OUT Iile. It contains the total lamination loss ,and its break up into individual
loss types. The relevant portion oI the OUT Iile is shown below:
***************************************************
** The Lamination Losses at operating frequency *** 25.0000
Total Losses for Lamination Body Number: 1
Total Lamination Losses: 0.42444E+03
Total Hysteresis Losses: 0.57443E+01
Total Eddy Current Losses: 0.40313E+03
Total Extra (STRAY) Losses: 0.15563E+02
***************************************************
The above losses are same in each increment.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
lamcase.proc Mentat procedure Iile
loss_curve.dat Loss input data
lam_init.mud Mentat model Iile
1900 Marc Users Guide: Part 3
CHAPTER 6.11
Chapter 6.12: Magnetic Levitation of a Ferromagnetic Sphere
6.12
Magnetic Levitation of a
Ferromagnetic Sphere

Summary 1902

Magnetic Levitation of a Ferromagnetic Sphere 1903

Results and Discussion 1912

Input Files 1913

New Mentat Menus 1914

1902 Marc Users Guide: Part 3
CHAPTER 6.12
Summary
Title Magnetic levitation oI a Ierromagnetic sphere
Problem Ieatures Coupled magnetostatic-structural analysis
Geometry
Material properties Typical properties oI the air, and Ior the iron sphere
Analysis type Magnetostatic-Structural
Boundary conditions The outer boundary oI the air has a Iixed magnetic potential oI 0. A current oI 8796.46A
is applied to one node to represent the coil. One node oI the iron sphere is Iixed.
Element type Planar axisymmetric
FE results
x
F
Coil
h
R
2
R
1
R
c
I

R = 0.035m
R = 0.05m
R = 0.07m
h = 0.03m
= 500
= 1.2566e-6H/m
I = 20000A
1
2
c
r
0

r
500
Method Total Force
virtual work 364.6N
maxwell stress 349.7N
analytical 372.9N
1903
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
In this example, a coupled magnetostatic-structural analysis is demonstrated. A coil is placed around the top part oI a
hollow iron sphere. The current in the coil generates a magnetic Iield in the iron sphere which pulls this sphere towards
the coil. By tuning the current in the coil, this pulling Iorce can be made equal to the gravitational Iorce acting on the
iron sphere, and thereIore creating levitation. The problem has an analytical solution which is taken Irom
Reference 6-1.
The layout oI some oI the Mentat menus have been redesigned in the 2010 version. This is done to better guide the
user in which data must be entered. For the materials, loadcase, and job selection, the user must Iirst speciIy which
analysis class will be perIormed. DiIIerent analysis classes are, Ior example, Structural, Thermal, Thermal/Structural,
Piezoelectric, Magnetostatic/Structural, or Fluid/Structural. Once this is set, Mentat shows the appropriate menus Ior
material data, loadcase options, and job options. More inIormation about this is available at the end oI this chapter.
The procedure Iile to demonstrate this example is called hollowsphere.proc under
mentat2009/examples/marc_ug/s6/c6.12. The procedure requires an initial mud Iile, sh_mesh.mud.
Magnetic Levitation of a Ferromagnetic Sphere
Figure 6.12-1 shows the representation oI the iron sphere which is analyzed. The diIIerent dimensions are also shown
in this Iigure. The inset shows a 3-D representation oI the model, while here an axisymmetric analysis is perIormed.
Note that, in Marc, the symmetry axis is the x-axis.
Figure 6.12-1 Model of the hollow sphere
Mesh Generation
The mesh is already deIined in the Iile hs_mesh.mud. The model starts by opening this Iile. The current will be
applied to an extra node which is not connected to the mesh. This node is added Iirst.
MESH GENERATION
NODE: ADD
x
F
Coil
h
R
2
R
1
R
c
I

R = 0.035m
R = 0.05m
R = 0.07m
h = 0.03m
= 500
= 1.2566e-6H/m
I = 20000A
1
2
c
r
0
1904 Marc Users Guide: Part 3
CHAPTER 6.12
Material Properties
The material properties oI the two materials are given below:
1. Air
a. Structural
Young`s modulus: E 20000Pa
Poisson`s ratio: v 0.3
b. Magnetostatic
Magnetic permeability H/m
2. Iron sphere
a. Structural
Young`s modulus: E Pa
Poisson`s ratio: v 0.3
Magnetostatic
Magnetic permeability H/m
Note that Ior easy modeling a high stiIIness Ior air is taken. It is also possible to let the air elements not be active in
the structural pass. To do this they must be deIined as magnetostatic elements.
0.07 0.03 0
MAIN
MATERIAL PROPERTIES
MATERIAL PROPERTIES
ANALYSIS CLASS
MAGNETOSTATIC/STRUCTURAL
NEW
STANDARD
NAME
air
STRUCTURAL
YOUNGS MODULUS
20000
POISSONS RATIO
0.3
OK
1.2566
6
10
2.0 10
11

r

0
500 1.2566
6
10 8796.46
1905
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
Boundary Conditions
One node inside the iron sphere is Iixed, so that the reaction Iorce calculated here balances the magnetic Iorce. When
this Iorce is higher than the gravitational Iorce (not calculated here), the magnetic Iield levitates the iron sphere. The
magnetostatic potential is set to 0 at the outer boundary and a point current is prescribed to the node which represents
the coil. This point current is Am.
MAGNETIC PERMEABILITY
MU
1.2566e-6
OK
SELECT
SELECT SET
sphere
OK
RETURN
ELEMENTS ADD
ALL: SELEC.
NEW
STANDARD
NAME
iron
STRUCTURAL
YOUNGS MODULUS
2e11
POISSONS RATIO
0.3
OK
MAGNETIC PERMEABILITY
MU
0.0006283
OK
ELEMENTS ADD
ALL: UNSEL.
MAIN
I l 20000 2 t 0.07 8796.46
1906 Marc Users Guide: Part 3
CHAPTER 6.12
Modeling Tools and Contact
Two methods are available to calculate the magnetic (Lorentz) Iorce on a body. The Maxwell Stress Tensor (MST)
method calculates this Iorce using a surIace surrounding the body. In Marc, this surIace is represented as a contact
interIace, so the elements on both sides oI this interIace must be disconnected. On this surIace, local Iorces are
calculated, where the summation oI these Iorces is the Lorentz Iorce. This Iorce is printed in the .out Iile. The second
BOUNDARY CONDITIONS
STRUCTURAL
NAME
fix
FIXED DISPLACEMENT
DISPLACEMENT X
DISPLACEMENT Y
OK
NODES ADD
159 #
RETURN
NEW
NAME
fix_A
MAGNETOSTATIC
FIXED POTENTIAL
POTENTIAL
OK
NODES ADD
697 714 731 680 663 646 629 612 595 578 561 544 527 510 493 476 459 442 425 #
NEW
NAME
load
POINT CURRENT
CURRENT
8796.46
OK
NODES ADD
446
MAIN
1907
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
method is the Virtual Work Method. This method is energy based, where a group oI elements onto which the Lorentz
Iorce is calculated gets a delta displacement in the x-, y-, and Ior 3-D z-direction. From energy diIIerences, the Lorentz
Iorce is calculated. For the method, the group oI elements onto which the Lorentz Iorce will be calculated is also
deIined as a contact body, but, in this case, the mesh does not need to be disconnected. This calculated Lorentz Iorce
is prescribed in the structural pass as a volumetric load, and is also printed in the .out Iile.
Best results are obtained when a layer oI air surrounds the part onto which the Lorentz Iorce will be calculated. For
the VWM method, this layer can be speciIied, but, Ior the MST method, the user has to create this. To Iacilitate this,
Mentat has a tool called MATCHING BOUNDARIES. This tool splits a mesh into two disconnected meshes. To Iind the
layer oI air surrounding the sphere, we start by selecting the elements oI the sphere and the elements inside the sphere.
Then, select all nodes belonging to these elements. Next, select all elements which contain one or more oI the just
selected nodes. Now, one layer oI elements is added. To add more layers repeat this procedure. With these selected
elements, the mesh is split using the MATCHING BOUNDARIES, and the two parts are added as contact bodies in the
CONTACT section. These contact bodies are 'glued together using GLUE Ior the interIace condition between the two
bodies in the CONTACT TABLE.
MODELING TOOLS
MATCHING BOUNDARIES
NEW
2-D
SELECT
CLEAR SELECT
SELECT SET
sphere_and_inside
OK
SELECT BY
NODES BY: ELEMENTS
ALL: SELEC.
ELEMENTS BY: NODES
ALL:SELEC.
NODES BY: ELEMENTS
ALL: SELEC.
ELEMENTS BY: NODES
ALL: SELEC.
RETURN
RETURN
SPLIT MESH
ALL:SELEC
MAIN
1908 Marc Users Guide: Part 3
CHAPTER 6.12
Links
The current is applied to a node to represent the coil, but without doing anything special, this current is not transIerred
to the mesh since this node is not connected to the mesh. To connect this node with the mesh, the INSERT option is
used. With this option, Marc creates tyings between the host entities (Iinite element mesh) and the embedded entities
(the node carrying the current).
CONTACT
CONTACT BODIES
NAME
sphere_air
ELECTROMAGNETIC
OK
ELEMENTS ADD
ALL: UNSEL.
NEW
NAME
outer_air
ELECTROMAGNETIC
OK
ELEMENTS ADD
ALL: SELEC.
RETURN
CONTACT TABLES
NEW
PROPERTIES
FIRST 1 SECOND 2
CONTACT TYPE
GLUE
OK
OK
MAIN
LINKS
INSERTS
HOST ENTITIES CONTACT BODIES
CONTACT BODIES
air_outer
OK
1909
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
Loadcases and Job Parameters
Two loadcases will be created: one demonstrating the Maxwell Stress Tensor method and the other demonstrating the
Virtual Work Method. Note that Ior the virtual work method, the user can speciIy an extra number oI layers
surrounding the body onto which the Lorentz Iorce will be calculated. For this example, it is zero since it is already
included in the body. In general, care should be taken that the extra layers which are created by Marc do not contain
any applied currents or diIIerent materials. This can be checked during post processing where a new SET called
VWM_body is available. This set contains all the elements which will get the delta displacement.
The steps which are described here are shown in Figure 6.12-2.
EMBEDDED ENTITIES NODES
ADD
732 #
MAIN
ANALYSIS
LOADCASES
ANALYSIS CLASS
MAGNETOSTATIC/STRUCTURAL
NEW
STEADY STATE/STATIC
NAME
mst
PROPERTIES
CONTACT
CONTACT TABLE
ctable1
OK
LORENTZ FORCE CALCULATION
LORENTZ FORCE
METHOD MAXWELL STRESS
LORENTZ FORCE BODIES: sphere
OK
STEPPING PROCEDURE # STEPS
1
OK
NEW
STEADY STATE/STATIC
NAME
1910 Marc Users Guide: Part 3
CHAPTER 6.12
vwm
PROPERTIES
CONTACT
CONTACT TABLE
ctable1
OK
OK
LORENTZ FORCE CALCULATION
LORENTZ FORCE
METHOD VIRTUAL WORK
# ELEMENT LAYERS SURROUNDING BODIES
0
LORENTZ FORCE BODIES sphere
OK
STEPPING PROCEDURE # STEPS
1
OK
MAIN
JOBS
NEW
MAGNETOSTATIC/STRUCTURAL
ELEMENT TYPES
ANALYSIS DIMENSION
AXISYMMETRIC
STRUCTURAL SOLID
10
OK
ALL: EXIST.
RETURN
PROPERTIES
AVAILABLE mst
JOB RESULTS
1st Comp of Magnetic Induction
2nd Comp of Magnetic Induction
1st Comp of Magnetic Field Intensity
2nd Comp of Magnetic Field Intensity
OK
OK
1911
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
Figure 6.12-2 Mentat Menus for Loadcase Creation
1912 Marc Users Guide: Part 3
CHAPTER 6.12
Save Model, Run Job, and View Results
AIter saving the model, the job is submitted and the resulting post Iile is opened.
Results and Discussion
Figure 6.12-3 shows a contour plot oI the x-component oI the magnetic induction. Here you can see that the
magnetization mostly stays inside the iron sphere.Note that in this plot, the node is also indicated onto which the
current is applied. The results oI the total Iorce Ior the Maxwell Stress Tensor analysis and Ior the analysis using the
Virtual Work Method are compared to the analytical solution in the Iollowing table. The analytical result is taken Irom
Reference 6-1.
FILES
SAVE AS
hs_model
OK
RUN
SUBMIT
MAIN
RESULTS
OPEN DEFAULT
NEXT
SCALAR
1st Comp of Magnetic Induction
OK
CONTOUR BANDS
Method Total Force
virtual work 367.6N
maxwell stress 352.6N
analytical 372.9N
1913
CHAPTER 6.12
Magnetic Levitation of a Ferromagnetic Sphere
Figure 6.12-3 Contour Plot of the Magnetic Induction
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
References
6-1. Z. Ren, Comparison of Different Force Calculation Methods in 3D Finite Element Modelling, IEEE
Transactions on Magnetics, Vol. 30, No. 5, September 1994
File Description
hs_run.proc Mentat procedure Iile
hs_mesh.mud Mentat input model Iile
Node to which
current is applied
1914 Marc Users Guide: Part 3
CHAPTER 6.12
New Mentat Menus
For Marc/Mentat2010 some oI the menus have been redesigned. The purpose is to better guide the user through the
diIIerent processing steps by showing only relevant inIormation. To obtain this a new selection button ANALYSIS
CLASS is introduced (Figure 6.12-4). This button is available in the sections MATERIAL PROPERTIES,
LOADCASES, and JOBS. BeIore a user can enter data in these sections, the ANALYSIS CLASS must be selected.
Once this is done, Mentat shows the menus appropriate Ior this analysis class. So Ior the
MAGNETOSTATIC/STRUCTURAL analysis class, Mentat shows in the MATERIAL PROPERTIES section buttons Ior
GENERAL, STRUCTURAL, and MAGNETIC PERMEABILITY. So only material data which is required Ior a
magnetostatic-structural analysis can be entered in these menus. Note that when the ANALYSIS CLASS is changed
data already entered is stored, so there is no loss oI inIormation. Once the ANALYSIS CLASS is set in one oI the
sections, it will be the deIault Ior the other sections, but can be overruled at any time.
Figure 6.12-4 Mentat Menus Showing the new ANALYSIS CLASS Button
Section 7: Mentat Features and Enhancements
Section 7: Mentat Features
and Enhancements
Marc Users Guide -1916
Chapter 7.1: Past Enhancements in Marc and Mentat
7.1
Past Enhancements in Marc
and Mentat

Chapter Overview 1918

Preprocessing Enhancements 1918

Postprocessing Enhancements 1944

Input Files 1948

Animation 1949
1918 Marc Users Guide: Part 3
CHAPTER 7.1
Chapter Overview
This chapter demonstrates various enhancements to Mentat. The most important improvements made in preprocessing
are:
Inclusion oI the Patran Mesh-on-Mesh surIace mesher
Inclusion oI the Patran tetrahedral mesher
A new and more consistent attach concept
New combined mesh generation commands that move, duplicate or expand a mixed list oI mesh entities
(nodes, elements), geometric entities (points, curves, surIaces, solids) and links (nodal ties, servo links,
springs) simultaneously
Improved handling oI links
New selection methods Ior selecting items within a certain distance oI a point, curve or surIace and Ior box
selection in the user coordinate system (that allows Ior selection in cylindrical and spherical coordinate
systems)
Multi-dimensional tables
The most important postprocessing enhancements are:
MPEG and AVI animations
Automatic execution oI a procedure Iile when a post Iile increment is read
Preprocessing Enhancements
This section demonstrates some oI the Mentat preprocessing enhancements. The new attach concept is discussed and
the beneIits in combination with initial conditions and boundary conditions applied to geometric entities (points,
curves, and surIaces) are stressed. The combined move, duplicate, symmetry, and expand commands that operate on
a mixed list oI mesh entities, geometric entities and links are introduced and the improved handling oI links is
elaborated. Furthermore, the Patran tetrahedral mesher that has been incorporated in Mentat and the new selection
methods are described in this chapter.
These new capabilities are illustrated by means oI a tire modeling example. Two Iinite element models oI a tire are
created, one consisting oI 20-node hexahedral elements and the other consisting oI 10-node tetrahedral elements. The
tire is loaded by an internal pressure, while the rim oI the wheel is Iixed.
In addition, the use oI the new multi-dimensional tables is shown in a separate example.
New Attach Concept
The outline oI the cross-section oI a tire is imported Irom an IGES Iile. The Iile contains only the right halI oI the
cross-section (Figure 7.1-1). AIter the Iile has been imported, the end points oI the curves are merged with a SWEEP
POINTS operation.
1919
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-1 Right-half of the Cross-section of the Tire
The region is meshed using the planar advancing Iront automatic mesher. The average curve division length is set to
6 and an even number oI curve divisions is Iorced on detected loops. The later is done to insure that a mesh with all
quadrilateral elements will be Iormed. The resulting mesh is displayed in Figure 7.1-2.
MESH GENERATION
FILES
IMPORT
IGES
tire.igs
RETURN (twice)
SWEEP
POINTS
ALL: EXIST.
RETURN
AUTOMESH
CURVE DIVISIONS
FIXED AVG LENGTH (on)
AVG LENGTH
6
1920 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-2 Finite Element Mesh of the Cross-section of a Tire
All automatic mesh generators attach the mesh to the geometry, according to the new attach concept. With that scheme:
a node can be attached to a point; and
an element edge can be attached to a curve; and
an element Iace can be attached to a surIace.
Nodes which are attached to a point always have the same position as the point. Nodes oI edges which are attached to
a curve always lie on that curve and nodes oI Iaces which are attached to a surIace always lie on that surIace. Note that
this implies that the common node oI two edges which are attached to diIIerent curves must lie on the intersection oI
RESTRICTION FORCE EVEN DIV
APPLY RESTRICTION TO DETECTED LOOPS
APPLY CURVE DIVISIONS
ALL: EXIST.
RETURN
2D PLANAR MESHING
QUAD MESH!
ALL: EXIST.
RETURN (thrice)
Note: The attaching two edges (1:1 and 1:2) that share a common node (3) to diIIerent curves causes the common node to be
placed automatically on the intersection oI the curves.
Node attached to point
Element edges attached to curves
1921
CHAPTER 7.1
Past Enhancements in Marc and Mentat
the curves (Figure 7.1-3). Similarly, the common nodes oI two Iaces which are attached to diIIerent surIaces must lie
on the intersection oI the two surIaces. The automatic meshers and the mesh generation commands that modiIy either
the mesh or the geometry guarantee that this is always the case. For example, iI one oI the curves is moved or otherwise
changed, the common node is repositioned automatically to the new point oI intersection. II that point cannot be Iound,
the operation is not permitted and an error message is issued.
Figure 7.1-3 Common Node with two Edges
Nodes, element edges, and element Iaces can also be attached or detached manually using the commands in the MESH
GENERATIONATTACH menu (Figure 7.1-4). Attached nodes are displayed as small circles and attached edges are
by deIault drawn in orange (Figure 7.1-2). Attached Iaces are plotted in a dark blue color. Recall that unattached nodes
are represented as squares, unattached edges in white, and unattached Iaces in a light blue. These colors can be changed
using the VISUALIZATIONCOLORS menu. The actual curve to which an edge is attached can be visualized by
switching on the edge labels and activating the attach inIormation in the PLOT menu:
PLOT
ELEMENTS
SETTINGS
EDGES
LABELS
ATTACH INFO
1922 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-4 The ATTACH Menu
II the attach inIormation is enabled, the label includes the curve number to which the edge is attached, separated Irom
the edge number by an -sign (Figure 7.1-2). Similar options are available Ior displaying the point and surIace to
which nodes and element Iaces are attached.
In the present example, the mesher automatically attaches the element edges on the boundary oI the mesh to the
appropriate curves. It also attaches the nodes that lie on the end points oI the curves to these points.
Expansion to 3-D
The three-dimensional model is obtained by expansion oI the axisymmetric model to 3-D. This operation is developed
especially Ior cases in which the axisymmetric analysis is perIormed Iirst, Iollowed by a Iull three-dimensional
analysis. It requires that the mesh consists oI axisymmetric elements.
JOBS
ELEMENT TYPES
MECHANICAL
AXISYMMETRIC SOLID
10
OK
ALL: EXIST.
RETURN (thrice)
1923
CHAPTER 7.1
Past Enhancements in Marc and Mentat
The expansion operation expands the two-dimensional axisymmetric elements into three-dimensional solid elements,
according to the speciIied angles and repetitions. In this example, the expansion is perIormed in 18 steps oI 20. In
addition, the command revolves points to which nodes are attached into circles and revolves curves into surIaces oI
revolution.
Attach relations between the axisymmetric mesh and the axisymmetric geometry are automatically transIerred to the
three-dimensional solid mesh. II a node oI the axisymmetric mesh is attached to a point, the element edges that arise
Irom expansion oI the node are attached to the circle that results Irom revolving the point. Similarly, iI an edge oI the
axisymmetric mesh is attached to a curve, the Iaces that arise Irom expansion oI the edge are attached to the surIace
that results Irom revolving the curve. Since all edges on the boundary oI the axisymmetric mesh are attached to the
curves, all Iaces on the surIace oI the three-dimensional mesh will be attached to the surIaces oI revolution
(Figure 7.1-5).
Figure 7.1-5 Three-dimensional Mesh of a Tire
MESH GENERATION
EXPAND
AXISYMMETRIC MODEL TO 3D
1 ANGLE
20
1 REPETITIONS
18
EXPAND MODEL
RETURN (thrice)
Note: This three-dimensional model is obtained by expansion oI the axisymmetric model. All Iaces on the surIace oI the mesh
are attached to the surIaces oI revolution.
1924 Marc Users Guide: Part 3
CHAPTER 7.1
Boundary Conditions on Geometric Entities
Any initial or boundary conditions applied to geometric entities are inherited by mesh entities attached to the geometry.
For example, a point load applied to a point is inherited by the nodes attached to the point, an edge load applied to a
curve is inherited by the edges attached to the curve, and a Iace load applied to a surIace is inherited by the Iaces
attached to the surIace. Moreover, iI a boundary condition that is normally applied to a node (such as a Iixed
displacement boundary condition) is applied to a curve or a surIace instead, then the nodes oI the edges or Iaces
attached to the curve or surIace inherit the boundary condition. The advantage is that loads can be speciIied
independent oI the Iinite element mesh.
MSC.Mentat 2003, by deIault, draws boundary and initial conditions that are applied to the geometry on the geometric
entities. Previous MSC.Mentat versions always draw the boundary and initial conditions on the mesh entities that
inherit Irom the geometric entities. The old pre-2003 behavior can be restored using DRAW BOUNDARY CONDS ON
MESH Irom the BOUNDARY CONDITIONS menu and DRAW INITIAL CONDS ON MESH Irom the INITIAL
CONDITIONS.
The tire is inIlated by a pressure oI 2 MPa. The pressure is applied to the interior surIace (surIace 3) oI the tire.
Furthermore, the displacements oI the rim are suppressed (Figure 7.1-6), by applying a Iixed displacement boundary
condition to curve 2.
BOUNDARY CONDITIONS
NEW
NAME
fixed
MECHANICAL
FIXED DISPLACEMENT
X DISPLACEMENT
Y DISPLACEMENT
Z DISPLACEMENT
OK
CURVES ADD
2
END LIST (#)
NEW
NAME
pressure
FACE LOAD
PRESSURE
2
OK
1925
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-6 Boundary Conditions applied to the Geometry (left) and inherited by the
Attached Mesh Entities (Faces and Nodes, right)
Combined Mesh Generation Commands
The Iull model is obtained by duplication oI the existing model using a symmetry operation with respect to the
yz-plane and by changing the linear elements to 20-node quadratic solid elements. The existing symmetry commands
either duplicate the elements, curves or the surIaces, but not both. This means that even though the mesh and the
geometry can be duplicated, the attach relations that exist between the original mesh and the original geometry, are
lost Ior the copies. Any boundary conditions applied to the duplicates oI the curves and surIaces are not transIerred to
the copy oI the mesh.
The new COMBINED SYMMETRY operation overcomes this problem. It operates on a mixed list oI items (nodes,
elements, points, curves, surIaces, etc.). These items are duplicated in the same way as the normal symmetry
commands and, in addition, any attach relations that exist between original mesh and geometry are duplicated Ior the
copies oI the mesh and the geometry.
The kind oI items that are accepted by the COMBINED SYMMETRY command are controlled by the toggles in the
COMBINED section oI the SYMMETRY menu (Figure 7.1-7). Only active types are accepted and only items oI these
types are graphically pickable using the usual single pick, box pick, and polygon pick methods iI the COMBINED
SYMMETRY command is executed. This allows to simultaneously duplicate elements and surIaces, but no curves, Ior
example. Wildcards like ALL: EXIST. and ALL: SELECT. can also be used with this command to indicate all existing
or all selected items oI the active types.
SURFACES ADD
3
END LIST (#)
RETURN
1926 Marc Users Guide: Part 3
CHAPTER 7.1
Similar operations exist in the DUPLICATE, EXPAND, and MOVE menus.
Figure 7.1-7 The SYMMETRY Menu with the COMBINED Section and Finite Element
Mesh of the Full Model
MESH GENERATION
SELECT
SELECT BY
FACES BY SRFS
4
END LIST (#)
RETURN (twice)
ATTACH
DETACH FACES
ALL: SELECT.
RETURN
SRFS REM
4
END LIST (#)
SYMMETRY
COMBINED SYMMETRY
ALL: EXIST.
RETURN
SWEEP
ALL
RETURN
1927
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Change Class
The linear 8-node solid elements are converted into quadratic 20-node solid elements using the new CHANGE CLASS
TO QUADRATIC ELEMENTS operation (Figure 7.1-8). This is a special conversion that converts linear elements to
quadratic elements, regardless oI their class. The CHANGE CLASS TO LINEAR ELEMENTS command does the
opposite operation.
All commands in the CHANGE CLASS menu (including the existing conversions Irom one class to another) that create
new nodes (such as the conversion Irom linear to quadratic elements) now generate unique nodes on coinciding edges
and Iaces. This implies that a sweep operation to remove any duplicate nodes is no longer required aIter such a
conversion. Moreover, new midside nodes are positioned on the curve or surIace to which the edge or Iace is attached.
The mid-edge nodes lie exactly halIway the edge.
Figure 7.1-8 CHANGE CLASS Menu with TO QUADRATIC ELEMENTS Operation and
Final Finite Element Mesh
Improved Links Handling
Handling oI links has been improved. Links (nodal ties, servo links, and springs) are graphically pickable now, using
the usual single pick, box pick, and polygon pick methods. Links can be duplicated and moved just like elements, and
commands have been added to the LINKSNODAL TIES, LINKS SERVO LINKS and
LINKSSPRINGS/DASHPOTS Ior removing either all or a list oI nodal ties/servo links/springs.
CHANGE CLASS
TO QUADRATIC ELEMENTS
ALL: EXIST.
RETURN (twice)
1928 Marc Users Guide: Part 3
CHAPTER 7.1
These new Ieatures are illustrated by replacing the boundary condition on the rim with a set oI rigid links (tying type
80). First oI all, the boundary condition is removed and a single nodal tie is created between a node on the rim and two
new retained nodes on the axis oI the tire.
Next, the nodal tie is duplicated 35 times by rotation around the axis oI the tire about an angle oI 10 per step. The
resulting ties are duplicated by symmetry with respect to the yz-plane and a Iinal sweep operation merges the duplicate
nodes on the rim and on the axis oI the tire. The resulting model is depicted in Figure 7.1-9.
BOUNDARY CONDITIONS
EDIT
fixed
MECHANICAL
REMOVE CURVES
ALL: EXIST.
RETURN (twice)
MESH GENERATION
ADD NODES
60.8 0 0
0 0 0
RETURN
LINKS
NODAL TIES
NEW
TYPE
80
TIED NODE
74
RETAINED NODE 1
19315
RETAINED NODE 2
19316
RETURN (twice)
MESH GENERATION
DUPLICATE
ROTATION ANGLES
10 0 0
REPETITIONS
1929
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-9 Full Model with Rigid Links (Tying Type 80)
Patran Tetrahedral Mesher
The mesh is deleted, and a new mesh is created using the new (Patran) tetrahedral mesher. Tetrahedral meshing is, as
always, done in two steps: Iirst, a triangular mesh is created on the surIaces enclosing the volume to be meshed; next
a tetrahedral mesh is created using the nodes oI the surIace mesh.
35
TIES
link1
END LIST (#)
RETURN
SYMMETRY
TIES
ALL: EXIST.
RETURN
SWEEP
TOLERANCE
0.1
NODES
ALL: EXIST.
RETURN (twice)
1930 Marc Users Guide: Part 3
CHAPTER 7.1
Note that aIter the Iirst step, all triangular elements have their Iace attached to one oI the surIaces. AIter the second
step, the resulting element Iaces on the surIace oI the mesh are attached to one oI the surIaces. As a result, the pressure
boundary condition on surIace 3 is automatically inherited by the attached element Iaces.
LINKS
NODAL TIES
REM TIES
ALL: EXIST.
RETURN (twice)
MESH GENERATION
CLEAR MESH
AUTO MESH
CHECK/REPAIR GEOMETRY
TOLERANCE
0.001
CHECK SURFACES
ALL: EXIST.
CLEAN SURFACE LOOPS
ALL: EXIST.
CHECK SURFACES
ALL: EXIST.
RETURN
CURVE DIVISIONS
AVG LENGTH
20
APPLY CURVE DIVISIONS
ALL: EXIST.
RETURN
SURFACE MESHING
SURFACE TRI MESH! (ADV FRONT)
ALL: EXIST.
RETURN
SOLID MESHING
OUTLINE EDGE LENGTH
TOLERANCE
1
SWEEP OUTLINE NODES
1931
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Once again, the linear elements are changed to quadratic elements.
Figure 7.1-10 The AUTOMESH SOLIDS Menu and Finite Element Mesh
New Select Methods
Two new selection methods are shown in this section, where tied nodes Ior RBE2`s are selected. The Iirst one is the
USER BOX method, which selects all entries that Iall entirely within a box, speciIied in the current user coordinate
system. In this case, a cylindrical coordinate system is used to select nodes which have a radial coordinate oI 185. The
second method shown is the CURVE DIST method, which selects all entries within a given distance Irom a curve. The
new methods POINT DIST and SURFACE DIST are similar but are not discussed here.
OUTLINE EDGE LENGTH
SOLID TET MESH!
ALL: EXIST.
RETURN (twice)
CHANGE CLASS
TO QUADRATIC ELEMENTS
ALL: EXIST.
RETURN
Note: Figure 7.1-10 was generated by the Patran tetrahedral mesher and the change class conversion to 10-node
tetrahedral elements.
1932 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-11 Select Method Menus
COORDINATE SYSTEM SET
GRID
CYLINDRICAL
U DOMAIN
0 200
U SPACING
5
ROTATE
0 90 0
RETURN
ADD NODES
0 0 0
RETURN
LINKS
RBE2S
NEW
RETAINED NODE
36810
SELECT
METHOD
USER BOX
RETURN
SELECT NODES
185-0.01 185+0.01
0 360
-100 100
CLEAR SELECT
1933
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-12 RBE2S Menu and Final Tetrahedral Model with Nastran RBE2
METHOD
CURVE DIST.
SELECT DISTANCE
0.5
RETURN
SELECT NODES
9
23
RETURN
TIED NODES ADD
ALL: SELECT.
DOF 1
DOF 2
DOF 3
DOF 4
DOF 5
DOF 6
RETURN (twice)
1934 Marc Users Guide: Part 3
CHAPTER 7.1
New Domain Decomposition Methods
Domain Decomposition Ior DDM has been enhanced by three new methods:
Metis Element Based
Metis Node Based
Metis Best (combined Metis Element Based and Metis Node Based)
Here, the Metis Best method is used to decompose the tire model:
Figure 7.1-13 Metis Best Domain Decomposition
JOBS
DOMAIN DECOMPOSITION
GENERATE!
8
ID DOMAINS
PLOT
NODES
POINTS
RBE2S
SHORTCUTS
GRID (off)
1935
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Multi-Dimensional Tables
The tables in Mentat have been enhanced to allow multiple independent variables. The number oI independent
variables ranges Irom 1 to 4, each variable having a diIIerent table type (physical meaning).
This section shows various ways to create tables starting with the simple one-dimensional table. The button sequences
below start Irom the TABLES menu, which can be accessed in many places; Ior example, via MATERIAL
PROPERTIES.
NEW
1 INDEPENDENT VARIABLE
FILL
NAME
E_t
TYPE
temperature
ADD
-100 1
1000 .1
FIT
FUNCTION VALUE F MIN
0
FILLED (on)
MORE
INDEPENDENT VARIABLE V1 LABEL
Temperature
FUNCTION VALUE F LABEL
Youngs Modulus
PREVIOUS
1936 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-14 Creating a One-dimensional Table
In this example, the yield stress is a Iunction oI the gasket closure distance (Iirst independent variable) and the
temperature (second independent variable). There are seven gasket closure values and two temperatures, hence, the
number oI yield stress values deIined is 7 x 2 14. The data in the table Iormat appears as:
This would be manually entered as Iollows.
For each independent variable, the table type is set. In addition, labels are deIined to be displayed along the axes oI
the table.
Temperature
Gasket Closure Distance
0 0.027 0.054 0.081 0.108 .135 .175
-100 0 2.08 8.32 18.72 33.28 52 56
1000 0 2.08 832 1.872 3.328 5.2 5.6
NEW
2 INDEPENDENT VARIABLES
NAME
E_d_t
TYPE
gasket_closure_distance
INDEPENDENT VARIABLE V1
INDEPENDENT VARIABLE V2
1937
CHAPTER 7.1
Past Enhancements in Marc and Mentat
The second independent variable is selected to be displayed along the X-axis. Note that Ior a table with multiple
dimensions it may be helpIul to rotate the plot. The table data is stored in an external Iile.
TYPE
temperature
ADD ALL POINTS
7
2
0 .027 .054 .081 .108 .135 .175
-100 1000
0 2.08 8.32 18.72 33.28 52 56
0 .208 .832 1.872 3.328 5.2 5.6
FIT
FILLED (off)
MORE
INDEPENDENT VARIABLE V2 LABEL
Temperature
INDEPENDENT VARIABLE V2
INDEPENDENT VARIABLE V1
INDEPENDENT VARIABLE V1 LABEL
Closure Distance
FUNCTION VALUE F LABEL
Youngs Modulus
PREVIOUS
X-AXIS: V1
X-AXIS: V2
FILLED (on)
RX+
RY-
RX+
RY-
FILL
RESET VIEW
FILL
WRITE
E_d_t.tab
1938 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-15 Creating a Two-dimensional Table
A diIIerent way to create a multidimensional table is by multiplying tables. First, a new one-dimensional table E_d is
created. Next, this table is multiplied by table E_t which was created earlier.
NEW
1 INDEPENDENT VARIABLE
NAME
E_d
TYPE
gasket_closure_distance
ADD
0 0
.027 2.08
.054 8.32
.081 18.72
.108 33.28
.135 52
.175 56
FIT
MORE
INDEPENDENT VARIABLE V1 LABEL
Closure Distance
FUNCTION VALUE F LABEL
Youngs Modulus
1939
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-16 Result of Table Multiplication
PREVIOUS
MULTIPLY TABLE
E_t
FILLE
D
(off)
NAME
E_d_t_2
MORE
INDEPENDENT VARIABLE V1
INDEPENDENT VARIABLE V2
FUNCTION VALUE F LABEL
Youngs Modulus
PREVIOUS
X-AXIS: V1
X-AXIS: V2
FILLED (on)
RX+
RY-
RX+
RY-
FILL
1940 Marc Users Guide: Part 3
CHAPTER 7.1
Creation oI a table with three independent variables is now shown using a Iormula to generate the data points. Note
that the independent variables are designated by v1, v2, v3, and v4. The Iormula is evaluated depending on the ranges
and the number oI steps oI the independent variables.
The user may now select which independent variable is displayed along the X-axis, and which along the Y-axis. For
the third independent variable, a Iixed value is taken, namely the i-th data point value Ior this independent variable.
The index i can be set with the FIX button and ranges Irom 1 to the number oI data points oI the independent variable.
RESET VIEW
FILL
NEW
3 INDEPENDENT VARIABLES
FORMULA
ENTER
.1+v1^2+sqrt(v2)+sin(v3*pi)
FIT
RX+
RY-
RX+
RY-
FILL
Y-AXIS: V2
Y-AXIS: V3
X-AXIS: V1
X-AXIS: V2
Y-AXIS: V3
Y-AXIS: V1
X-AXIS: V2
X-AXIS: V3
Y-AXIS: V1
Y-AXIS: V2
FIX V1
6
FIX V1
11
1941
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Figure 7.1-17 Creating a Three-dimensional Table
Especially Ior use in the EXPERIMENTAL DATA FIT menus, Mentat allows creation oI tables with one independent
and two dependent variables. In previous versions, this could only be done by reading raw table data. Now, such a table
can be created, edited, and displayed like any other table.
FILL
X-AXIS: V3
X-AXIS: V1
FIX V3
6
NEW
1 INDEP. & 2 DEP. VARIABLES
RESET VIEW
FILL
ADD
-4/3 -8 0.9605
-1 -6 0.9703
-2/3 -4 0.9801
-1/3 -2 0.9900
0 0 1
1942 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-18 Creating a Table with 2 Dependent Variables
SCALE
0.01 10 1
FIT
TYPE
experimental_data
MORE
INDEPENDENT VARIABLE V1 LABEL
Strain
FUNCTION VALUE F LABEL
Stress
PREVIOUS
Z-AXIS: F
Z-AXIS: F2
MORE
FUNCTION VALUE F
FUNCTION VALUE F2
FUNCTION VALUE F2 LABEL
Vol/Vol0
PREVIOUS
1943
CHAPTER 7.1
Past Enhancements in Marc and Mentat
User-defined Text Input
User-deIined text may be added to the parameter, model deIinition, or history deIinition sections oI the data Iile. The
JOBS menu contains the links to the parameter and model deIinition menus, and the LOADCASES menu contains the
link to the history deIinition menu.
Figure 7.1-19 Additional Input File Text Menu
64-bit Version of Mentat
MSC.Mentat 2003 was enhanced to be a Iull 64-bit compliant application Ior use on the platIorms show in Table 7.1-1.
The perIormance oI the 32-bit and 64-bit versions are the same; however, the memory requirements are much higher.
For example, a 1.1 million element model using the 32-bit version takes about 1.27GB, while the 64-bit version needs
about 1.87GB. ThereIore, in most cases you may want to use the 32-bit version Ior models under 1 million elements.
Python
The ability to obtain a user-deIined string Irom Mentat in a Python script has been added in this release. The user can
speciIy the string using the PARAMETERS menu, and the Python script obtains the value using the py_get_string
routine. The Iollowing example uses the model Iile in Chapter 7 oI the Marc Python Tutorial and prints out the number
oI sets in the model. The steps Ior this example are:
Browse to the Python examples directory.
SpeciIy the name oI the model Iile that we want to check.
Run the Python script.
Table 7.1-1 64-bit Supported Platforms
Vendor OS Hardware
HP-Compaq OSF1 4.0D Alpha
HP HP-UX 11.0 PA 2.0
IBM AIX 5.1 RS/6000
Linux 2.4.9 Intel Itanium
SGI IRIX64 6.4 Mips4
Sun Solaris 2.8 Ultra III
1944 Marc Users Guide: Part 3
CHAPTER 7.1
The Python script is as Iollows:
1 from py_mentat import *
2
3 def main():
4 fn = py_get_string(filename)
5 s = *open_model %s" % fn
6 py_send(s)
7 n = py_get_int("nsets()"
8 print "Sets found: ",n
9 return
10
11 if __name__ == '__main__':
12 main()
The output oI the script will be printed in the terminal window:
Sets found: 8
Postprocessing Enhancements
MPEG and AVI Animations
Mentat can now create an MPEG animation Iile or an AVI (Windows NT/2000/XP only) animation Iile. It is accessed
Irom the RESULTSANIMATION submenu. The settings are preset to typical deIault values so that Ior most users,
only one button needs to be pressed to start the creation oI the animation Iile.
UTILS
CURRENT DIRECTORY
mentat2010/examples/python/tutorial/c07
OK
PARAMETERS
(NAME)
filename
(EXPRESSION)
sets.mfd
OK
PYTHON
RUN
nsets.py
1945
CHAPTER 7.1
Past Enhancements in Marc and Mentat
The MPEG and AVI animation menus are very similar. The BASE FILE NAME is automatically set to the name oI the
post Iile. The GENERATE ANIMATION FILES button enables or disables the creation oI the intermediate display list
Iiles that are read and displayed when selecting the PLAY button in the ANIMATION main menu. In most cases, you
want to have this option selected unless you are assembling an animation Irom various increments in the post Iile. The
buttons under the INCREMENTS section are the same as in the RESULTS main menu. The ATTRIBUTES menu
provides shortcuts to the LEGEND settings, RANGE and COLORMAP buttons. The CLEAN FILES button removes all
the intermediate display list Iiles and the PPM image Iiles used to create an MPEG movie.
Figure 7.1-20 MPEG and AVI Animation Menus
The DELAY button in the MPEG menu duplicates Irames (increment images) in the MPEG movie since some MPEG
players attempt to play the movie in real time. For example, iI there are 100 increments, some MPEG players skip
Irames to try and play the entire movie in 100/24Ips 4 seconds.
When the MAKE MPEG MOVIE button is pressed, the intermediate display list Iiles are generated, then they are played
back and images are created Irom each oI the increments and stored in the PPM graphic Iiles. Then the MPEG
encoding program, mpeg_encode.exe in Mentat`s bin directory, is run in the background.
Note that there is no Ieedback Irom this program back to Mentat to indicate that the MPEG encoder has completed.
The most reliable way to detect this is to use the ps command on Unix or the Windows Task Manager on Windows
NT. You can also monitor the size oI the MPEG Iile. When it is no longer growing in size, the encoder has completed
generating the Iile.
The COMPRESSION DIALOG button in the AVI menu allows you to select the compression method Ior the AVI Iile.
In most cases, you should not select the deIault oI Full Frames (Uncompressed) but select Microsoft Video 1 as the
compression method.
When the MAKE AVI MOVIE button is selected, it perIorms tasks similar to that Ior the MPEG movie. The intermediate
display list Iiles are generated, and then they are played back and images are created. However, these images are not
Note: Do not use the CLEAN FILES button until you have successIully viewed the resulting animation Iile.
1946 Marc Users Guide: Part 3
CHAPTER 7.1
saved to a Iile. They are Ied immediately to the AVI movie generator. When all oI the display list Iiles have been
displayed and images created, the AVI movie generator will write the AVI Iile to disk.
Creating a Movie
The Iollowing example displays how to make an MPEG movie. The technique used Ior generating an MPEG movie
is very similar to that Ior generating an AVI movie. This example uses the HELPRUN A DEMO
PROBLEMRUBBER REZONING example to generate the post Iile.
When the run completes, perIorm the Iollowing steps:
AIter the Iollow message appears:
Creating ppm files....
the MPEG encoder is started and runs in the background. Note that no message appears in the dialogue area when the
process is complete. Check its status using the Windows Task Manager or use the ps command on UNIX.
The PLAY MPEG button may be selected when the MPEG encoder has been started. It starts the mpeg_window
(mpeg_window.bat on Windows NT) script which waits until the MPEG encoder has Iinished beIore attempting to
play the MPEG movie. Note that on UNIX the mpeg_window script must be modiIied to use the application on your
system that supports playing MPEG movies. The movie players are not supplied with the product. On Windows NT
systems, the deIault is to use the application associated with MPEG movies, which is originally Windows Media
Player. This can be changed by either modiIying the mpeg_window.bat script, or by associating a diIIerent
application to MPEG movie Iiles.
Postprocessing in 3-D
New commands have been added named *set_post_procedure on/off (menu button POST PROCEDURE) and
its associated command *post_procedure_file <procedure filename> (menu button FILE) in the RESULTS
menu. These commands allow Ior the speciIication oI a procedure Iile whose contents are executed as each increment
HELP
RUN A DEMO PROBLEM
RUBBER REZONING
RESULTS
MOR
E
ANIMATION
MPEG MOVIE
MAKE MPEG MOVIE
1947
CHAPTER 7.1
Past Enhancements in Marc and Mentat
is read. This is most useIul when a 2-D analysis has been run and a 3-D model is desired to be viewed based on
symmetry.
Figure 7.1-21 Postprocessing Results Menu
For example, the Marc Users Guide, Chapter 3.31 problem oI a tire analysis produces a 2-D section oI the tire. To
build a Iull 3-D model, place the Iollowing commands in a procedure Iile and select it using the FILE button:
*clear_mesh
*set_expand_rotations
20 0 0
*set_expand_repetitions
18
*symmetry_elements
all_existing
*expand_elements
all_existing
To enable its use, select the POST PROCEDURE button.
See Figure 7.1-22 oI the original analysis on the leIt, and the Iull 3-D model is on the right.
1948 Marc Users Guide: Part 3
CHAPTER 7.1
Figure 7.1-22 Views of 2-D and Full 3-D Model
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
attach.proc Mentat procedure Iile
md_table.proc Mentat procedure Iile
tire.igs Iges input Iile
1949
CHAPTER 7.1
Past Enhancements in Marc and Mentat
Animation
Click on the Iigure below to play the animation.
1950 Marc Users Guide: Part 3
CHAPTER 7.1
Chapter 7.2: Importing a Model
7.2
Importing a Model

Chapter Overview 1952

Background Information 1952

Detailed Session Description 1952

Input Files 1964

1952 Marc Users Guide: Part 3
CHAPTER 7.2
Chapter Overview
This chapter describes the process oI importing a geometric or Iinite element model Irom a supported CAD or FEM
program. The process is illustrated through a sample session that involves importing and meshing a geometric model
speciIied in IGES Iormat.
Background Information
Description
The structure you are importing is a seal made out oI rubber that will undergo large deIormations caused by coming
into contact with other parts. The structure is modeled using a boundary representation oI straight lines and curves.
AIter reading the IGES Iile, you will select a portion oI the model and transIorm it into a Iinite element mesh. This
process is described in the steps listed below.
The IGES Iile will be Iound in the Mentat installation directory, in the subdirectory examples/marc_ug and is named
seal.igs.
Overview of Steps
Step 1: Import IGES file.
Step 2: Eliminate all duplicate points and curves using SWEEP processor.
Step 3: Create two sets of the upper and lower parts of the model.
Step 4: Hide upper part of model and scaled the lower part to fill the graphics area.
Step 5: Use of the 2-D planar meshers from the AUTOMESH processor to complete the meshing of the
model.
Detailed Session Description
Step 1: Import IGES Iile.
Assume you are already in Mentat and in the directory where the Iile you wish to import is located.
Use the Iollowing button sequence to read the IGES Iile. Click on the FILL button located in the static menu area to
scale the model to Iill the graphics area. The scaled model that appears in the graphics area is shown in Figure 7.2-1.
MAIN
FILES
IMPORT
1953
CHAPTER 7.2
Importing a Model
Step 2: Eliminate all duplicate points and curves using SWEEP processor.
Prior to manipulating the model in any way, you are advised to eliminate all duplicate points and curves using the
SWEEP processor. Use the Iollowing button sequence to sweep the model oI all duplicate entities.
Mentat responds by sweeping all duplicate points and curves, respectively.
Figure 7.2-1 Scaled Model of Imported IGES File
IGES
seal.igs (select file)
OK
FILL
MAIN
MESH GENERATION
SWEEP
sweep POINTS
all: EXIST.
sweep CURVES
all: EXIST.
1954 Marc Users Guide: Part 3
CHAPTER 7.2
To improve the quality oI the display, change the deIault plot settings to a higher accuracy.
Step 3: Create two sets oI the upper and lower parts oI the model.
Assume you only need to mesh the lower part oI the model shown in Figure 7.2-1. It is useIul to store the upper and
lower parts oI the model in two separate sets as it makes it much easier to reIerence when working with only part oI
the model. An option in Mentat that aids you in Iocusing on the part oI the model you want to mesh is the VISIBLE
option which is used to hide extraneous inIormation.
You are going to use the automatic overlay meshing Ieature which requires a closed boundary description. The lower
part oI the geometry thereIore needs an additional line segment. Create this line in the vicinity oI the lower neck oI the
model using the Iollowing button sequence.
MAIN
MESH GENERATION
PLOT
curves SETTINGS
predefined settings HIGH
REGEN
RETURN (twice)
MAIN
MESH GENERATION
ZOOM BOX
(zoom in on the base of the neck)
crvs ADD (use the default curve type)
235 (pick point)
188 (pick point)
1955
CHAPTER 7.2
Importing a Model
Figure 7.2-2 Curved added at base of Neck
Use the Iollowing button sequence to create the two sets: one Ior the upper part oI the model, the other Ior the lower
part. Due to the awkward shape oI the model, it is best to use the polygon pick method (CTRL key ML~) described
in List Specification oI the Introduction section, select the members Ior each set.
Repeat this operation Ior the lower part oI the model and save the set as lowerpart. A suggestion Ior the contour oI
the polygon pick is depicted in Figure 7.2-3 and Figure 7.2-4.
MAIN
PLOT
draw POINTS (off)
REGEN
RETURN
FILL
MESH GENERATION
SELECT
crvs STORE
upperpart (the curve set name)
OK
(use the Polygon Pick Method to select the curves)
END LIST (#)
1956 Marc Users Guide: Part 3
CHAPTER 7.2
To veriIy that you have created two sets, click on the sets SELECT SET button. A pop-up menu appears over the
graphics area listing the currently deIined sets. Both lowerpart and upperpart should be listed. Click on OK to return
to the SELECT menu.
Figure 7.2-3 Polygon Pick Contour for Upper Part
Step 4: Hide upper part oI model and scaled the lower part to Iill the graphics area.
To Iocus on the lower part oI the structure, use the Iollowing button sequence to hide the upper part oI the model.
MAIN
MESH GENERATION
SELECT
SELECT SET
lowerpart
OK
MAKE VISIBLE
FILL
1957
CHAPTER 7.2
Importing a Model
Figure 7.2-4 Polygon Pick Contour for Lower Part
The upper part oI the model is hidden and the lower part scaled to Iill the graphics area as is shown in Figure 7.2-5.
Figure 7.2-5 Lower Part of Model scaled to fill the Graphics Area
A closer look at Figure 7.2-5 reveals there is an extra curve in the geometry that interIeres with the boundary
description oI the part. This curve must be removed beIore the automatic meshing Ieature is invoked. The curve is
located in the inner part oI the seal on the right hand side oI the model. Use the Iollowing button sequence to remove
this curve.
1958 Marc Users Guide: Part 3
CHAPTER 7.2
Figure 7.2-6 Lower Part of Model with Curve removed
Step 5: Use oI the 2-D planar meshers Irom the AUTOMESH processor to complete the meshing oI the model.
The model is ready to be meshed using the 2-D PLANAR MESHING Irom the AUTOMESH processor Irom the MESH
GENERATION menu. First, the overlay mesher is used. Due to the intricate shape oI the model, it is necessary to use
a density oI 70 elements in both the X and Y direction. A setting oI less than 70 causes holes to appear in the mesh.
Make sure you speciIy all: VISIBLE curves Ior the Enter overlay curve list: prompt. Use the Iollowing button
sequence to mesh the model. Keep in mind that it takes the program some time to generate the model due to the number
oI divisions speciIied.
MAIN
MESH GENERATION
crvs REM
180 (pick curve)
END LIST (#)
MAIN
MESH GENERATION
AUTOMESH
2-D PLANAR MESHING
quadrilaterals (overlay) DIVISIONS
70 70
quadrilaterals (overlay) QUAD MESH!
all: VISIBLE
1959
CHAPTER 7.2
Importing a Model
It is helpIul to turn oII some oI the plot entities to produce a cleaner view oI the mesh.
Figure 7.2-7 shows the resulting mesh that should appear in the graphics area.
Figure 7.2-7 Mesh generated with OVERLAY Mesher
In order to demonstrate the use oI the other 2-D planar meshers, the element mesh is removed and the display oI the
curves and points id activated.
MAIN
PLOT
draw NODES (off)
elements SETTINGS
FACES (off)
RETURN
draw CURVES (off)
REGEN
RETURN
SAVE
MAIN
MESH GENERATION
AUTOMESH
2-D PLANAR MESHING
1960 Marc Users Guide: Part 3
CHAPTER 7.2
Since we have already assured that a closed loop exist Ior the curves, we do not have to enter the REPAIR GEOMETRY
menu in AUTOMESH. Instead, we go to CURVE DIVISIONS directly. Meshers other than the overlay mesher require
a curve division. We Iirst determine the distance between the two parallel curves. Based on this distance, a proper
curve division is set. Note that the Advancing Front QUAD mesher (Figure 7.2-10) requires an even division on
the loops.
CLEAR MESH
PLOT
draw POINTS (on)
draw CURVES (on)
REGEN
RETURN
MAIN
MESH GENERATION
AUTOMESH
CURVE DIVISIONS
UTILS
DISTANCE
(click two points on the parallel curves)
RETURN
AVG LENGTH
0.3
restriction FORCE EVEN DIV
apply restriction LOOPS (select DETECTED LOOPS)
APPLY CURVE DIVISIONS
ALL VISIBLE
1961
CHAPTER 7.2
Importing a Model
Figure 7.2-8 Determine Distance between Parallel Curves
Figure 7.2-9 Apply Curve Divisions
The mesh based using the Advancing Front QUAD masher is now obtained with:
MAIN
MESH GENERATION
AUTO MESH
2-D PLANAR MESH
1962 Marc Users Guide: Part 3
CHAPTER 7.2
Figure 7.2-10 Mesh generated with Advancing Front QUAD Mesher
Clear the mesh and repeat the meshing with the Delaunay triangular mesher.
quadrilaterals (adv frnt) QUAD MESH!
ALL VISIBLE
PLOT
draw CURVES (off)
draw POINTS (off)
REGEN, RETURN
MAIN
MESH GENERATION
AUTOMESH
2-D PLANAR MESHING
CLEAR MESH
PLOT
draw CURVES (on)
REGEN
RETURN
triangles (delaunay) TRI MESH!
ALL VISIBLE
PLOT
draw CURVES (off)
1963
CHAPTER 7.2
Importing a Model
Figure 7.2-11 Mesh generated with a Delaunay Triangular Mesher
Clear the mesh and repeat the meshing with the Advancing Front triangular mesher.
REGEN
RETURN
MAIN
MESH GENERATION
AUTOMESH
2-D PLANAR MESHING
CLEAR MESH
PLOT
draw CURVES (on)
REGEN
RETURN
triangles (adv frnt) TRI MESH!
ALL VISIBLE
PLOT
draw CURVES (off)
RETURN
REGEN
1964 Marc Users Guide: Part 3
CHAPTER 7.2
Figure 7.2-12 Mesh generated with a Advancing Front Triangular Mesher
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
seal.proc Mentat procedure Iile
seal.igs Iges input Iile
Chapter 7.3: HyperMesh Results Interface
7.3
HyperMesh

Results Interface

Chapter Overview 1966

About Postprocessing of Results 1966

About Preprocessing 1967

Mentat Preprocessing for HyperMesh 1967

Postprocessing using HyperMesh 1976

1966 Marc Users Guide: Part 3
CHAPTER 7.3
Chapter Overview
Marc can create a binary Iile oI the results that may be postprocessed using HyperMesh. The writing oI the HyperMesh
results Iile is invoked by the HYPERMESH model deIinition option described in detail in Volume C: Program Input.
The binary Iile created as a result has the title jobid.hmr where jobid is the name oI the Marc data Iile Ior the job.
This Iile, as well as the data Iile (Ior most cases) can be read by HyperMesh Ior postprocessing. II the model was
originally created using HyperMesh, the HyperMesh geometry database can be used instead oI the data Iile. The data
that may be postprocessed ranges Irom elemental quantities, such as stress and strain, to nodal quantities, such as
displacement, acceleration, temperature, and eigenmodes.
About Postprocessing of Results
Interfacing Analysis and Postprocessing
Postprocessing oI Iinite element analysis results usually reIers to the graphical interpretation oI results, perIormed by
way oI a graphics capable computer program. This program usually is the same one used Ior the preprocessing
(modeling) phase, but this is not always necessary. By the use oI a postprocessor, the user can visualize the response
oI a Iinite element model as obtained Irom an analysis oI the model. Such response may include, but is not limited to,
the deIormed shapes, stress and/or strain contours, temperature distribution, and mode shapes.
The results to be postprocessed are normally generated by a Iinite element analysis computer program. These results
are then usually written into a results Iile Ior reasons oI saving the inIormation in a semi-permanent manner. This is
not absolutely necessary as the soItware may be designed to pass the results directly to a postprocessor without saving
them, which is not advisable Ior obvious reasons. As a third choice, the analysis program may pass the results directly
to the postprocessor, while at the same time saving them in a Iile. Marc can interIace with Mentat in the Iirst manner
where the results are Iirst written into the Marc post Iile, which is then read by Mentat Ior postprocessing. This
interIace is transparent when Marc is run Irom Mentat, and intermediate results can be postprocessed while the job is
still running.
For the case oI HyperMesh, the manner oI interIacing is similar. The results Iile, jobid.hmr, is written out by Marc,
and is then later read in by HyperMesh Ior postprocessing. The HyperMesh results Iile is binary, so it is not readable
by a text editor. AIter reading in the results Iile, HyperMesh can be used to create graphic displays oI the results as is
explained later herein.
Data Written into the HyperMesh Results File
The types oI data that may be selected Ior writing into the HyperMesh results Iile are listed in Volume C: Program
Input, under the HYPERMESH model deIinition option. These types oI data are classiIied into two categories: element
results and nodal results. The element quantities (stresses, strains, etc.) written into the results Iile are both the
component values and the invariant values. They are each an average value within the element. Stresses and strains at
nodes are values extrapolated Irom the integration points and based on a weighted average. The other nodal quantities
include results such as displacements, accelerations, reactions, temperatures, and eigenmodes.
1967
CHAPTER 7.3
HyperMesh Results Interface
About Preprocessing
A Marc Iinite element model is usually created using Mentat. When the model is created, it is then written into a Marc
data Iile, which is to be used Ior analysis. The model may also be created by another Iinite element preprocessor which
has the capability to write out a Marc Iile. HyperMesh has this capability regarding the geometric model, Ior most
cases. Thus, the model can be created within HyperMesh, then written out in the Iorm oI a Marc data Iile.
Mentat Preprocessing for HyperMesh
The creation oI a HyperMesh results Iile using Marc is invoked by the HYPERMESH model deIinition option (see
Volume C: Program Input) in the Marc data Iile. We will now review the procedure Ior entering this option into the
data Iile, with the help oI a simple Iinite element model contained in a data Iile initially named x243.dat. This model
is composed oI Iour quadrilateral shell elements (element 75). Both large displacement and Iree vibration eigenvalue
analyses will be perIormed on the model.
There are two ways in which the HYPERMESH option may be entered into the data Iile:
1. By way oI a text editor program, Iollowing the instructions Ior the HYPERMESH option in Volume C: Program
Input.
2. By way oI Mentat, Iollowing the procedure given below:
a. Create the Iinite element model, or iI model is already available, read it into Mentat as described below:
To read the data Iile, start up Mentat. The Main menu shows up (Figure 7.3-1).
Figure 7.3-1 Mentat Main Menu
Press the FILES button (Figure 7.3-1) along the bottom row (static buttons) to access the FILE I/O screen
(Figure 7.3-2).
1968 Marc Users Guide: Part 3
CHAPTER 7.3
On the leIt hand side, under MARC INPUT FILE, press READ to reach the READ MARC INPUT FILE
submenu (Figure 7.3-3).
Figure 7.3-2 The FILE I/O Menu
Figure 7.3-3 READ MARC INPUT FILE Submenu
Type in the selection in the window provided or select using the available access buttons. In this case, the
Iile Ior example, is to be read in, so you have to press the appropriate button in the directory list oI contents.
This operation will result in the Iilename being added to the selection Iield window near the bottom.
Pressing OK on the screen or the <RETURN> key on your keyboard activates the program to read in the data
Iile. When the reading is completed, the deIault view oI the model appears on the screen (Figure 7.3-4).
To see the entire model, press FILL (second row oI the static buttons along the bottom oI the screen,
Figure 7.3-4) to Iill the screen (Figure 7.3-5).
1969
CHAPTER 7.3
HyperMesh Results Interface
Figure 7.3-4 Default View of the Plate Model
Figure 7.3-5 Full View of the Plate Model
The arrows indicate the Iixed boundary conditions which may be seen better iI the model is appropriately
rotated (Figure 7.3-6) by using the RX, RY, RZ and/or DYN. MODEL buttons on the bottom two rows (static
buttons). The plate is acted upon by an increasing distributed load.
b. Go back to the Main menu (Figure 7.3-7) by pressing the MAIN button.
1970 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-6 Rotated View of the Plate Model
Now click on the JOBS button to bring up the JOBS menu (Figure 7.3-8).
Figure 7.3-7 Mentat Main Menu and the JOBS Button
c. Click on the ANALYSIS CLASS type; in this case, MECHANICAL (Figure 7.3-8) and a pop-up menu appears
on the screen (Figure 7.3-9).
1971
CHAPTER 7.3
HyperMesh Results Interface
Figure 7.3-8 Mentat JOBS Menu
d. Now click on JOB RESULTS (Figure 7.3-9) to reach the Job Results menu (Figure 7.3-10), which is
essentially the Marc post Iile related data entry screen.
Figure 7.3-9 Mechanical Analysis Class Pop-up Submenu
e. On the top, your right-hand side, oI the screen, click on the HYPERMESH button (Figure 7.3-10) in order to
access the HyperMesh results Iile related data entry screen (Figure 7.3-11).
1972 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-10 JOB RESULTS Submenu
I. The Irequency (i.e., every how many increments) with which results are to be output into the HyperMesh
results Iile is controlled by the FREQUENCY button at the top leIt oI the screen (highlighted in
Figure 7.3-11) Click on the button to change the deIault Irequency by entering the appropriate number in the
dialogue area. In this case, we wish Ior results every 3rd increment; thus, simply type 3 and press
<RETURN>.
Figure 7.3-11 HYPERMESH RESULTS Submenu
1973
CHAPTER 7.3
HyperMesh Results Interface
g. For ELEMENT RESULTS, select the types oI results desired, by clicking, and thus turning on, the related
options Irom among those available (stress through plastic strain). In this case, assume that the stresses are
to be written into the results Iile Ior layer 3 only. Thus, you should Iirst click on the stress button under
ELEMENT RESULTS.
h. Now select the layers Ior which results are to be output by using the buttons LAYERS through DEF (deIault)
towards the right. II you wish to enter speciIic layer numbers, you can do this by Iirst clicking the LAYERS
button, then by entering the layer numbers, separated by commas or spaces, in the dialogue area, Iollowed
by <RETURN>. In this particular case, type 3 in the dialogue area, then press <RETURN> twice to reach the
command prompt.
i. For NODAL RESULTS, the choices range Irom nodal displacements to eigenmodes (reached at by means oI
the slider bar to the right). Simply click on the desired types oI output to turn them on. In this case, we wish
to save the displacements and eigenvectors into the results Iile. Thus, now you should click on the button
displacement (Iirst button) and on the button eigenmode (last button, highlighted in Figure 7.3-12).
The Iinal appearance oI the screen is shown in Figure 7.3-12.
Figure 7.3-12 Final Appearance of HYPERMESH Submenu Screen
Note that the little squares Ior the switched on buttons show slightly darker in Figure 7.3-12 (e.g. stress
under ELEMENT RESULTS and eigenmode under NODAL RESULTS).
1974 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-13 JOBS Menu with the Finite Element Model
j. Click OK to complete the task. This takes you back to the JOB RESULTS menu (Figure 7.3-10), which
should be Iilled only iI you are also requesting a Marc post Iile to be written out at the end oI analysis.
k. Click OK on each oI the previous two submenus to arrive back at the JOBS menu (Figure 7.3-13).
Important Data Preparation Considerations Regarding Eigenmodes
In case eigenvectors Ior buckling or eigenIrequency analysis are to be written into the HyperMesh results Iile, it is
important to note that the corresponding Marc Iile should have the BUCKLE INCREMENT or MODAL INCREMENT
model deIinition option, as appropriate, together with the associated BUCKLE or DYNAMIC parameter. The history
definition options BUCKLE, MODAL SHAPE, and RECOVER are not to be used.
Relation to other Types of Results Files
Marc has the capability also to write Intergraph and SDRC I-DEAS results Iiles, at the same time as the HyperMesh
results Iile and Marc post Iiles. The writing oI these additional results Iiles is invoked by the IRM and SDRC model
deIinition options, respectively. II the HYPERMESH option is used simultaneously with either or both oI the IRM and
SDRC options, the program internally treats the data in a cumulative manner. For example, iI stresses are requested
Ior the SDRC Universal (results) Iile and creep strains are requested Ior the HyperMesh results Iile, both quantities
are output into both Iiles.
l. Now, press the FILES button (Figure 7.3-13) again to proceed to write an Marc data Iile containing the
HYPERMESH option (Figure 7.3-14).
1975
CHAPTER 7.3
HyperMesh Results Interface
m. Press WRITE under MARC INPUT FILE (Figure 7.3-14) to access the appropriate submenu (Figure 7.3-15).
Simply type in the path and name oI the data Iile to be written (x243_hm.dat in this case, to diIIerentiate
Irom the input Iile that was read in), then press <RETURN> on the keyboard. The updated Marc input Iile
will be written to the indicated directory.
Figure 7.3-14 File I/O Menu with WRITE Button to be Pressed
Figure 7.3-15 WRITE MARC INPUT FILE Submenu
1976 Marc Users Guide: Part 3
CHAPTER 7.3
Postprocessing using HyperMesh
The HyperMesh results Iile, jobid.hmr, contains only the results oI the Iinite element analysis perIormed by Marc.
However, postprocessing oI analysis results requires that the geometry data also be available. At the time the analysis
results are available in the jobid.hmr Iile, the Iinite element geometry is available in the Marc jobid.dat data Iile,
and possibly in a HyperMesh database Iile. We review here the case where the geometry is to be read in Irom the Marc
data Iile.
Since HyperMesh allows only one deIormed shape plot per simulation, each eigenvector oI an eigenvalue analysis is
saved as a separate simulation. Thus, when using HyperMesh, these eigenvectors can be plotted by skipping to the
next simulation rather than to the next data type oI a simulation. The contour plots can be obtained Ior all data types
including eigenvectors.
In case the number oI requested eigenvalues is more than the number extracted, the data type in the HyperMesh
deformed screen inIorms you regarding those modes that have not been extracted. The next button may need to be
clicked to see the data type in the 'deIormed mode oI plotting.
AIter running a job with Marc using the HYPERMESH model deIinition option in the jobid.dat data Iile, you will
obtain a binary HyperMesh results Iile named jobid.hmr. HyperMesh can now be invoked to postprocess the results
contained in jobid.hmr. This process will be illustrated with the help oI the analysis results Ior x243_hm.dat.
The Iirst operational menu oI HyperMesh at start-up is shown in Figure 7.3-16. For better visualization, the Iont size
and background colors have been modiIied using the options menu at the bottom right-hand side oI the screen.
In this case, you see that the Geom option on the right is selected as the deIault. Clicking on the files button at the
upper-leIt corner brings you the next screen (Figure 7.3-17).
Figure 7.3-16 HyperMesh Main Menu
1977
CHAPTER 7.3
HyperMesh Results Interface
Figure 7.3-17 HyperMesh Default hm file Menu
Choosing import Irom the choices on the leIt, you can then proceed to the File Import menu (Figure 7.3-18).
Figure 7.3-18 HyperMesh File Import Menu
This screen now has a choice Ior the type oI input Iile. Double clicking on translator = brings up the Translator menu
oI the various data types which may be read in (Figure 7.3-19).
1978 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-19 HyperMesh Translator Menu
Click on marc to select it. This also brings you back to the previous menu, with your selection now entered
(Figure 7.3-20).
Figure 7.3-20 Entry of Marc into Window
Now double click on the filename = to browse the directory listings. The Iirst menu will show some oI the Iiles in the
current directory and will also provide an option to go up one level (Figure 7.3-21).
1979
CHAPTER 7.3
HyperMesh Results Interface
Figure 7.3-21 Browsing the Directory
In this particular case, we advance through the directory contents by using the next button until we see the required
Iile x243_hm.dat (Figure 7.3-22).
Figure 7.3-22 File Selection
Clicking on the Iile name, x243_hm.dat, brings you back to the Import menu, but with the required Iile name recorded
in the window (Figure 7.3-23).
1980 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-23 Data File Selection Complete
To read in the data Iile, you now click on import. The deIault view oI the Iinite element model appears on the screen
when the reading is completed (Figure 7.3-24).
Figure 7.3-24 Default View of the Finite Element Model
You now need to select the results option at the bottom leIt oI the menu in order to prepare Ior reading in the analysis
results Iile. This operation takes you to the Results File menu (Figure 7.3-25).
1981
CHAPTER 7.3
HyperMesh Results Interface
Figure 7.3-25 Results File Menu
Double clicking on results file = brings you again to the Iiles in the current directory (Figure 7.3-21). In this particular
case, we advance with the next button until we see the x243_hm.hmr Iile; i.e., the HyperMesh results Iile Ior the job
x243_hm.dat. This Iile is obtained as a result oI a Marc run Ior the x243_hm.dat job.
Clicking on x243_hm.hmr now returns you to the Results File menu with the appropriate Iile name recorded in the
window (Figure 7.3-26).
Figure 7.3-26 Results File Selection Complete
1982 Marc Users Guide: Part 3
CHAPTER 7.3
Now you are ready to go into the postprocessing phase. Click on the return button at the bottom right oI the menu
(Figure 7.3-26). This takes you to the initial deIault screen, but this time with the Iinite element model showing
(Figure 7.3-27).
Figure 7.3-27 Main Menu with Finite Element Model
At this point, select the post option at the bottom right oI the menu (Figure 7.3-27) to advance to the postprocessing
screen (Figure 7.3-28).
Figure 7.3-28 HyperMesh Postprocessing Menu
1983
CHAPTER 7.3
HyperMesh Results Interface
By means oI this menu, you can process the data in the results Iile in various ways. These Ieatures are better Iollowed
through the literature available on HyperMesh. Here, we only show several representative examples to indicate how
the data Irom a Marc analysis run can be processed.
You can now use the deformed button near the middle to proceed with plots oI deIormed geometry, either due to
displacements or eigenvectors. Pressing this button takes you to the Deformed Shape screen (Figure 7.3-29).
The eigenvalue analysis results were saved in the results Iile Ior increment 0. However, no displacements were saved
since increment 0 was trivial in terms oI stress analysis. Thus, the Iirst screen Ior deIormed shape has 'Increment 0
Mode 1" as the Iirst simulation (Figure 7.3-29).
Figure 7.3-29 Deformed Shape Screen
To pick up the related data Irom the results Iile, press the next button across Irom data type =. This brings the word
eigenvector to the small window and the data is available. By clicking the next button across Irom simulation =, you
can reach the results Ior other nodes and displacements, including those in other increments as well. For purposes oI
illustration, we now do this once to arrive at the second mode. You can use the a button at right to rotate the model in
drag mode, then use I to Iill the screen. Now set model units = to 1.0 to obtain a reasonably scaled deIormed shape
(eigenvector), then press the deform button to obtain the shape Ior the second mode at increment 0 (Figure 7.3-30).
1984 Marc Users Guide: Part 3
CHAPTER 7.3
Figure 7.3-30 Second Mode of Free Vibration at Increment 0
II the eigenvector is Ior Iree vibration, as in this case, you can now press the modal button Ior animating the mode
shape.
Going back to Figure 7.3-30, iI the next button across Irom simulation = is pressed twice more, you reach the window
shown in Figure 7.3-31.
Figure 7.3-31 Advancing to Results for Increment 3
1985
CHAPTER 7.3
HyperMesh Results Interface
You have now arrived at the stress analysis results Ior increment 3. Note that the plot does not change during these
moves. To get to the displacement data, press the next button across Irom data type = until the data type window shows
the word displacements. Then click on deform to obtain the deIormed shape Ior increment 3 (Figure 7.3-32).
Figure 7.3-32 Displacement Plot for Increment 3
To obtain contour plots, press the return button at the bottom right. This takes you back to the Postprocessing menu
oI Figure 7.3-28. Pressing the contour button takes you to the Contour screen.
You can now get a contour plot oI the results quantities, such as the displacement plot in Figure 7.3-33.
Figure 7.3-33 Contour Plot of Displacements for Increment 3
1986 Marc Users Guide: Part 3
CHAPTER 7.3
or the second stress invariant oI layer 3 in increment 6, the contour plot in Figure 7.3-34.
Figure 7.3-34 Contour Plot of Stress Invariant
Chapter 7.4: Translators
7.4
Translators

Chapter Overview 1988

New Mentat Writers 1988

New Mentat Readers 1989

Improved Readers 1992

1988 Marc Users Guide: Part 3
CHAPTER 7.4
Chapter Overview
This chapter highlights the output oI a model to Iour standard Iormats: dxfout, stlout, vdaout, and vrmlout (which is
not a standalone program). Also we have two new readers: c-mold and stl.
New Mentat Writers
dxfout:
This is a brand new writer which will output an ASCII DXF Iile based on AutoCAD 2000.
stlout:
This is a brand new writer which will output an ASCII StereoLithography InterIace speciIication (STL) Iile based on
Oct 1989's standard.
vdaout:
This is a brand new writer which will output an ASCII VDA-FS Iile based on VDA-FS Revision 2.0.
vrmlout:
This translator is embedded in Mentat. It's based on VRML97 (a.k.a ISO VRML or VRML 2.0). It will NOT output
any geometric entities Irom Mentat; instead, the output is based on the graphical entities and view settings. The Iile
Iormat is in ASCII.
Use the button sequence or the new Mentat writers and see the sample menu (Figure 7.4-1).
MAIN
FILE
EXPORT
1989
CHAPTER 7.4
Translators
Figure 7.4-1 Sample of the EXPORT Menu
New Mentat Readers
c-mold:
The current version oI the interIace supports C-MOLD versions 98.7 to 99.1. It reads data Irom Iour C-MOLD Iile
types:
the parameter Iile (extension .par or .PAR)
the Iinite element mesh Iile (extension .Iem or .FEM)
the material properties Iile (extension .mtl or .MTL)
the results Iile oI the C-MOLD stress analysis (extension .ppt or .PPT)
These Iiles should reside in the same directory. You must speciIy the name oI one oI these Iiles. The names oI the
others are automatically derived Irom it.
Part oI the data is imported directly into Mentat. The other data (most notably, the residual stresses, the elastic and
thermal properties, and material orientations, which are all layer and element dependent) is written to a Marc post Iile
that can be viewed directly Irom the RESULTS menu. This post Iile data is read at the start oI a Marc job. This requires
that the user subroutine cmold2marc.I in the Mentat bin directory is used. The Iollowing data is extracted Irom the
C-MOLD Iiles:
1990 Marc Users Guide: Part 3
CHAPTER 7.4
Parameter file (.par or .PAR):
Finite element mesh file (.fem or .FEM):
Material properties file (.mtl or .MTL):
Results file (.ppt or .PPT):
Use the Iollowing button sequence Ior the new Mentat reader C-MOLD and see the sample menu (Figure 7.4-2).
Data Set T-CODE Description
PRMT 100 Number oI layers across the Iull-gap thickness
620 Fibre orientation analysis option
TITL Title oI the model (currently not used)
Data Set T-CODE Description
EPRO 30100 Thickness oI triangular elements
NODE Coordinates oI the nodes
QUAD Connectivity Ior quadrilateral element
TITL Title oI the model (currently not used)
TRI Connectivity Ior triangular element
Data Set T-CODE Description
MTRL 1600 Isotropic material properties
1602 Orthotropic material properties
1700 Isotropic thermal expansion coeIIicient
1702 Orthotropic thermal expansion coeIIicients
TITL Title oI the model (currently not used)
Data Set T-CODE Description
ELDT Layer-based residual stresses and material properties Ior Iibre-Iilled analyses
TITL Title oI the model (currently not used)
TSDT Layer-based residual stresses Ior unIilled analyses; material properties are taken
Irom Material properties Iile
MAIN
FILE
IMPORT
1991
CHAPTER 7.4
Translators
Figure 7.4-2 Sample of the IMPORT Menu with the C-MOLD Button Highlighted
stl:
This new reader which reads both ASCII and binary version oI Stereo Lithography InterIace speciIication (STL) Iiles.
Use the Iollowing button sequence Ior the new Mentat reader STL and see the sample menu (Figure 7.4-3).
Figure 7.4-3 Sample of the IMPORT Menu with the STL Button Highlighted
MAIN
FILE
IMPORT
1992 Marc Users Guide: Part 3
CHAPTER 7.4
Improved Readers
SigniIicant works have been put in the Iollowing readers: acis, dxf, ideas, nastran, and patran.
acis:
In this release, it can read ACIS R13 and earlier.
dxf:
In this release, it can read AutoCAD 2000 (or earlier) ASCII/Binary DXF Iiles and DWG Iiles.
ideas:
This version oI I-DEAS reader has been re-written Irom scratch. It's based on MS7.
nastran:
This version oI nastran reader has been re-written Irom scratch. It's based on Nastran v69. A Iew oI v70 entities were
also included.
patran:
This version oI patran reader has been rewritten Irom scratch. It's based on Patran v8.
Chapter 7.5: Sweep Nodes on Outlines
7.5
Sweep Nodes on Outlines

Chapter Overview 1994

Background Information 1994

Detailed Session Description 1994

Input Files 1998

1994 Marc Users Guide: Part 3
CHAPTER 7.5
Chapter Overview
This chapter describes the usage oI the SWEEP NODES button on Outlines in Mentat. One box with six surIaces will
be created to explain how to use the Iunction.
Background Information
In Mentat, 3-D models are composed by nurb surIaces bounding a closed volume. The surIace mesh is created on every
individual surIaces. In order to create 3-D mesh, the nodes on the outlines oI each surIace mesh should be merged with
the closest nodes on their neighboring outlines. The merging process is controlled by sweep tolerance.
Overview Steps
Step 1: Create six flat surfaces
Step 2: Create surface mesh
Step 3: Sweep the nodes on outlines
Detailed Session Description
Step 1: Create six Ilat surIaces
Use the Iollowing button sequence to create six Ilat nurb surIaces to Iorm a closed box.
MAIN
MESH GENERATION
srfs ADD
point (0.4,0.4,0.0)
point (-0.4,0.4,0.0)
point (-0.4,-0.4,0.0)
point (0.4,-0.4,0.0)
1
2
point (-0.4,0.4,0.6)
point (0.4,0.4,0.6)
3
2
5
point(-0.4,-0.4,0.6)
1995
CHAPTER 7.5
Sweep Nodes on Outlines
Figure 7.5-1 Six Surfaces Created to Form a Closed Box
4
3
7
point (0.4,-0.4,0.6)
4
1
6
8
6
5
7
8
INTERSECT
TRIM OUTER
all: EXIST.
VIEW
SHOW VIEW 4
1996 Marc Users Guide: Part 3
CHAPTER 7.5
Step 2: Create surIace mesh
In step 2, apply a curve division on surIace trimming curves and create the surIace mesh on all six surIaces.
Figure 7.5-2 The Nodes are repeated on the Outlines of the Surface Mesh
MAIN
MESH GENERATION
AUTOMESH
CURVE DIVISIONS
FIXED AVG LENGTH
APPLY CURVE DIVISIONS
all: EXIST.
RETURN
SURFACE MESHING
triangles (delaunay)
SURFACE TRI MESH!
all: EXIST.
PLOT
elements SOLID
REGEN
FILL
1997
CHAPTER 7.5
Sweep Nodes on Outlines
Step 3: Sweep the nodes on outlines
Now use the SWEEP NODES button on all outlines oI mesh. The ALIGN SHELL option is also necessary to make sure
the all elements have the same orientation. Finally, check iI no Iree outlines are leIt. Use the Iollowing button
sequences Ior the Iinal result.
Figure 7.5-3 Repeated Nodes on the Outlines are Meshed
You may wish to run Mentat procedure Iiles that are in the
examples/marc_ug/c7.5.proc subdirectory
under Mentat. The procedure Iile c7.5.proc builds, runs, and postprocesses this simulation.
MAIN
MESH GENERATION
AUTOMESH
SOLID MESHING
SWEEP OUTLINE NODES
EXIST
ALIGN SHELL
481
OUTLINE EDGE LENGTH
1998 Marc Users Guide: Part 3
CHAPTER 7.5
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
outlines.proc Mentat procedure Iile
Chapter 7.6: Transition Parameter for Meshing
7.6
Transition Parameter for
Meshing

Chapter Overview 2000

Background Information 2000

Detailed Session Description 2000

Input Files 2004

2000 Marc Users Guide: Part 3
CHAPTER 7.6
Chapter Overview
This chapter describes the usage oI mesh coarsening parameter. This parameter is used to control the mesh density
transition Irom the boundary to the domain center.
Background Information
As the parameter value is bigger than 1, the element size at domain center is bigger. As the value is smaller than 1, the
element size at domain center is smaller. The TRANSITION parameter applies to 2-D, surIace advancing Iront,
Delaunay meshers, and 3-D Delaunay mesher.
Overview Steps
Step 1: Create a close 2-D boundary
Step 2: Create mesh with default transition parameter value 1
Step 3: Create mesh with the value bigger than 1
Step 4: Create mesh with the value smaller than 1
Detailed Session Description
Step 1: Create a close 2-D boundary
First, use the Iollowing button sequences to create six curves to Iorm a closed 2-D meshing domain (Figure 7.6-1).
MAIN
MESH GENERATION
crvs ADD
point (-.5, .8,0.0)
point (-.9, .3,0.0)
2
point (-.3, 0.0,0.0)
3
point (-.3, -.7,0.0)
4
point (1.0, -.7,0.0)
5
point (1.0, 1.0,0.0)
2001
CHAPTER 7.6
Transition Parameter for Meshing
Figure 7.6-1 A 2-D Bound Domain to be meshed
Step 2: Create mesh with deIault transition parameter value 1
In this step, curve division on curves and create quad mesh with deIault transition parameter value 1.0 (Figure 7.6-2).
6
1
MAIN
MESH GENERATION
AUTOMESH
CURVE DIVISIONS
AVG LENGTH
0.2
APPLY CURVE DIVISIONS
all: EXIST.
RETURN
2D PLANAR MESHING
quadrilaterials (adv frnt)
QUAD MESH!
all: EXIST.
2002 Marc Users Guide: Part 3
CHAPTER 7.6
Figure 7.6-2 Quad Mesh with Transition Value at 1.0
Step 3: Create mesh with the value bigger than 1
In this step, change the transition parameter value to 1.5, and create a quad mesh (Figure 7.6-3).
MAIN
MESH GENERATION
AUTOMESH
2D PLANAR MESHING
CLEAR MESH
TRANSITION
1.5
quadrilaterials (adv frnt)
QUAD MESH
all: EXIST.
2003
CHAPTER 7.6
Transition Parameter for Meshing
Figure 7.6-3 Quad Mesh with Transition Value at 1.5
Step 4: Create mesh with the value smaller than 1
With this Iinal step, change transition parameter value to 0.5, and create quad mesh (Figure 7.6-4).
MAIN
MESH GENERATION
AUTOMESH
2D PLANAR MESHING
CLEAR MESH
TRANSITION
0.5
quadrilaterials (adv frnt)
QUAD MESH
all: EXIST.
2004 Marc Users Guide: Part 3
CHAPTER 7.6
Figure 7.6-4 Quad Mesh with Transition Value at 0.5
You may wish to run Mentat procedure Iiles that are in the
examples/marc_ug/meshing_param.proc subdirectory
under Mentat. The procedure Iile, c7.6.proc, builds, runs, and postprocesses this simulation.
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
meshing_param.proc Mentat procedure Iile
Chapter 7.7: Mentat Features 2001 and 2003
7.7
Mentat Features 2001 and
2003

Chapter Overview 2006

2001 Features 2006

2003 Features 2013

Input Files 2016

2006 Marc Users Guide: Part 3
CHAPTER 7.7
Chapter Overview
In the 2001 Mentat release, the new Ieatures described below were implemented in the Mentat program.
2001 Features
Optimized Element Graphics Generation
In Mentat, elements are plotted Ior display in a much more eIIicient manner. The positions oI their nodes are used to
produce the lines and polygons representing the elements. Previously, the element's shape Iunctions were used to
evaluate the geometry oI each element. Also, the method by which the visible edges and Iaces oI each element are
determined has been greatly streamlined. These changes have resulted in cutting the element regeneration time in halI.
Optimized Entity Recoloring
All entities (elements, curves, nodes, etc.) are now recolored in a more eIIicient manner when they are picked, selected,
or need to have their color changed Ior other reasons such as identiIying sets.
Previously, when recoloring oI an entity was desired, all the graphical primitives (lines, polygons, etc.) were scrapped
and replaced by a completely new set oI primitives with the correct new colors. Now, Mentat bypasses this costly
approach, and instead simply changes the color oI the existing primitives. This change has resulted in large time
savings.
Post Reader Optimization
The low level code Ior reading in post Iiles has been optimized to deal with larger blocks oI data Irom the Iile. Now
post Iiles read into Mentat in much less time than in earlier releases.
2007
CHAPTER 7.7
Mentat Features 2001 and 2003
Flowline Plotting
Flowlines can now be computed by Marc and displayed in Mentat. When using global remeshing, the mesh is no
longer attached to the material. To visualize how the material Ilows, the original mesh is used below to Iorm the
Ilowlines. Open the gui.mud Iile; then:
This turns on the calculations oI the Ilowlines that are attached to the material. Figure 7.7-1 shows the selection menu
to turn on the calculations oI the Ilowlines in Marc. The model results superimposed in Figure 7.7-1 show the original
and Iinal mesh. The original mesh is a uniIorm rectangle oI 70 elements. Global remeshing changes the mesh during
the analysis to over 300 elements.
Figure 7.7-1 Request for MSC.Marc to Compute Flowlines Submenu
JOBS
JOB RESULTS
FLOWLINES
Body_1
2008 Marc Users Guide: Part 3
CHAPTER 7.7
Submit the job and open the post Iile. The Ilowlines are automatically plotted until turned oII. Controls are available
Ior selecting which Ilowline edges are plotted, and whether or not to restrict them to the model outline or surIace. Use
the Iollowing button sequence to get to the FLOWLINES submenu to change the plot controls.
Figure 7.7-2 Flowlines from Original Mesh
Figure 7.7-2 shows the Ilowlines on top oI the deIormed mesh at the end oI the analysis. Since the original mesh was
used Ior the undeIormed Ilowline grid, the Ilowlines in Figure 7.7-2 allow us to see how distorted the original mesh
becomes and the necessity oI global remeshing.
Particle Tracking
Particle tracking can be also requested. Trajectories oI material particles are computed along with values oI equivalent
stress and total plastic strain. The request Ior Marc to compute these trajectories are made in JOB RESULTS and can
be seen in Figure 7.7-1 on the top panel under TRACKING FILE. Here, you are prompted Ior a set oI nodes whose initial
position will determine which material particles are tracked.
RESULTS
MORE
FLOWLINES
2009
CHAPTER 7.7
Mentat Features 2001 and 2003
Figure 7.7-3 Particle Tracking Trajectories with Equivalent Stress Magnitudes
During postprocessing, the *post_tracks_stress command plots the trajectories as shown in Figure 7.7-3 Ior
those particles at the original nodes on the boundary.
PostScript Thin Lines Option
A new option has been added to the raster (deIault) PostScript plot capability in Mentat. This THIN LINES option
speciIies that all drawn lines have a width oI one dot or pixel. This can be desirable Ior high resolution images that
have many lines (such as a mesh with many thousands oI elements). Note, this comes in handy Ior very large meshes,
and aIIords a level oI detail which would otherwise be impossible. When this option is oII, a thicker line width is used,
which compensates Ior varying resolutions.
Use the Iollowing button sequence to get to the THIN LINES option (Figure 7.7-4):
UTILS
SETTINGS
THIN LINES
2010 Marc Users Guide: Part 3
CHAPTER 7.7
Figure 7.7-4 Postscript Settings Menu with THIN LINES Command
Curve Direction
The parameterized direction oI curves within Mentat are now optionally displayed by an arrow.
This command toggles the drawing oI an arrowhead on each curve, which points in the direction the curve is deIined
in. Thus, Ior a given curve, the arrowhead points in the direction its curve is traversed when that curve is evaluated in
an increasing direction in parametric space.
Use the Iollowing button sequence to get to the CURVE DIRECTION option (Figure 7.7-5):
PLOT
MORE
MORE
CURVE DIRECTION
2011
CHAPTER 7.7
Mentat Features 2001 and 2003
Figure 7.7-5 Plot Settings (Cont.-2) Menu with CURVE DIRECTION Command
New Viewing Capability
Two new viewing commands, SET ANGLES and SET TRANSLATIONS, have been added to Mentat.
The SET ANGLES command sets absolutely the viewing rotation angles Ior the model, while leaving the viewing
model scale and translations alone. All camera settings remain unchanged by this command. You must speciIy separate
X, Y, and Z rotation angles in degrees.
Use the Iollowing button sequence to get to both viewing commands (Figures 7.7-6 and 7.7-7):
The SET TRANSLATIONS command allows you to set the model's viewing displacement Irom the view space origin.
All camera settings remain unchanged by this command. Any pre-existing viewing translation is replaced by the given
translation.
This command also acts on all the currently active views.
VISUALIZATION
VIEW
MANIPULATE MODEL
Note: This command acts on all the currently active views.
2012 Marc Users Guide: Part 3
CHAPTER 7.7
Figure 7.7-6 Manipulate Model Menu with SET ANGLES Command
Figure 7.7-7 Manipulate Model Menu with SET TRANSLATION Command
2013
CHAPTER 7.7
Mentat Features 2001 and 2003
2003 Features
The new Ieatures described below have been implemented in Mentat 2003.
User Defined Variable Names
The names Ior User DeIined Nodal Quantities and User DeIined Element Scalars may now be edited through Mentat.
Select the Iollow Mentat buttons to go to the JOB RESULTS menu:
The submenu Ior the available names is shown in Figure 7.7-8.
Figure 7.7-8 List of Available Post Quantities
Simply click the button adjacent to the name oI one oI the User Nodal Quantity values desired to select it, and
then click inside the edit box and type in a new name. The same is available Ior the User DeIined Element Scalar
values.
JOBS
MECHANICAL
JOB RESULTS
2014 Marc Users Guide: Part 3
CHAPTER 7.7
Status File Information
The JOBS RUN menu now displays more status inIormation Irom a Marc job as shown in Figure 7.7-9. The number
oI cycles, number oI separations, number oI cut backs, and the number oI remeshes that were perIormed. Also
displayed is the ANALYSIS TIME, which is the current loadcase time value. This data is printed into a Iile named
jobname.sts.
Figure 7.7-9 The RUN JOB Menu displaying the New JOB STATUS Information
DCOM Server Support for Windows NT
The Marc DCOM Server allows you to run jobs on a remote Windows NT machine without actually being logged into
that machine. Unlike Marc Parallel, it will only run a single CPU job. See the Marc and Mentat Installation and
Operations Guide for Windows NT Ior inIormation on installing and conIiguring the Marc DCOM Server.
A remote machine may be speciIied Irom Mentat using the RUN JOB menu as shown in Figure 7.7-10. Select the
DCOM button, then click inside the adjacent text box and type the name oI the machine you wish to run the job on.
Note that you aree not able to monitor the progress oI the job using the MONITOR button Irom the RUN JOB menu.
You may monitor the post Iile results Irom the MONITOR button in the RESULTS menu.
Figure 7.7-10 The RUN JOB Menu displaying the DCOM Server Option
2015
CHAPTER 7.7
Mentat Features 2001 and 2003
The Iiles used Ior a DCOM job must be located in a shared directory. To share a directory, go to My Computer and
browse to the directory where the job Iile is located. A directory higher up in the path may be shared instead.
For example, iI the Iile is located in a directory named d:\projects\data\dynamics, the directory d:\projects,
it may be shared. When you browse through and reach the directory to be shared, right click on the icon, select Sharing,
and then enter a share name.
The job may also be run using the run_marc script Irom the command line. The syntax Ior running the job is:
run_marc -pc computername -j jobname
The computername may be any Windows NT computer on the network that has the Marc DCOM Server loaded and
conIigured properly.
User-defined NUMERIC Format
The appearance oI the numeric inIormation displayed with the RESULTSNUMERIC option has been updated to
allow a user-deIined Iormat. To access the user-deIined numerics menu, go to the RESULTS(SCALAR
PLOT)/SETTINGS menu, and then select NUMERICS.
The deIault setting is AUTOMATIC. You note that the PRECISION button is grayed out Ior this Iormat as shown in
Figure 7.7-11.
Figure 7.7-11 The NUMERICS SETTINGS Menu
The button displaying AUTOMATIC is a roller button that cycles through Iour options which are:
AUTOMATIC Mentat uses the deIault precision Ior the mantissa determined by the Iloating point Iormat
oI 'g (generally six digits). The exponent is displayed.
EXPONENTIAL The mantissa precision to the right oI the decimal point may be speciIied by the user using
the PRECISION button. There is one digit to the leIt oI the decimal point. The exponent
is displayed.
FLOATING The exponent is not displayed and the precision may be adjusted as with the
EXPONENTIAL option.
INTEGER The numbers is displayed as integers.
2016 Marc Users Guide: Part 3
CHAPTER 7.7
Previous and Last Increment Buttons
Two new buttons have been added to the RESULTS menu in Mentat which go to the previous (PREV) and the last
(LAST) increment on the post Iile as shown in Figure 7.7-12.
Figure 7.7-12 The PREVIOUS and LAST Increment Buttons
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
gui.mud Mentat model Iile
Chapter 7.8: Generalized XY Plotter
7.8
Generalized XY Plotter

Chapter Overview 2018

Background Information 2018

Detailed Session Description 2018

Input Files 2023

2018 Marc Users Guide: Part 3
CHAPTER 7.8
Chapter Overview
This chapter describes the usage oI XY plotting. Three history plots are collected into one XY plot to demonstrate the
XY plot Ieature.
Background Information
Generalized XY-Plot allows users to put multiple plots associated with diIIerent jobs into one plot. For example, users
can compare the computing results Irom diIIerent methods on one model by overlaying the plots. Generalized XY plot
has the ability to collect plots Irom various plotters: History plot, Response Gradient/Design Variable plot, Path plot,
Table and Xcurve plot.
Overview Steps
In the example, three jobs are used to describe the XY plot Ieature. All three Iorming jobs are on one model. The Iirst
two jobs use shell and membrane elements respectively, and the third one uses 2-D plane strain continuum elements.
The three history plots Ior each job is created on one node at the same location.
Step 1: Read post file, create history plot, and move into XY plot
Step 2: Repeat the first step for another two jobs
Step 3: Obtain XY plot on the three curves
Detailed Session Description
Step 1: Read post Iile, create history plot, and move into XY plot
Read the Iirst Iorming job post Iile. Create one history plot oI process pressure over time on node 40. Move the history
plot into XY plot by selecting the >XY button in the HISTORY PLOT menu. By doing so, the plot is not lost when a
user starts working on the second history plot. In the Iollowing examples, the data Iiles are located in the directory
mentat2010/examples/marc_ug/s7/c7.8.
MAIN
RESULTS
OPEN
xy_plotter_a.t16
OK
FILL
HISTORY PLOT
SET NODES
40
(click the right mouse button for # | End of List)
2019
CHAPTER 7.8
Generalized XY Plotter
Procedure Iile is:
*post_open xy_plotter_a.t16
*fill_view
*set_history_nodes
40
*history_collect
0 500 20
*history_add_var
Time
Process Pressure
*history_fit
*get_history_plots
Figure 7.8-1 History Plot for Post File 1
COLLECT DATA
0 500 20
NODES/VARIABLES
ADD VARIABLE
Time
Process Pressure
FIT
RETURN
generalized xy plot COPY TO
2020 Marc Users Guide: Part 3
CHAPTER 7.8
Step 2: Repeat the Iirst step Ior another two jobs
Repeat step 1 to create another two history plots.
Figure 7.8-2 History Plot for Post File 2
MAIN
RESULTS
OPEN
xy_plotter_b.t16
OK
HISTORY PLOT
SET NODES
40
(click the right mouse button for # | End of List)
COLLECT DATA
0 500 20
NODES/VARIABLES
ADD VARIABLE
Time
Process Pressure
FIT
RETURN
generalized xy plot COPY TO
2021
CHAPTER 7.8
Generalized XY Plotter
Figure 7.8-3 History Plot for Post File 3
MAIN
RESULTS
OPEN
xy_plotter_c.t16
OK
FILL
HISTORY PLOT
SET NODES
196
(click the right mouse button for # | End of List)
COLLECT DATA
0 500 20
NODES/VARIABLES
ADD VARIABLE
Time
Process Pressure
FIT
RETURN
generalized xy plot COPY TO
2022 Marc Users Guide: Part 3
CHAPTER 7.8
Step 3: Obtain XY plot on the three curves
AIter the three history plots are moved into XY plot, users can compare the results Irom diIIerent approaches on the
same model.
Figure 7.8-4 Three History Plots Displayed in One XY Plot
You may wish to run Mentat procedure Iiles that are in the examples/marc_ug/s7/ c7.8/xy_plotter.t16
subdirectory under Mentat. The procedure Iile xy_plotter.proc builds, runs, and postprocesses this simulation.
UTILS
GENERALIZED XY PLOT
FIT
Note: Three t16 Iiles in the directory examples/marc_ug/ should be copied to the current working directory in
order to run the procedure Iile properly.
2023
CHAPTER 7.8
Generalized XY Plotter
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
xy_plotter.proc Mentat model Iile
xy_plotter_a.t16 Marc post Iile
xy_plotter_b.t16 Marc post Iile
xy_plotter_c.t16 Marc post Iile
2024 Marc Users Guide: Part 3
CHAPTER 7.8
Chapter 7.9: Beam Diagrams Example
7.9
Beam Diagrams Example

Chapter Overview 2026

Background Information 2026

Detailed Session Description 2027

Input Files 2034

2026 Marc Users Guide: Part 3
CHAPTER 7.9
Chapter Overview
The sample session described in this chapter demonstrates the procedure oI displaying generalized stresses plots along
the axial direction oI the beam elements. The goal oI the demonstration is to show the simplicity oI these procedures.
These diagrams are particularly important Ior the design and analysis oI Irame structure, which is usually composed
oI several connected members which are either Iixed or pinned-connected at their ends. By selecting appropriate post
codes Ior corresponding generalized stresses, we can postprocess plots such as shear Iorce, axial Iorce, bending
moment, or even torque and bi-moment iI supported, along the axial direction oI the elements.
In this chapter, a simple Irame structure is analyzed and the procedure to display the shear Iorce and bending moment
diagrams is demonstrated (Figure 7.9-1).
Figure 7.9-1 Simple Frame Structure Subject to Concentrated and Distributed Loads
Background Information
A simple Irame structure consists oI two members is used to illustrate how to generate inIormation required by, and
manipulate the settings oI, the beam diagrams. Two members with a length oI 4 m and 5 m, respectively, are connected
at an angle oI 53.1. The leIt end is only allowed to rotate along Z axis and the right end is allowed move along X axis
in addition to rotate along Z axis. The horizontal member is subjected to a point load oI 80 kN in the middle, and a
distributed load oI 40 kN acts along the inclined member. Figure 7.9-1 shows the model analyzed.
Overview of Steps
Step 1: Create the model
Step 2: Apply appropriate boundary conditions
Step 3: Apply material and geometric properties to elements
Step 4: Select the post codes and submit the job
Step 5: Postprocess the results
4 m 4 m
80 kN
2 m
3 m
4
0

k
N
/
m
A
B C
2027
CHAPTER 7.9
Beam Diagrams Example
Detailed Session Description
Step 1: Create the model
The Irame structure described earlier is modeled by 16 two-noded elements, which are generated by converting two
lines into two Iinite elements and subdividing them into 16 elements.
Step 2: Apply appropriate boundary conditions
The Iollowing sequence speciIies the loading on the Irame as well as the boundary conditions. Figure 7.9-2 shows the
loading and boundary conditions.
MESH GENERATION
nodes ADD
-4 0 0
0 3 0
4 3 0
FILL
ELEMENT CLASS
LINE (2)
RETURN
elems ADD
1 2 2 3
SUBDIVIDE
DIVISIONS
8 1 1
ELEMENTS
all: EXIST.
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
MAIN
BOUNDARY CONDITIONS
NEW
2028 Marc Users Guide: Part 3
CHAPTER 7.9
MECHANICAL
FIXED DISPLACEMENT
DISPLACEMENT X (on)
DISPLACEMENT Y (on)
DISPLACEMENT Z (on)
ROTATION X (on)
ROTATION Y (on)
OK
nodes ADD
1
#
NEW
FIXED DISPLACEMENT
DISPLACEMENT Y (on)
DISPLACEMENT Z (on)
ROTATION X (on)
ROTATION Y (on)
OK
nodes ADD
3
#
NEW
POINT LOAD
FORCE Y (on)
FORCE
-80
OK
nodes ADD
14
#
NEW
GLOBAL LOAD
FORCE X (on)
FORCE
24
OK
2029
CHAPTER 7.9
Beam Diagrams Example
Figure 7.9-2 Finite Element Model and Boundary Conditions
Step 3: Apply material and geometric properties to elements
The Irame members are modeled as an isotropic material.
elements ADD
1 2 3 4 5 6 7 8
#
MAIN
MATERIAL PROPERTIES
NEW
ISOTROPIC
YOUNG'S MODULUS
50000
POISSON'S RATIO
0.2
OK
elements ADD
all: EXIST.
MAIN
2030 Marc Users Guide: Part 3
CHAPTER 7.9
Step 4: Select the post codes and submit the job
The key to the successIul processing oI beam diagrams is to select the necessary post codes Irom the JOB RESULTS
buttons. The post code Ior beam orientation must be selected in order to postprocess any beam diagrams.
Select the result to be written on the post Iile and submit the job. Figure 7.9-3 shows the available buttons to select
Irom Ior the beam diagram.
GEOMETRIC PROPERTIES
NEW
3-D
ELASTIC BEAM
AREA
0.1
Ixx
0.01
Iyy
0.01
VECTOR DEFINING LOCAL X-AXIS: Z
1
OK
elements ADD
all: EXIST.
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
3-D TRUSS/BEAM
98
OK
all: EXIST.
RETURN (twice)
NEW
MECHANICAL
JOB RESULTS
BM_ORIENT
BM_AXI_FOR
2031
CHAPTER 7.9
Beam Diagrams Example
Figure 7.9-3 Job Results Submenu
Step 5: Postprocess the results
The results oI the Ilaring process analysis have been saved in a post Iile. Use the Iollowing button sequence to open
the Iile. Under the Beam Diagram subscreen shown in Figure 7.9-4, select the appropriate diagram you wish to view.
BM_BND_MOM_X
BM_BND_MOM_Y
BM_SHR_FOR_X
BM_SHR_FOR_Y
OK (twice)
RUN
SUBMIT 1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
MORE
BEAM DIAGRAM: SETTING
OPTIONS: SCALE FACTOR
3
RETURN
SHEAR FORCE
BEND. MOMENT
AXIAL FORCE
2032 Marc Users Guide: Part 3
CHAPTER 7.9
Figure 7.9-4 Postprocessing Results Menus
Figure 7.9-5 shows the shear force diagram, Figure 7.9-6 shows the bending moment diagram, and Figure 7.9-7 shows
the axial force diagram.
Figure 7.9-5 Shear Force Diagram
Diagrams
2033
CHAPTER 7.9
Beam Diagrams Example
Figure 7.9-6 Bending Moment Diagram
Figure 7.9-7 Axial Force Diagram
2034 Marc Users Guide: Part 3
CHAPTER 7.9
Input Files
The Iiles below are on your delivery media or they can be downloaded by your web browser by clicking the links (Iile
names) below.
File Description
beam_diagrams.proc Mentat model Iile