PREFACE

This laboratory manual is prepared to supplement the course of Computer-Aided

Design of Mechanical Systems. It outlines the basic procedure for data preparation,
program execution and output interpretation for NASTRAN, PATRAN, GENESIS and
PRO/E studied in the course. Examples and homework assignments are included in each
chapter for practice.

Chapter 1 of the manual familiarizes students with UNIX systems. Chapter 2

introduces a popular finite element code, NASTRAN, starting with static analysis. It is
followed by the introduction of PATRAN in Chapter 3 that considers finite element
modeling. The rest of the manual discuss various applications of finite element methods;
vibration analysis (Chapter 4), transient analysis (Chapter 5) and NASTRAN analysis of
laminated composite panels (Chapter 6). Chapter 1 as well as the rest of Chapters that
cover subjects such as design optimization, thermal analysis Pro-E, Pro-Mechanica are not
included in this particular document.

The Computer-Aided Engineering (CAE) laboratory was initiated by Dr. E. A.

Thornton in 1984 and was officially formed in 1986 with an equipment grant from Control
Data Corporation. With this facility and the support from the college and the department,
the course of Computer-Aided Design of Mechanical Systems was first introduced in 1987.
The first draft of the laboratory manual was used in the class in 1988. Since then the
hardware and software in the CAE laboratory, the curriculum and, of course, the laboratory
manual have gone through a great deal of upgrading and revision due to the ever-changing
world of computers. The current form of the laboratory and the manual has been achieved
as endless efforts of many former and current students; Chris Barton, David Coley, Jim
Dishman, Gopi Koganti, Amidu Oloso, Bryan Marz, Yang Wang, Arun Satyanarayana,
Srinivas Ravi, Daniel Hou, Kalyani Soman, Shantanu Ghate, Nick Huynh and my closest
colleague, Dr. Sebastian Bawab. Special thanks given to Timothy Watjen, Jonathan
Degroff and Jin Kim, particularly to Timothy who took the leadership to revise Chapter 3
in its entirety. His dedication and professionalism made the document more useful,
readable and enjoyable.

Special thanks to Dr. Vanderplaats of VMA Engineering and Mr. Price of Altair
Corporation whose donation of their first class software, GENESIS and HYPERMESH,
respectively, were invaluable in putting the curriculum together. Finally, it is also my
pleasure to acknowledge Tony Demato, Jamal Green and Mark Gilford and other staff
system engineers at University Computer Center whose tireless efforts have kept the
computer facility in the laboratory run smoothly.

Gene Hou
Professor

Revised August 2007

Revised August 2010
Revised August 2017

i
TABLE OF CONTENTS
PREFACE i

II NASTRAN STATIC ANALYSIS ......................................................................................................................................2.1

CONTENTS ......................................................................................................................................................................... 2.1
II.1 OBJECTIVE ................................................................................................................................................................. 2.2
II.2 INTRODUCTION ........................................................................................................................................................ 2.2
II.3 RUNNING NASTRAN .................................................................................................................................................. 2.3
II.4 CONCEPTS AND COMMANDS ................................................................................................................................... 2.4
II.5 EXAMPLES: ............................................................................................................................................................. 2.31
II.6 SUPPORT AND JOINT CONDITIONS ................................................................................................................................... 2.59
II.7 NASTRAN TO PATRAN .............................................................................................................................................. 2.64
II.8 NASTRAN TO PATRAN (NEW VERSION) : .DB AND .XDB GENERATION AND VISUALIZATION............................................ 2.66
II.9 FORCE VECTOR ....................................................................................................................................................... 2.68
II.10 HOMEWORK ......................................................................................................................................................... 2.74

III. MSC/PATRAN ........................................................................................................3.1

III.1 OBJECTIVE ............................................................................................................ 3.3
III.2 INTRODUCTION ................................................................................................... 3.3
III.3 LAB. SESSION 1: BUILDING A WIREFRAME MODEL USING POINT, CURVE & SURFACE .......... 3.9
III.4 LAB SESSION 2: FINITE ELEMENT MESH MODEL OF PLATE PROBLEM ............................. 3.28
III.5 LAB SESSION 3: ANALYSIS OF THE 2D PLATE PROBLEM................................................ 3.35
III.6 LAB SESSION 4: SPATIAL FRAME PROBLEM USING BAR STRUCTURE ............................... 3.54
III.7 - BUILT-UP STRUCTURES WITH RBE2 .......................................................................... 3.61
III.8 LAB SESSION 8: HOMEWORK - 45 DEGREE WEDGE FOR STATIC AND VIBRATION ANALYSES3.108
III.9 .DB AND .XDB GENERATION AND VISUALIZATION .............................................. 3.117
IV. NORMAL MODES ANALYSIS.......4.1
IV.1 OBJECTIVE.. ..4.1
IV.2 NTRODUCTION......4.1
IV.3 COMMANDS..........4.4
IV.4 EXAMPLES..........4.6
IV.5 NORMAL MODES ANALYSIS USING P3/PATRAN...4.8
IV.6 HOMEWORK....4.10
V. DYNAMIC TRANSIENT ANALYSIS ..5.1
V.1 OBJECTIVE.... .5.1
V.2 TIME DOMAIN ANALYSIS.....5.2
V.3 FREQUENCY DOMAIN ANALYSIS.....5.16
VI. ANALYSIS OF LAMINATED COMPOSITE PANELS..6
VI.1 INTRODUCTION.6.1
VI.2 MAT8 AND PCOMP CARDS .....6.2
VI.3 EXAMPLE PROBLEM ....6.8
VI.4 HOMEWORK ASSIGNMENT.....6.13

ii
 II NASTRAN STATIC ANALYSIS

Contents
II NASTRAN STATIC ANALYSIS ............................................................................................................... 1
CONTENTS ................................................................................................................................................... 1
II.1 OBJECTIVE .......................................................................................................................................... 2
II.2 INTRODUCTION .................................................................................................................................. 2
II.3 RUNNING NASTRAN ........................................................................................................................... 3
II.4 CONCEPTS AND COMMANDS ............................................................................................................. 4
NASTRAN Input Data File ..................................................................................................................... 4
Executive Control ................................................................................................................................. 5
Case Control ......................................................................................................................................... 5
Output Control ..................................................................................................................................... 6
Bulk Data ............................................................................................................................................. 8
II.5 EXAMPLES: ..................................................................................................................................... 31
Example 1: Cantilever Beam with CBARs ........................................................................................... 31
Example 2: Cantilever Beam with CQUADs ........................................................................................ 42
Reactions at the Single Point Constraints .......................................................................................... 49
Verification ........................................................................................................................................ 49
Example 3: Built-Up Structure with Offset ......................................................................................... 50
Example 4: Pin-Flag and Balance of Joint Internal Forces ................................................................. 52
Another Example ............................................................................................................................... 56
II.6 SUPPORT AND JOINT CONDITIONS ........................................................................................................... 59
Singularity .......................................................................................................................................... 59
St. Venant Principle ............................................................................................................................ 60
Symmetry ........................................................................................................................................... 61
Multiple Point Constraint ................................................................................................................... 62
Rigid Element Type 2 ......................................................................................................................... 62
II.7 NASTRAN TO PATRAN ....................................................................................................................... 64
II.7.1 Pre-Process ............................................................................................................................... 64
II.7.2 Post-Process .................................................................................. Error! Bookmark not defined.
II.8 NASTRAN TO PATRAN (NEW VERSION) : .DB AND .XDB GENERATION AND VISUALIZATION .................... 65
II.9 FORCE VECTOR ................................................................................................................................. 67
II.10 HOMEWORK ................................................................................................................................... 72
Homework 1 ...................................................................................................................................... 72

2.1
 Homework 2 ...................................................................................................................................... 72
Homework 3 ...................................................................................................................................... 73
Homework 4 ...................................................................................................................................... 73
Homework 5 ...................................................................................................................................... 74
Homework 6 ...................................................................................................................................... 75
Homework 7 ...................................................................................................................................... 75
Homework 8 ...................................................................................................................................... 76

II.1 OBJECTIVE
The objective of this brief manual is to familiarize the users with NASTRAN, a commercially
rated Finite Element Analysis (FEA) program. This document will deal primarily with static
analysis. The concepts and commands given here will contain most of the basic information
necessary for the use of NASTRAN. Students may refer to NASTRAN/PATRAN published
manual or tutorial materials for details.
II.2 INTRODUCTION
NASTRAN or NAsa STRuctural ANalysis was developed initially in 1966 under the sponsorship
of NASA based on known requirements of the aerospace industry for structural analysis. The
version used here, MSC/NASTRAN, has been enhanced and maintained by the MacNeal-
Schwendler Corporation which has been involved with NASTRAN since its inception.
MSC/NASTRAN is a general purpose FEA program capable of solving a wide variety of
engineering problems. The program is mainly written in FORTRAN and currently contains over
700,000 program statements. Some of the solving capabilities of MSC/NASTRAN include:
Linear static analysis
Static analysis with geometric and material nonlinearities
Transient analysis with geometric and material nonlinearities
Vibration and buckling analysis
Direct and modal complex eigenvalue analysis
Direct and modal frequency analysis and random response
Direct and modal transient analysis - including response spectral analysis
Linear static and vibration analysis with cyclic symmetry
Linear and nonlinear steady-state heat transfer
Transient heat transfer
Aero-elasticity
Multilevel super-elements
Design sensitivity
Optimization/resizing
2.2
 II.3 RUNNING NASTRAN
NASTRAN always uses an extension of a filename to differentiate between files containing data
and commands. For example,

filename.bdf The NASTRAN input file.

filename.f06 The NASTRAN output results file in text-format.
filename.log The NASTRAN file lists information about program execution.
filename.op2 The NASTRAN results file in binary format ( read by PATRAN during
translation).

NASTRAN usually assigns a filename by placing a period sign(.) between the filename and its
extension. For example, the extension, bdf, denotes the NASTRAN input file. Therefore, the
file, temp.bdf, indicates an NASTRAN input file with job name, temp. Once the input file,
temp.bdf, is constructed, it can be submitted to NASTRAN for analysis. Open Nastran from the
desktop. This will cause a window to open and will prompt for a *.bdf file. Once the file has
been chosen, select the Open command; the file temp.bdf will then be submitted for analysis.
Wait until NASTRAN finishes processing your job. Once done, the screen will indicate that the
NASTRAN job "temp" is completed. At this moment, NASTRAN should have produced three
new files; temp.f06, temp.f04, and temp.log. Among these, the most important one is temp.f06 (or
the f06 file) which lists the results of the NASTRAN analysis. One may open this file using
Notepad to check the output. The end of the file should find the statement, *** END OF
JOB***. Now take a look to the lines just above this statement. If there is no mention of the
word ERROR then your file, temp.bdf, has been successfully analyzed by NASTRAN and the
temp.f06 file contains the required output such as the displacements and stresses. This file can
be later used in viewing the results graphically (called post-processing) using PATRAN. If there
is an error, the f06 file often points out the source of error. Correct the error and re-submit for
NASTRAN analysis.

2.3
 II.4 CONCEPTS AND COMMANDS
The key element in finite element analysis is to discretize the distributed properties of a
structure into a finite number of simple substructures or elements. The physical properties such
as stiffness, inertia, conductivity, etc, are assigned to each of the elements. Grid points connect
these elements together, where point loads and constraints can be applied.

NASTRAN Input Data File

The NASTRAN input file (the bdf file) is made up of five sections. The first two sections are
optional. The required three sections are separated by the delimiters CEND, BEGIN BULK
and END DATA. The sequence of the input file structure is given as follows:
NASTRAN Definitions (Optional)
...
File Management Statements (Optional)
...
Executive Control Statements
...
CEND
Case Control Commands
...
BEGIN BULK
Bulk Data Entries
...
ENDDATA
CEND in the above file structure designates the end of the Executive Control Section, while
BEGIN BULK designates the beginning of the Bulk Data entries. The final line in the input
file must read ENDDATA.
The first optional section, NASTRAN Definitions, is only used for unusual actions. The second
optional section, File Management Statement provides means for database initialization and
management, along with job identification and restart conditions. The NASTRAN File
Management can be used to create a custom-made output file for graphic display of results that
are of interest to the user.

2.4
Executive Control
This section is used to select the type of analysis to be performed. It also controls other general
job functions such as maximum job time and analysis alterations to a DMAP. A DMAP (Direct
Matrix Abstraction Programming) allows a user to input user-defined matrices, manipulate the
data base and output the data in the form preferred by the user. Some executive control
statements are given below.

ALTER: Used to designate the addition or deletion or both of statements in an existing

SUBDMAP.

CEND: Used to designate the end of the Executive Control Section. This statement
is required unless an ENDJOB FMS statement appears in the Executive
Control Section.

COMPILE: Used to specify which NDDL, or SUBDMAP, to compile.

ECHO: Used to control print out of the Bulk Data deck.

ENDALTER: Used to designate the end of an ALTER.

ID: Used to designate a job-ID for a specific run.

SOL: Specifies the DMAP main program to be executed in this run.

TIME: Sets the maximum execution time of a job.

Case Control
The Case Control Section has several basic functions. Specifically,
it selects the loads and constraints,
it requests printing, plotting, and/or punching of Input and Output data ( Plotter
commands are described in Section 4 of the NASTRAN User's Manual ), and
it defines the subcase structure for the analysis.
The whole purpose of the case control section is to allow the repeated analysis of the same
structure subjected to various loadings and boundary conditions specified by the subcase

2.5
statements. Subcases are begun by the command SUBCASE. All items placed ahead of the
first subcase will be used for all following subcases unless overridden in the individual subcase.
For static analysis, it is possible to combine subcases. This is useful for studying different
loading combinations. Some of the Case Control commands are listed below:

SUBCASE n
Example: SUBCASE 2
Description: Delimits and identifies a subcase.

LOAD = n
Example: LOAD = 300
Description: Selects the external static load set to be applied to the structural model where n
corresponds to at least the ID number of a load entry in the Bulk Data section.

SPC = n
Example: SPC = 400
Description: Selects the single-point constraint set to be applied to the structural model.
Parameter n refers to the ID number of a constraint entry in the Bulk Data section. A single-point
constraint assigns a fixed value to a translational or rotational component at a geometric grid or
scalar point. NASTRAN will use the SPC conditions to eliminate unwanted degrees of freedom.
One may use PARAM, AUTOSPC, YES to eliminate all singularities found by the Grid
Point Singularity Processor (GPSP). However, this option must be used with caution as it may
cover up true design flaws. Another helpful parameter command is PARAM, K6ROT,
YES, which will eliminate the singularity of the CQUAD element due to lack of resistance to
the in-plane rotation.

MPC = n
Example: MPC = 500
Description: Selects the multipoint constraint set to be applied to the structural model. An
MPC is described by a linear equation involves more than one degree of freedom.

Output Control
TITLE = text
Example: TITLE = Analysis for a Cantilever Beam

2.6
Description: The text described in the TITLE line will appear on the first line of every output
page in f06 file. If placed inside a subcase, the title will appear for that subcase only.

SET n= a, b, c, THRU, s, EXCEPT, h, i, j, THRU, q

Example: SET 1 = 5, 6, 7, 8, 9, 10, THRU, 55
Description: Defines lists of point numbers, element numbers, frequencies or time steps. These
sets can be used with the Physical Set Output Requests.

DISPLACEMENT = n or ALL
Example: DISPL = 20
Description: Requests output of displacements for all or subset of physical points.

FORCE = n or ALL
Example: FORCE = 10
Description: Requests output of forces for all or subset structural elements.

STRESS = n or ALL
Example: STRESS = 30
Description: Requests output of stresses for all or subset of structural elements. Default output
is the Von Mises stress. The command, STRESS (MAXS) = n or ALL, requests output of
the maximum shear stress.

Notice that n in the above requests refers to the n in the SET definition. This could be used if, for
example, only the output of the displacement of a few nodal points was desired instead of the
entire model. Other useful output requests are:

GPSTRESS: Stresses at grid points

STRAIN: Strain for plate elements
GPFORCE: Force balance for a set of grid points
SPCFORCES: Single-point forces of constraint for a set of points
PRESSURE: Hydroelastic pressure for a set of points

You will be able to see in the examples that Case Control commands usually reference one or
more data entries in the Bulk Data section.

2.7
Bulk Data
The Bulk Data Section starts with the delimiter, BEGIN BULK. Bulk Data entries contain all
of the information necessary to define the model. This includes geometry, element connectivity,
material properties, constraints, loadings, etc. The user must specify the geometric location of
each grid point. This is normally done with respect to a global rectangular coordinate system.
Local coordinate systems can be used to define the grid point locations, which are directly or
indirectly related to the global system. The six possible local coordinate systems are CORD1R,
CORD2R, CORD1C, CORD2C, CORD1S and CORD2S. They are referred to as the
rectangular, cylindrical and spherical systems. The following sections explain in details the Bulk
Data entries.

Grid Points
Grid points are used to specify:
Model geometry,
Degrees of freedom at a point,
Locations of the constrained displacements,
The applied loads and
The desired output quantities.
Two coordinate systems are used in defining a GRID entry. The first coordinate system is used
to locate the grid point. The second coordinate system is used to define the components of the
output displacements and constraint reactions. Each grid point can move in six different ways,
called degrees of freedom. Each of the movements is represented by one of integers ranging from
1 to 6. The degrees 1, 2, and 3 correspond to translation in the x, y, and z directions, respectively;
i.e., u x , u y and u z . And the degrees 4, 5, and 6 correspond to rotation about the x, y, and z
directions, respectively; i.e., q x , q y and q z .

Bulk Data Entry: GRID- Grid PointDescription:

Defines the location of a geometric grid point of the structural model, the vector components of
its displacement, and its permanent single-point constraints.

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
GRID ID CP X Y Z CD PS SEID

2.8
Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
GRID 2 3 1.0 -2.0 3.0 316

or
GRID,2,,1.0,-2.0,3.0
This data card is written in the commonly used format, called the free format, in which a comma
is used to separate each of the data entries. This example Grid card states that the location of
Grid Point number 2 is (1.0, -2.0, 3.0 ), which is measured in terms of the global coordinate
system. The entry in column 8 indicates that the x- and z-translations and rotation about z-axis of
Node 2 are constrained.

Field Contents
ID Grid point identification number (1,000,000 > Integer > 0)

CP Identification number of coordinate system in which the location of the grid point is
defined (Integer 0 or blank; blank indicates that the global coordinate system is used
here.)

X,Y,Z Location of the grid point in the coordinate system CP (Real)

CD Identification number of the coordinate system by which the displacement vectors

and constraint reactions are measured. (Integer 0 or blank, blank indicates that the
global coordinate system is used here.)

PS Permanent single-point constraints associated with the grid point of concern (any of
the digits 1-6 with no imbedded blanks) (Integer>0 or blank)

SEID Superelement identification number (Integer>0 or blank)

2.9
Single Point Constraints
The function of the SPC has been discussed in the Case Control Section. The SPC entries in the
Bulk Data Section contain the data necessary to define how certain grids will be constrained.
Note that subcases can use two or more SPC entries to impose different constraining conditions
at the same grid point. However, only one SPC entry would be referenced at a time in a subcase.
It should be noted that SPC and SPC1 listed below are used as bulk data entries. Their use is
different from that in the Case Control command, SPC=n. The integer n in SPC=n has to be the
SID number of the SPC or SPC1 card in the Bulk Data deck. Only those SPC and SPC1
cards, whose SIDs are equal to n will be involved in the analysis.

Bulk Data Entry: SPC - Single-Point Constraint

Description:
Defines sets of non-homogeneous single-point constraints imposed upon a single degree of
freedom.

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC SID G C D G C D

Examples:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC 2 32 3 -
2.65

or
SPC,2,32,3,-2.6,5
The SPC card 2 imposes the condition that the z-displacement of node 32 is equal to 2.65.

2.10
Field Contents
SID Identification number of single-point constraint set ( Integer > 0)

G Grid or scalar point identification number (Integer > 0)

C Component number of global coordinate (6 Integer 0; up to six unique digits

may be placed in the field with no embedded blanks).

D Value of enforced displacement for all coordinates designated by G and C (Real)

Bulk Data Entry: SPC1 - Single-Point Constraint, Alternate Form

Description:
Defines sets of homogenous single-point constraints

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC1 SID C G1 G2 G3 G4 G5 G6
G7 G8 G9 etc..

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC1 3 2 1 3 10 9 6 5
2 8

Alternative Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC1 SID C GID1 THRU GID2

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
SPC1 313 12456 6 THRU 32

or

2.11
SPC1,313,123456,6,THRU,32
The SPC1 card number 313 states that the six degrees of freedom of nodes from 6 to 32 are
completely fixed.

Field Contents
SID Identification number of single-point constraint set (Integer > 0)

C Component number of global coordinate (any unique combination of the digits 1-

6 with no embedded blanks, when point identification numbers are grid points;
must be null if point identification numbers are scalar points)

Gi,GIDi Grid or scalar point identification numbers (Integer > 0)

Note that SPC1 is able to define constraints for more grids per entry than SPC is. Thus, it is
more generally used. SPC is needed only when the enforced displacement is not zero. Linear
combination of various boundary conditions can be imposed by the SPCD entry. Details of
SPCD can be found in the NASTRAN User Manual.

Multi-point Constraint
Bulk Data Entry: MPC - Multipoint Constraint

Description:
The MPC card defines a linear constraint involving more than one degree of freedom. A
multipoint constraint equation can be expressed in the form

A u
j
j j =0

where uj represents the involved degree of freedom at grid or the scalar point Gj.

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MPC SID G1 u1 A1 G2 u2 A2 blank
G3 u3 A3 -etc.-

2.12
Example 1:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MPC 3 28 3 6.0 2 4 -4.0
1 5 -2.5
or
MPC, 3, 28, 3, 6.0, 2, 4, -4.0, ,,ABC
+BC, 1, 5, -2.5

In the above example, A1=6.0, A2=-4.0 and A3= -2.5, while u1 is u z at node 28, u2 is q x at node 2
and u3 is q y at node 1. The example MPC card implies that
6.0u1-4.0u2-2.5u3 = 0

Field Contents
SID Set identification number (Integer > 0)

Gj Identification Number for the Grid-point or the scalar Point. (Integer > 0)

Cj Component numbers (Any one of the integers 1 thru 6 that represents the translational
and rotational components for grid-points; blank or zero for the scalar points.)

Aj Coefficient (Real; Default = 0.0 except A1 must be non-zero)

Remarks
1. Multipoint Constraint sets must be selected with Case Control command MPC= n.
2. The first degree of freedom (G1, C1) in the sequence is defined to be the dependant
degree of freedom. A dependent degree of freedom assigned by one MPC entry cannot be
assigned as a dependent one by another MPC entry or by rigid element.
3. The m-set degrees of freedom specified on this entry may not be specified on the other
entries that define mutually exclusive sets.

Example:
A crank is modeled as a beam connected nodes 1 and 3. The node 3 represents a slider that can
only move parallel to the surface AB. Therefore, the component of the resultant displacement at
node 3 that is normal to the surface AB has to be zero. Thus, u x sin q + u y cosq = 0 , where q

2.13
is the angle measured from the surface AB to the x-axis. If q is set to be 300, the MPC
constraint is then expressed as 0.5u x + 0.866 u y = 0 and the corresponding MPC card is stated
as MPC,2,3,1,0.5,3,3,0.866.

N
A
Uy

B
Y
Ux
X

Figure 2.1 MPC at the Slider Joint

Loads
There are many different types of loading entries that can be used for the static loading case. The
most commonly used entry is FORCE. This entry simply applies a point force at a specified grid.

Bulk Data Entry: FORCE - Static Load

Description:
Defines a static load at a grid point

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FORCE SID G CID F N1 N2 N3 AXI

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FORCE 2 5 2.9 0.0 1.0 0.0

or
FORCE,2,5,,2.9,0.0,1.0,0.0

2.14
It states that a force with a magnitude of 2.9 units being applied in the positive y-direction at grid
point 5. Note that the CID entry is blank which indicates that the force vector is measured in
terms of the global coordinate system.

Field Contents
SID Load set identification number (Integer > 0)

G Grid point identification number (Integer > 0)

CID Coordinate system identification number (Integer 0, or blank; Default = 0)

F Scale factor (Real)

N1,N2,N3 Components of the force vector measured in the coordinate system defined by
CID (Real; must have at least one non-zero component)

AXI Indicates an axisymmetric loading (BCD: "AXI" or blank)

Bulk-Data Entry: MOMENT - Static torque or twisting moment.

Description:
Point torsion or twisting moments applied at a grid point

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MOMENT SID G CID M N1 N2 N3

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MOMENT 2 5 2.9 0.0 1.0 0.0
or
MOMENT,2,5,,2.9,0.0,1.0,0.0

2.15
It states that a moment with a magnitude of 2.9 units being applied in the positive y-direction at
grid point 5. Note that the CID entry is blank which indicates that the moment vector is measured
in terms of the global coordinate system.

Field Contents
SID Load set identification number (Integer > 0)

G Grid point identification number (Integer > 0)

CID Coordinate system identification number (Integer 0, or blank; Default = 0)

M Scale factor (Real)

Remarks:
1. The static force and the moment applied to a grid-point are given by
! ! ! !
f = FN m = MN
!
where N is the direction vector defined in the fields 6, 7 and 8 and F and M are the
magnitudes of the force and the moment, respectively.
2. In the static solution sequence, the load set ID must be selected by the Case Control
command LOAD= SID.
3. A CID of zero or blank (default) references the global coordinate system.

The following is a summary of loading types and their corresponding Bulk Data entries.
Type of LOAD Bulk Data Entries

Forces applied directly to Grid points FORCE, FORCE1, FORCE2

Moments applied directly to Grid points MOMENT, MOMENT1, MOMENT2
Loads on scalar points SLOAD
Loads on line elements PLOAD1
Pressures and Tractions on Surfaces PLOAD, PLOAD2, PLOAD3,
PLOAD4, PLOADX
Gravity GRAV (Plus Specification of Mass)
Centrifugal Force RFORCE (Plus Specification of Mass)

2.16
 Thermal Expansion TEMP, TEMPD, TEMPP1, TEMPP3, TEMPRB
(Plus Specification of Thermal Expansion Coefficients)
Enforced Extensional Deformation DEFORM elements (Except BEND)
Enforced Displacement at Grid Points SPCD, SPC, GRID
Linear Combination of Load Sets. LOAD
Transient Load TLOAD1

The user must understand that a comprehensive explanation of the above load statements is
beyond the scope of the manual. However, the user can obtain sufficient information regarding
their usage in Volume I of the NASTRAN User's Manual.

Material Properties
Definitions of various structural and thermal properties of isotropic, anisotropic and orthotropic
materials are available in NASTRAN. Definitions are also available for plastic and fluid
materials. The most commonly used entry for STATIC analysis is the MAT1 entry.

Bulk Data Entry: MAT1 - Material Property Definition, Form 1

Description:
Defines the material properties for linear, temperature-independent, isotropic materials

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MAT1 MID E G NU RHO A TREF GE
ST SC SS MCSID

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MAT1 17 3.e+7 0.33 4.28
20e+4 15e+4 12e+4
or
MAT1,17,3.e+7,,.33,4.28,,,,ABC
+BC,20e+4,15e+4,12e+4

2.17
The Youngs modulus of the material is 3.e+7 units, Poissons ratio 0.33 and mass density 4.28.
The allowable tensile stress, the allowable compressive stress and the allowable shear stress are
20e+4, 15e+4 and 12e+4 units.

Field Contents
MID Material identification number (integer >0)

E Youngs Modulus (Real or Blank)

G Shear Modulus (Real or Blank)

NU Poissons Ratio (-1.0 < Real 0.5 or blank)

RHO Mass density (Real)

A Thermal Expansion Coefficient.(Real)

TREF Reference temperature for the calculation of a) thermal loads, or b) a temperature-

dependant thermal expansion coefficient.

GE Structural element damping coefficient.

ST,SC,SS Stress Limits for tension, compression, and shear (Real). Used only to compute
the margins of safety in certain elements; they have no effect on the
computational procedures. NOTE: Margin of Safety = Factor of Safety - 1.

MCSID Material coordinate system identification number (used only in CURV

module processing; Integer 0 or blank)

Remarks:
1. The material identification number must be unique for all MAT1, MAT2, MAT3,
and MAT9 entries.
2. The following rules apply to the values of E, G, and NU .
E and G may not both be blank.
2.18

If NU and E, or NU and G, are both blank, then both are set to 0.0.

If only one E, NU, or G is blank, then it will be computed from the equation:
E
G=
2(1 + n )
Filed 2 or MID is referenced by an element property card, which defines the geometric or
sectional properties of a structural element.

Structural Elements
There are many different types of elements available in NASTRAN. Only the connection and
property entries for the 1-D bar element and the 2-D quadrilateral element are discussed here.

Bulk Data Entry: CBAR - Simple Beam Element Connection

Description:
Defines a simple beam element (BAR) of the structural model

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CBAR EID PID GA GB X1,G0 X2 X3
PA PB W1A W2A W3A W1B W2B W3B

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CBAR 2 39 7 3 1. 0. -1.
513

or
CBAR,2,39,7.0,3.0,13,,,,ABC
+BC,,513
The beam element 2, whose sectional properties will be defined by PBAR card number 39,
!
connects Nodes 7 to 3. The orientation vector V , is a position vector of (1.0, 0., -1. ) originated
at node 7 and measured in terms of the global coordinate system. Furthermore there is no axial

2.19
force, z-direction shear force and y-direction bending moment transmitted into the beam element
through Node 3.

Field Contents
EID Unique element identification number (Integer > 0).

PID Identification number of a PBAR entry (Integer > 0)

GA, GB Grid point identification numbers of connection points. (Integer > 0;

GA GB).

!
X1, X2, X3 Components of the vector, V , originated from GA and measured in terms of
the global coordinate system (Real).

!
G0 Alternate method to supply the orientation of vector, V , by connecting GA to
the grid point G0. (Integer > 0 ).

PA, PB Pin Flags for the bar ends GA and GB, respectively, used to remove the rigid
connections at the common grid points joining the bar elements along the
specified degrees of freedom. The degrees of freedom are defined in the
element coordinate system. (See Figure 2.2). The bar element of concern must
have non-zero stiffness associated with pin flag degree of freedom. For
example, if PA=4 is specified the PBAR entry, the beam must have a non-zero
value for the torsional stiffness. (up to 5 of any integers 1 through 6 with no
embedded blanks; Integer > 0).

! !
W1A,W2A,W3A Components of the offset vectors W A and W B , respectively (see
W1B,W2B, W3B in Figure 2.2) in the displacement coordinate systems at
point GA and GB, respectively. (Real or blank)

Remarks:
1. Element identification numbers must be unique with respect to all other element
identification numbers.
!
2. Figure 2.2 defines the bar element geometry. The orientation vector, V , should be
specified by users to generate a local coordinate system (Xe, Ye, Ze), for any beam
2.20
 element, with which the moments of inertia, I1, I2 and I12 and the locations of the stress
recovery points (C1, C2), (D1, D2), (E1, E2) and (F1, F2), defined by PBAR card, can be
determined. Here is how a local coordinate system, Xe, Ye and Ze be defined with help
!
of vector, V ;
i) Xe-axis is defined by the vector connecting grid point GA to grid point GB
ii) Ze-axis is parallel to the result of cross product of Xe-axis
! ! ! !
and V ;i.e., Z e X e Ve
iii) Ye-axis is then defined as the direction normal to Xe Ze Plane.
3. The continuation may be omitted if there are no pin flags or offsets.
4. For additional information please refer to the NASTRAN User's Manual Section 1.2.1.

Orientation Vector V
Xe

End B
Ye Z
Plane 1
Offset Wb Y
qy

Grid Point B X
qx

2 1
Plane 2
End A

Ze
Z
qz
Offset W a
3 4
Y
X
Grid Point A
Fig. 2.2(a) Coordinate System of a Beam (CBAR) Element

2.21
 ye
Z
Y
P1
xe X
GA
P2 GB
Global System

ze
Fig. 2.2(b) Local and Global Coordinate Systems

Bulk Data Entry: PBAR - Defines the property of a simple beam element (CBAR entry).

Description:
Defines the sectional and material properties of a simple beam (CBAR), which is used to create
bar elements and stress output points

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PBAR PID MID A I1 I2 J NSM
C1 C2 D1 D2 E1 E2 F1 F2
K1 K2 I12

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PBAR 39 6 2.9 2.4 5.97 1.75
2.0 4.0

or
PBAR,39,6,2.9,2.4,5.97,1.75,,,ABC
+BC,,,2.0,4.0
The material property of the PBAR card number 39 is referred to MAT1 card number 6. The
cross-sectional area of the bar is 2.9 units, the moments of inertia, I1 and I2, are 2.4 and 5.97

2.22
units, respectively. The torsional rigidity of the bar is 1.75 units. The stresses are reported at
point (2.0,4.0) according to the local Ye-Ze axes on the cross sections at both ends of the bar.

Field Contents

PID Property Identification Number (Integer > 0)

MID Material Identification number. (Integer > 0)

A Area of bar cross section. (Real)

2
I1, I2, I12 Area moments of Inertia. (Real I1 0.0 , I2 0.0 , I1* I2 > I12 ), (I1=IZeZe,
I2=IYeYe)

J Torsional Constant. (Real)

NSM Non-structural mass per unit length. (Real)

K1, K2 Area factor for shear in planes 1 and 2. See Remark 3. (Real)

Ci, Di, Ei, Fi Stress Recovery Coefficients where stress values at theses locations are displayed.
(Real; Default = 0.0).

Remarks:
1. For structural problems, PBAR entries may only reference MAT1 material entries.
2. For heat transfer problems, PBAR entries may reference MAT4 or MAT5 material
entries.
3. I1 and I2 are the bending moments of inertia with respect to the local coordinates, Ze and
Ye axes, respectively. I12 defines the product inertia, IYeZe.
4. The torsional rigidities, J, of common sections are given in Table 2.1 in which
q = M JG . Alternatively, one may approximate J of a solid cross-section as
0.025 A 4
J where A is the cross-sectional area and Ip is the polar moment inertia;
Ip
4W 2t
I p = I 11 + I 22 , and of a closed single-celled, thin walled section, J where t is the
S
thickness of the wall, W is the area enclosed by the wall and S is the circumferential
2.23
 length of the section. For various open thin-walled sections, additional formulae for the
torsional rigidity are given in Fig. 2.3.

Figure 2.3 Torsional Rigidity for Open Sections

5. The transverse shear stiffness per unit length in plane1 and plane 2 (shown in Figure 2.2) are
computed by K1 A G and K 2 A G , respectively, where G is the shear modulus, A is the
cross-sectional area and K1 and K2 are called shear factors. The default values for K1 and K2
are infinite; in other words, the transverse shear flexibilities are set equal to zero. The effects
of K1 and K2 are ignored if I120.

6. The stress recovery coefficients C1, C2, D1, D2 etc are the Ye and Ze coordinates in the bar
element coordinate system of a point at which the stresses are computed. The stresses are
computed at both ends of the bar element. At most, 4 stress values can be reported at each
end of the bar element. Particularly, C1, D1, E1 and F1 are coordinates of stress recovery
point; C, D, E and F in Ye axis and C2, D2, E2 and F2 are coordinates of the same point, in
Ze axis.

Table 2.1 Torsional Rigidity for Commonly Used Sections

2.24
 Ye
F2
E F
F1

Ze

C D

2.25
Figure 2.4 Stress Recovery Points at the Four Corners of the Section.

7. The Shear displacement yv of a beam is yv = VL / KAG where ( 1 / KAG ) is the shear flexibility
of the beam, V is the transverse shear load in the element z direction, L beam length, K shear
factor whose value depends on the geometry of the cross section, A is beam cross sectional
area, G is beam shear modulus that NASTRAN uses two values , K1 and K2 to define shear
factors of a beam with a nonsymmetrical cross-section. Value K1 contributes to shear
resisting transverse force in the element Ye axis direction and value K2 contributes to shear
resisting transverse force in the element Ze axis direction. The values of K1 and K2 are
infinite by default, i.e. beam is infinitely rigid in shear. The value ok K is equal to 1/F where
the value of F is given in the following table.

Table 2.2 F Factors for Shear Effects

Shape Cross Section Value of F

Rectangular 6/5
Solid Circular 10/9
Thin Walled Hollow Cylinder 2
I beam or Box Section A/Aweb

or F = 1 +
( )
3 D22 - D12 t 2 4 D22
- 1
2
2 D23 t1 10r

where D1 is the distance from the neutral axis to the nearest surface of the flange, D2 is the
distance from the neutral axis to the extreme fiber, t1 is thickness of the web (or webs in box
beams), t2 is the width of the flange and r is the radius of gyration of section with respect to the
neutral axis.

Bulk Data Entry: CQUAD4 - Quadrilateral Element Connection

Description:
Defines a quadrilateral plate element (QUAD4) of the structural model. This is an isoparametric
membrane-bending element.

2.26
Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CQUAD4 EID PID G1 G2 G3 G4 THETA/MCID ZOFFS
T1 T2 T3 T4

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
CQUAD4 111 203 31 74 75 32 2.6 0.3
1.77 2.04 2.09 1.80

or
CQUAD4,111,203,31,74,75,32,2.6,0.3,ABC
+BC,,,1.77,2.04,2.09,1.80

Field Contents
EID Unique element identification number (Integer > 0).

PID Identification number of a PSHELL entry (Integer > 0 or blank)

G1,G2,G3,G4 Grid point identification numbers of the connection points. (integers >0,
all unique)

THETA Material property orientation angle in degs. (Real; Default = 0)

MCID Material coordinate system identification number. The X-axis of the

material coordinate system is determined by projecting the X-axis of the
MCID coordinate system (defined by the CORDij entry or zero for the
basic coordinate system) onto the surface of the element. See Fig. 2.5.
(Integer 0; if blank then THETA = 0.0 is assumed).

ZOFFS Offset from the surface of the grid points to the element reference plane.

2.27
T1,T2,T3,T4 Membrane thickness of element at grid points G1 through G4. (Real
0.0 or blank, not all zero.)

y element

G3

x element
x material
G4 a
b +g
a q a = -----
2

zelement g
b
G2
G1
Fig. 2.5: Coordinate System for CQUAD Element

Remarks:
1. Element identification numbers must be unique with respect to all other element
identification numbers.
2. Grid points G1 through G4 must be ordered consecutively around the perimeter of the
element.
3. All the interior angles must be less than 180 degrees.
4. The continuation or the second line in the CQUAD4 entry above is optional. If it is not
supplied, the T1 through T4 will be set equal to the value of T on the PSHELL entry.
5. The stresses are output in the element local coordinate system.
6. The elements maybe offset from the connection points by means of ZOFFS. Other data,
such as material matrices and stress fiber locations, are given relative to the reference plane. A
positive value of ZOFFS implies that the element reference plane is offset a distance of ZOFFS
along the positive Z-axis of the element coordinate system. The use of ZOFFS will result in the
incorrect resullts in buckling analysis and differential stiffness. If the ZOFFS field is used, then
MID1 and MID2 fields must be specified on the PSHELL entry specified by the PID.

2.28
Bulk Data Entry: PSHELL - Shell Element Property

Description:
Defines the membrane, bending, transverse shear, and coupling properties of thin shell
elements.

Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PSHELL PID MID1 T MID2 12I/T3 MID3 TS/T NSM
Z1 Z2 MID4

Examples:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PSHELL 203 204 1.90 205 1.2 206 0.8 6.32
+.95 -.95

or
PSHELL,203,204,1.90,205,1.2,206,0.8,6.32,ABC
+BC, +0.95,-0.95

Field Contents
PID Property identification number (Integer >0)

MID1 Material identification number for the memrane (Integer 0 or blank)

T Default value for the membrane thickness. (Real) (For a membrane that is
uniform in its thickness )

MID2 Material identification number for bending. (Integer or blank)

12I/T3 Bending stiffness parameter. (Real >0.0; Default = 1.0)

2.29
MID3 Material identification number for transverse shear. (Integer 0 or blank
unless MID2 > 0)

TS/T Transverse shear thickness divided by the membrane thickness. (Real > 0.0;
Default = .833333)

NSM Non structural mass per unit area. (Real).

Z1, Z2 Fiber distances for the stress calculations. The positive direction is determined
by the right hand rule and the order in which the grid points are listed on the
connection entry. See Remark 9 for defaults. (Real or blank)

MID4 Material identification number for membrane-bending coupling. See Remark

6 and 13. (Integer > 0 or blank, must be blank unless MID1>0 and MID2
>0, may not equal MID1 or MID2).

Remarks:
1. All PSHELL properties must have unique identification numbers.
2. The structural mass is calculated from the density using the membrane thickness and
membrane material properties.
3. The results of leaving the MID field blank (or MID2 = -1) are :
MID1 No membrane or coupling stiffness.
MID2 No bending, coupling, or transverse shear stiffness.
MID3 No transverse shear flexibility.
MID4 No membrane coupling unless ZOFFS is specified on the
connection entry. See remark 6 .
NOTE: MID1 and MID2 must be specified if the ZOFFS field is also
specified on the connection entry.
4. The continuation entry is not required.
5. The structural damping (for dynamic solution sequences) uses the values defined for the
MID1 material.
6. The MID4 field should be left blank if the material properties are symmetric with respect
to the middle surface of the shell. If the element centerline is offset from the plane of
the grid points but the material properties are symmetric, the preferred method for
modeling the offset is by use of the ZOFFS field on the connection entry. Although

2.30
 MID4 may be used for this purpose, it may produce ill conditioned stiffness matrices
(negative terms on the factor diagonal) if done incorrectly.
See the NASTRAN User's Manual Section 1.3.5 for further information.
7. For structural problems, the PSHELL entries may reference MAT1, MAT2, or MAT8
material property entries.
8. For heat transfer problems, the PSHELL entries may reference the MAT4 or MAT5
material property entries.
9. The default for Z1 is -T/2 and for Z2 is +T/2. T is the local plate thickness defined by T
on this entry or by membrane thicknesses at the connected grid points, if they are input
on the connection entries.
10. For plane strain analysis, Set MID2 = -1 and set MID1 to reference a MAT1 entry.

II.5 EXAMPLES:
In the following examples, the entire input file including the Executive Control section, the Case
Control section and the Bulk Data section will be presented. For more complicated geometries,
modeling software, such as PATRAN, will be used to create the model. Then, using a translator,
the .bdf file will be automatically generated and ready for analysis.

Example 1: Cantilever Beam with CBARs

In this example, , a cantilever beam of a rectangular section is subjected to a single 10 lbs of load
applied at the free end, along the Z-direction. The length of the beam is 72 inches. The Youngs
modulus and the Possions ratio of the beam are 10 10 6 psi and 0.33, respectively. Other
dimensions of the beam are shown in the figure. The entire data deck of the nastran bdf file is
shown below as well. We will model a cantilever beam by placing four BAR elements. Subcases
are used here to consider the responses under different sets of loadings. The beam to be modeled
is shown below.
z
F z
y
2"
1 2 3 4 x y
4"
18" 18" 18" 18"
1 2 3 4 5

2.31
Fig. 2.6 First Beam Example
Load on the bar:
10 lbs., 20 lbs. and 30 lbs. are applied separately at node 5.
Model:
To model the beam we will use four 18-inch BAR elements connected end-to-end. Open a new
file, and name it beam1.bdf using the EDIT utility. Enter the following data in this file.
$ EXECUTIVE CONTROL SECTION BEGIN BULK
ID MEM441, BEAM1BDF $ BULK DATA SECTION
SOL 101 SPC1,100,123456,1
TIME 10 FORCE,100,5,,10.,0.,0.,-1.
CEND FORCE,200,5,,20.,0.,0.,-1.
$CASE CONTROL SECTION FORCE,300,5,,30.,0.,0.,-1.
TITLE=CANTILEVER BEAM USING BAR ELEMENTS GRAV, 14, , 386.14, 0., 0., -1.0
DISPL=ALL LOAD, 400, 1.0, 1.0, 14, 1.0, 300
STRESS=ALL SPC1,200,246,2,3,4,5
SPC=100 SPC1,200,123456,1
SUBCASE 1 SPC, 200, 5, 3, -0.04
LOAD=100 GRID,1,,0.,0.,0.
SUBTITLE=10 LB. LOAD CASE GRID,2,,18.,0.,0.
SUBCASE 2 GRID,3,,36.,0.,0.
SUBTITLE=20 LB. LOAD CASE GRID,4,,54.,0.,0.
LOAD=200 GRID,5,,72.,0.,0.
SUBCASE 3 CBAR,1,100,1,2,1.,0.,1.
SUBTITLE=30 LB. LOAD CASE CBAR,2,100,2,3,1.,0.,1.
LOAD=300 CBAR,3,100,3,4,1.,0.,1.
SUBCASE 4 CBAR,4,100,4,5,1.,0.,1.
SUBTITLE = GRAVITATIONAL LOADING CASE PBAR,100,10,8.,10.667,2.667,7.676,,,123
LOAD=400 +23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
SUBCASE 5 +56,.8333,.8333
SUBTITLE = PRESCRIBED DISPLACEMENT PROBLEM MAT1,10,3.+7,,0.33
SPC = 200 ENDDATA

Input Data Explanation:

Executive Control Section:

The DMAP, or solution sequence, used is SOL 101 for STATICS. TIME 10 implies that the job
is set to terminate if the solution is not reached in 10 CPU minutes. This job should take less than
one minute.

Case Control Section:

Outputs requested for all subcases are the displacements at all grids and the stresses in all
elements. This explains the statements DISPL = ALL and STRESS = ALL. The same single-
point constraints will be used for all subcases. Here the two SPC1 entries in the bulk data section
whose ID numbers are 100 are referenced. Each subcase refers to a different FORCE entry. For
example, in SUBCASE 3 of the Case Control Section, LOAD=300 references the

2.32
FORCE,300 entry in the Bulk Data Section. Unlike the first three subcases, the fourth
subcase refers to the LOAD entry in the Bulk Data Section. Thus, the LOAD = n, entry in the
Case Control Section can refer to any static loading condition defined in the Bulk Data Section.
In the last subcase, an enforced displacement solution is sought. In this subcase, the SPC card
200 is selected from the Bulk Data Section. The previous SPC number (= 100) will be ignored.

Bulk Data Section:

GRID:
The second field of each GRID entry gives the grid ID. The third field, which is empty, gives a
second coordinate system ID. The last three fields give the x, y and z global coordinates of the
grid.

FORCE:
The second field of each FORCE entry gives a unique ID number, one for each subcase. The
third field gives the grid to which the force is applied. The fourth field is a coordinate system ID.
Here, the default system which is the global coordinate system is used. The fifth field is the force
magnitude. The last three fields give the components of the force vector.

GRAV:
Used to define the gravitational forces applied to the structure. The second field of the GRAV
entry gives a unique ID to the entry. This GRAV entry, with an ID = 14, defines a gravitational
acceleration constant of 386.14 units being applied in the negative z direction.

LOAD:
Used to combine two or more load case entries. It must be used to apply the gravitational force.
The LOAD card here is combining the gravitational load and the FORCE entry 300. The
multiplication factors for both the gravitational force and the force entry are 1.0. This means that
the forces defined by the FORCE entry 300 and the GRAV card are multiplied by 1 and then
added together. The load equation is of the form
p = S (S i PLi )
i

where S = 1.0, S1 = 1.0, PL1 = 14, S2 = 1.0, PL2 = 300.

SPC1:

2.33
The second field of the SPC1 entry is the ID referred to in the case control section. The third
field gives the degrees of freedom that are to be constrained. The fields, which follow, give the
grid numbers to which the SPC applies. Notice that grid one is constrained in all six degrees of
freedom making the end fixed. These are directions in which no displacement or rotation takes
place. All of the other grids have not been constrained.

SPC:
Used to specify displacements on degrees of freedom. This card here is different from SPC1,
which specifies only zero displacements. In this example, a displacement of 0.04 units is
specified on node 5 in the z-direction.

CBAR:
The second field of the CBAR entry is the ID for the BAR. The third field refers to the ID
number of a PBAR entry. The fourth and fifth fields give the grid-points that are the two ends
of the BAR. The last three entries are the components of a vector, which defines a local
coordinate system for each bar. This local coordinate system is needed to define moments of
inertia, shear factors and points on the cross section for stresses to be output. The figures in
Fig. 2.2 shows how the local coordinate system is set up. Grid points A and B are shown at offset
positions and with local coordinate systems defined. The local coordinate system for the BAR
element is shown by xe, ye and ze. The CBAR element geometry is defined as follows:

(1) Element X-axis: Always defined positive from GA to GB as given on the CBAR
entry.
!
(2) Element X-Y plane: Determined by V which lies in the xe - z e plane. Plane1 is always
! !
the xe - z e plane.
! !
(3) Element Z-axis: Cross-product xe V . Plane 2 is always the xe - z e plane.

!
The vector V is defined with respect to the global coordinate system. In our example, the last
three fields of the CBAR entries define the element coordinate system as shown below. Notice
that field three of the CBAR entries refers to the PBAR entry.

PBAR:
The PBAR entry defines all the geometric properties for the BAR element. Field three
references the material property entry MAT1. Field four is the cross-sectional area. Fields
five and six give the area moments of inertia for bending in plane one and plane two,
respectively. In the example, these planes are defined by the element local coordinate system.
2.34
Notice that the PBAR entry is continued for two more lines. The first continued line gives the ye
and ze coordinates of points on the BAR cross-section where stress outputs are desired. The
coordinates of the stress recovery points, given as C, D, E and F, are given in Fig. 2.7. The next
continued line gives the area factors for shear in planes one and two. The continuation entries are
in field one of the continued lines. These fields contain the symbol + followed by the same seven
characters found in Field 10 of the continuing line. (Note: Each field can contain only eight
characters. The first character in Field 10 of the line to be continued is ignored).

MAT1:

In the example, a Young's modulus of 3 x 107 psi and a Poisson's ratio of .33 are used. Shear
modulus is left out and will automatically be calculated when the program is run as explained in
the MAT1 entry earlier.

After entering all of the data, check to make sure that all real number data fields have decimal
points. Degrees of freedom and ID numbers are integers and do not need decimals.

Z,Ye
C D

Y
Ze
4

F 2.35 E
2
Figure 2.7 Stress Recovery Points

NASTRAN Execution:

To execute your file in NASTRAN, open the MSC Nastran program from the desktop. You will
then be prompted to open a *.bdf file. Choose the location where you have beam1.bdf saved,
select the file and select Open. After a few moments Nastran will generate several files in the
same location. The one of interest to you is the *.f06 file. This file will contain all the calculated
data that was requested in the *.bdf file.

OUTPUT:

Epsilon & External Work

The first piece of output information is epsilon and external work. The quantity external work
is the work done by the external load. External Work terms greater than 500 could indicate
incompatible stiffnesses in adjacent regions of a model. Epsilon is the work done by the
unbalanced forces. When this term is greater than 10-6 it could mean that elements are poorly
modeled.

Bar output:

Displacement
For the displacement results vector columns, T1, T2 and T3 give translation displacements
in the x, y and z directions respectively. Quantities R1, R2 and R3 give rotation displacements
about the x, y and z directions, respectively. Notice that there is one set of output for each
subcase.

Stress

Columns 2:5
The first four columns of the output for stresses in bar elements give normal stresses at the four
points on the bar cross-section defined in the PBAR entry. There are two normal stress values
given for each BAR. These stresses are at grid A and grid B of each BAR element. For

2.36
example, in this problem, SA1 and SB1 are the stresses at point C on the cross-sections at grids
A and B of each element.
Columns 6,7
The sixth and seventh columns give maximum and minimum stresses. These are combined
bending and axial stresses.
Column 8
The eighth column gives the margin of safety for tension and compression. This defaults to one
unless stress limits are specified on the material property entry. If margin of safety is less than
zero, then failure has occurred. Margin of safety is defined as the difference between the ratio of
s
the calculated stress and the yielding stress and 1; i.e., M .S . = - 1.
s yd
Specifically, the bending stresses are reported at 4 stress recovery points at each of the grid
points at the end of the beam element are calculated based upon the internal bending moment and
the axial force at the grid point of the concern. The formula for the axial stress is given below,
which is caused by the internal bending moment, M , and the internal axial force, P,
P Mc
s xe =
A I
where A, c and I denote the bending moment, the fiber distance measured from the centroid of
the beam section to the stress recovery point and the moment of inertia, respectively. Note that
the magnitudes or the signs of P, M, c and I depend upon the local beam axis system. An
example output is given in Table 2.2.

Table 2.2 The Stress Output for Case 1

ELEMENT SA1 SA2 SA3 SA4 AXIAL SA-MAX SA-MIN M.S.-T
ID. SB1 SB2 SB3 SB4 STRESS SB-MAX SB-MIN M.S.-C
1 2.699662E+02 -2.699662E+02 -2.699662E+02 2.699662E+02 0.0 2.699662E+02 -2.699662E+02 1.1E-01
2.024747E+02 -2.024747E+02 -2.024747E+02 2.024747E+02 2.024747E+02 -2.024747E+02 1.1E-01

As shown in the table, the stresses are reported in two rows for Element 1. The first row data
are associated with the stresses at the grid point GA and the second the grid point GB. The
bending stresses at the four stress recovery points are reported in the first four columns of the
output in an order corresponding to C, D, E and F. That is, SA1 is the bending stress at C of grid
point GA. The axial stress is reported in the fifth column. Among the four stress outputs at a grid
point, the maximal and the minimal values are reported at the sixth and the seventh columns,
respectively. For example, the values under SA-MAX and SA-MIN give the maximal and the

2.37
minimal axial stresses at the grid point GA. The
margins of safety under tension and compression are
reported at the last column, which is calculated by
Note that the values of the Stress Limit
for Tension and Compression are input
M .S - T = ( Stress Limit for Tension / Max Tension
in through ST and SC entries in the
Stress)-1
MAT1 card.

M .S - C = ( Stress Limit for Compression / Max Compression Stress)-1

As an illustrative example, the cantilever beam of a rectangular section given in Example 1 is

repeated here. The magnitude of the load is limited to 10 lbs. The length of the beam is 72
inches. The Youngs modulus and the Possions ratio of the beam are 10 10 6 psi and 0.33,
respectively. Other dimensions of the beam are shown in Figure 2.5. The local coordinates of
three different definitions of C, D, E and F selected for the NASTRAN runs. The coordinate
values of the stress recovery points in terms of the local coordinates are tabulated in Table 2.3.

Table 2.3 Local Coordinates of Stress Recovery Points in inches

Cases C1 C2 D1 D2 E1 E2 F1 F2
1 2. 1. 2. -1. -2. -1. -2. 1.
2 2. 0. 0. -1. -2. 0. 0. 1.
3 0. 2. -1. 0. 0. -2. 1. 0.

The NASTRAN analyses were then conducted for the stress recovery points listed in Table 2.3.
The results are tabulated in Tables 2.2, 2.4 and 2.5.

Table 2.4. The Stress Output for Case 2

ELEMENT SA1 SA2 SA3 SA4 AXIAL SA-MAX SA-MIN M.S.-T

ID. SB1 SB2 SB3 SB4 STRESS SB-MAX SB-MIN M.S.-C
1 0.0 -2.699662E+02 0.0 2.699662E+02 0.0 2.699662E+02 -2.699662E+02 1.1E-01
0.0 -2.024747E+02 0.0 2.024747E+02 2.024747E+02 -2.024747E+02 1.1E-01

Table 2.5. The Stress Output for Case 3

ELEMENT SA1 SA2 SA3 SA4 AXIAL SA-MAX SA-MIN M.S.-T

ID SB1 SB2 SB3 SB4 STRESS SB-MAX SB-MIN M.S.-C
1 5.399325E+02 0.0 0.0 -5.399325E+02 0.0 5.399325E+02 -5.399325E+02 -4.4E-01
4.049494E+02 0.0 0.0 -4.049494E+02 4.049494E+02 -4.049494E+02 -4.4E-01

2.38
 These bending stresses at grid points 1 and 2 can be verified based upon Eq.(1), where P=0, I
= 2.667 inch4, E=30x107 psi and the bending moment at point 1, M1=72x10=720 lbs-in and at
point 2, M2=54x10=720 lbs-in. After simplification, the bending stresses at grid point 1 and 2 are
computed by
s 1 xe = 269.96 c
s 2 xe = 202.47 c
where c is the fiber distance whose value can be found in Table 2.6 for various cases. The
computed stresses are in an excellent agreement with those reported in Tables 2.2, 2.4 and 2.5.
Table 2.6. The Fiber Distances for Stress Recovery Points in inches

Cases C D E F
1 1. -1. -1. 1.
2 0. -1. 0. 1.
3 2. 0. 0. -2.

Remark:
Use PATRAN to check the V vector.

V vector has to be defined so that NASTRAN can correctly interpret the sectional geometry of
the beam.

Users Procedure to define CBAR and PBAR

!
1. Define the local xe by connecting GA point to GB point.
! !
2. Identify the principal axes, y e and z e , of the cross-section of a given beam element,
which will be the local beam axes.
3. Calculate the cross-sectional area (A), the moments of inertia, ( I z e z e and I ye ye ), and the
torsional rigidity (J) of the beam element. Enter these information to the PBAR card.
4. Select the four corners to report the stress. Identify the coordinates of these four points
! !
on the y e - z e plane. Enter these coordinates for (C1,C2), (D1,D2),(E1,E2) and (F1,F2) of
the CBAR card.
! ! ! !
5. Set the V vector on the known xe - y e plane. Enter the components of the V vector in
terms of the global coordinate system to X1, X2 and X3 of the CBAR card.
This module

2.39
 ! !
NASTRANs Procedure to define y e and z e , based upon CBAR
!
1. Define the local x e by connecting GA point to GB point.
!
2. Read in the components of V , the orientation vector, from CBAR card.
! ! ! !
! !
( )
3. Recognize the z e by cross-product of x e to V ; i.e., z e // xe Ve .
! ! !
4. Determine e
y = z e xe
! ! !
5. The sectional properties, I, J, C1, C2, et al are all read in based upon so determined xe , ye , z e

The solid model generated by PATRAN can be used to show whether the V vector is correctly
defined by the beam element. For example, the V-vector for the fifth element added to the end of
the following beam structure

Case I.
Set the V-vector as ! = (1,0,1). The NASTRAN bdf statement is given below

CBAR,5,200,5,6,1.,0.,1.
PBAR,200,10,8.,2.667,10.667,7.676,,,ABC
+BC,2.,1.,2.,-1.,-2.,-1.,-2.,1.,CDF
+DF,.8333,.8333

In the PATRAN solid model, it shows the fifth beam is twisted.

2.40
Case II.
Now set the V-vector as ! = (1,1,0). One then prepares the NASTRAN input as below.

CBAR,5,200,5,6,1.,1.,0.
PBAR,200,10,8.,2.667,10.667,7.676,,,ABC
+BC,2.,1.,2.,-1.,-2.,-1.,-2.,1.,CDF
+DF,.8333,.8333

The PATRAN displays the correct orientation of the fifth element.

2.41
Example 2: Cantilever Beam with CQUADs

In this example we will model the same cantilever beam from example one, but we will use
the CQUAD4 element, which is a 2-D plate element. The length, the depth and the thickness of
the beam are 72, 6 and 0.6, respectively. The beam is discretized into 4 CQUAD4 elements
and 10 nodes. The far left of the beam, nodes 1 and 2 are fully constrained and two point loads,
100 lbs each, are applied at nodes 9 and 10, along the Y-direction. The Youngs modulus of the
beam is 30x106 psi and the Poissons ratio is 0.33.

z
z
y
2 3 5 7 9 0.6"

x y
6"
1 2 3 4

1 18" 4 18" 6 18" 8 18" 10

F
.
Figure. 2.8 Second Beam Example
2.42
$ EXECUTIVE CONTROL SECTION BEGIN BULK
ID MEM441, BEAM2BDF GRID,1,,0.,0.,-3.
SOL 101 GRID,2,,0.,0.,+3.
TIME 10 GRID,3,,18.,0.,3.
CEND GRID,4,,18.,0.,-3.
$CASE CONTROL SECTION GRID,5,,36.,0.,3.
TITLE=CANTILEVER BEAM USING QUAD4 GRID,6,,36.,0.,-3.
ELEMENTS GRID,7,,54.,0.,3.
SET 1=9,10 GRID,8,,54.,0.,-3.
DISPL=1 GRID,9,,72.,0.,3.
STRESS=ALL GRID,10,,72.,0.,-3.
SPC=100 SPC1,100,123456,1,2
SUBCASE 1 SPC1,100,56,3,THRU,10
SUBTITLE=10 LB. LOAD CASE SPC1,200,123456,1,2
LOAD=100 SPC1,200,5,3,THRU,10
SUBCASE 2 PLOAD2, 500, 10., 4
SUBTITLE=20 LB. LOAD CASE FORCE,100,9,,100.,0.,1.,0.
LOAD=200 FORCE,100,10,,100.,0.,1.,0.
SUBCASE 3 FORCE,200,9,,200.,0.,1.,0.
SUBTITLE=30 LB. LOAD CASE FORCE,200,10,,200.,0.,1.,0.
LOAD=300 FORCE,300,9,,300.,0.,1.,0.
SUBCASE 4 FORCE,300,10,,300.,0.,1.,0.
SUBTITLE = PRESSURE LOAD CQUAD4,1,100,1,4,3,2
SPC = 200 CQUAD4,2,100,4,6,5,3
LOAD = 500 CQUAD4,3,100,6,8,7,5
SUBCOM 5 CQUAD4,4,100,8,10,9,7
SUBTITLE = COMBINATION LOADS PSHELL,100,10,.6,10
SUBSEQ = 1.0, 1.0, 0.0, 0.0 MAT1,10,3.+7,,0.33
$ BULK DATA SECTION ENDDATA

Explanation
Executive Control Section:
There are no changes here. It is similar to the one explained in Example 1

Case Control Section:

Here we define SET 1 as grids 9 and 10. The next line indicates that we want displacement
output for SET 1. This will cause the displacements to be output only at grids 9 and 10. In the
Subcase 4, a new set of LOAD, the pressure load is applied and a new set of SPC is also
selected. The final subcase SUBCOM is a method to combine the loads. SUBCOM is a
subcase, but gets its loads from the previously defined subcases. The coefficients are selected
from the SUBSEQ statement. In the example problem, the loads defined in Subcase 1 and

2.43
Subcase2 are combined for SUBCOM 5. You might want to compare the results of the
SUBCASE 3 and SUBCOM 5.

Bulk Data Section:

SPC1:
Notice here that for the second SPC1 entry we used the THRU option. We specified that
the degree of freedom 5 of every grid from 3 through 10 is constrained. This statement implies
that the rotation about the y-direction is contrained.. This was necessary because the QUAD4
element offers no resistance to rotation about the direction perpendicular to its surface.

CQUAD4:
The CQUAD4 entry refers to the element property entry PSHELL. The last four fields set
the grid ID numbers for G1, G2, G3 and G4. For purposes of interpreting the output, the
element local coordinate system is shown in Fig.2.4. Since the CQUAD element used here is
rectangular in shape. The element x-axis is defined at the plate centroid and parallel to the line
connecting G1 and G2. The element y-axis is perpendicular to the x-axis and is in the plane of
the plate.

PSHELL:
As with the PBAR entry, this entry refers to a property entry. The only numerical data in this
entry is the plate thickness of .6. The MAT 1 entry has not been changed from Example 1.

PLOAD2:
This load statement applies uniformly distributed pressure load acting normal to the plane of the
element. The PLOAD2 defined in the example problem defines a load of 10 units acting
normal to the plane of the CQUAD4 element number 4. For further information about the
PLOAD2 command, refer to the MSC/NSATRAN user's manual to get information about the
fields and its usage.

Execution
Run this job as you did for the first example. Once delete files not needed and look at your
results in the f06 file.

2.44
NASTRAN OUTPUT:
Numerical Error Indicator
EPSILON e = 1.729E-12
In linear static analysis, NASTRAN aims to find the displacement vector, x , by solving the
finite element matrix equation, Kx = f , where K is the stiffness matrix and f is the given force
vector. The error indicator, e , is calculated by
e = x T (Kx - f )
In general, the error is greater than 10-6, e 10 -6 , the accuracy of the solution is not acceptable.

Displacement Vector

The displacement vector lists the values of 6 degrees of freedom, ( u, v, w, q x , q y , q z ), of each

node. The displacement table reported here is the result of the input Case Control Deck, SET 1
= 9, 10 and DISPL=1. It prints out only the displacement vectors at nodes 9 and 10.

D I S P L A C E M E N T V E C T O R

POINT ID. TYPE T1 T2 T3 R1 R2 R3

9 G 0.0 7.499607E+00 0.0 2.721649E-04 0.0 1.576648E-01

10 G 0.0 7.499607E+00 0.0 -2.721649E-04 0.0 1.576648E-01

The node type indicates that the result is associated with a grid point, G, not a temperature, T.
The values in the columns of T1, T2 and T3 are the translational displacements
corresponding to u, v, and w, respectively, while those of R1, R2 and R3 corresponding to
q x , q y , and q z , respectively. Note that the units of the translational displacement vectors are
the same as that of the coordinates and the units of the rotational degrees of freedom are in
radians. In this example, the vertical deflection at the tip is 7.5 inches and the slopes with respect
to the X and Z axes are 0.00027 and 0.157 radians, respectively.
State of Stresses in CQUAD

The stresses in every CQUAD4 elements are printed out, as required by the Case
Control Deck, STRESS=ALL.

S T R E S S E S I N Q U A D R I L A T E R A L E L E M E N T S ( Q U A D 4 )
ELEMENT FIBRE STRESSES IN ELEMENT COORD SYSTEM PRINCIPAL STRESSES (ZERO SHEAR)
ID. DISTANCE NORMAL-X NORMAL-Y SHEAR- XY ANGLE MAJOR MINOR VON MISES

2.45
1 -3.000000E-01 4.839433E+03 3.500000E+04 -1.364242E-09 -90.0000 3.500000E+04 4.839433E+03 3.284874E+04
3.000000E-01 -4.839433E+03 -3.500000E+04 1.364242E-09 0.0000 -4.839433E+03 -3.500000E+04 3.284874E+04
2 -3.000000E-01 -2.500000E+04 1.318299E+03 3.637979E-10 90.0000 1.318299E+03 -2.500000E+04 2.568453E+04
3.000000E-01 2.500000E+04 -1.318299E+03 -3.637979E-10 0.0000 2.500000E+04 1.318299E+03 2.568453E+04
3 -3.000000E-01 -1.500000E+04 -4.337629E+02 2.311632E-09 90.0000 -4.337629E+02 -1.500000E+04 1.478789E+04
3.000000E-01 1.500000E+04 4.337629E+02 -2.311632E-09 0.0000 1.500000E+04 4.337629E+02 1.478789E+04
4 -3.000000E-01 -5.000000E+03 4.167526E+02 2.523848E-09 90.0000 4.167526E+02 5.000000E+03 5.220866E+03
3.000000E-01 5.000000E+03 -4.167526E+02 -2.523848E-09 0.0000 5.000000E+03 -4.167526E+02 5.22086E+03

As shown in the first row of the attached table, the state of stresses in the first CQUAD4 element
is
s x = 4,839 , s y = 35,000 and t x y = -1.36 E - 9 0 (2.1)
which can be shown on an infinitesimal element located at the center of CQUAD4 as

y = 35 Kpsi
xy ~ 0

x = 4.8 Kpsi

Note that these results are the stresses at the center of the element and corresponding to Fiber
Distance, -0.3, as indicated in the second column of the stress output table. Besides their
magnitudes, there are many questions associated with these numbers; the units, the meanings of x
and y axes, Fiber Distance, etc.

Units
Units of the stresses are defined based upon the input units of the load and the length. If
the load is measured by pounds and the length is measured by inches, the unit of a stress quantity
is psi.

Orientation of the Stresses

The orientation of the stresses is measured based upon the element coordinate system, xe
and ye, assigned to each of the element. This coordinate system depends upon the connectivity of
the CQUAD4 element, as shown in Fig. 2.5 (p.2.24)

2.46
 y element

G3

x element
x material
G4 a
b +g
a q a = -----
2

zelement g
b
G2
G1

The input CQUAD4 card for the first element in our example is stated as
CQUAD4,1,100,1,2,3,4.
Therefore, G1=1, G2=2, G3=3 and G4=4, which result in an elemental coordinate with Xe
pointing to the global Z axis, Ye to the global X axis and Ze to the global Y axis. Therefore, the
state of stresses of CQUAD4 is shown in the following figure. Note that the fiber distance
denoted in the output indicates the distance measured from the central plane of the CQUAD4
element along the Ze axis. Thus, the stresses reported in Eq.(2.1) are located on the tensioned
surface of the beam.

Xe

2 s x= 0.48 3
Kpsi
Ye

s y = 3.5 Kpsi
4
1

If the input CQUAD4 card is changed to

CQUAD4,1,100,1,4,3,2
the resultant stresses are reported in the following table.

2.47
S T R E S S E S I N Q U A D R I L A T E R A L E L E M E N T S ( Q U A D 4)
ELEMENT FIBRE STRESSES IN ELEMENT COORD SYSTEM PRINCIPAL STRESSES (ZERO SHEAR)
ID. DISTANCE NORMAL-X NORMAL-Y SHEAR- XY ANGLE MAJOR MINOR VON MISES

1 -3.000000E-01 -3.500000E+04 -4.839433E+03 1.364242E-09 90.0000 -4.83943E+03 -3.50000E+04 3.284874E+04

3.000000E-01 3.500000E+04 4.839433E+03 -1.364242E-09 0.0000 3.50000E+04 4.83943E+03 3.284874E+04

The first line of the output gives the state of stresses as

s x = -3,500 , s y = -483 .9 and t x y = 1.36 E - 9 0 (2.2)
which can be shown on an infinitesimal element located at the center of CQUAD4 with a fiber
distance -0.3 as

y = -4.8 Kpsi
xy ~ 0

x = -35.0 Kpsi

The output in Eq.(2.2) is quite different from that in Eq.(2.1). This difference is caused by the
change of the connectivity information, G1=1, G2=4, G3=3 and G4=2, which results in the
change in the local element coordinate system. In this case, the elemental coordinate system, (Xe
, Ye, Ze), is corresponding to the global axis system, (X,Z,-Y). Therefore, the fiber distance -0.3
inches places the infinitesimal element on the compressed surface of the beam.

Ye

2 3
y = -4.8 Kpsi
Xe

x= -35.0 Kpsi

4
1

2.48
Reactions at the Single Point Constraints
It is sometimes necessary to know the reactions at the constrained points. These reactions
can be used to investigate the strength of joining mechanism.
For this example problem, nodes 1 and 2 are fully constrained. As reported in the following
table, the reaction force along the Y-axis, Ry, and the reaction moments along the X and Z axes,
Mx and Mz , at Node 1 can be found as
R y = -100 lbs, M x = 3,821lb-in. and M z = -7,200 lb-in.

F O R C E S O F S I N G L E - P O I N T C O N S T R A I N T
POINT ID. TYPE T1 T2 T3 R1 R2 R3
1 G 0.0 -1.000000E+02 0.0 3.821196E+03 0.0 -7..2000E+03
2 G 0.0 -1.000000E+02 0.0 -3.821196E+03 0.0 -7.2000 E+03

Verification

It is always helpful to verify the NASTRAN result with hand calculation based upon
simplified model. The example problem may be analyzed based upon the beam model. The
tension stress at the center of CQUAD4 element 1 on the tensioned side of the beam surface can
be computed by the relation
MC
sx =
I
where the bending moment, M, the fiber distance, C and the moment of inertia, I , are given as
M=200 lbs 72 inches=14,400 lb-in.
C=0.3 in.
bh 3 6 (0.6)
3
I= = = 0.108 in4
12 12
Therefore, the tensile stress is computed as
MC 14400 0.3
sx = = = 40,000 psi
I 0.108
And the deflection and the slope at the free end is computed by the equation,

200 (72 )
3
PL3
v= = = 7.68 in.
3EI 3 3 10 7 0.108

200 (72 )
2
PL2
q = v = = = 0.16 radians
2 EI 2 3 10 7 0.108

2.49
The maximal error of 2.3% is observed in the tip displacement.

Example 3: Built-Up Structure with Offset

Two new concepts are applied in the following example. The first is the use of offset for the
BAR element. The second is the use of subcases for different boundary conditions instead of
different loadings. The figure shows a plate with a stiffener subjected to loading. Constraints
for the first case is shown below with Points 1, 3, 4 and 6 completely fixed.

Ye

.5"

2.25"
3" Ze
4"

.5"
.5"

Figure. 2.9 A Stiffened Plate

2.50
The second case imposes roller supports to Points 3 and 6. Point 3 allows to roll along Y-
direction, while Point 6 along X-direction. The NASTRAN input file below pertains to the above
problem and you can copy and execute the program.

$EXECUTIVE CONTROL SECTION SPC1, 200, 23, 3

ID NAME, FILENAME SPC1, 200, 123456, 1, 4
TIME 10 FORCE, 100, 2,,20.,0.,0.,-1.
SOL 101 FORCE, 100,2,,30.,-1.,0.,0.
CEND GRID,4,,0.,0.,0.
$CASE CONTROL SECTION GRID,1,,36.,0.,0.
DISP=ALL GRID,2,,36.,12.,0.
STRESS=ALL GRID,3,,0.,12.,0.
LOAD=100 GRID,5,,36.,24.,0.
TITLE = PLATE WITH A STIFFENER GRID,6,,0.,24.,0.
SUBCASE 1 CBAR,1,100,2,3,1.,0.,1.,,+B1
SUBTITLE = FIXED - FREE CASE +B1,,,0.,0.,-2.25,0.,0.,-2.25
SPC = 100 CQUAD4,1,100,1,2,3,4
SUBCASE 2 CQUAD4,2,100,2,5,6,3
SUBTITLE = SIMPLE - ROLLER CASE PBAR,100,10,5.5,10.375,2.2812,12.6562,,,+P1
SPC = 200 +P1,-2.,0.
BEGIN BULK PSHELL,100,10,.5,10
$BULK DATA SECTION MAT1,10,3.+7,,.33
SPC1, 100, 123456, 1, 3, 4, 6 ENDDATA
SPC1, 200, 13, 6

Explanation:

Case Control Section:

Notice carefully that the two force entries with ID 100 are referenced by LOAD = 100 in the
Case Control section. This is done before the subcase level so these two forces are applied for
all cases. In addition to constraining the degrees of freedom dictated by the boundary conditions,
the degree 6 is also constrained. Direction 6 is constrained for the CQUAD4 elements because,
from the theory of FEA, these elements fail to handle in-plane rotation. Directions 2 and 4 may
be constrained to prevent translation in y-direction and rotation about x-direction.
Bulk Data Section:
On the second line of the CBAR entry, offset vectors are defined at each end of the beam.
Without offset, the local coordinate system of a CBAR is defined by GA and GB on the CBAR
entry. If this were the case in our example, then the neutral axis of the beam would be defined on
the central plane of the plate. This would define the neutral axis of the CBAR is using Grids 2
and 3, which lies on the central plane of CQUAD4 elements. So, the offset vectors act as rigid
links between the grids and the neutral axis of the beam Notice that both offset vectors have the
components (0,0, -2.25) in the global coordinate system.

2.51
Notice that the CBAR and PBAR entries have continuation fields in the tenth field of their first
line and in the first field of the continued line. The fields are eight columns long. In field one of
the continued card, the symbol + must be added in column one. The first letter in field ten is
ignored. So, the CBAR entry could also have been given as:

CBAR,1,100,2,3,1.,0.,1.,,AB1
+B1,,,0.,0.,-2.25,0.,0.,-2.25

The A in AB1 is ignored. Note that the letters assigned to a continuation field can not be
duplicated for another continuation field. In the PSHELL entry, a material ID of 10, referring to
MAT1,10, was added because this plate needs stiffness for bending. (See MID2 field in
PSHELL definition).

Example 4: Pin-Flag and Balance of Joint Internal Forces

The CBAR card provides a mechanism to define various joint conditions at GA and GB of the
concerned bar member. The integers given in PA or PB, ranging from 1 to 6, indicate the
degrees of freedom at which the corresponding joint can move freely. Note that the components
given in PA or PB are defined based upon the local coordinate system of the bar member. For
example, PA = 1 and PB = 456 indicates that the CBAR member can slide freely along
the local X-axis at Joint GA and rotate freely about local X, Y and Z-axes at Joint GB. In other
words, Joint GA and Joint GB are a sliding joint and a spherical joint, respectively, for the bar
member of concern. On the other hand, PB = 34, indicates that the CBAR member can
translate freely along the local Z-axis and rotating freely along the local X-axis at Joint GB. The
pin-flag option enables CBAR be used to analyze a truss structure. The following example
demonstrates such application.

A three bar truss structure is shown in Fig. 2.10. The cross-section of the truss is 2x4 rectangle,
same as that in Example 1. Find the force balance at each joint and check whether member fails
due to buckling.

2.52
Figure 2.10: 3-Bar Truss

The NASTRAN bdf file for the truss problem is given below. Note that the orientation
vectors of those bar members are defined in such a way that the local Z-axis is perpendicular to
the paper. As a result, to represent a hinge joint freely rotating with respect to the local Z-axis,
the pin-flags are assigned to be 6 for the end nodes of all members. Furthermore, the command
lines, GPFORCE=ALL and SPCFORCE=ALL, are added to the CASE CONTROL DECK
which require the internal forces applied to each of the GRID points and the single support
points are printed out, respectively.

$ EXECUTIVE CONTROL SECTION GRID,1,,0.,0.,0.

ID MEM441,Hinge.BDF GRID,2,,18.,0.,0.
SOL 101 GRID,3,,37.,,-9.
TIME 10 GRID,4,,9.,,-9.
CEND CBAR,1,100,1,2,1.,0.,-1.,,ABC
$CASE CONTROL SECTION +BC,6,6
DISP=ALL CBAR,2,100,2,3,1.,0.,-0.5,,EDF
STRESS=ALL +DF,6,6
SPC=100 CBAR,3,100,2,4,-0.5,0.,-1.,,HGJ
SPCFORCE=ALL +GJ,6,6
GPFORCE=ALL PBAR,100,10,8.,10.667,2.667,7.676,,,123
LOAD=100 +23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
BEGIN BULK +56,.8333,.8333
$BULK DATA SECTION MAT1,10,3.+7,,0.33
SPC1,100,123456,1,3,4 ENDDATA
FORCE,100,2,,100.,0.,0.,-1.

The member stresses are listed in the following table. Note that zero or near zero bending
stresses are found for each member in the table. This indicates that the structure is made of two-
force members. Thus, the structure is indeed a truss.

2.53
 In order to check the buckling failure, the axial force of each member has to be computed to
compare with the critical buckling load,
p 2 EI
Pcr = 2
l

S T R E S S E S I N B A R E L E M E N T S ( C B A R )

ELEMENT SA1 SA2 SA3 SA4 AXIAL SA-MAX SA-MIN M.S.-T

ID. SB1 SB2 SB3 SB4 STRESS SB-MAX SB-MIN M.S.-C
1 0.0 0.0 0.0 0.0 2.424548E+00 2.424548E+00 2.424548E+00
0.0 0.0 0.0 0.0 2.424548E+00 2.424548E+00
2 6.8049E-16 6.8049E-16 -6.8049E-16 -6.80490E-16 -7.5651E+00 -7.56515E+00 -7.565152E+00
2.0217E-15 2.0217E-15 -2.0217E-15 -2.02172E-15 -7.56515E+00 -7.56515E+00
3 0.0 0.0 0.0 0.0 -1.3097E+01 -1.3098E+01 -1.30976E+01
3.1729E-15 3.1729E-15 -3.1729E-15 -3.172995E-15 -1.3097E+01 -1.30976E+01

The axial force may be calculated as a product of members cross-sectional area and its axial
stress; P = A s . The results of the three bar truss are given in the following table. The results
show that members 2 and 3 which are under compression will not buckle under the applied load.
In this example, A = 8. , E = 30E 7 and I = 2.667 are used for calculation.

Member Length Stress P Pcr

1 18.00 2.424548 19.39 -
(tension)
2 21.02 -7.565152 60.52 1.78E6
3 12.72 -13.09768 104.8 4.88E6

Because of the Case Control Card, GPFORCE=ALL, the f06 file will print out the internal
forces applied to each of the GRID points, including the constrained ones in a table titled,
GRID POINT FORCE BALANCE, as shown in the following table.

G R I D P O I N T F O R C E B A L A N C E

POINT-ID ELEMENT-ID SOURCE T1 T2 T3 R1 R2 R3

1 F-OF-SPC -1.939638E+01 0.0 0.0 0.0 0.0 0.0
1 1 BAR 1.939638E+01 0.0 0.0 0.0 0.0 0.0
1 *TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0
2 APP-LOAD 0.0 0.0 -1.00000E+02 0.0 0.0 0.0
2 1 BAR -1.939638E+01 0.0 0.0 0.0 0.0 0.0
2 2 BAR -5.469531E+01 0.0 2.59083E+01 0.0 0.0 0.0
2 3 BAR 7.40917E+01 0.0 7.40917E+01 0.0 0.0 0.0
2 *TOTALS* 0.0 0.0 -1.42108E-14 0.0 0.0 0.0
3 F-OF-SPC -5.469531E+01 0.0 2.59083E+01 0.0 0.0 0.0
3 2 BAR 5.469531E+01 0.0 -2.59083E+01 0.0 0.0 0.0
3 *TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0
4 F-OF-SPC 7.40917E+01 0.0 7.40917E+01 0.0 0.0 0.0
4 3 BAR -7.40917E+01 0.0 -7.40917E+01 0.0 0.0 0.0

2.54
 4 *TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0

The information provided by the table can be used to find the internal forces in each member
of joint. For example, Joint 2 connects members 1, 2 and 3 and is subjected to external load, -100
lbs. Therefore, there will be four sources contributing the forces to Joint 2; internal forces from
members 1, 2 and 3 and the external load. Specifically, based upon the table of GRID POINT
FORCE BALANCE, one can summarize the forces applied to Joint 2 as

Source X-component Y-component Resultant Force

Applied Load 0.00 -100. 100
Member 1 -19.3 0. 19.3
Member 2 -54.7 25.9 60.52
Member 3 74.0 74.0 104.2

Note that the resultant force calculated here for each member is the same as that calculated by
P = As .
Next, assume that the hinge joint, Joint 2, is held together by a 0.5 unhardened dowel pin with a
shearing yield strength, 64 kpsi. Will the pin sustain the applied without shearing failure? What
will be the maximal allowable load that the pin can support without shearing failure? To answer
this question, one needs to find the maximal shear force applied to Joint 2. Based upon the above
table, the maximal X-component and Y-component forces are 74.0 and 100., respectively.
Therefore, the maximal shear force is
S = 74 2 + 1002 = 124.4 lbs.
which yields a shear stress
124.4
t= = 158.4 psi
p 0.5 2
which is far below the shearing yield strength. In fact, the structure is safe from shearing, as long
as the force is less than F = (64000 / 158.4) 100 / 158.4 = 40400 lbs. The free body diagrams of
the forces applied at Joint 1 and Joint 2 are shown in Figs. 2.11 and 2.12.

Figure 2.11. Free Body Diagram at Joint 1

2.55
Figure 2.12. Free Body Diagram at Joint 2

Another Example
Repeat Example 4 on P.2.48. However, in this assignment, the grid point 4 is moved to a new
location (37,0,10). Modify the given bdf file, submit for Nastran run. Find the global location
and the value of the maximal stress. Answer the following questions:

a) Draw the free body diagram to show the balance of the forces at Grid Point 2. ( a picture
similar to Fig. 2.12)
b) Report the global location and the value of the maximal stress and the maximal displacement,
respectively.

Solution:
The input bdf file is of the form given below. Grid point 4 has the coordinates (37.,0.,10.)
which is 37 in. for the global x location, 0 in. in the global y location, and 10 in. in the global z
location. Pay attention to the V vector set up for CBAR ID 3. Furthermore,
C1,C2,D1,D2,E1,E2,F1,F2 in PBAR were not correctly defined ( Why?)

2.56
 However, all members are truss members, subjected to no bending. Thus, bending stresses are
all zero ( Can you confirm this statement by checking the stress values reported under SA1, SA2,
P
SA3, SA4, SB1, SB2, SB3 and SB4?) The non-zero stresses are all associated with axial stress, ,
A
where P is the axial member force and A is the cross-sectional area of the member of concern.

$EXECUTIVE CONTROL SECTION GRID,1,,0.,0.,0.
ID MEM441, Hinge.BDF DEGROFF GRID,2,,18.,0.,0.
SOL 101 GRID,3,,37.,0.,-9.
TIME 10 GRID,4,,37.,0.,10.
CEND CBAR,1,100,1,2,1.,0.,-1.,,ABC
$CASE CONTROL SECTION +BC,6,6
DISP= ALL CBAR,2,100,2,3,-1.,0.,-0.5,,EDF
STRESS= ALL +DF,6,6
SPC= 100 CBAR,3,100,2,4,1.,0.,0.,,HGJ
SPCFORCE= ALL +GJ,6,6
GPFORCE= ALL PBAR,100,10,8.,10.667,2.667,7.667,,,123
LOAD= 100 +23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
BEGIN BULK +56,.8333,.8333
$BULK DATA SECTION MAT1,10,3.+7,,0.33
SPC1,100,123456,1,3,4 ENDDATA
FORCE,100,2,,100.,0.,0.,-1.

The output in the f06 file is given in the form below:

D I S P L A C E M E N T V E C T O R

POINT ID. TYPE T1 T2 T3 R1 R2 R3

1 G 0.0 0.0 0.0 0.0 0.0 0.0
2 G 1.343066E-07 0.0 -2.214564E-05 0.0 0.0 0.0
3 G 0.0 0.0 0.0 0.0 0.0 0.0
4 G 0.0 0.0 0.0 0.0 0.0 0.0
1 OCTOBER 1, 2016 MSC Nastran 7/ 3/15 PAGE 12

F O R C E S O F S I N G L E - P O I N T C O N S T R A I N T

POINT ID. TYPE T1 T2 T3 R1 R2 R3

1 G -1.790755E+00 0.0 0.0 0.0 0.0 0.0
3 G -9.905750E+01 0.0 4.692197E+01 0.0 0.0 0.0
4 G 1.008483E+02 0.0 5.307803E+01 0.0 0.0 0.0
1 OCTOBER 1, 2016 MSC Nastran 7/ 3/15 PAGE 13

S T R E S S E S I N B A R E L E M E N T S ( C B A R )
ELEMENT SA1 SA2 SA3 SA4 AXIAL SA-MAX SA-MIN M.S.-T
ID. SB1 SB2 SB3 SB4 STRESS SB-MAX SB-MIN M.S.-C
0 1 0.0 0.0 0.0 0.0 2.238443E-01 2.238443E-01 2.238443E-01
0.0 0.0 0.0 0.0 2.238443E-01 2.238443E-01
0 2 0.0 0.0 0.0 0.0 -1.370108E+01 -1.370108E+01 -1.370108E+01
0.0 0.0 0.0 0.0 -1.370108E+01 -1.370108E+01
0 3 0.0 0.0 0.0 0.0 1.424542E+01 1.424542E+01 1.424542E+01
-1.151685E-15 -1.151685E-15 1.151685E-15 1.151685E-15 1.424542E+01 1.424542E+01
1 OCTOBER 1, 2016 MSC Nastran 7/ 3/15 PAGE 14

2.57
 G R I D P O I N T F O R C E B A L A N C E

POINT-ID ELEMENT-ID SOURCE T1 T2 T3 R1 R2 R3

1 F-OF-SPC -1.790755E+00 0.0 0.0 0.0 0.0 0.0
1 1 BAR 1.790755E+00 0.0 0.0 0.0 0.0 0.0
1 *TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0
0 2 APP-LOAD 0.0 0.0 -1.000000E+02 0.0 0.0 0.0
2 1 BAR -1.790755E+00 0.0 0.0 0.0 0.0 0.0
2 2 BAR -9.905750E+01 0.0 4.692197E+01 0.0 0.0 0.0
2 3 BAR 1.008483E+02 0.0 5.307803E+01 0.0 0.0 0.0
2 *TOTALS* -1.421085E-14 0.0 -1.421085E-14 0.0 0.0 0.0
0 3 F-OF-SPC -9.905750E+01 0.0 4.692197E+01 0.0 0.0 0.0
3 2 BAR 9.905750E+01 0.0 -4.692197E+01 0.0 0.0 0.0
3 *TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0
0 4 F-OF-SPC 1.008483E+02 0.0 5.307803E+01 0.0 0.0 0.0
4 3 BAR -1.008483E+02 0.0 -5.307803E+01 0.0 0.0 0.0
*TOTALS* 0.0 0.0 0.0 0.0 0.0 0.0

a) Draw the free body diagram to show the balance of the forces at Grid Point 2.
Find all forces in the Table of GRID POINT FORCE BALANCE that are applied at Point-ID 2. And
mark the X and Y components of those forces at the free body of Grid Point 2. Note that all
three members connected at Grid Point 2 are all truss members. Thus, no internal bending
moments applied to the joint, Grid Point 2.

b) Report the global location and the value of the maximal stress and the maximal displacement,
respectively.

The maximum x-component of deflection is (-1.34*10^-7) and the maximum y-

component of deflection is (-2.21*10^-5) in. Therefore the maximum deflection is

2.58
 -1.34 10-( ) + -2.21 10-, ) 2.21 10-, in. The location of the maximum deflection
is at node 2 which is located at the global coordinates (18.,0.,0.) in.
The maximum axial stress is approximately 14.25 psi located in member 3 and
reported at nodes 2 (18.,0.,0.) in. and 4 (37.,0.,10.) in. Since there is no bending stress, the
stress values are zero at the stress recovery points. So the maximum stress is 14.25 psi.

II.6 Support and Joint Conditions

The finite element model of a machine or a mechanical system is made of elements that are
connected through common nodes. The forces are applied at the nodes and the boundary
conditions are imposed at the nodes as well. The connecting conditions between nodes play
important role in the quality of the finite element analysis result. In this section, the issues related
such connecting conditions, such as Singularity, St. Venant Principle, Symmetry, Rigid Element
and Multiple Constraint are discussed.

Singularity
For the user of finite element analysis (FEA), its important to keep in mind that every finite
element node has six degrees of freedom; three translations, (u, v, w); and three rotations,
( )
q x ,q y ,q z , and joints and support conditions are put in place to restrain displacements with
respect to certain degrees of freedom at the node of concern. For example, two beam elements
share the same 6 degrees of freedom at the common node, whereas two truss elements share only
the 3 translation degrees of freedom at the common node. Forces or moments are always
generated to achieve the displacement constraints. Such forces are called as the reaction forces at
the constraints or the internal forces at the joints. Again, for example, three internal forces (axial
and two shears) and three internal moments (torsion and two bending) will be generated at the
beam joint so as to maintain the continuity of the displacements and slopes. On the other hand,
only three internal forces are generated at a truss joint, because only the translational degrees of
freedom are constrained to be the same. The nodal degrees of freedom for different elements are
listed below for reference.
Element Degrees of Freedom
Truss translation in X, Y, Z
Beam translation and rotation in X, Y, Z
Brick translation in X, Y, Z
Plate translation in X, Y, Z; rotation in two directions but no rotation about in-plane
vector.
The static analysis of a structural component is done by solving the following matrix equation,
2.59
Kx = p where
K is the stiffness matrix,
x the displacement vector and
p the applied load.
The problem is very similar to a spring under a load, kx=p, where k is the spring constant. Note
that the above equation has a unique solution only if K is invertible or not singular. NASTRAN
computes and printed out the value of Epsilon, e , to indicate whether the model suffers for
singularity or near singularity. The value of Epsilon is equal to e = x T (Kx - p) . A finite
element analysis suffers for singularity or near singularity if the associated Epsilon value is
greater than 10-6 .
The stiffness matrix, K, can become singular, if the following
If a NASTRAN run fails to
occurs:
produce analysis results due to
1. The entire structure or part of the structure undergoes
singularity, one may add the
rigid body motion. The rigid body motion can be
command,
eliminated by constraining 6 independent degrees of PARAM,AUTOSPC,YES, to
freedom. This can be achieved by completely constraining
the CASE CONTROL deck and
a single finite element node, as a node has six degrees of
resubmit the bdf file for
freedom. This practice is not recommended though,
NASTRAN run. This command
because it produces a case with high reaction forces
will eliminate the singularity and
concentrated at a single point. As a result, a non-realistic
enable NASTRAN to produce
high stress is created around the single constrained point.
results. However, this result is not
2. Some of elements in K have zero values; for example; the
correct for sure, which is generated
degree freedom with in-plane rotation, the beam with zero
to help the user to identify the
rigidity, the plate without material properties,
modeling mistake.
unconnected element or group of elements.
For example, the finite element stiffness matrix of a CQUAD4
element is not provided a stiffness to resist the rigid body motion associated with the rotation
degree of freedom that is perpendicular to the plate. To eliminate such singularity, the user can
add the command, PARAM,K6ROT,100., to the CASE CONTROL Deck.

St. Venant Principle

Finite element discretization introduces not only numerical error, but also modeling error. As
an example, the distributed edge force applied to a plate has to be converted into a set of finite
forces applied to nodes that divide the edge of the plate into elements. Furthermore, a
continuous, welded seam between a pair of plates has to be represented by a set of finite nodes
2.60
that connect the plates. However, the errors created by such approximations become
insignificant, if the stresses and the displacements of concern are sufficiently far away from the
area of joining or loading approximation. This observation is supported by the St. Venent
Principle:

If some distribution of forces acting on a portion of the surface of a body is replaced by a

different distribution of forces acting on the same portion of the body, then the effects of the two
different distributions on the parts of the body sufficiently far removed from the region of
application of the forces are essentially the same, provided that the two distributions of forces are
statically equivalent.

Symmetry
Symmetry helps to reduce the size of the problem to be solved. The difficulty is to specify
the boundary conditions along the symmetric line. If the applied load, the material and the
geometry of a structure are all symmetric with respect to certain axis, s , then the displacement
that is normal to the symmetric line, un, and the rotation (i.e., slope ) with respect to the
symmetric line, q s should be zero. For example, the plate problem given below does not provide
an obvious boundary condition. It is noted though that the load and the geometry are symmetric
with respect to the center vertical and horizontal lines. Therefore, the problem can be solved with
one quarter of the plate. Furthermore, this quarter plate is now subjected to the boundary
conditions along the mid-sections, resulted from the symmetric conditions. The distributed load
P is 10 lbs/in.

E = 3x10 7 psi
n = .33
Thickness = 0.4"

P 20" P
10"

20"

2.61
Multiple Point Constraint
The Multiple Point Constraint defines a constraint condition that involves more than one
degree of freedom. In the following example, an inclined support is added at Node 5. Assume
that the slope of the inclined support be 300 about X-axis. Since the resultant displacement that is
vertical to the inclined surface has to be zero, this inclined support imposed to Node 5 yields an
MPC constraint
- 0.5u5 + 0.866w5 = 0
where u5 is the first degree freedom of displacement at Node 5 and w5 is the third. The bdf file of
the given example is given below, which is the same as that presented for the cantilever beam
given in Example 1, except those associated with MPC.

z
F z
y
2"
1 2 3 4 x y
4"
18" 18" 18" 18"
1 2 3 4 5
Fig. 2.6 Beam Example
$ EXECUTIVE CONTROL SECTION GRID,4,,54.,0.,0.
ID BEAM AT INCLINED SUPPORT GRID,5,,72.,0.,0.
SOL 101 SPC1,100,123456,1
TIME 10 FORCE,100,4,,10.,0.,0.,-1.
CEND CBAR,1,100,1,2,1.,0.,1.
$ CASE CONTROL SECTION CBAR,2,100,2,3,1.,0.,1.
DISPL=ALL CBAR,3,100,3,4,1.,0.,1.
STRESS=ALL CBAR,4,100,4,5,1.,0.,1.
SPC=100 PBAR,100,10,8.,10.667,2.667,7.676,,,123
MPC=100 +23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
GPFORCE=ALL +56,0.8333,0.8333
LOAD=100 MAT1,10,3.+7,,0.33
$ BUCK DATA SECTION MPC,100,5,1,-0.5,5,3,0.866
BEGIN BULK ENDDATA
GRID,1,,0.,0.,0.
GRID,2,,18.,0.,0.
GRID,3,,36.,0.,0.

Rigid Element Type 2

The rigid element is a line element that connects the master node to the slave nodes through a
rigid link and forces the slave nodes to move in the same prescribed motion as that of the master
node. The NASTRAN statement for a simple rigid link is RBE2. The NASTRAN description of
RBE2 is given below.

Bulk Data Entry: RBE2 Rigid Body Element, Type 2

2.62
Description: Defines a rigid link between the dependent degrees of freedom to those of
independent degrees of freedom.
Format:
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
RBE2 EID GN CM GM1 GM2 GM3 GM4 .

Examples:
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
RBE2 2 32 123 2 4

or RBE2,2,32,123,2,4
The three translational degrees of freedom of nodes 2 and 4 are determined by the translational
degrees of freedom of Node 32 based upon the rigid link relation,
u2 = u32 + r 32
where r is the distance between nodes 2 and 32.
Field Contents
EID Element identification number ( Integer > 0)
GN Grid or scalar point identification number of the independent node
(Integer > 0)
CM Component numbers of the dependent degrees of freedom in the global
coordinate system (6 Integer 0; up to six unique digits may be
placed in the field with no embedded blanks).
GMi Grid or scalar point identification numbers of the dependent nodes
(Integer > 0)

An example problem is given in the figure where a rigid block is loaded to press down a slab.
The slab is modeled by 4 CQUAD elements. The applied load, 1000 lbs, presses the rigid block
down vertically. Since Nodes 1, 2 and 3 are in full contact with the slab, their vertical
displacements have to be the same as that of the rigid block. To achieve this condition, an
additional node, Node 10, is introduced that is placed at the center of the rigid block. An RBE2
element is then introduced that connects Nodes 1, 2 and 3 to the center of the rigid block (i.e.,
Node 10) and requires the connected nodes move with the same z-displacement. Note that the
master point in RBE2 has to be Node 10. This is because the applied load is applied to Node 10.

2.63
$ EXECUTIVE CONTROL SECTION GRID,4,,0.,0.,5.
ID RIGID PLATE GRID,5,,5.,0.,5.
SOL 101 GRID,6,,10.,0.,5.
TIME 10 GRID,7,,0.,0.,0.
CEND GRID,8,,5.,0.,0.
$ CASE CONTROL SECTION GRID,9,,10.,0.,0.
DISPLACEMENT=ALL GRID,10,,5.,0.,15.
STRESS=ALL SPC1,100,123456,7,THRU,9
SPCFORCES=ALL FORCE,100,10,,1000.,0.,0.,1.
GPFORCES=ALL CQUAD4,1,100,7,8,5,4
SPC=100 CQUAD4,2,100,8,9,6,5
LOAD=100 CQUAD4,3,100,4,5,2,1
PARAM,KGROT,1.0 CQUAD4,4,100,5,6,3,2
$ BUCK DATA SECTION PSHELL,100,10,0.6,10
BEGIN BULK MAT1,10,3.+7,,0.33
GRID,1,,0.,0.,10. RBE2,100,10,3,1,2,3
GRID,2,,5.,0.,10. ENDDATA
GRID,3,,10.,0.,10.

II.7 NASTRAN to PATRAN

Once NASTRAN bdf file has been created, PATRAN can be used to display either the input
finite element model or the output results graphically. Some modifications of the NASTRAN bdf
are required in order to post-process the result file.

II.7.1 Pre-Process

PATRAN can be used to graphically investigate the finite element model built by the
NASTRAN bdf file. The key of this process is to create a PATRAN database file, db.

Step 1:
Open PATRAN using the command at the prompt, cae-patran2010.

2.64
Step 2:
Create a new PATRAN database file, db, using the following command strings,

FILE > IMPORT > Model/Source/MD NASTRAN INPUT

Step 3:
Select the required bdf file in the newly opened window with a list of directories
and files.

PATRAN will respond with a table summarizing the finite element model data that has been
successfully selected to build the db file. With this newly created db file, PATRAN will display
the finite element model built through the NASTRAN bdf file.

II.8 NASTRAN to PATRAN (new version) : .DB and .XDB GENERATION

and VISUALIZATION

Purpose: Once the dbf file is created and its f06 file has been checked to be correct, one can then
follow the procedure given below to visualize the finite element model and the analysis results
through the use of PATRAN. This achieves in two steps. The first one is to generate .db file for
visualization of the finite element model and the second to generate .xdb file for visualization of
the analysis results.

Task 1. Create .db File

Step 1:

Open PATRAN. Click on File > New. In the New Database window manual. Find the proper
directory and key in the file name with an extension .db for the db file in the space, then OK.

Step 2:
Click on File again. Select Import to import the manually created bdf into PATRAN. As shown
in the following picture. Be sure that on the right, select MSC. NASTRAN Input in the Source.
Go to the proper directory and select the desired bdf file. Click Apply

2.65
A .db file is now created and the grid points and elements of the finite element are displayed now
in the PATRAN window.

Task 2. Create .xdb File

Step 1:
Open PATRAN, click on Analysis. As shown in
the following picture, select Analyze in Action,
Entire Model in Object and Full Run in Method. If
desirable, you may assign the newly created xdb
file with a new name keyed in in space under Job
Name. See the following figure. Click Apply. The
bdf file will be submitted to NASTRAN run to
generate the f06 and xdb files.

Step 2:
Open NATRAN to analyze the bdf file just created
by PATRAN in the above steps. A xdd file with
the assigned name should now be present within
the folder.

Task 3. Visualization of the Analysis

Results

Step 1:

2.66
 Open PATRAN, click on Analysis. Select
Access Results in Action, Select Attach XDB in
Object and Result Entities in Method as shown
in the following picture.

Step 2:

Attach the xdb file by selecting the xdb file just created. Click on Select Results
File to identify the desirable xbd file. Then OK to close up the result file
selection. Click Apply

Step 3:
Click Results. Select the Result Cases. Click
Displacement- Translational in both windows
of the Select Fringe Result and the Select
Deformation Result. See the following figure.
Click on Apply. The PATRAN window will
display the stress contour plot which is imposed
upon the deformation plot.

Note that you may skip Step 2 by adding a line

PARAM,POST,0 directly under BEGIN BULK.
That is,

After submitting the bdf file to NASTRAN run, this PARAM statement will ensure the
generation of the .xdb file in addition to the .f06 file. The former can be used to generate the
stress contour plot and the deformation plot follow

II.9 FORCE VECTOR

The point force is a vector. The FORCE statement in NASTRAN is given below which requires
the input, F, the magnitude of the force, and N1, N2 and N3 the direction of the force.
Format:

2.67
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FORCE SID G CID F N1 N2 N3 AXI

As an example, if the force vector is given by ( 200., -50., 30 ). The magnitude of the force is
given by
F = 2002 + (-50) 2 + 30 2 = 208.3.
The unit director of the force is then given by
200 - 50 30
n= = (0.96 - 0.24 0.144)
208.3 208.3 208.3
Thus, the force vector statement can then be written as
FORCE,SID,G,CID,208.3,0.96,-0.24,0.144
The alternative is to set the magnitude of the force as 1 while the direction vector
is n = (200 - 50 30). The force vector statement in NASTRAN is then given by
FORCE,SID,G,CID,1.,200.,-50.,30.

Example Problem:
Find the location and the magnitude of the maximal stress in the beam, which a point load, (100.,
50., -20.) lbs is applied at Point 4.

The magnitude of the force is found to be F = 1002 + (50) 2 + (-20) 2 = 113.58 and the unit
vector of the force is given by

2.68
 100 50 - 20
n= = (0.88 0.44 - 0.176)
113.58 113.58 113.58
One may enter the FORCE in two different ways,

Case I:
Use FORCE,100,4,,1.,100.,50.,-20.
The bdf file is given below.
$ EXECUTIVE CONTROL SECTION CBAR,2,200,2,3,0.,1.,1.
ID MEM441, BEAM2BDF CBAR,3,300,3,4,0.,0.,1.
$ Jinhyuk Kim
SOL 101 PBAR,100,10,24.,72.,32.,79.754,,,123
TIME 10 +23,2.,3.,2.,-3.,-2.,-3.,-2.,3.,456
CEND +56,.8333,.8333
$CASE CONTROL SECTION
TITLE=CANTILEVER BEAM USING QUAD4 ELEMENT PBAR,200,10,12.,9.,16.,20.73,,,789
DISPL=ALL +89,1.5,2,1.5,-2.,-1.5,2.,1.5,2.,ABC
SPCFORCE=ALL +BC,.8333,.8333
GPFORCE=ALL
STRESS=ALL PBAR,300,10,48.,256.,144.,331.776,,,CDE
SPC=100 +DE,3.,4.,-3.,4.,-3.,-4.,3.,-4.,EFG
LOAD=100 +FG,.8333,.8333
$ BULK DATA SECTION
BEGIN BULK SPC1,100,123456,1
GRID,1,,0.,0.,0. FORCE,100,4,,1.,100.,50.,-20.
GRID,2,,100.,0.,0. MAT1,10,3.+7,,0.33
GRID,3,,100.,0.,50. ENDDATA
GRID,4,,130.,-30.,58.
CBAR,1,100,1,2,1.,1.,0.

The 3D model created by PATRAN shows that the V-vector is correctly defined.

2.69
The maximal displacements are found to be at Node 4, u = 0.0994 , v = 0.0804 and
w = -0.0621 . All in inches. The maximal stress in psi is -2.138E+2 in bending and -1.666 in
axial stress which results in -2.140E+2. The location of the maximal stress is found to be at the F
corner point of Node 2 of beam element 2. In the local coordinate system, the F point is (1.5, 2.).
Since the global coordinate of Node 2 is (100., 0., 0.) and the orientation vector is (0.,1.,1.), the
F point in the global coordinate system is (102, 1.3, 0). Units are all in inches.

2.70
Case II:
Use FORCE,100,4,,113.58.,0.88,0.44,-0.176

The maximal displacements are found to be at Node 4, u = 0.09937 , v = 0.08036 and

w = -0.06215 inches. The results reported here are the same as those in Case I.

2.71
 II.10 HOMEWORK
Homework 1
Add the force of 50 lbs to Example 2 along the X-direction at Nodes 9 and 10.

1. Repeat the same analysis with four CQUAD elements with the changed load. Verify your
solution with a classical solution.
2. Divide each CQUAD of the original model into four CQUADs. Prepare the NASTRAN
input file and submit it for NASTRAN run.
3. Answer the following questions:
a) Report the location and the value of the maximal stress.
b) Draw a square element picture to show the state of stresses of Element 1 in both
cases. Be sure to indicate the relationship between the element coordinate system
and the global coordinate system in the picture.

Homework 2
Consider the two beams shown in Figs. 2.8 and 2.9. Prepare a common Bulk Data Deck for
both the beams and distinguish them in your case control deck. Generate two sets of data. In
the first set, let the length of each element be 7.5 inches and divide the beams uniformly. In the
second case, let the length of each element be 2.5 inches. Find the following:

1. NASTRAN input and output files for the 7.5 inch elements
2. NASTRAN input and output files for the 2.5 inch elements
3. Disp. .vs. length graph for beam in Fig. 2.8.
4. Disp. .vs. length graph for beam in Fig. 2.9.
5. Stress .vs. length graph for beam in Fig. 2.8.
6. Stress .vs. length graph for beam in Fig. 2.9
7. Find the factor of safety of the beams with 2.5 inch elements through NASTRAN.
q = 40 lb/in E = 3 x 10 7 psi

C x .25"
A
B .5"
15" 20"
y
Fig. 2.13: Continuous Simply Supported Beam

2.72
 q = 40 lb/in E = 3 x 10 7 psi

C x .25"
A
B .5"
15" 20"
y
Fig. 2.14: Continuous Clamped - Simply Supported Beam

Homework 3
Model the plate shown below. Take advantage of the symmetry so that only a quarter of the plate
is constructed. Make two plates, one using two elements and one using eight elements. In
constructing the eight elements use your own judgment for their size and placement. Find
stresses in all elements and displacements only along the right edge. Run each plate for two
subcases of loadings, one for P=10 lb/in and one for P=30 lb/in. Find the factor of safety.

E = 3x10 7 psi
n = .33
Thickness = 0.4"

P 20" P
10"

20"

Fig. 2.15 A Plate with a Circular Hole

Homework 4
Consider the two stiffened plates shown in the Fig. 2.16. One end is fixed while the other is
free. The thickness of the plates as well as the beams is .05". The dimension of the plates are
38"x26". Each of the plates are divided into 4 equal CQUAD4 elements. Each of the three
applied forces is 100 lbfs; two vertical ones at Nodes 3 and 6 and a horizontal one at Node 6.

2.73
Find the locations and the magnitudes of the max stress in the plate and in the beam. Note that
E= 200 Gpa, n= 0.3.
Y

13

10
14

X
11 1 4 7
15

2 5 8
0.4"
12
0.6"

6 9
3
0.6"

Z
0.4"

Fig. 2.16: Angled Plates with Stiffener

Homework 5
Use NASTRAN to analyze the 2-bar truss structure
shown in Fig. 2.17. The section is assumed to be round
and solid. Let the radius of the top truss member be
17.84 mm and the radius of the bottom truss member be
19.95 mm. Material properties are the same ast those
given in Figure 2.10 of Example 4. Find the values of the
maximal stress and the deflection and the member that is
subjected to the maximal stress. Draw the free body
diagram of forces balanced at Point A. Will member AB
buckle under the given load?

E= 200 Gpa, n= 0.3

Figure 2.17: 2-Bar Truss

2.74
Homework 6
A piece of sheet metal is connected to a solid rod, which is welded to the wall, as shown in
Figure 2,18. The sheet metal transmits a point load, F of 300 lbs, to the wall. Construct a finite
element model of the problem. The dimensions are given as: l=48", d=4", h=8" and
thickness=0.5". The structure is made of standard steel. Find the location and the magnitude of
the maximal stresses in the sheet metal. Note that E= 200 Gpa, n= 0.3. (Hint: Use PATRAN to
model the bar as a beam and the arm in CQUAD4s.)

Figure 2.18: Connecting Rod

Homework 7
An axial load P=385 KN is applied to the composite block shown in Fig. 2.19. Determine the
stress in each materials. The whole system can be modeled by 5 nodes and two bar elements. A
node is assigned at the point where the load is applied.

Figure 2.19 Two Material Bars under a Loaded Rigid Plate

2.75
Homework 8

A rigid and weightless bar is suspended by two metal cables. Those cables are modeled as two
truss members. The rigid beam in Fig. 2.20 is level before the load was applied. The entire
system is modeled by five nodes and two bar elements. A node is assigned to the point where the
load is applied. Find the stresses in the members.

Figure 2.20 A Rigid and Weightless Bar Suspended by Two Cables

2.76
 III.5.3 Mesh Generation ................. 41
III.5.4 Material Selection ................ 44
Contents: III.5.5 Element Properties ............... 46
III.5.6 NASTRAN Analysis ............... 47
TABLE OF FIGURES ... ERROR! BOOKMARK NOT III.5.7 Access Results ...................... 51
DEFINED.
III.5.8 Homework ........................... 53
III.6 LAB SESSION 4: SPATIAL FRAME
III. MSC/PATRAN ................................... 3 PROBLEM USING BAR STRUCTURE ............. 54
III.1 OBJECTIVE ..................................... 3 III.6.8 LAB HOMEWORK: Testing the
III.2 INTRODUCTION ............................. 3 Mini-Baja for Vibration Analysis .... 60
III.2.1 Engineering Functionality ...... 4 III.7 - BUILT-UP STRUCTURES WITH RBE2 .. 61
III.2.2 Graphical User Interface ........ 5 III.7.1 LAB SESSION 5: Homework
III.2.3 Files ........................................ 5 Analysis of a Bolted Plate .............. 64
III.2.4 Analysis Preferences .............. 6 III.7.2 LAB SESSION 7: Using MPCs
III.2.5 Results Visualization .............. 7 and Equivalence ............................. 68
III.2.6 On-Line Help/Documentation 7 III.7.3 LAB SESSION 6: Homework A
III.2.7 Overview of Mouse ................ 8 Simple Bridge ................................. 85
III.3 LAB. SESSION 1: BUILDING A III.7.4 LAB Developing a Strategy to
WIREFRAME MODEL USING POINT, CURVE & Build ............................................... 98
SURFACE ................................................. 9 III.8 LAB SESSION 8: HOMEWORK - 45
III.3.1 Menu Components of DEGREE WEDGE FOR STATIC AND VIBRATION
MSC/PATRAN ................................... 9 ANALYSES ........................................... 108
III.3.2 Modifying the View .............. 14 Case 1. Static Case ....................... 110
III.3.3 How to print the contents of Case 2. Vibration Case ................. 113
the current viewport. ..................... 16 III.9 .DB AND .XDB GENERATION AND
III.3.4 Creating the Geometry ........ 17 VISUALIZATION ................................ 117
III.4 LAB SESSION 2: FINITE ELEMENT MESH Task 1. Create .db File .................. 117
MODEL OF PLATE PROBLEM ..................... 28 Task 2. Create .xdb File ................ 118
Homework: Lug Problem ............... 34 Task 3. Visualization of the Analysis
III.5 LAB SESSION 3: ANALYSIS OF THE 2D Results ......................................... 118
PLATE PROBLEM ..................................... 35
III.5.1 Boundary Conditions ............ 36
III.5.2 Applied Loads ....................... 38

3.2
 III. MSC/PATRAN

III.1 OBJECTIVE

The objective of this manual is to introduce the users to the

finite element modeling capability of MSC/PATRAN and to give a
hands on, end-to-end guided tour of the software. PATRAN is a pre-
and post-process supporting finite element analysis software such as
NASTRAN. The way PATRAN building a finite element mesh model
is different from other solid modelers. It is based upon the wireframe
modeling technique in which the model is built up from point, curve,
patch and then volume. Other CAD software, such as SolidWorks,
usually build the model using the solid modeling technique in which
the analysis model is directly built from solid elements through union
and subtraction.

This manual shows users how to

Start MSC/PATRAN,
Create the geometry model,
Create the finite element mesh,
Define the element and material properties,
Create load case and the loads and boundary conditions,
Prepare data for the analysis,
& Post-process the results.

This manual does not replace the MSC/PATRAN User's manual

(which can currently be browsed through in the online help menu),
but provides sufficient information to start with.

III.2 INTRODUCTION

The finite element method is a proven technique for using

computers to model a wide variety of engineering problems. Its

3.3
application in the real world was hindered, however, by the amount of
time spent both in producing the raw data to feed a Finite Element
Analysis (FEA), and in interpreting the usually large volumes of
results from the analysis. MSC/PATRAN is a software developed to
provide a systematic approach towards making FEA modeling fast and
accurate. It uses a simple step-by-step approach that helps to create,
analyze and interpret mathematically realistic model of the structure.
This approach is built around geometric modeling, interactive
computer graphics, and current finite element theory.

III.2.1 Engineering Functionality

The capabilities of MSC/PATRAN include:

1) A full set of tools for the creation of parameterized model

geometry. In addition, MSC/PATRAN has Single Geometric
Model (SGM) capability. SGM accesses geometry data, topology,
and evaluators from the CAD (Computer Aided Design/Drafting)
system and establishes and maintains associativity with the
corresponding MSC/PATRAN finite element entities throughout
the entire design and analysis process.

2) Finite modeling tools for analysis model creation and verification,

including mapped meshing, automatic surface meshing, and
automatic tetrahedral solid modeling.

3) A complete set of functional (loads, boundary conditions, and

material/element properties) assignment capabilities, including the
capability to assign these directly to the geometry or the finite
element model. In application, this means that the finite element
mesh can be deleted and the geometry re-meshed without having to
reapply the functional assignments. All functional assignments can
be collected into load cases and named, modified or deleted at the
user's discretion.

3.4
4) PATRAN Command Language (PCL) for the customization of
MSC/PATRAN, the performance of variance and design
sensitivity studies, and automation of routine, repetitive tasks.

III.2.2 Graphical User Interface

The MSC/PATRAN user interface is form driven and truly

intuitive; minimizing the time required to learn the system. At startup,
after creating a new database or opening an existing one, the
MSC/PATRAN user interface consists of a control panel, graphics
window (view port), and a command/history window. The control panel
controls the parameters for system tasks such as managing files, setting
preferences, manipulating groups and graphics, as well as switching
between interactive applications. The information displayed in the
graphics window includes geometry, finite elements, applied loads,
boundary conditions, analysis results and user defined annotation. The
command/history window contains a record of all commands executed
during a MSC/PATRAN session, as well as enabling commands to be
entered from the keyboard.

III.2.3 Files

MSC/PATRAN modeling begins with opening a file (new or

existing) by using the FILE pick in the upper left corner of the main
menu. File management options available under the FILE pick include:

create a new database,

open or close an existing database,
access CAD geometry,
import IGES models or PATRAN Neutral Files,
record

3.5
 and playback session/journal files.

The MSC/PATRAN relational database defaults to a .db

extension and contains a complete record of user created geometry,
finite elements, load cases, load and boundary conditions, materials,
element properties, preferences, and results. MSC/PATRAN database
creation is memory expensive. Each database requires approximately
6MB of disk space. Therefore old database should be compressed using
the Compact Database option under the FILE pull down menu.

The record of commands shown in the command/history

window during a MSC/PATRAN modeling exercise are stored in
editable files that can later be run through MSC/PATRAN to perform
parametric or sensitivity analyses. Two files are created by
MSC/PATRAN to store these commands:

A session file (.ses extension) which contains the record of all PCL
commands for each interactive session.

A journal file (.jou extension) which contains a record of all PCL

commands for a particular PATRAN database.

III.2.4 Analysis Preferences

MSC/PATRAN provides for the user unique analysis

preferences i.e. a user can readily switch between commercial analysis
solvers and proprietary in-house codes. Thus the user is not faced with
having to learn the command and syntax of each particular analysis
solver. Upon setting the new preference, MSC/PATRAN automatically
updates common element properties, material properties, loads,
boundary conditions, and analysis code forms to those of the newly
selected solver. Note that in this manual, our analysis solver will be
MSC/NASTRAN.

3.6
 III.2.5 Results Visualization

MSC/PATRAN post processing is state-of-the-art in its ability

to display, sort, combine, scale, and query in a general way a single
results database. After execution, analysis results are loaded directly
into the MSC/PATRAN relational database and can be sorted by time
step, frequency, temperature or spatial location. MSC/PATRAN post
processing enables the engineer to filter engineering results
simultaneously by:

Material
Element type and property
Node and element IDs
Thresholds of results
etc.,.

Display types include but are not limited to:

Deformed
Fringe
Vector
Tensor
Engineering X_Y plots

Shear and bending moment diagrams are also available for beam
results and a sophisticated text report writer is available to print out all
results in a user defined sorting sequence and format. The insight
application within MSC/PATRAN condenses mountains of raw
numerical data into graphical tools and displays for complete, accurate
interpretation of finite element analysis results.

III.2.6 On-Line Help/Documentation

The entire manual of MSC/PATRAN can be browsed through

by clicking the left mouse button on the HELP button of the Top Menu
Bar of the Control Panel. All interactive features, functions, and
applications in MSC/PATRAN can be obtained from the completely
topical and context sensitive help on-line. Hypertext links throughout

3.7
the on-line system allow for instant retrieval of complex information.
The MSC/PATRAN on-line help system eliminates the need for printed
documentation, because the MSC/PATRAN printed documentation is
identical to on-line help files.

III.2.7 Overview of Mouse

MSC/PATRAN uses a three-button mouse. Depending

on what you are doing, each of the three buttons will accomplish
a specific task in MSC/PATRAN. The following is a summary
of what you will need to know to complete the guided tour.

Three Buttons

The left mouse button is used to select model entities. To select more
than one entity, hold down the Shift key while selecting the entities with
the left mouse button.

The right mouse button is used to deselect an entity that was selected in
error. To deselect more than one entity, hold down the Shift key while
selecting the entities with the right mouse button.

The center mouse button is used to change the view by rotating,

translating, or zooming. Keep the button pressed down and drag the
mouse to move the model.

Cursor Selecting Shortcuts

To cursor select model entities with a rectangular box, hold down the
left mouse button at the upper left hand point and drag the cursor to the
lower right hand point. Then, release the mouse button.

To cursor select entities with a polygon shape box, hold down the
Control Key (Ctrl) and use the left mouse button and start picking at
arbitrary points A, B, C, D, to define a closed polygon box.

3.8
 III.3 LAB. SESSION 1: Building a Wireframe Model using
Point, Curve & Surface
Topics Covered:
Menu Components Break

Using the viewport Fillet

Using the mouse Surfaces

Printing Correcting Mistakes

Creating Geometry Transforms

Points Mirroring

Curves Translating

III.3.1 Menu Components of MSC/PATRAN

MSC/PATRAN Control Panel

MSC/PATRANs Control Panel appears at the top of the screen

(example below). The selections on this panel are divided by a
horizontal line. The selection above the line is MSC/PATRANs Top
Menu Bar, which includes MSC/PATRANs heartbeat and the on-line
help system. The items below the line are a group of radio buttons and
a group of Application radio buttons.

Figure 1 - PATRAN Control Panel

3.9
 The Heartbeat icon.
MSC/PATRANs heart indicates whether Patran is busy or
waiting for input from you.

A stationary MSC globe with a grey background on Windows

machines
means ready and waiting. (As Shown)
Figure 2 - Patran
A spinning MSC globe with a blue background means busy, but
Heartbeat
can be interrupted.
A spinning MSC globe with a red background means busy, cannot
be interrupted.

The Menu Bar

The Menu Bar selections
The menu bar contains menus to help setup and operate your
are pull-down menus
model.
which will be explained
later in using the mouse. File
This menu includes basic file options:

New
Open
Close
Save
Print
and others

The menu also includes more advanced options such as

Import
& Images

Group
Using Groups in this course is not necessary to complete any of the
assignments, however, it can be a valuable tool to save time and effort
where tedious actions can be avoided by copying and mirroring.

3.10
Groups separate you model into parts that allow you to have greater
accessibility to parts of your model that would have otherwise been
hidden or blocked.

Viewport
In creating a new *.db file, a viewport is automatically created. This
menu allows you to create, modify, delete or tile your viewports. Use
of these options are pretty self-explanatory.

Viewing
Contains advanced options for controlling the viewport. The most
common methods are found elsewere: [see Toolbar > Home >
Viewport Radio Buttons]

Display
This useful menu includes methods to:

Adjust the colors of the different entities

o Points, curves, nodes, forces etc
Display the specific labels for the entities
Change the view between 1-D, 2-D, & 3-D
Add a title to the display

Preferences & Tools [Not Necessary]

Preferences includes advanced functions for user interface, while
Tools contains advanced analysis tools that are beyond the scope of
this course. These two menus will not be needed.

Help
This menu allows access to the online database of the full technical
operation manual for PATRAN as well as an index, technical support
and tutorials.

Utilities

Getting Started on a Project

To start

The Toolbar

3.11
 The toolbar is a collection of organized tools which will be
used in the construction of your FEM. The categories are:

Home
These buttons are divided as such:

The Defaults group includes radio buttons for filing and

general PATRAN operation. Such as:
o New, Open & Save File
o Print
o Undo (Last Action)
o Abort (Current Process)
o Copy to Clipboard (viewport)
Figure 3 - Defaults group o Refresh Graphics
o And others

The Transforms group includes mouse operations for

controlling the viewport:
(All these operations can be accessed by
using a combination of the middle
mouse button, Shift and Ctrl.)
Figure 4 - Transforms
o Pan
Group o Rotate X/Y
o Rotate Z
o Zoom

The Viewport group

(These buttons help with the Viewport
display)
View Corners
Figure 5 - Viewport Fit View
Group Rotation Center

Radio buttons are mutually exclusive choices in that only

one radio button within a group can be pressed in at one
time. They work like the radio buttons in your car. Only
one button on your car radio can be pressed in at one time.

3.12
 MSC/PATRAN Application Form and Subordinate
Form

The radio buttons on the Control Panel will bring up a vertical

form on the right side of the screen. These forms are called the
Application forms. There are also smaller forms that are sometimes
displayed which are associated with the specific Application form, and
they are called Subordinate forms.

MSC/PATRAN Command Window

MSC/PATRAN also has a Command Window that appears at

the bottom of the screen (see below). You can manually enter
commands in this window, but mostly this window is used to view the
commands MSC/PATRAN generates when a menu form is executed,
and to view errors or information messages.

Figure 6 - Command Line

The history window. You can use the
The command line. Commands scroll bars at the bottom and right sides
can be entered here to scroll back or to scroll to the right.

PATRAN Help

MSC/PATRANs on-line help is an easy to use, hypertext-

based documentation system. You can enter the on-line help either by
pressing Help on the Menu Bar, or by placing the cursor anywhere on
a MSC/PATRAN form and pressing the F1 key to get help on that
particular form. If your cursor selects anything that is in magenta in

3.13
the on-line help, MSC/PATRAN will quickly jump to another page
in the on-line help that it is referencing.

III.3.2 Modifying the View

There are several ways to alter the view of your model. Below
are some of the more common ways.

Using the Middle Mouse Button

As mentioned earlier, you can rotate, translate, or zoom in on

your model by dragging the mouse, while holding down the middle
mouse button.

Hold down the middle mouse button (MMB) to rotate

the X/Y axis.
Hold down Ctrl and the MMB to zoom.
Hold down Shift and the MMB to Pan the X/Y view.
Hold down Ctrl, Shift and the MMB to rotate the Z
axis.
To change these settings:
o Select Preferences
Mouse
Mouse Preferences

View/Select Corners

You can zoom in on a particular part of your model, and at the same
time, define a new viewing center by selecting from the Menu Bar:

MSC/PATRAN will turn the cursor into a cross-hatch. Define a

new viewport by dragging the cursor diagonally and defining a
rectangular box. Drag the cursor along a diagonal direction to select a
portion of the model you want to view up close. Select View/Fit View
if you want to return the view back to the entire model.

3.14
 View/Transformations

You can rotate or translate the view incrementally by selection from

the Menu Bar.

A subordinate Transformations form will appear with a matrix of

buttons. By pressing any one of the buttons, you can translate, zoom,
or rotate your model incrementally.

For example by pressing a button, MSC/PATRAN will rotate

your model about the screen or models Y axis

To perform a Fit View, where MSC/PATRAN will resize the

view of the model to fit within the viewport, you would simply press
the assigned button.

More view/transformation buttons are shown on the next page.

Show the entire model with hidden lines

Show the entire model without hidden lines

Show the shaded solid view of the model

Show the end view in the Y_X plane

Show the end view in the Z_X plane

Show the end view in the Z_Y plane

Show the isometric view

Show all the labels (set labels on)

Do not show all the labels (set labels off)

3.15
 III.3.3 How to print the contents of the current
viewport.

To print a hardcopy output of the contents of a viewport, the user

should click the left mouse button on the 'File' option of the top menu
bar of the control panel and select 'Print' in the pulled down menu. The
following application form is shown on the screen:

If you are printing

from class then use the
classroom designated
printer.

If you are printing

from MOVE.odu then
change the printer to
PDF. This will allow
you to save the
document as a pdf and
can email it to print
elsewhere.

Figure 7 - Print Window

3.16
 III.3.4 Creating the Geometry

Starting a new project

Items Covered in this exercise:

Creating a new Database Creating Title
Creating Geometry: Undo
Points Delete
Curves Radio buttons:
Surfaces ZYX view
Solids Display Lines
Translating points, curves

File > New > (this pulls up the New Database window):
Note that one should
click on Modify
Preference to check
whether
MSC/NASTRAN is
selected as the analysis
before proceeding. In
the form shown above,
A new data base name
LabManual has been
entered. The user can
supply his/her own
name.
Select OK

Figure 8 - New Database

3.17
 After the database has been created, the following form
for model preferences is automatically displayed:

This form is completed for tolerance limit,

Approximate maximum dimension of the model
to be created,
& analysis code of choice and analysis type.

The form shown here has been completed for structural

analysis with MSC/NASTRAN as the analysis code.
As said earlier (see p.2), our analysis code will be
MSC/NASTRAN. After the form is completed,OK
button is selected.

Figure 9 - New Model Preference Window

For displaying your drawing

geometry go to:

Display > Geometry > to check

Point, Curve, and Surface as
shown in Fig 11.

Figure 11 - Display Menu

Figure 10 - Geometric Attributes

3.18
 Creating Geometry
The next step is to start the creation of a geometrical model. In this lab
session, the reader is taken through creation of basic geometrical
entities such as point, curve (1_D), surface (2_D), and solid (3_D).

Create Points
The creation of all geometrical entities is initiated by clicking
the left mouse button on the GEOMETRY radio button of the control
panel (see Figure 12). A typical form that needs to be completed is
shown below (The left mouse button is used to move through the form).
Note that in MSC/PATRAN, the input format of the coordinate of a
point, (x, y, z) is given by [x, y, z] , which is different from that of a
vector given by < x, y, z >.

The form on the left has been completed for the creation of a point at
(x,y,z) = [0,0,0]. It should be mentioned here that there are several
options under Action, Object, and Method. The choice of
combination of options for these three is based on the result desired. Figure 12 - Geometry Window

Further examples will be seen in the course of this manual. After a

geometrical entity has been created, the figure shown below (Visible
Entities) is seen at the bottom of the screen. This figure highlights
geometrical entities that can be selected by clicking of the left mouse
button on the desired entity. The second point is at [5 0 0].

The figure below shows the view port after two points have been
created.

Figure 13 - Create 2 Points

3.19
 Create Curves
The next stage is to create a curve. The form below has been
completed for the creation of a 2_point curve using the two
points created earlier:

The result of applying the above form is shown below:

Figure 14 - Geometry Create

Figure 15 - Created Curve

Next, we illustrate the creation of a 3_point arc. The third point is at

[2.5, 2.5, 0]. For this purpose the completed form is shown below:

Dont select Auto Execute. Instead, click on Apply. It

gives the user a little room to correct typo errors. The result of
applying the above form is the arc (curve 2) shown in the figure
below:
2
3

Figure 16 - Creating Arc 1 1 2

Z X

Figure 17 - Created Arc

3.20
 Creating Surface
The next stage is to create a surface. An example is given here
of how the surface can be created from two curves.
A surface can be created
from 2, 3, 4, or N curves
where N is a number to
First, the 3_point arc should be removed.
be supplied by the user.

2 Methods to Correct a Mistake

Delete
Select delete under action
Click the 3_point arc in the viewport.
o Because the object is Any, anything you click on
will be deleted.
o Observe the different Objects that can be selected
from the pull down menu.

Undo
On the quick access toolbar you can click the undo button

Next, the form on the left is completed for creating a second curve by
translating the first 2_point curve:

Figure 18 - Delete Action

Figure 19 - Undo Command

3.21
 Translating Curve
In the displayed form,
Take note: The direction
Direction Vector < 0 5 0> vector can be of any
Direction Magnitude 5 value. The only data that
Curve List Curve 1 is collected is the direction
that the vector points to.
The Direction Magnitude
contains the value for the
The result of applying this form is
distance of Translations
shown here:

3 2 4

Z X 1 1 2

Figure 20 - Translated Curve

Take note: If you want only to move

the selection and not have a copy,
select the Delete Original

3.22
After the two curves have been created, the form here is completed for
the creation of a surface:

Applying the above form, we have the result shown below:

3 2 4

Y
1 1 2
Z X

Figure 21 - Create Surface

The Option we selected for this example is 2

Curve. This option is the most common and
the only one we will need in this course.

3.23
 Creating a Solid
The next stage is to create a solid. Again, there are several ways
of creating a solid. In this example, we are using the 2 Surface option.
Hence we need to create another surface.

Translating a Surface
This is done by translating the previous surface 5 units of distance in the
Z direction (i.e. Translation Vector: <0 0 5>). The form completed for
surface translation is similar to that of curve translation with Surface
as the Object. After surface translation, we have the result below:

8
2
6

3
2
4
1
1

Display Line, Clicking Z

Y
1
2
X
this button will show
cause lines to appear on
the generated surfaces.

8
2
6

3
2
4
1
1

Z 1
Y 2
X

Figure 22 - Display lines command

3.24
 The rectangular solid is created by completing the form shown:

Applying the form on the right, we have the result shown below.

5
Solid Model

8
2
6

3
2
4
1 Figure 23 - Geometry create
1
window
1
Z
Y 2
X Figure 24 - Created Solid

Inserting a Title

To write a note or legend

on the model, click on
Display, then Title
and follow the
instructions to fill out
the proper information.
Click on Apply. In the
following picture, the
title is given as Solid
Model.

3.25


III.3.4.1 Exercise: Break, Trim and Fillet

In this exercise, we will use break, trim and fillet functions to
create a circle with 2.5 units in radius out of a 55 square.

Step1: Create 4 points: point 1 [0 0 0], point 2 [0 5 0], point 3

[5 0 0], point 4 [5 5 0], with which to create 4 curves.

Step2: Use the break option to bisect each curve into 2. New
curve IDs are created along with new nodes.

Step3: Select Fillet for Method in Geometry to construct

Figure 25 - Edit Curve window

Figure 26 - Delete caution window

fillets. See
the menu on
the left.

3.26
A fillet will be tangent to two intersected curves. PATRAN will
require you to select these two curves and their intersected point.
Turn on both boxes of Trim Original Curves and Auto Execute in
the manu. Assign the curve ID for the fillet and give its radius.
Click on the first selected curve, the cross point and then the second
curve. The result is displayed hereafter. Repeat three more times to
complete a circle.

Figure 30 - Creating Fillet Figure 27 - Inserting Fillet

Step4: Use method Trim to remove the original curves that form the
circle. To remove a curve, click on the first end point of the curve to
be removed, the curve itself and then the other end point. Be sure
that Auto Execute is on. Repeat it four times to remove the original
box.

Figure 28 - Trimming curve

Figure 29 - Trimming

3.27
 III.4 LAB SESSION 2: Finite Element Mesh Model of Plate
Problem
Topics Covered:
Practice using: Transforms

Geometry Mirroring

Break Translating

Fillet

In this lab session, try to create the plate shown below. The
steps used in generating the plate are illustrated. You are not limited to
using these steps. You can obtain the same results using any other
combination of steps if you are confident of doing so.

Step 1: Create a new database named 'classwk1' (or any other name
you prefer) by clicking on 'File/New Database' of the control panel.
The following form is displayed. It is completed by supplying the new
data base name.

Figure 29 - Plate Problem

Step 2: Set the model parameters by completing the model preferences

form shown below. This form is automatically displayed after a new
database has been created or an existing one opened.

3.28
Step 3: Start geometry creation by clicking on the 'Geometry'
button of the radio panel. The following form is displayed. The
form is completed for point 1 (at [x y z] = [0 0 0]) as shown.

Step 4: Similar to point 1, create points 2 and 3 at [x y z] of [5 0

0] and [0 3 0] respectively.

Step 5: Create point 4 by translation method. This is done by

completing the form shown below. The form has been
completed for creation of point 4 by translating point 3 through
a vector of <5 0 0>. The Point ID List is shown to be 5 because
the 'Apply' button has been clicked on and point 4 has been
created.

Step 6:

Similar to point 4, create point 5 by translating point 2 through

<0 1 0>.

Create point 6 by translating point 4 through <0.5 0 0>.

Create point 7 by translating point 5 through <0.5 0 0>.

Create point 8 by translating point 7 through <2 0 0>.

Create point 9 by translating point 6 through <2 0 0>.

Create point 10 by translating point 9 through <2.5 0 0>.

Create point 11 by translating point 8 through <2.5 0 0>.

Create point 12 by translating point 10 through <2.5 0 0>.

Create point 13 by translating point 11 through <2.5 0 0>.

Figure 30 - Translating geometry

3.29
Step 7: This step begins the creation of curves. Create a curve
by selecting 'Curve' as 'Object' in the 'Geometry' form. The form
is shown below completed and already applied (hence 2 is
shown in the Curve ID List) to create curve 1. Curve 1 is created
using the 2_point method and the points used are 1 and 2.

Step 8: Similar to curve 1, create curves 2 through 11 by using

'Starting Point List' [3, 4, 2, 5, 6, 7, 8, 8, 11, 13] and 'Ending
Point List' [4, 6, 5, 7, 9, 8, 9, 11, 13, 12] respectively. In other
words, curve 11, for example, is created by joining point 13 and
point 12.

Step 9: Create curve 12 by using the '2D ArcAngles' option

under 'Method'. A form to be completed for this is shown below.
This form has been completed and applied for curve 12 (hence
13 is displayed in the 'Curve ID List'). For this curve, the
following information was supplied: Radius (1.0), Start Angle
(180), End Angle (225), Center Point List (Point 10).

Step 10: Similar to curve 12, create curves 13 through 15 using

the same radius and center point list as curve 12 but with start
angles 225, 270, 315 and end angles 270, 315, 360 degrees.

Step 11: Create curve 16 by using the 'Fillet' option under

'Method'. The region of fillet is shown in the plate figure. The
form displayed to be completed is shown to the left. This form
has been completed for the fillet curve. The following
information was supplied: Fillet Radius (0.5), Curve/Point 1 List
(click on curve 5, then point 5), Curve/Point 2 List (click on
curve 4, then point 5). After the form is completed, P3 asks if
the original curves should be trimmed. A window is displayed
for this (see below). The option 'Yes' or Yes For All' is
selected.

3.30
A new point 19 and a new curve 16 are created. The result is shown:

Figure 31 - Trimming, Before and After

3.31
 Step 20: Trim the curve between points 5 and 7 by using Edit
under Action. Select Trim under method and point under
option. Select Trim under method and point under
option. Select point 5 and curve 5 and point 7 for Trim Point
List and Curve/Point List. Click on Apply after the form is
completed. Also, Trim the curve between point 5 and point 19

Step 21: Create Surface 1 using the 2_curve option. This is done by
completing the form shown below. The form has been completed
and applied for surface 1 (hence it is ready for surface 2, i.e.
Surface ID List = 2). The 'Starting Curve List' for surface 1 is
'Curve 16' and the 'Ending Curve List' is 'Curve 3'.

Figure 32 - Create Surface, 2-Curve

method

Figure 33 - Adjusting Geometry

3.32
 Figure 36 - model with surfaces

Step 22: Similar to surface 1, create surfaces 2 through 7 with 'Starting

Curve List' 2, 6, 12, 13, 14, 15 and 'Ending Curve List' 1, 17, 8, 9, 10,
11. For example, surface 3 is created from curves 6 and 17.

Step 23: Before we continue, we want to set the 'Display Property'

such that only surface numbers will be displayed, the curve and
point numbers will be hidden. This is done by clicking on the
Display' button on the control panel then choose Geometry option.
This option allows the surfaces to be displayed with the fewest
number of parametric lines. Next, all label buttons in the 'Entity
Types' section of the form are deactivated except the surface label
button. Select Apply.

3.33
 Figure 37 - Model after mirroring

Step 24: Finally to complete our plate geometry creation, we mirror

the figure in the viewport which is half of our geometry. This done by
completing the form below. The mirror plane normal is defined as {[0
3 0][0 0 0]}. The 'Surface List' is selected to include all existing
surfaces (i.e surfaces 1 through 7).

Step 25: The created geometry can then be printed by following the
example shown on p-3.16. The user is free to supply his/her own file
name.

Homework: Lug Problem

Now as your homework, generate the geometry of the LUG

shown below. It is not necessary
for the patch numbers to match
those given in the figure. The left
edge of the lug is completely
fixed and two 1000 lbs point
loads applied at the top and the
bottom of the hole as shown in the
following figure. The material
properties and the thickness of the
plate are all the same as the
example problem discussed in
Section III.4.
Figure 34 Lug Problem

3.34
III.5 LAB SESSION 3: Analysis of the 2D Plate Problem
Topics Covered:
Cquad Structure: Materials

Loads Analysis

Boundry Conditions Accessing ResultS

Meshing

In this lab session, we will analyze the plate of Lab Session 2.

In particular we will learn how to use the following radio buttons:

1. Load/BCs: To assign boundary

conditions and loads

2. Load Cases: To create and load cases

3. Finite Element: To generate finite

element mesh.

4. Materials: To assign material properties.

5. Element Props: To assign element

properties.

6. Analysis: Patran generates bulk data file

(file_name.xdb) and access Results for
analysis automatically. Figure 35 - Open Database

7. Results: To graphically view analysis results.

To accomplish the above, we will follow the steps shown below. Load
the geometry of the plate generated in Lab Session 2.

Open the MSC/PATRAN database saved in the Lab Session 2

o Click on File/Opening an existing database.
o This database is automatically
displayed in 'Database List'.
o The database can then be
selected by clicking on its
name in the list.

3.35
 III.5.1 Boundary Conditions

The next step is to assign boundary conditions. The desired

boundary condition is to fix the left edge of the plate, that is, to
constrain all six degrees of freedom for the left edge.

To achieve this, the 'Load/BCs' radio button is clicked on.

o This will bring a display of the form shown below on the left.
This form is completed as indicated.
The 'New Set Name' for the boundary condition we will call:
'fixed'.
The new set name is just for the user to identify the data, it is not
used in NASTRAN input file.

To complete the boundary condition

specification, we have to 'Input Data' and
'Select Application Region'.

To input data:

Click the 'Input Data...' button

o The form on the right is displayed.
This form is completed as indicated.
'Translations' vector is <T1 T2 T3> = <0 0 0>
'Rotations' vector is <R1 R2 R3> = <0 0 0>.
o This implies that the six degrees of freedom
are constrained.

Figure 36 - Loading Boundry

conditions
Figure 37 - Inserting B/C Data

3.36
To select the application region:

Click the 'Select Application Region...' button

o This will bring a display of the form shown:
Select 'Geometry'
o Since we are assigning the boundary conditions before the
finite element mesh is generated.
Click inside the 'Select Geometry Entities' text region.
o At this point, since we want to apply the boundary condition to
the left edge of the plate, we first click on the edge icon in the
'Visible Entities' display.
o The select menu will then state 'Select a Curve'.
At this point, one can move the cursor to select the left edge of the plate
by using a rectangular cursor selection.
After this is done, Point 1, Point 3, and Point 23 are shown in the 'Select
Geometry Entities' window.
We then click on 'Add' and Point 1, Point 3, and Point 23'' is now shown
in the 'Application Region' window. Figure 38 - Inserting B/C
After this is completed, click on 'OK' to conclude this part. Application Region
Then press 'Apply' on the 'Load/BCs' Application Menu form to
conclude the whole step.

Notice that this

boundary
condition is
applicable to any
future finite
element mesh
generated for the
plate, that is, all Figure 43 - Model demonstrating
fixed edge

Notice the 'cones' at the left edge of

the plate at the display lines' point
location. These cones represent the
three fixed translational and three
At this point, you should have the figure shown below.

3.37
 III.5.2 Applied Loads

The next step is to assign loads to the plate. In this example, we

will have two point (nodal) loads, one in the negative Y-direction and
the other in the positive X-direction, the lines of action of both forces
passing through the center of the circular hole in the plate. To achieve
this we complete the same set of forms as for the boundary conditions
for each of the point loads, one after the other, but with the following
modifications.

The 'Object' is 'Force' rather than 'Displacement'.

For the first point load:
o 'New Set Name' of '1000_pound_down' can be used and
o for the second point load, '1000_pound_right' can be
used.
These names are just for the user's benefit, they
are not used in NASTRAN's input file.
In the 'Input Data...' form, the following force vectors are
provided: <F1 F2 F3> = <0 -1000 0> for the first point load and
<F1 F2 F3> = <1000 0 0> for the second point load.
For both loads, the moment vector is <M1 M2 M3> = <0 0 0>
since no external moment is applied.
In the 'Select Application Region...' form, when the cursor is in
the 'Select Geometry Entities' window, the 'Point' icon is
selected instead of the 'Curve' icon.
For the first load, move the cursor to select the point at the top
of the line dividing surfaces 12 and 13.
For the second load, the cursor selects the point at the middle of
the right edge of the plate.
After 'Apply' is pressed for each of the loads in the 'Application
Menu' form, the figure below is the result.

3.38
 Figure 39 - Model demonstrating load conditions

Notice the arrows drawn

at the selected points
Now a load case will be created that combines the two point signifying the
loads and also the specified boundary conditions. This is done by 1000_pound force in the
clicking on the `Load Cases' button of the control panel. The negative Y_direction
following form is then displayed. Initially under `Existing Load and 1000_pound force
Cases', we have the `Default' load case listed. This load case is always in the positive
present in any analysis, irrespective of how many more load cases are X_direction. Each arrow
created by the user. Then we specify the new `Load Case Name' as is displayed as soon as
`lcase_1'. We make the new load case current. The user is free to give the load corresponding
some description for the new load case under `Description'. This is to it is created.
optional. The next step is to go to the list of `Assigned Load/BCs Sets'
and then select the combinations of loads and boundary conditions that
will constitute the load case being created. For our example, we select
all the loads and boundary conditions in this list to constitute our load
case. This means that our load case is made up of `Displ_fixed',
`Force_X_direction' and `Force_Y_direction'. After the form is
completed, we press on `Apply' to conclude the creation of the load case.
By now, we should see under `Existing Load Cases' our newly created
load case i.e. `case_1'.

3.39
 Figure 41 - Inserting data for Load Condition

Figure 40 - Creating Load Case

3.40
 III.5.3 Mesh Generation

The next step is to create the finite element mesh for our plate.
This step is divided into two, the first is to create mesh seeds for the
curves bounding our geometry. Mesh seeds allow us to define exactly
how many elements we want on a selected curve or edge of a surface
or solid. The number of mesh seeds on a curve is equal to the number
of elements on that curve. An example is given here of how to create
10 mesh seeds on curve 1 of our geometry. The radio button
'Elements' is clicked on and the form on the left is displayed. The
form is completed as indicated. Mesh seeds are created for other
curves one after the other such that surfaces 2 and 9 have 5x10
elements each, surfaces 1 and 8 have 3x4 elements each, surfaces
3 to 7 and 10 to 14 have 4x4 elements each.

After the mesh seeds have been created, the next thing is to create
mesh for our geometry. This is achieved by selecting mesh as the
'object' in the form shown in the middle below. The 'type' of element
is chosen as surface since we are dealing with a 2_D problem. The
'element topology' selected is 'Quad4', i.e., quadrilateral elements
with four nodes. We select 'Isomesh' as our 'Mesher'. All these
selections are then applied to all the surfaces 1 to 14 as indicated in Figure 42 - Creating Mesh seed
the 'Surface List'. To select all surfaces, we can use the window
cursor after we have clicked on 'Surface' icon in the 'Visible Entities'
display. After the apply button is pressed, the entire geometry is
subdivided into 'Quad4' elements.

3.41
 Before concluding the process of mesh generation, some
modifications are made to the nodes at the interface of
surfaces 8 and 9 and also the interface of surfaces 1 and 2.
The modification involves making, at the interface, the
nodes for surface 9 coincident with the corresponding nodes
for surface 8 and the nodes for surface 2 coincident with the
corresponding nodes for surface 1. This is achieved by
moving the nodes to their new positions. An example is
shown in the form above on the right where 'Node 75' is
moved to coincide with 'Node 4'.

Figure 43 - Creating Mesh Surface

Figure 44 - Manually Moving Nodes

3.42
 After all the modifications are completed, the plate with the
generated element mesh looks as shown below.

The process of mesh generation is completed by doing the

'Equivalence' and 'Optimization'.

Equivalencing forces nodes that are within some specified tolerance

limit to coincide. The tolerance used here is 0.005. This means that any
two or more nodes within 0.005 unit of distance from one another will
be forced to coincide at the position of the node with the lowest node
identification number within the group. Examples of nodes affected by
Equivalencing are the nodes along the symmetry line and along lines
common to adjacent surfaces. The form beside is completed for
'Equivalence'.

3.43
 Optimization is done to minimize the memory requirement of
the global stiffness matrix by rearranging the elements of the
matrix such that the storage bandwidth is reduced. The form
below has been completed to 'Optimize'. The table below the
form is a result of the optimization when the 'Apply' button is
pressed. Notice the drastic reduction in the bandwidth after
optimization! Pressing 'OK' will make the table disappear.

Figure 53 - Model after equivalencing, Mashing is

now aligned

Figure 45 - Equivalence Meshing Figure 46 - Optimizing

window

III.5.4 Material Selection

In this section, we will create the structural material that constitutes our
elements.

We start by clicking on the 'Materials' button on the control

panel. The form shown below on the left side is displayed.

Figure 47 - After Optimitzation

3.44
 We complete the form as shown.
o Our material is isotropic and the properties will be
input manually.
We supply the material name under 'Material Name'.
After this we click on the 'Input Properties' button and we
have the form shown below.
Select the constitutive model which in this case is 'Linear
Elastic'.
We also fill out the property sheet as shown:
o 'Elastic Modulus'
o 'Poisson Ratio'
o 'Density'.
o press 'Apply'
We then press 'Cancel' to complete the process.
We now see in the left side form under 'Existing Materials'
the name 'Steel'.

Figure 48 - Creating Material

Figure 49 - Data input for Material

3.45
 III.5.5 Element Properties

This section deals with specification of properties of our

elements. This is started by clicking on the 'Element Props' button
on the control panel. This gives a display of the form shown on the
left below. This form is completed as indicated. The elements are
2_dimensional in geometry. Hence the element type chosen is
'shell'. The name `plate_shell' is given to the property set being
created. Under 'Option(s)' , the material is chosen to be
`Homogeneous'.

The next step is to input the element properties. This is done by

clicking the left mouse button on `Input Properties...'. The form
shown on the right below is then displayed. The form is completed
as shown. The material name is 'steel' where steel is the material
created in Step 5. Notice that 'm:' precedes the material name. This
indicates that the elements whose properties are being created are
made of material `steel'. The element thickness is specified to be
`0.125'. For the problem we want to solve, the information
supplied so far is sufficient. Thus the element properties creation
is concluded by pressing `OK' on this form and then `Apply' in the
Figure 50 - Element Properties left form.

Select Members box in Application Region. (Note that you

can not type in the member numbers
directly into Application Region
box.) One can either type in the
members selected (e.g. Node 1:10,
Surface 1:14, etc.) or use the mouse
to select the members by boxing out
the members of concerns. Next, click
on Add button. The ID numbers of
the members selected will appear in
the Application Region box and then
Apply in the left form.
Figure 51 - Element Proerties Data input

3.46
 III.5.6 NASTRAN Analysis

It should be mentioned here that MSC/PATRAN

automatically generates the NASTRAN bulk data file (.bdf) and
result file (.XDB). The way we do this is shown here. The first
thing is to click on the `Analysis' button of the control panel.
We have the form shown on the left below. We want to analyze
the `Entire Model' in `Full Run'. The `Job Name' we supplied
is `classwk1'. The user is free to use any job name. Notice that
the analysis code to be used is MSC/NASTRAN and the
solution type is `Structural'.

When the `Translation Parameters...' button is clicked on, we

have the form displayed on the right above. All default values
in this form are accepted. Notice that we have `XDB and Print'
under `Data Output' This ensures that results will be printed
both in the XDB format and the .f06 ascii format.

Figure 52 - Analyzing Model

3.47
 Note:
For newer PATRAN version access to XDB and Print Option
can be found under:

"Solution Type",
o "Solution Parameters",
then "Results Output Format",
finally click "OK" button to choose the
desired option.

Clicking on the `Solution Type...' button will give the form

shown below. The form is accepted as defaulted by pressing OK.

Figure 53 - Setting Solution Type Figure 54 - Setting Solution Parameters

3.48
 The next step is to create subcases for the analysis. This is done by
clicking on `Subcase Create...' which then displays the form on the left
below. To make lcase_1 (i.e. the load case created earlier) a possible
subcase, we click on lcase_1 under `Available Load Cases' and then
press apply. This will place lcase_1 under `Available Subcases'. To
make requests for the variables we want in our output for this subcase,
we click on `Output Requests...' and we have the form on the right
below. In this form, we click on `Displacements', `Element Stresses' and
`Constraint Forces' one after the other under `Select Result Type'.
Pressing `OK' will place these three output requests under `Output
Requests' as indicated. The indicated output requests will be printed
both in the .f06 and .xdb files. We then press 'Cancel' to conclude this
selection. We then press `Apply' in the form on the left to conclude
`Output Requests' and `Cancel' to conclude `Subcase Create'.

Figure 55 - Subcases

3.49
 The last step is to select the subcases (out of the subcases created
above) that will be used for analysis. This is done by clicking on
`Subcase Select...'. This will display the form shown below. We click
on lcase_1 under `Subcases For Solution Sequence: 101' and then press
`OK'. This will place lcase_1 under `Subcases Selected'. Thus lcase_1
has been selected for analysis. We then press `Cancel' to conclude this
step.

After all these have been concluded, we press `Apply' on the

main application menu form and MSC/PATRAN starts the creation of
classwk1.bdf (the bulk data file for the job `classwk1'). At this point, the
user waits until this process is completed. After the creation of
classwk1.bdf, MSC/PATRAN automatically submits the file for
analysis to NASTRAN without the user
intervention. At the conclusion of
NASTRAN analysis, we will now have
in the current directory, among the files
generated by NASTRAN,
classwk1.xdb and also classwk1.f06
(an ascii result file). At this point we are
set for post processing to see
graphically our results.

Figure 56 Subcases Select

3.50
 III.5.7 Access Results

Change action to Access Results,

check Object to attach XDB file.
Select the file on the available Jobs
Click Apply.

When the translation process is completed, we are ready to view our

results. This is achieved by clicking on the 'Results' radio button on the
control panel. We then have the
form shown below on the left in
which we have chosen 'Basic' as the
'Form Type'. The result for our
load case, i.e., load_case_1, is
automatically placed under 'Select
Result Cases'. We then click on our
load case name as shown on the
right below and the results that can
be displayed by fringe plotting are
indicated under 'Select Fringe
Result' while those that can be
displayed line plotting are indicated
under 'Select Deformation Result'.

Figure 58 - Accessing Results Figure 57 - Plotting results of

Analysis

3.51
 Clicking on '2.1 _ Displacements, Translational' under 'Select
Deformation Result' will lead to the display of the deformed shape
shown below.

Figure 59 - Color Shows Deformation

Clicking on '3.1 _ Stress Tensor' under 'Select Fringe Result' will

lead to a display of the fringe plot of the stress tensor superimposed on
the deformed shape as shown below.

Figure 60 - Color Shows Stress

3.52
 III.5.8 Homework

For the lug shown on p.Error! Bookmark not defined. whose

geometry you have generated), specify the loads and boundary
conditions indicated in the figure. Also generate a suitable finite element
mesh and perform NASTRAN analysis. Submit a print out of the
deformed shape and the fringe plot of the Von Misses Stresses.

3.53
 III.6 LAB SESSION 4: Spatial Frame Problem Using Bar
Structure

Topics Covered:
Beam Structure: Materials

Loads Analysis

Boundry Conditions Accessing Results

Meshing

The objective of this lab session is to demonstrate the use of the

'Element Props' radio button for problems involving one dimensional
bar or beam elements. The example problem is the frame shown below
which is made up of 1D straight curves. Geometry creation, boundary
conditions and loads assignment, load case creation, finite element mesh
generation, material properties assignment, submission of job to
NASTRAN for analysis and post processing can be handled in a manner
similar to the last example. Hence, enough information will only be
provided to guide us through the example. Necessary detailed
explanations will be given under 'Element Props'

Figure 61 - Session 4

3.54
 III.6.1 Geometry

Points 1 to 8 in the above figure are at (x y z) of [0 1 0], [1 1

0], [1 1 1], [0 1 1], [0 0 0], [1 0 0], [1 0 1], [0 0 1] respectively.
2_Point straight curves are generated as shown by connecting points
(1 & 2), (2 & 3), (3 & 4), (1 & 4), (1 & 5), (2 & 6), (3 & 7), (1 & 4),
and finally, (4 & 8).

III.6.2 Load/BCs

All six degrees of freedom are constrained for points 5

through 8, that is 'Translations' and 'Rotations' vectors are both
equal to <0 0 0>. Force vectors of <0 -120 50> and <30 40 0> act at
points 1 and 3 respectively. For both points, the moment vector is <0
0 0>. A load case with any name the user wants is created from the
specified boundary conditions and the imposed point forces or you
can use Default load case.

III.6.3 Mesh Generation

First mesh seeds are created, one mesh seed per curve. Thus
each curve is represented by one element. After this, a mesh is
created to connect the nodes created by the mesh seed operation. Figure 62 - Creating Meshing
During mesh creation, in the 'Application Menu' form, 'Curve' is
chosen under 'Type' and 'Bar2' is chosen under 'Element Topology'.
Under 'Curve List', the entire geometry is chosen, i.e. Curves 1 through
8. 'Equivalencing' and 'Optimization' steps are carried out as usual.

3.55
 III.6.4 Material Selection

An isotropic material 'Input Properties...

o 'Elastic Modulus' of 3.E7
o 'Poisson Ratio' of .33
o You can specify any name for the material under
'Material Name'.

III.6.5 Element Properties

Proceed with element properties' specification as in the previous

example. However in the 'Element Props Application Menu' form,
Object should be '1D' and 'Type' should be 'Beam'. Under 'Options'
select 'General Section' In this example, even though the section
properties of the bar elements are the same, the elements belong to three
different groups, each group having its own orientation vector.
Therefore, three property sets will be assigned, each set for each
orientation vector. The first orientation vector is for the elements in the
two curves parallel to the x_axis, the second is for the elements in the
two curves parallel to the z_axis, and the third is for the elements in the
four curves parallel to the y_axis. The user is free to supply 'Property
Set Name' of choice for each of these sets. For the first property set,
'Input Properties...' is clicked on and the appropriate form is displayed.
A portion of this form is shown below.

3.56
Figure 63 - Creating Beams

The information required to be supplied in the above form are

listed below. Where particular information is applicable to our example,
the required value (real scalar or vector) is given in front of the
information. The values given are for the first property set, i.e. for
curves parallel to the x-axis.

Material Name: m:steel

(Note: Material name 'steel' is used for
illustrative purpose)
Bar Orientation: <1 1 0>
(Note: This is the orientation vector, V.
See p. 2.25 of the manual).
[Offset @ Node 1]: Not applicable, skip.
[Offset @ Node 2]: Not applicable, skip
[Pinned DOFs @ Node 1]: Not applicable, skip.

3.57
 [Pinned DOFs @ Node 2]:
Area: 8.0 (Note: This is a real scalar)
[Inertia 1, 1]: 10.667 (Note: See p. 2.16 of
the manual).
[Inertia 2, 2]: 2.667
[Inertia 2, 1]: Not applicable, skip.
Torsional Constant: 13.334 (Note: This is 'J'.
See p. 2.17 of the manual)
[Shear Stiff, Y]: 0.8333 (Note: See p. 2.17 of
the manual).
[Shear Stiff, Z]: 0.8333
[Nonstructural Mass]: Not applicable, skip.
[Y of Point C]: 2. (Note: See p. 2.17 of the
manual).
[Z of Point C]: 1.
[Y of Point D]: 2.
[Z of Point D]: -1.
[Y of Point E]: -2.
[Z of Point E]: -1.
[Y of Point F]: -2.
[Z of Point F]: 1.
Station Distances: Not applicable, skip.

For the curves parallel to the y-axis, the vector of 'Bar

Orientation' is <1 1 0> while for the curves parallel to the z-axis, the
vector is <0 1 1>. Every other information is the same as for the curves
parallel to the x-axis.

III.6.6 NASTRAN Analysis

Submit your job for NASTRAN analysis following the steps in III.5.6
Figure 64 - Creating a Quick Plot
of the plate problem. Be sure to confirm the load cases and boundary
conditions before submitting job for analysis.

3.58
 III.6.7 Post Processing

In the 'Result Menu Form' shown below, select 'Von Mises'

(stress quantity to be reported) as the option under 'Result Quantity'
and select '6_At Center' (the position where stress quantity will be
reported) under 'Position' at point C which implies that the stress at
Point C defined in Section 6.5 will be reported.

The fringe plot of the Von Mises stresses is superimposed on the plot of
the deformed shape as shown below.

Fringe results appear as

the color variations of
the model.

The Deformation results

appear as the skew from
the original model

3.59
 III.6.8 LAB HOMEWORK: Testing the Mini-Baja for
Vibration Analysis

Grid X Y Z
1 108.0 1.0 23.0
2 108.0 23.0 23.0
3 108.0 1.0 0.0
4 96.0 -2.0 26.0
5 108.0 23.0 0.0
6 96.0 26.0 26.0
7 96.0 -2.0 0.0
8 60.0 -6.0 78.0
9 72.0 -8.0 32.0
10 96.0 26.0 0.0
11 60.0 30.0 78.0
12 72.0 32.0 32.0
13 72.0 -8.0 0.0
14 36.0 -2.0 78.0
Figure 65 - Mini Baja Frame
15 48.0 -4.0 28.0
16 72.0 32.0 0.0 Construct the model baja frame shown in Figure 65 - Mini
17 36.0 26.0 78.0
Baja Frame. The coordinates of the grids are given.
18 48.0 28.0 28.0
Connect these with bar elements. These numbers can be
19 48.0 -4.0 0.0
20 24.0 0.0 24.0 used for grid numbers and node numbers. If you prefer,
21 48.0 28.0 0.0 you can use your own numbering scheme. These numbers
22 24.0 24.0 24.0 have already been optimized for connectivity.
23 24.0 0.0 0.0
24 0.0 0.0 24.0 Grids 7,10,23 and 25 are points of connection for the
25 24.0 24.0 0.0 wheels.
26 0.0 24.0 24.0
27 0.0 0.0 0.0
o Constrain these points in the x,y,z, qx and qz
28 0.0 24.0 0.0 directions.

Figure 66 - Baja Frame Coordinates The qy direction is left free because of the wheels.
The material used is steel with E= 10x106 psi, n=.33
and the weight is 0.098 lbs per cubic inches.
Compute and plot the first two mode shapes for the
frame.

3.60
 III.7 - Built-up Structures with RBE2
Topics Covered:
MPCs Optimizing

RBE2 Creating Groups

Equivalence Transfoming Groups

This section will provide two examples of built-up structures

which require proper use of rigid link, RBE2 to model a joint and
proper definition of the orientation vector of a beam in space.

What is RBE2?
Rigid body with independent DOF at one node, and dependent DOF at
an arbitrary number of nodes.

- One independent node (triangle)

Figure 74 - Relationship between Dependent and Independent Nodes

-
Multiple dependent nodes and DOFs (circles)

How to Create RBE2

3.61
 Step 1 Define depednet nodes (can select multiple nodes)

(1) Elements: Create/MPC/RBE2

(2) Click on Define Terms
(3) Disable Auto Execute
(4) Select the nodes
(5) Select all six degrees of freedom in the DOFs list
(6) Click Apply

Step 2 Define independent node (can select only one node)

(1) Select node

(2) Click Apply
(3) Click cancel
(4) Click Apply in the Finite elements form

Figure 67 - Creating MPC , RBE2 Dependent Terms

Figure 68 - RBE2 representing bolt.

3.62


Figure 69 - Creating MPC , RBE2 Independent Terms

3.63
 III.7.1 LAB SESSION 5: Homework Analysis of a
Bolted Plate

In this homework, we will use the RBE2 elements to model a

bolted plate. The steel plate is 550 mm long with a cross section of
15200 mm. The plate is bolted to a solid channel with four M162
bolts. A single point load of 16 kN is applied at the free end of the plate.
The example problem is taken from Example 8-7, pages 457-459 in
Shigleys Mechanical Engineering Design, 9th edition, by R. G.
Budynas and J. K. Nisbett.

3.64
3.65
The shear stress in the bolt and the bearing stress and the bending
stress in the plate are calculated analytically and presented in the
example. The assignment here is to use NASTRAN/PATRAN to
model the bolted plate and to compute the stress distribution in the
plate and find the reactions applied to each of the bolts. Use E =
30Mpsi and n = 0.25 .

The following outlines the required steps to build the finite element
analysis model of the problem.

3.66
 Step 1: Geometry Model

Model the plate as a 2D plane stress problem. Discretize the plate

through its central plane into 9 2D surface patches to build a geometry
model of the plate. Note that there must have nodes place at the
locations of the bolt joints. They are nodes 7 10 in this example.

Figure 70 - Location of Nodes to be Selected

In the model, add 4 nodes underneath the corresponding bolt joints.

These four nodes represent the ground points. The ground here
represents the rigidity of the C-channel to which the rectangular plate
is bolted to. These four nodes are 17 20 as shown in the following
figures. The distance between these ground nodes and the
corresponding bolt joints is equal to the half thickness of the plate.

3.67
Figure 79 - Result after Analysis

Step 2: Mesh Model and Boundary Conditions

Now place the meshes in each of the surfaces. Be sure that the
interface nodes are all connected. Connect the ground nodes and the
corresponding bolt joint nodes with RBE2 elements. Four RBE2
elements in total in this case. The ground nodes are fully
constrained.

Step 3: Finite Element Analysis

The stress contour plot is given below. Now, how to get the f06
file that includes the reaction forces at the nodes and support points?

III.7.2 LAB SESSION 7: Using MPCs and

Equivalence

3.68
 In this lab session, we will create a rectangular plate with a
square hole. Then we will weld a smaller right angle plate at some
angle to the larger plate and attach a bar from the large plate to the
smaller plate using RBE2 elements between the bar and the large plate.

Figure 71 Session

Step 1:
Create the geometry points for the large plate with a square hole in it
first. It will look like the geometry in figures 1 and 2 without the
curves.

(1) Create point 1 at (0,0,0).

(2) Create point 2 at (20,0,0).
(3) Create point 3 at (20,10,0).
(4) Create point 4 at (0,10,0).
(5) Create point 5 at (0,2,0).
(6) Create point 6 at (5,0,0).
(7) Create point 7 at (13,0,0)
(8) Create point 8 at (15,0,0).
(9) Create point 9 at (20,2,0).
(10) Create point 10 at (20,5,0).

3.69
 (11) Create point 11 at (20,8,0).
(12) Create point 12 at (15,10,0).
(13) Create point 13 at (13,10,0).
(14) Create point 14 at (5,10,0).
(15) Create point 15 at (0,8,0).
(16) Create point 16 at (0,5,0).
(17) Create point 17 at (5,2,0).
(18) Create point 18 at (13,2,0).
(19) Create point 19 at (15,2,0).
(20) Create point 20 at (15,5,0).
(21) Create point 21 at (15,8,0).
(22) Create point 22 at (13,8,0).
(23) Create point 23 at (5,8,0).
(24) Create point 24 at (5,5,0).
(25) Create point 25 at (13,5,0).

Step 2:
Create the curves for the large plate with a square hole in it first. It
will look like the geometry in figures 1 and 2.

(1) Create curve 1 with start point 1 and end point 6.

(2) Create curve 2 with start point 6 and end point 7.
(3) Create curve 3 with start point 7 and end point 8.
(4) Create curve 4 with start point 8 and end point 2.
(5) Create curve 5 with start point 2 and end point 9.
(6) Create curve 6 with start point 9 and end point 10.
(7) Create curve 7 with start point 10 and end point 11.
(8) Create curve 8 with start point 11 and end point 3.
(9) Create curve 9 with start point 3 and endpoint 12.
(10) Create curve 10 with start point 12 and end point 13.
(11) Create curve 11 with start point 13 and end point 14.
(12) Create curve 12 with start point 14 and end point 4.
(13) Create curve 13 with start point 4 and end point 15.
(14) Create curve 14 with start point 15 and end point 16.
(15) Create curve 15 with start point 16 and end point 5.
(16) Create curve 16 with start point 5 and end point 1.
(17) Create curve 17 with start point 5 and end point 17.
(18) Create curve 18 with start point 17 and end point 18.
(19) Create curve 19 with start point 18 and end point 19.
(20) Create curve 20 with start point 19 and end point 9.
(21) Create curve 21 with start point 16 and end point 24.
(22) Create curve 22 with start point 24 and end point 25.
(23) Create curve 23 with start point 25 and end point 20.
(24) Create curve 24 with start point 20 and end point 10.
(25) Create curve 25 with start point 15 and end point 23.

3.70
 (26) Create curve 26 with start point 23 and end point 22.
(27) Create curve 27 with start point 22 and end point 21.
(28) Create curve 28 with start point 21 and end point 11.
(29) Create curve 29 with start point 6 and end point 17.
(30) Create curve 30 with start point 17 and end point 24.
(31) Create curve 21 with start point 24 and end point 23.
(32) Create curve 32 with start point 23 and end point 14.
(33) Create curve 33 with start point 7 and end point 18.
(34) Create curve 34 with start point 18 and end point 25.
(35) Create curve 35 with start point 25 and end point 22.
(36) Create curve 36 with start point 22 and end point 13.
(37) Create curve 37 with start point 8 and end point 19.
(38) Create curve 38 with start point 19 and end point 20.
(39) Create curve 39 with start point 20 and end point 21.
(40) Create curve 40 with start point 21 and end point 12.

Figure 73: Geometry of Large Plate with Square Hole

Figure 72 - Geometry of Large Plate with Square Hole and Geometry Labels

3.71
 Step 3:
Now create the geometry points and curves for the small plate. It will
look like the geometry in figure 3.

(1) Create point 26 at (5,0,5).

Create point 27 at (5,2.12132,7.12132).
(2) Create point 28 by translating point 26 through a vector of
<8,0,0> (Refer to Translating Curve on page 22).
(3) Create point 29 by translating point 27 through a vector of
<8,0,0>.
(4) Create point 30 by translating point 28 through a vector of
<2,0,0>.
(5) Create point 31 by translating point 29 through a vector of
<2,0,0>.
(6) Create curve 41 with start point 17 and end point 26.
(7) Create curve 42 with start point 26 and end point 27.
(8) Create curve 43 with start point 27 and end point 24.
(9) Create curve 44 with start point 26 and end point 28.
(10) Create curve 45 with start point 28 and end point 29.
(11) Create curve 46 with start point 29 and end point 27.
(12) Create curve 47 with start point 28 and end point 30.
(13) Create curve 48 with start point 30 and end point 31.
(14) Create curve 49 with start point 31 and end point 29.

Figure 74 - Points and Curves of Large and small plates

3.72
 Step 4:
Now create the surfaces for both plates. The geometry will look like
figures 4 and 5.

(1) Create surface 1 with start curve 1 and end curve 17.
(2) Create surface 2 with start curve 17 and end curve 21.
(3) Create surface 3 with start curve 21 and end curve 25.
(4) Create surface 4 with start curve 25 and end curve 12.
(5) Create surface 5 with start curve 2 and end curve 18.
(6) Create surface 6 with start curve 18 and end curve 22.
(7) Create surface 7 with start curve 22 and end curve 26.
(8) Create surface 8 with start curve 26 and end curve 11.
(9) Create surface 9 with start curve 3 and end curve 19.
(10) Create surface 10 with start curve 19 and end curve 23.
(11) Create surface 11 with start curve 27 and end curve 10.
(12) Create surface 12 with start curve 4 and end curve 20.
(13) Create surface 13 with start curve 20 and end curve 24.
(14) Create surface 14 with start curve 24 and end curve 26.
(15) Create surface 15 with start curve 26 and end curve 9.
(16) Create surface 16 with start curve 41 and end curve 43.
(17) Create surface 17 with start curve 44 and end curve 46.
(18) Create surface 18 with start curve 47 and end curve 49.

Figure 75 - Generated Plates

3.73
 Step 5:
Create geometry for the CBAR element referred hereafter as
angled_bar.

(1) Create point 32 at (14,6.5,0).

(2) Create point 33 using the extraction method.
(a) Under Geometry,:
a. Action,
i. Create
b. Object,
i. Point
c. Method.
i. Extract
(b) There will be three pictures underneath the Method
option. Choose the center picture. This will be the single
point on a surface.
(c) Keep parametric options as they are since this point will be
in the center of the surface 18.
(d) Now left click on surface 18.
(e) See Figure 76 to the left for reference.

(3) Create curve 50 with start point 32 and end point 33.

Step 6:
Create the mesh now. To do this, refer to III.5.3 Mesh Generation
on p-3.41. This section shows how to do the mesh for the surfaces
on the plate. To create the mesh on the curves, change Surface to
Curve under Type.

(1) Mesh Seed

(a) For the plate with the square hole, use a mesh seed of 2
for all of the curves.
(b) For the right angle plate, use a mesh seed a mesh seed
of 2 for the vertical sides. This will give the right angle
plate a matching mesh to the plate with the square hole.
Use a mesh seed of 2 for the horizontal sides.
(c) For curve 50, use a mesh seed of 2.
(2) Mesh
(a) Mesh the plates using Surface under Type (Figure
77 Meshing of Surfaces).
(b) Mesh the plates using Curve under Type (Figure 78
- Meshing of Curves).
Figure 76 - Extract Command Box

3.74
Figure 77 Meshing of Surfaces Figure 78 - Meshing of Curves

(3) Equivalence and optimize

(a) Equivalence is to delete duplicate nodes so that the
geometry fits together.
(b) Optimizing is to renumber nodes after deletion of the
duplicate nodes.

3.75
 Figure 80 - Finite Elements (CBAR and CQUAD).

Step 7:
Now create material properties. The material properties for the
plate are given in III.5.4 Material Selection on p-3.44, and they are
the same for the CBARs.

Step 8:
Next, create and assign element properties. For assigning element
properties to the plate, refer to Section III.5.5 Element Properties on
p- 3.46. For assigning material properties to the CBAR elements,
refer to the steps given below.

(1) Under the Element Properties tab, change 2D under

Object to 1D.
(2) Make sure Beam is selected under Type.
(3) Create Property Set Name. Type bar_1 for the first
CBAR.

Figure 79 - Create 1D Beam (1) Click on Input Properties near the bottom of the box.
"angled_bar." This will give the Input Properties form below.

3.76
(2) Properties form below.

Figure 81 - CBAR Properties "angled_plate."

(3) Steel must be input for the Material Name.

(4) The Bar Orientation is another way of asking for the V-
vector. You will recall that the beam starts at coordinates
(14,6.5,0) and ends at coordinates (13,1.5,12). In order for
the beams cross-section to remain level with the global x-
axis, the V-vector will have to be parallel to the local xe-
axis when viewed in the global xz-plane. However, you
must ensure that the angle between the V-vector and the

3.77
 local xe-axis is greater than zero.i
It is clear that a vector of <0,1,1> will define an appropriate
V-vector.
(5) To create a section refer to III.6.5 Element Properties on
page-3.56 as an example. In this instance the properties
will be different. However, there is an easier way to do
this. Click on Create Sections Beam Library. A form
will appear as below. Fill in the information as given in the

Figure 82 - Defining Section in 'Beam Library'

form below.
(6)
Figure 11: Defining Section Properties in "Beam Library."

i
The local ze-axis is parallel to the cross product of the V-vector and the local xe-
axis. In mathematical symbols, || ($&') $&' = $ &' sin (.). This
means that the angle between V and the xe-axis is close to zero, then the ze-axis
will be undefined since sin 0 = 0.

3.78
 (7) Then press apply at the bottom of the form (not shown in
Fig.).ii
(8) Then close out of the Beam Library and click on the I-
beam icon to the left of the Section Name. This should
open another window that shows the section you created
from the Beam Library. Select the CBAR1 you
created.
(9) To make sure the CBAR is oriented the right way with
respect to the coordinate axes, then left-click on the
Calculate/Display button. This will give the window
below.

Figure 83 - Section Properties and Orientation in "Calculate/Display" Window

(10) Now click OK.

(11) Then Select Application Region. The application
region should be curve 44. Click Add, press OK, then Figure 84 - Prop. Attributes
Apply.

ii
If you are completing this step on your laptop using remote desktop, it will be
necessary to tab down 5 times and press enter on the keyboard to Apply. The
name CBAR1 will then appear in the Existing Sections box if it was successfully
applied. Those using the lab desktops will not run into this issue.

3.79
 (12) To check if the beam is of the proper orientation follow
the steps below.
(a) Click on the Display tab at the top of the screen.
(b) Then click on Load/B.C./Elem. Props
(c) Scroll down until you find Beam Display. It should
currently have 1D:Line selected in the drop down
menu directly below it. Select 3D:Fullspan in the
drop down menu and Apply.
(d)

Figure 85 - View of Completed Geometry in 3D:Fullspan.

Step 9:
The next step is to create a rigid body connection in the form of a
multipoint constraint or (MPC). This will connect bar_1 to the plate at
the square hole.
(1) First under the Meshing tab, select Create under
Action, MPC under Object, and RBE2 under
Type.

3.80
 (1) Next, go to Define Terms and you will see the Define
Terms box shown below.

(1) There are two categories of nodes defined in and RBE2,

and they are the dependent term(s) and the independent
term.iii
(2) You will also see the DOFs at the bottom of the box.
Select all of them by left-clicking on the UX. You can
either select each DOF by left clicking on each one
individually, or hold down the shift key and press the down
arrow on the keyboard until all DOFs are highlighted.
Uncheck the Auto Execute.
(3) Then define the independent term first by left clicking
beside Create Independent and clicking on node 79.
Highlighted in the Fig. below in the red circle.
(4) Next, left-click Apply.

Figure 86 - "Define Terms" Window

Figure 87 - Node 79 on the CBAR highlighted in Red.

Figure 88 - Creating MPC

iii
The phrase- independent term- was left was left purposely singular because
there can only be one node defined as such. The same is not true for the
dependent term in which case there can be multiple nodes defined in this category.

3.81
 Figure 89 - Finished RBE2 Element.

Step10:
Now apply the same Boundary Conditions and Loads you
created in III.5.1 Boundary Conditionsand and III.5.2 Applied
Loads on pages 3.36 through 3.38.

(1) Apply the boundary conditions to curves 13 through 16.

(2) Apply the 1000 lb. down to node 68 and the 1000 lb. right
to node 68. In this instance, you will have to change
Geometry to FEM under Select Application Region.

Figure 90 - "Select Application

Region" for selecting Nodes.

3.82


Figure 91 - Forces (in blue) and Boundary Conditions (in green).

3.83
 Step 11:
Now run the analysis. Refer to III.5.6 NASTRAN Analysis &
III.5.7 Access Results on p 47 & 51.

Figure 93 - von Mises Stress

Figure 93 - Displacement

3.84
 III.7.3 LAB SESSION 6: Homework A Simple
Bridge

This example shows how to model and mesh a built-up structure

and how to use RBE2s to connect structural members.

The structure example is shown in Figure 104, which is made of three

hollow beams and 4 plates. The ends of three hollow, rectangular
beams are welded to the bolted plate assemblies. The beams are 20
inches long. Each plate assembly is made of two plates, a short one,
4 3 inches and a long one, 9 3 inches. The beams are welded to the
shorter plate, which is then bolted to the longer plate by three bolts.
The footing of the longer plate is fixed on the ground. The locations
of beam-plate welded connections and the plate-plate bolt joints are
shown in Figure 96. Dimensions of the plates are shown in Figure 95
while those of the beams in Figure 97. Note that the thickness of each
plate is 1 and the thickness of each beam is 0.02. A point load of 200
lbs is applied to the mid-point of each of the top beams.

Create a PATRAN model of the built-up structure. Report the

magnitudes and the locations of the maximal displacement and the
max stress of the built-up structure subjected to the given two point
loads. The material is steel with Youngs modulus of 30 10 6 psi and
the Poissons ratio, 0.3.

Figure 94 Beams welded to Plate Assembly

3.85
 Figure 96 - Locations for Beam-Plate Connection (Boxes)
and Bolt Joints (Circles)
Figure 95 - Dimensions for the Plates

Figure 97 - Dimension for the beams

3.86
 Node Placement on the Plates
One has to incorporate all points of concern in the geometry model
as part of the surface model in order to include
those points as GRID POINT in the finite element
models.

As examples, the outer plate, Plate 1 is bolted

with the inner plate, Plate 2 at 3 designated points:
(pts 5, 6 and 7 shown in Figure 77 (a)). In
geometry modeling, the lines are created to go
through those points of interesting as shown in
Figure 77 (b) so that these points of concern will
be part of corner points of surfaces (4-edge
patches or c-quads).

12 surfaces in total along with 20 nodes are

created as shown in Figure 77 (b). Similarly, Plate
2 will be divided into 24 surfaces along with 35
Nodes.

The latter are located at the centers of the Figure 98 - Geometry Model of Plate
beams. It is worthwhile mentioning that additional
lines and surfaces have to be created in the model, as
Nodes 44 and 31 are not lined up along the Y-axis; Note that 3 bolt joints will go
same for Nodes 40 and 32. Figure 79 shows the front through Plate 2 at Nodes 29, 31
view of the geometry model of the built-up structures. and 32 and 3 beams will be welded
to Plate 2 at Nodes, 36, 40 and 44.

Figure 99 - Final Geometry Model of the Built-up Structure

3.87
 Gap and RBE2
The finite element model of the built-up structure is shown in Figure
100 & Figure 101. Three special features of this model in Figure 99
are worthwhile mentioning.

1. One inch gap presents between two end plate models.

2. A half-inch gap presents between the ends of the beams and the
plate model.

This is because a plate is modeled by its central plain in the finite

element mesh. As the plate of concern is one inch in thickness, the gap
between two adjacent central plains is then exactly equal to one inch.
Moreover, as the end of the beam is welded to the surface of the plate,
the distance between the end of the beam and the central plane of the
joined plate is half-inch, More detailed explanation are given later.

3. 3 RBE2s are used to model the bolt joints between two bolted
plates and 3 more RBE2s to model the welded joints between
the beams and the plate.
As shown in Figure 101, the central plains of both plates are
connected through RBE2s. The RBE2 is a rigid rod that
represents the bolt. The gap between the two central plains is
(t1 + t 2 )
where t1 and t 2 are the thicknesses of the plates.
2

200.0000
200.0000

123456
123456
123456

Z
X

Figure 100 The Finite Element Model Figure 101 - The Side View of the Finite
Element Model

3.88
a) Plate Jointed by Bolts
b) Central Planes of the Plates with RBE2

In the geometry model, the bolt joints connect Node 7 of Plate

1 to Node 29 of Plate 2, Node 5 of Plate 1 to Node 31 of Plate 2
Node 6 of Plate 1 to Node 42 of Plate 2. The result is shown in Figure 102 - Sketch of 2 plates
Figure 104

Figure 104 - RBE2 for Bolt Connection between two plates

Figure 103 - Sketch of spacing
between 2 plates

3.89
 1. A half-inch gap presents between the ends of the beams and
the plate model

Since the beams are welded to the surface of the shorter

t
plate, a gap of 2 presents between the ends of the beams
2
and the finite element model of the plate which lies in its
central plain. The thickness of the plate is t 2 .
The situation is demonstrated in Figure 106. The forces that
are applied to the RBE2 represents the amounts of the forces
Figure 105 - Sketch showing spacing that have to be sustained
for RBE2 by the welded joint. In
other words, the MPC
forces reported at the
RBE2 can be used to
Note that RBE2, the design the weld joint that
rigid link, is used to includes weld length, the
connect the beam to the weld material and the
central plain of the plate throat length. The
to model the welded RBE2s that model the
joint. plate-beam connections
are shown in Figure 107.

Figure 106 - RBE2 spacing

Figure 107 - Ends of Beams Connected to the Central Plain through RBE2

3.90
 The entire finite element model is shown
in Figure 109. The stress contour plot of the
built-up structure is shown in Figure 111
with max stress of 43.8 psi at the fixed end
and the max deflection of 0.0166 inches.

Figure 108 - RBE2s for three plate-beam connection

Figure 109 - Finite Element Mesh

Figure 110 - RBE2s representing bolts and welds

3.91
Figure 111 - The Displacement and Stress Contour Plots

3.92
 The result as shown in the below pictures:

Figure 112 - Finite Elemnt Model

Figure 113 - Side view showing placement of RBE2s

3.93
 How to Translate Plates

You have to select what you want to transform in your drawing.

In this example, the surface will translate.

1. Click Geometry
2. Select Transform in Action
3. Select Surface (you can select node or curve too)
4. Choose direction vector
5. Put length - the distance between plates
6. Select the surface(s) you want to translation
7. Click apply

3.94
 How to Create a Group
Grouping is advantageous because instead of having to translate
Surfaces, then Curves, then Meshes, then MPCs and so on, the user
can transform many different types of elements at once.

Two 5x5 squares with 2 RBE2s are used to explain group and
function of translate or mirror.

Step 1 Create geometric and RBE2 as shown in pictures

Figure 114 - 2 Plates with Corresponding RBE2s

3.95
Step 2 Click Group Menu Tab on top of screen

Step 3 Group/Create/Select Entity

Step 4 Type name of group

Step 5 Select Entity Selection

Step 6 Click Apply

Now, you can translate or

mirror geometrics and RBE2

Step 7 Click Transform

Step 8 Select group what you want to translate

or mirror

Step 9 Select translate or mirror in Method. In

this example, translate method was used.

Step 10 Put correct direction vector and

magnitude

In this example, the direction vector is <0 0 1>

and magnitude is 20

Step 11 Click apply

3.96
 Figure 115 - Using Transform on a Group

The result for translate is as show the below picture

Figure 116 - Group translation:

3.97
 III.7.4 LAB Developing a Strategy to Build

A simple bridge with center plates

This example show how to model and mesh a built up structure and
how to use RBE2s to connect structural members. The structure
example is shown in Figure 1, which is made of three hollow beams
and 6 plates. The ends of three hollow, rectangular beams are welded
to the bolted plate assemblies. The beams are 20 inches long. Each
plate assembly is made of two plates, a short one, 4x3 inches and a
long one, 9x3 inches. The beams are welded to the shorter plate, which
is then bolted to the longer plate. The footing of the longer plate is
fixed on the ground. The locations of beam-plate welded connections
and the plate-plate bolt joints are shown in Figure 1.

Note that the thickness of each plate is 1 and the thickness of each
beam is 0.02.

A point load of 150 lbs (total load is 600lbs) is applied to the mid-
point of each of the top beams.

Create a PATRAN model of the built up structure. Report the

magnitude and the locations of the maximal displacement and the max
stress of the built up structure subjected to the given two point loads.

The material is steel with Youngs modulus of 30 102 345 and the
Poissons ratio, 0.3.

Simple Bridge - 3D Images

3.98
Figure 117 Beams welded to plates

Figure 118 Dimensions for the 9"X3" plate

3.99
 Figure 120 Locations for beam-plate
Figure 119 Dimensions for beams and 0.02 thickness
connection and bolt joint

Drawing Geometric Model

1. Drawing 9 X 3 Plate

12 surfaces in
total along with 20
nodes are created as
shown in Figure 5.

2. Drawing 4 X 3
Plate

Similarly, 4X3 plate

will be divided into 24
surfaces along with 35
Figure 121 Geometry model of 9" x 3" plate
nodes.
Figure 122 Geometry model of 4" x 3"
Plate. The three grid points in the circles
3. Side View of geometric model of 9x3 and 4x3 plates
are for beams. The three grid points in the
diamonds are for bolt joints.
A plates I modeled by its central plain in the finite element mesh. As
the plate of concern is one inch in thickness, the gap between two
adjacent central plains is then exactly equal to one inch in thickness, the gap between adjacent
central plains is then exactly equal to one inch.

3.100
The example structure shown in Figure 1
can be built up by two different ways.

1st Method - Using

Translate only

1) Translate surfaces four times

2) Translate nodes for beams

Figure 123 The side view

Figure 125 Example for translate nodes.
of geometry model of 9"
Node 251, 252, and 253 are from node 48, x 3" and 4" x 3" plates.
53, and 41.

Figure 124Example for translate surfaces

3.101
 Figure 128 Example for translate nodes. Node 251, 252, and 253 are from node 48, 53, and 41.

Figure 127 The side view of the example structure without beam

Figure 126 Final Geometry model of the built-up structure

3.102
 2nd Method Using Translate & Mirror

1) Translate surface to build the half of the example structure

Figure 129 The half of the example structure are built by translation

2) Mirroring Surfaces for plates and mirroring curve for bars

Mirror surfaces of plates (Left Side)

Define Mirror Plane Normal

{[0 0 22.5][0 0 1]} {[coordinate of position of
point][indicate normal 6 to the plane]}

3.103
 Figure 130 The example of translate surfaces

Mirror curves

Figure 131 The final geometry model of the example structure

3.104
Built Finite Element Model using geometry model

Mesh
1. Mesh surface for plates

2. Mesh curve for beams

(a)

(b)

(c)

(d)

Figure 132 Finite element model. (a) Entire structure finite element mesh, (b) 9 x 3
plate finite element mesh (c) 4 x 3 plate mesh finite element mesh, and (d) 1 x 0.25
and 0.25 x 0.25 beam finite element mesh.

3.105
 Using RBE2 to model the bolt joint between two bolted plates and to
model the welded joint between the beams and plates (Total 21
RBE2s)

In the finite element model, the bolt joints In the finite element model, the welded
connect (Boxes) joints connect (Ovals)
node 5 to node 23 node 53 to node 252
node 6 to node 24 node 48 to node 251
node 7 to node 25 node 41 to node 253

Figure 133 RBE2s representing bolts and welds

The entire finite element model with RBE2s is shown in Figure 17.

Figure 134 Ends of beams connected to the central plain through RBE2

3.106
Apply Boundary condition and Forces (600 lbs)

The results as shown in the below pictures (Figure 18).

Figure 135 The displacement and stress contour plots

3.107
 III.8 LAB SESSION 8: Homework - 45 Degree Wedge for
Static and Vibration Analyses
Topics Covered:
Vibrations

Figure 136 - Wedge Model

In this homework, we will create a 3D wedge to conduct Static

and Vibration analysis. The wedge is shown in the following figures.
The major steps used in generating the wedge are illustrated. The 450
wedge is 19.573 inches in depth with a flange of 2.121 inches. The
wedge is made of a plate with 0.25 inches thickness. Four of 20.250
inches by 18.875 inches plates are bolted to the top of the wedge with 6

3.108
 bolts, three on each opposite edge. The thickness of the plate is also 0.25
inches. The material of the wedge is steel with the Youngs modulus
2.9E7 psi, Poissons ratio 1.124E7. The weight density and the thermal
expansion coefficient are 0.28 lb/in3 and 0.29, respectively.

8.254

10.824

Figure 137 - Wedge Dimensions

3.109
 The surfaces of the wedge include four top plates, one 45 degree
wedge and two triangular plates on the sides. The example shown below
is made of 1232 grid points, 1053 CQUAD elements and 18 CTRIA3
elements. 12 RBE2 cards are used to model the bolt joints connecting
the 4 layers of top plates together at six points.

Case 1. Static Case

The geometry model of the wedge is shown below. The detailed

steps used in generating the plates have been discussed in Chapter 5.
Only major steps are discussed here. It should be noted that the
triangular side plates are discretized with triangular patches, while the
surfaces of the rest of the model are discretized with 4 edge patches. The
interfaces between the end triangles and the faces of the wedges should
be aligned properly with the same number of the nodes, placed at the
same locations. Thus, equivalence has to be applied around these
common interfaces. Also, 6 RBE2 are used to model the bolts to connect
the four top plates and the wedge as shown the picture. One master node
is placed at the top plate connected to 4 slave nodes, each of the slave
nodes is at the center of the rest of the plates. A point load of 1,000 lbs
is imposed at the center of the top plate pointing downward. The wedge
is completely fixed along its bottom edge.

Once the geometry mode is built, one can proceed to create

surfaces. Then create the meshes of quads for 4-edge patches and
triangles for triangular patches. The Tables below show how to create
triangular surfaces and meshes.

3.110
Figure 138 - Finite Element Model of Wedge

3.111
The finite element model is shown below, once the geometric and mesh
processes completed. Again, you should carefully match mesh lines and
nodes between the wedge and triangular plates.

Figure 139 - Wedge, Meshed

The deformation plot of the wedge under the given point load is shown
below. The maximal displacement is 0.0853 inches.

Figure 140- Deformation Plot

3.112
 Case 2. Vibration Case

The method of the modified householder method is used for

eigenvalue analysis in this example, which requires the output of
frequencies between 0 to 1000 Hz for up to 21 eigenvalues. Each of the
eigenvectors is normalized with respect to the mass matrix of the
structure. In this example, the finite element model used for static
analysis is used here again for vibration analysis. In this case, however,
no boundary conditions and forces are applied to the model. Thus, only
a few changes are needed to get vibration results. The first six modes
should be rigid body modes with their frequencies are close to zero. The
first elastic mode is the 7th mode with a frequency of 87.873 Hz.

Figure 141 - Vibration Plot

3.113
 The two pictures listed below show what options should be changed for
vibration analysis. First, the solution type should be changed from
Linear Static to Normal Modes. In the Solution Parameters, select
Automatic Constrains. Then, select Modified Householder in Solution
Parameters.

Figure 143 - Solution Type Figure 143 - Solution Parameters

3.114
Next, put the lower and upper values of the output frequencies in the
Eigenvalue Extraction table as shown below.

Figure 145 - Eigen Value Extraction

Figure 144 - Eigen Val. Translational

3.115
After finished the process, we will get a bdf file when you click analyze
on the Patran. You can see only bdf file on your folder. Run the bdf
file using the Nastran. Then, we will see the f06 file and xdb file.
Access the xdf file on Patran. Finally, we can see the results displayed
by Patran as shown picture.

The bdf file created by the PATRAN procedure discussed above is

listed below for reference.

$ NASTRAN input file created by the Patran 2014.1 64-Bit input file
$ translator on February 09, 2016 at 13:36:52.
$ Direct Text Input for Nastran System Cell Section
$ Normal Modes Analysis, Non-database
SOL 3
CEND
TITLE = MSC.NASTRAN JOB CREATED ON 06-NOV-15 AT 15:00:40
ECHO = NONE
$ Using Nastran default values for RESVEC
SUBCASE 1
$ Subcase name : Default
SUBTITLE=Default
METHOD = 1
VECTOR(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
BEGIN BULK
PARAM POST 0
PARAM* WTMASS .002588
PARAM PRTMAXIM YES
EIGR 1 MHOU 0. 1000. 21
MASS
$ Elements and Element Properties for region : PLATE
PSHELL 1 1 .25 1 1
$ Pset: "PLATE" will be imported as: "pshell.1"
CQUAD4 1 1 808 886 893 807 0. 0.
CQUAD4 2 1 893 807 806 885 0. 0.
CQUAD4 3 1 806 885 884 799 0. 0.
...
...
CQUAD4 1053 1 140 141 142 110 0. 0.
CTRIA3 1054 1 138 190 139 0. 0.
CTRIA3 1055 1 191 139 190 0. 0.
....
....
CTRIA3 1071 1 154 204 153 0. 0.
$ Referenced Material Records
$ Material Record : steel
$ Description of Material :
MAT1* 1 2.9+7 1.12403+7 .29
* .28
$ Multipoint Constraints of the Entire Model
$ ID conflict : the PATRAN MPC ID was 1072
RBE2 2155 365 123456 291 292 293 437
$ ID conflict : the PATRAN MPC ID was 1073
RBE2 2156 375 123456 309 310 311 446
$ ID conflict : the PATRAN MPC ID was 1074
RBE2 2157 372 123456 300 301 302 444
$ ID conflict : the PATRAN MPC ID was 1075
RBE2 2158 1184 123456 1147 1148 1149 1202
$ ID conflict : the PATRAN MPC ID was 1076
RBE2 2159 1196 123456 1172 1173 1174 1210
$ ID conflict : the PATRAN MPC ID was 1077
RBE2 2160 893 123456 803 804 805 973
RBE2 2149 364 123456 288 289 290 436
RBE2 2150 374 123456 306 307 308 445
RBE2 2151 371 123456 297 298 299 443
RBE2 2152 1183 123456 1141 1142 1143 1201
RBE2 2153 1195 123456 1164 1165 1166 1209
RBE2 2154 892 123456 800 801 802 972
$ Nodes of the Entire Model

3.116
GRID 1 14.3493 9.438 -9.60907
GRID 2 14.3493 7.8504 -9.60907
...
...
GRID 1231 8.74785 -.125 -14.6852
GRID 1232 8.01287 -.125 -13.2487
$ Loads for Load Case : Default
$ Referenced Coordinate Frames
ENDDATA 8e9fe7b4

III.9 .DB and .XDB GENERATION and VISUALIZATION

Purpose: Once the dbf file is created and its f06 file has been checked
to be correct, one can then follow the procedure given below to
visualize the finite element model and the analysis results through the
use of PATRAN. This achieves in two steps. The first one is to
generate .db file for visualization of the finite element model and the
second to generate .xdb file for visualization of the analysis results.

Task 1. Create .db File

Step 1: Open PATRAN. Click on File > New. In the New Database
window manual. Find the proper directory and key in the file name
with an extension .db for the db file in the space, then OK.

Step 2: Click on File again. Select Import to import the manually

created bdf into PATRAN. As shown in the following picture. Be
sure that on the right, select MSC. NASTRAN Input in the Source.
Go to the proper directory and select the desired bdf file. Click Apply

3.117
A .db file is now created and the grid points and elements of the finite
element are displayed now in the PATRAN window.

Task 2. Create .xdb File

Step 1: Open PATRAN, click on Analysis. As shown in the following

picture, select Analyze in Action, Entire Model in Object and Full Run
in Method. If desirable, you may assign the newly created xdb file
with a new name keyed in in space under Job Name. See the following
figure. Click Apply. The bdf file will be submitted to NASTRAN run
to generate the f06 and xdb files.

Step 2: Open NATRAN to analyze the bdf file just created by

PATRAN in the above steps. A xdd file with the assigned name should
now be present within the folder.

Task 3. Visualization of the Analysis Results

3.118
Step 1: Open PATRAN, click on Analysis. Select Access Results in
Action, Select Attach XDB in Object and Result Entities in Method as
shown in the following picture.

Step 2: Attach the xdb file by selecting the xdb file just created. Click
on Select Results File to identify the desirable xbd file. Then OK to
close up the result file selection. Click Apply

Step 3: Click on Results. Select the Result Cases. Click

Displacement, Translational in both windows of the Select
Fringe Result and the Select Deformation Result. See the
following figure. Click on Apply. The PATRAN window will
display the stress contour plot which is imposed upon the
deformation plot.

Note that you may skip Step 2 by adding a line PARAM,

POST, 0 directly under BEGIN BULK. That is,

After submitting the bdf file to NASTRAN run, this

PARAM statement will ensure the generation of the .xdb
file in addition to the .f06 file. The former can be used to
generate the stress contour plot and the deformation plot
following Task 3 described above.

3.119
 IV. NORMAL MODES ANALYSIS
IV.1 OBJECTIVE
The objective of this laboratory session is to review and to explain the principles and the
necessary input data to run normal modes analysis using MSC/NASTRAN or MSC/PATRAN.
Concepts and commands pertaining to MSC/NASTRAN will be first discussed in Sections IV.2
to IV.4, while those pertaining to MSC/PATRAN will be discussed in section IV.5.

IV.2 INTRODUCTION
First a few terms about vibration will be explained, followed by an example which
explains the solution procedure for normal modes analysis.

Equation of Motion
Consider the displacement of a system, which is subjected to a dynamic load, as a
function of time, e.g. u(t). as shown below.

m P(t)

c
u(t)

Fig. 4.1 Spring, Damper, and Mass

The equation of dynamic equilibrium can be established to solve u(t). This equation states that
all the forces including the inertial force generated by the motion of the system should be in
balance. In other words, the external force should be equal to the sum of the internal force, the
damping force and the elastic force. Symbolically, one has

p(t ) = Mu
+ Cu + Ku (4.1)
where p(t), M, C and K are the external load, the mass matrix, the damping coefficient matrix
and the stiffness matrix respectively. It is the purpose of transient analysis to solve Eq (4.1)
which is the equation of motion.

4.1
Free Vibration Problem
..
If p(t) = 0 and C = 0, Eq. (4.1) becomes M u+ Ku = 0 . Let u(t) represent a harmonic
motion. That is, u(t) = aeiwt, where a is the vector of displacement amplitudes and is the
frequency of vibration (rad./sec.). Equation (4.1) becomes (K-M)a=0 where =2 is the
eigenvalue, and a is the mode shape or eigenvector. The purpose of eigenvalue analysis is to
solve the following equation:

(K i M )a i for i = 1,2,.......n (4.2)

where n is the dimension of Eq. (4.2).
Note that eigenvalues, i eigenvectors, a i represent the dynamic and static characteristics of
a structure. For example, if a > b, where a and b indicate two different structures, one can
conclude that structure a is more stiff or less flexible than structure b. Furthermore, many
industrial applications such as system identification and control, dynamic or shock analysis, etc.,
use eigenvalues and eigenvectors of a system to characterize the vibration characteristics of the
system.

Example
A single example is given in the following to demonstrate a solution procedure for
eigenvalue analysis. The example has the stiffness and the mass matrices given as

6 2 2 0
K = and M =
2 4 0 1
The eigenvalue problem is then stated as

6 2 2 a1 0
( K i M )a i = 0 or =
4 a 2 0
(4.3)
2
To have a nontrivial solution, Eq. (4.3) should satisfy the condition:
6 2 2
det =0
2 4
which yields (6-2)(4-) - 4 = (-5)( -2) = 0. The eigenvalues which are the roots of the
polynomial are 1 = 2 and 2 = 5. If =2 then Eq. (4.3) becomes
2 2 a1 0
2 =
2 a 2 0
which results in a single equation 2a1-2a2 = 0, which implies a1= a2. Therefore, the eigenvector
corresponding to the lowest eigenvalue 1, is a1T = (1, 1) for an arbitrary constant .

4.2
Similarly, the eigenvector corresponding to the second eigenvalue, 2, is a 2T = (2, 1) for an
arbitrary constant .
Note that the eigenvalues are usually rearranged in an increasing order, i.e. 1 < 2 < < n .
Furthermore, eigenvectors cannot be specified uniquely. Some normalization measures should be
used to normalize the eigenvectors. Three approaches used in MSC/NASTRAN are the MASS,
MAX and POINT normalization. The MASS normalization requires the eigenvectors to be
normalized with respect to the mass matrix, i.e., a iT Ma i = 1 . The MAX normalization requires
the maximum component of an eigenvector to be 1. The POINT normalization normalizes an
eigenvector with respect to a specified component. Examples are presented here to illustrate the
first two approaches.
a) Application of MASS normalization for a1 results in
T 2 0 1
a1T Ma1 = 1 = [1 1] = 3 2
0 1 1
which gives = 1 / 3 . Therefore, the normalized eigenvector 1 is given as a1 = 1 / 3[1 1] .
T

b) Application of MAX normalization for a 2 results in 2 = 1, since the largest component

[ ]
in a 2 is 1. Therefore, = 1/2 and a 2T = 1 1 .
2
MSC/NASTRAN uses rigid format SOL 103 to solve a real eigenvalue problem, which
offers three basic methods to solve an eigenvalue problem:
- Inverse power method with shifts (INV)
- Givens triangularization method (GIV)
- Householder triangularization method (HOU)
Variations of these methods are also available. A brief discussion of these methods are given
below for reference. See the MSC/NASTRAN handbook for Dynamic Analysis, Section 2.2.4
for a detailed discussion of methods and method selection.

Inverse power method with shifts (INV)

The standard INV method first solves for eigenvalues based on the determinant,
det[K M ] = 0 . Eigenvectors are then solved with the given i as
{K i M }a i = 0 (4.6)
The INV method is applied best for large scaled problems in which few eigenvalues/vectors are
desired.
The SINV method is an enhanced version of the INV method. It uses Sturm sequence number
techniques to make it more likely that all roots in the range have been found. It is generally more
reliable and more efficient than the INV method. For the INV and SINV methods, convergence
is achieved at 10E-6. Convergence is not tested for other methods.

4.3
Givens triangularization method (GIV)
The GIV method is an example of a transformation method in which the real eigenvector,
a i , is mapped to a vector with reduced dimension, b, by the relation, a i = Ab , where Anxm is a
rectangular matrix with n<m. Then Eq. (4.2) becomes AT ( K i M ) Ab = 0 or
( K * i M * )b = 0 where K* and M* are defined as K* = ATKA and M* = ATMA. Note that the
dimension of the above equation is smaller than that of Eq. (4.2). Therefore, the above equation
should be easier to solve. However, the mass matrix for the analysis set must be positive define.
Otherwise, the MGIV method should be used.
The GIV methods are good for small systems requiring many eigenvalues/vectors.

Householder triangularization method (HOU)

The HOU and MHOU methods are generally faster on machines with vector processors. If
METHOD = AGIV or AHOU, the program automatically determines the need for a modified
method (MGIV or MHOU) and makes the proper selection.

IV.3 COMMANDS
Entries and values needed for eigenvalue analysis are as follows:
Executive section
SOL 103
Case control section
METHOD=n
where n corresponds with the SID field on the EIGR entry in the Bulk Data section.
SVECTOR = all or n
requests the output of the solution set of eigenvectors.

Natural frequencies
Bulk Data Entry: EIGR - Eigenvalue Extraction Entry
Description: Defines data needed to perform real eigenvalue analysis

Format:
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

4.4
 EIGR SID METHO F1 F2 NE ND
D
NORM G C

Example:
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
EIGR 13 SINV 1.9 15.6 12
POINT 32 4

or EIGR,13,SINV,1.9,15.6,12,,,,ABC
+BC,POINT,32,4

Field Contents
SID Set identification number (integer >0)
METHOD Method of eigenvalue extraction (character). The method can be
chosen from:
- Inverse power method (INV).
- Inverse power method with enhancements (SINV).
- Givens method of tridiagonalization (GIV).
- Modified Givens method (MGIV).
- Householder method of tridiagonalization (HOU).
- Modified Householder method (MHOU).
NORM Method for normalizing eigenvectors. (Character: MASS, MAX,
or POlNT; Default = MASS).
MASS Normalize to unit value of the generalized mass (Default).
MAX Normalize to unit value of the largest component in the analysis
set.
POINT Normalize to a positive or negative unit value of the component
defined in fields 3 and 4 (defaults to MAX if defined component is
zero).
G Grid or scalar point identification number. Required only if NORM
= POlNT. (Integer > 0).
C Component number. Required only if NORM = POlNT and G is a
geometric grid point. (1 Integer 6).

4.5
 METHOD Field

Field INV or SINV GIV, MGIV, HOU or MHOU

F1 Frequency range of interest. F1 must be Frequency range of interest. If ND is not
F2 input. If METHOD = SINV and ND is blank, F1 and F2 are ignored. If ND is
blank, then F2 must be input. (Real 0.0). blank, eigenvectors are found whose
The units of F1 and F2 are cycles/second. natural frequencies lie in the range
between F1 and F2. (Real 0.0; F1 < F2).
NE Estimate of number of roots in range Not used.
(Required for METHOD = INV). Not
used by SINV method. (Integer > 0).
ND Desired number of roots. If this field is Desired number of eigenvectors. If ND is
blank and METHOD = SINV, then all zero, the number of eigenvectors is
roots between F1 and F2 are searched and determined from F1 and F2. If all three are
the limit is 600 roots. (Integer > 0, Default blank, then ND is automatically set to one
is 3*NE for METHOD = INV only). more than the number of degrees of
freedom listed on SUPPORT entries.
(Integer 0; Default = 0).

Table 4.1. Relationship Between METHOD Field and Other Fields.

Mass Properties
To construct a mass matrix for normal mode analysis, mass density must be known and supplied
this is done on the material properties entry. For anisotropic material, the density would be
specified in the field labeled RHO in the MAT1 entry. If English units are being used for
density, then the weight unit of pounds needs to be converted to mass units. This is done by the
PARAM,WTMASS entry. If the unit of weight is lbs./ft3, 'PARAM,WTMASS, .031056' can be
used to change it to slug/ft3. On the other hand, 'PARAM,WTMASS, .002588' is used to
change the unit of weight, lbs./in3, to the unit of mass, slug/in3. This parameter statement divides
each term of the structural mass matrix by the gravitational constant. For example, when the
units are lb/ft/sec, each term is multiplied by .031056, which is 1/32.2 ft/sec2.

IV.4 EXAMPLES
In this example, the file beam1.bdf from Example 1 of the NASTRAN static analysis section will
be analyzed. We will extract the first two eigenvalues and their corresponding eigenvectors.
Open and edit the file beam1.bdf to effect the following changes:
ID NORMAL_MODE, ANALYSIS

4.6
SOL 103
TIME 10
CEND
TITLE=CANTILEVER BEAM USING BAR ELEMENTS
DISP=ALL
METHOD=1
SPC=100
BEGIN BULK
SPC1,100,123456,1
GRID,1,,0.,0.,0.
GRID,2,,18.,0.,0.
GRID,3,,36.,0.,0.
GRID,4,,54.,0.,0.
GRID,5,,72.,0.,0.
CBAR,1,100,1,2,1.,0.,1.
CBAR,2,100,2,3,1.,0.,1.
CBAR,3,100,3,4,1.,0.,1.
CBAR,4,100,4,5,1.,0.,1.
PBAR,100,10,8.,10.667,2.667,13.334,,,123
+23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
+56,.8333,.8333
MAT1,10,3.+7,,0.33,0.28356
PARAM,WTMASS,0.002588
EIGR,1,SINV,.1,1000., ,2
ENDDATA

Explanation
Executive Control Section
The solution sequence 103 is used here to perform the normal modes analysis.

Case Control Deck

DISP = ALL outputs all the eigenmodes.
METHOD = 1 selects the EIGR card with SID 1. The EIGR defines the method to be used to
perform the normal modes analysis.

4.7
Bulk Data Section
The EIGR entry requests the SINV method to find the eigenvalues/vectors with angular
frequencies ranging from .1 to 1000 cycles per second. The number of desired roots is two.

Output
After all user information messages, an Eigenvalue Analysis Summary is given. The method
used is listed along with the title. Next, a list of eigenvalues is given. Next to the eigenvalue, the
natural frequency in radians per second is given. Note that this value is the square root of the
corresponding eigenvalue. Next the natural frequency is given in cycles per second. The
T
generalized mass is defined as X M X which should be equal to an identity matrix, because the
mass is normalized by default on the EIGR entry. The generalized stiffness matrix is defined as
X T KX . Note also that the generalized stiffness is equal to the eigenvalues. This is shown by
observing the vibration equation:
( K M ) X = 0
T
which is premultiplied by X to obtain
X T ( K M ) X = 0 or X T KX X T MX = 0
Since X T MX = 1 here, one proves that X T MX = I where I is the identity matrix. Also note
that one more eigenvalue is given than was requested. This is because this method only ensures
the accuracy of n-1 eigenvalues computed. Next the eigenvectors are listed as displacements for
each grid.

IV.5 NORMAL MODES ANALYSIS USING P3/PATRAN

In this section, we will learn how to use MSC/PATRAN to set up the bulk data file (.bdf) for the
natural vibration problem (normal modes analysis) of NASTRAN.

Procedures For illustrative purposes, we will use the plate example problem of Section III.5 on
Page 3.32 of the manual. Thus, we only need to modify the database of Section III.5 for normal
modes analysis. Note that the point forces do not participate in normal modes analysis. The steps
involved are as follows.

Step 1: Open the database (.db) file from Section III.5.

Step 2: Modify the material properties by supplying a value for material density. This
information is needed to construct the mass matrix as required for normal modes analysis. To do
this,
(i) Click on Materials radio button.

4.8
(ii) Select Modify under Action.

(iii) Click on the Material whose properties are to be modified under Existing Materials. The
material clicked on will now appear under Material Name.

(iv) Click on Input Properties.... This will bring the material property Subordinate form which
should show existing values for Elastic Modulus (3.0e+7) and Poisson Ratio (0.33). On the row
marked Density, enter the value for density. For this example, use 0.28356. Click on Apply in
this form to register and apply the properties as modified. Click on Cancel to conclude this part.

Step 3: Select a new solution type and the corresponding parameters suitable for normal modes
analysis. To do this,
Type. This will bring the Solution Type Subordinate form. In this form, select the NORMAL
MODES
(i) Click on Analysis radio button. In the main form, click on Solution option. Doing this will
show 103 as the Solution Sequence. This solution sequence is for normal modes analysis.

(ii) Also in the Solution Type form, click on Solution Parameters. This will bring a display of
a second subordinate form called Normal Modes Solution Parameters. The only parameter
we need to change in this form is Wt.-Mass Conversion. This parameter is used to convert
density from one unit to another. The parameter has a default value of 1. This value should be
changed to 0.002588 (This constant converts from lb/in3 to slug/in3).

(iii) In the Normal Modes Solution Parameters form, we deselect the Database Run button
and then click on the Eigenvalue Extraction... button. This will bring a display of yet another
Subordinate form, the Real Eigenvalue Extraction subordinate form. In this form, we select
the method for extracting the eigenvalues in the frequency range we desire. We also specify how
many eigenvalues (out of the extracted ones) and the corresponding eigenvectors we want
printed in the output files (.f06 and .op2). Under Extraction Method, we select Enhanced
Inverse Power since we want only a few eigenvalues/vectors printed. Under Frequency Range
of Interest, we specify 0.1 as the Lower frequency and 100000.0 as the Upper frequency. For
Estimated Number of Roots, we enter 100. Actually we can accept the default value for this
parameter. For the Number of Desired Roots, we enter 5. This implies that only the first 6 (i.e.
5+1) of the extracted eigenvalues/vectors will be be output. Mass (the default) is accepted as

4.9
Normalization Method. Having completed this form, we click on OK to accept the values
specified. We then go back to the Solution Type form and click on OK to conclude this part.

(iv)At this point we go back to the Analysis main form and click on Apply. This will start the
process of preparing the required .bdf for NASTRAN. The .bdf file is eventually submitted to
NASTRAN for analysis as usual.

Step 4: To view the mode shapes from the normal modes analysis results, the .op2 result file is
translated as usual. When we go to result display (by using the Results radio button as usual), we
will have, among other results, six results for the mode shapes corresponding to the first six
eigenvalues/vectors. We can then select any of the mode shape results for display. Shown below
in Fig. 4.2 are plots of the first two mode shapes for our example problem.

(a)First Mode Shape (b) Second Mode Shape

Fig. 4.2: First and Second Mode Shapes of the Frame Example Problem

IV.6 HOMEWORK
1. Shown in Fig. 4.3 is an L-shaped cantilever beam which has a mechanical device that weighs
3000 lbs. The cross sectional area is rectangular (4"x6") skewed by an angle of 30o . The length
of all the elements are 10" each. The material used is steel with E= 30x106 psi,=.33, Sy=
32x103 psi, St=32x103 psi, Ssy=24x103 psi. The device is to operate at a range varying between
60Hz and 120 Hz. Optimize the structural shape to avoid resonance and failure; i.e. resonance
occurs when the output to input ratio is infinite in theory.

4.10
 y

y 6"
z

4"
1 2 3 4
x

z
5
y

3000 lbs
6" x

4"
6

Fig. 4.3: L-Shaped Cantilever Beam

2. Construct the model baja frame shown in Figure 4.2. The coordinates of the grids are given.
Connect these with bar elements. These numbers can be used for grid numbers and node
numbers. If you prefer, you can use your own numbering scheme. These numbers have already
been optimized for connectivity. Grids 7,10,23 and 25 are points of connection for the wheels.
Constrain these points in the x,y,z, x and z directions. The y direction is left free because of
the wheels. Use aluminum for the frame. The material used is steel with E= 10x106 psi, =.33
and the weight is 0.098 lbs per cubic inches. Compute and plot the first two mode shapes for the
frame.

4.11
 8 14

11 17

9
15
1 4 20 24
12
6 18
2 22 26
Z
3 7 13 19 23 17

X 5 10 16 21 25 28
Y
Grid X Y Z
1 108.0 1.0 23.0
2 108.0 23.0 23.0
3 108.0 1.0 0.0
4 96.0 -2.0 26.0
5 108.0 23.0 0.0
6 96.0 26.0 26.0
7 96.0 -2.0 0.0
8 60.0 -6.0 78.0
9 72.0 -8.0 32.0
10 96.0 26.0 0.0
11 60.0 30.0 78.0
12 72.0 32.0 32.0
13 72.0 -8.0 0.0
14 36.0 -2.0 78.0
15 48.0 -4.0 28.0
16 72.0 32.0 0.0
17 36.0 26.0 78.0
18 48.0 28.0 28.0
19 48.0 -4.0 0.0
20 24.0 0.0 24.0
21 48.0 28.0 0.0
22 24.0 24.0 24.0
23 24.0 0.0 0.0
24 0.0 0.0 24.0
25 24.0 24.0 0.0
26 0.0 24.0 24.0
27 0.0 0.0 0.0
28 0.0 24.0 0.0

Fig. 4.4 Mini Baja Frame

4.12
 V. DYNAMIC TRANSIENT ANALYSIS
V.1 OBJECTIVE
The objective of this lab session is to review the transient response analysis and to give the
necessary input data to run the transient analysis in NASTRAN and PATRAN.
Consider a dynamic system upon which a time or a frequency dependent force is acting. The
response of the system is changed as a function of time or frequency. The forces can be either
time dependent or frequency dependent. Each of them is explained separately here.

V.2 TIME DOMAIN ANALYSIS

The responses in the time domain can be computed by several methods such as Newmark-
and Wilson- methods. These methods seek the direct integration of the responses. An
alternative method, known as the Modal Analysis Method, first decouples the mass and stiffness
matrix and then use the Duhamel's integral to integrate the responses. In this context, anyhow,
only the direct approach, Newmark- method, is presented.

Consider the dynamic equation of motion, Eq. (5.1),

(t ) + Cu (t ) + Ku(t ) = p(t )
Mu (5.1)

with initial velocity and initial displacement being defined as u (0 ) = 0 and u(0 ) = 0 ,
respectively. The displacement, velocity and the acceleration vectors are all functions of time and
have to be evaluated at each time instance. The entire time period during which the responses
(displacement, velocity and accelerations) are to be evaluated, is discretized into smaller time
intervals and the responses are updated at each time instance. The selection of the time interval
plays an important part in the computation of the responses. If the time step is chosen incorrectly,
the error increases as the time increases and the responses at the later time stage will be
erroneous.

Using the central difference scheme, the velocity and the accelerations can be obtained in
terms of displacements as,

u n =
1
(u n+1 un1 ) (5.2)
2t
n =
u
1
(u n+1 2un + un1 ) (5.3)
2t 2

5.1
where the subscripts n, n-1 and n+1 indicate the quantities pertaining to the present, the last and
the next time instances, respectively. Substituting these equations into the dynamic equation of
motion, Eq. (5.1) can be re-written as
( )
M
(u n+1 2un + un1 ) + C (u n+1 2un1 ) + K (u n+1 + un + un1 ) = p n+1 + pn + pn1
2t 2
2t 3 3

where the displacement and the forces are averaged out over three adjacent time instances. In
compact, the above set of equations of motion can be written as,
A1 u n +1 = p + A2 un + A3 un 1 (5.4)
where
M C K
A1 = + +
2t 2
2t 3
p + p + p n 1
p = n +1 n
3
M K
A2 = 2
t 3
and
M C K
A3 = +
2t 2
t 3

Thus, with the knowledge of the displacements at the last and the present time instances, the
responses at the next time instance can be easily found out. Furthermore, Eqs. (5.2) and (5.3) can
be used to compute the velocities and the accelerations.

In the case of nonlinear problems, Eqs. (5.2) to (5.4) are solved iteratively as the mass, the
stiffness and the damping matrices are functions of responses.

V.2.1 Commands
The entries and commands needed for the transient analysis are as follows:

Executive Control Section

SOL 27 (69 or 109)
where SOL 27 is the primary set of sequence used to perform the transietnt analysis, SOL 69 is a
later addition and can handle superelements, and SOL 109 is the structured solution sequence and
is the most recent one.
Case Control Section
DLOAD = n

5.2
where n corresponds to the SID on the TLOADi entry in the bulk data deck. This selects the
dynamic loading entry.

TSTEP = n
where n corresponds to the SID on the TSTEP entry in the bulk data deck. This selects the time
step required for numerical integration.

IC = n
where n corresponds to the SID on the TIC entry in the bulk data deck. This selects the initial
conditions (if any, optional) to be applied to the displacements and the velocities. This entry is
not selected if the initial displacements and velocities are zero.
VELOCITY = ALL ( or a set number)
ACCELERATION = ALL ( or a set number)
where output are requested for all or the selected sets of nodal velocities and accelerations.
Bulk Data Section
In order for NASTRAN to analyze a problem of this nature, it needs a set of dynamic loading
vector, the time step, the time interval in which the responses are required and the initial
conditions. In the following section each of the specific NASTRAN commands related to these
data is discussed.

Bulk Data Entry: TLOAD1

Description: Defines the time dependant dynamic load or enforced motion of the form

p(t ) = af (t )
for use in transient response analysis, where is the amplitude and is the time delay of the
load. The DAREA entry defines the magnitude and location of the dynamic forces which are
related to , the DELAY entry defines the magnitude and location of the delay which is related
to , and TABLEDi entry defines the variation of the load vector with time, f (t ) .

FORMAT:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TLOAD1 SID DAREA DELAY TYPE TID

Example:

5.3
 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TLOAD1 5 7 13

or TLOAD1,5,7,,,13

Fields Contents
SID Set identification number. (Integer > 0)
DAREA Identification number of DAREA entry set or (thermal load set in
Heat Transfer Analysis) which defines . (Integer >0)
DELAY Identification number of the DELAY entry that defines (Integer
0 or blank)
TYPE Defines the nature of the dynamic excitation. See Remark 2
(Integer 0, 1, 2, 3 or blank)
TID Identification number of the TABLEDi entry that give f (t ) .
(Integer > 0)

Remarks
1) Dynamic load sets must be selected with the Case Control Command DLOAD = SID
2) The nature of the dynamic excitation is defined in Field 5 according to the following table:
3) If the DELAY is blank or zero then = 0.

Integer Excitation Function

0 or blank Force or Moment
1 Enforced Displacement
2 Enforced velocity
3 Enforced Acceleration
For further information regarding the above entry, refer to the MSC/NASTRAN Handbook for
Dynamic Analysis.

Bulk Data Entry: DAREA

Description: Defines the scale (area) factors for dynamic loads. DAREA is used in conjunction
with the RLOADi and TLOADi entries.

Format

5.4
 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DAREA SID P1 C1 A1 P2 C2 A2
Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DAREA 3 6 2 8.2 15 1 10.1

or DAREA,3,6,2,8.2,15,1,10.1

Fields Contents
SID Identification Number. (Integer >0). This should match with the
one on the TLOAD1 entry.
Pi Grid, extra, or scalar point identification number or the location
where the force is to be applied(Integer > 0)
Ci Component number that gives the direction in which the force is to
be applied(Integer 1 through 6 for the gid point; and blank for 0 or
for extra scalar point)
Ai Scale (Area) factor. (Real). This corresponds to the scale factor to
be applied to the forces defined in the TABLEDi entry.
(Corresponding to SID in TLOAD1 entry)
Remarks:
1) One or two scale factors may be defined on a single entry.
2) Component numbers refer to the displacement coordinate system.
3) Refer to the RLOAD1, RLOAD2, TLOAD1 or TLOAD2 entries for the formulas that define
the scale factor Ai.
4) The above example states that the scale factors for the load along the y-direction at node 6 and
the load along the x-direction at node 15 are 8.2 and 10.1, respectively.

Bulk Data Entry: DELAY

Description: Defines the time delay term in nthe equations of the dynamic loading function.
Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DELAY SID P1 C1 T1 P2 C2 T2

5.5
Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DELAY 5 21 6 4.25 7 6 8.1

or DELAY,5,21,6,4.25,7,6,8.1

Fields Contents
SID Identification number. (Integer >0). This should match with the
one on the TLOAD1 entry.
Pi Grid, extra, or scalar point identification number or the location
where the force is to be applied(Integer > 0)
Ci Component number that gives the direction in which the force is to
be applied (Integer 1 through 6 for the gid point; and blank for 0 or
for extra scalar point)
Ti Time delay (REAL) for the designated point Pi and the
component Ci. It specifies the amount of lag in seconds to be
applied to the Point/Component defined in the above DELAY.

Remark
1) One or two dynamic load time delays may be defined on a single entry.
2) SID must also be specified on the RLOAD1, RLOAD2, TLOAD1 or the TLOAD2 entry. See
those entry descriptions for the formulas that define the manner in which the time delay is
used.
3) A DAREA must also be defined for the same grid point and the component.
The above example specifies the delay for the Z-direction moments at nodes 21 and 7 to be 4.25
and 8.1 units, respectively.

Bulk Data Entry: TABLED1

Description: Defines a tabular function for the use of generatinf frequency-dependant and time
dependant dynamic loads.

Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TABLED1 TID

5.6
 x1 y1 x2 y2 x3 y3 -etc-

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TABLED1 32
-3.0 6.9 0.0 5.6 3.0 5.6 ENDT

or TABLED1,32,,,,,,,ABC
+BC,-3.0,6.9,0.0,5.6,3.0,5.6,ENDT

Fields Contents
TID Table Identification Number. (Integer > 0)
xi, yi Tablular values. (Real). Pairs of (time, forces). xi corresponds to
the time and the yi correspond to F (in the forcing function
equation). A linear interpolation/extrapolation is used to obtain
force function at values different from points xi. An ENDT column
is necessay to end the table.

Remarks
1) xi must be in either ascending or descending order, but not both.
2) Discontinuities may be specified between any two points except the two starting points or the
two end points. For example, in Figure 5.1 discontinuities are allowed only between points x1
through points x2. Also if y is evaluated at a discontinuity, then the average value of y is used.
3) At least one continuation must be specified.
4) For frequency dependant loads the x1 and xi are measured in cycles per unit time.
The above example describes a forcing function shown in Fig 5.2.

5.7
y
6.9
F(t)
5.6

x1 xc x2 -3.0 0.0 3.0

x t
Fig. 5.1 Discontinuity in Forcing Function Fig. 5.2: Forcing function

Bulk Data Entry: TIC

Description: Defines the values for the initial conditions of the variables used in the transient
analysis. Both displacement and velocity values may be specified at the independant degrees of
freedom.

Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TIC SID G C Uo Vo

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9)

TIC 1 3 2 5.0 -6.0

or TIC,1,3,2,5.0,-6.0

Fields Contents
SID Set identification number. (Integer >0)
G Grid, scalar, or the extra point identification number (Integer >0)
of the grid points where the initial conditions are applied.
C Component numbers or the direction in which the initial conditions
are applied. (Integer 1 through 6 for the grid point; and blank for 0
or for extra scalar point)
Uo Initial displacement . (Real)

5.8
Vo Initial velocity. (Real)
In the above example an initial displacement of 5 units and an initial velocity of -6.0 units are
applied to the node 3, in the Y-direction. The rest of the initial conditions are zero.
Remarks
1) Transient initial condition sets must be selected with the Case Control Command IC = SID.
2) If no TIC is selected in the Case Control Section then all the initial conditions are assumed to
be zero.
3) Initial conditions for the coordinates not specified on the TIC entries will be assumed to be
zero.
Bulk Data Entry: TSTEP
Description: Defines the time step intervals at which solution will be generated and output in
transient analysis.
Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TSTEP SID N1 DT1 NO1
N2 DT2 NO2
-etc-

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TSTEP 2 10 .001 5
9 .01 1

or. TSTEP,2,10,.001,5,,,,ABC
+BC,,9,.01,1

Fields Contents
SID Set identification number. (Integer > 0)
Ni Number of time steps of value DTi. (Integer 1)
DTi Time increment. (Real > 0.0)
NOi Skip factor for output. Every NOi-th step will be saved for output.
(Integer > 0)
Remarks

5.9
1) TSTEP entries must be selected with Case contraol Command TSTEP = SID.
2) Note that the entry permits in changes in size of time steps during the course of the solution.
Thus in the example shown, there are 10 steps of value .001 followed by 9 time steps of value
.01. Also, the usre has requested that the output be recorded for t = 0.0, 0.05, .01, .02, .03 etc...
3) Refer to NASTRAN Theoritical manual Sections 11.3, 11.4 for a discussion of considerations
leading to the selection of time steps.

V.2.2 Example
Determine the transient response of the cantilever beam structure excited by a force as shown in
Fig. 2.1. The forcing function is shown in Fig. 5.2. The force is delayed by 0.2 seconds and
applied in the Z direction at node 5. Evaluate the responses for 0.05 seconds in time steps of
0.005 seconds and report the results for every 0.01 seconds. The material and geometric
properties of the cantilever beam are the same as in Example Problem 1 in Chapter II.

ID MEM441, TRANSIENT
$STRUCTURED SOLUTION FOR TRANSIENT ANALYSIS
SOL 109
TIME 100
CEND
$ DYNAMIC LOADING
DLOAD=35
$TIME STEP
TSTEP=45
TITLE=TRANSIENT ANALYSIS, CLASSWORK EXAMPLE PROBLEM
$ OUTPUT THE DISPLACEMENTS, ACCELERATIONS AND
$ THE VELOCITIES
DISP=ALL
ACCELERATION=ALL
VELOCITY=ALL
SPC=100
BEGIN BULK
SPC1,100,123456,1
TLOAD1,35,29,31,3,40
DAREA,29,5,3,-100.0
DELAY,31,5,3,0.2
TABLED1,40,,,,,,,,123

5.10
+23,-3.0,6.9,0.0,5.6,3.0,5.6,ENDT
TSTEP,45,10,0.005,2
GRID,1,,0.,0.,0.
GRID,2,,18.,0.,0.
GRID,3,,36.,0.,0.
GRID,4,,54.,0.,0.
GRID,5,,72.,0.,0.
CBAR,1,100,1,2,1.,0.,1.
CBAR,2,100,2,3,1.,0.,1.
CBAR,3,100,3,4,1.,0.,1.
CBAR,4,100,4,5,1.,0.,1.
PBAR,100,10,8.,10.667,2.667,13.334,,,123
+23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
+56,.8333,.8333
MAT1,10,3.+7,,0.33
ENDDATA

Explanation
Executive Control Section:
The DMAP solution sequence 109 is used to perform the transient analysis. The solution 109 is
a structured (advanced ) sequence and has several optional alter statements included in it.

Case Control Section:

The displacement, velocity and acceleration outputs are requested at every nodal point. The
output for the above run will be in three groups. The first group consists of the displacements, the
second group consists of velocities and the third group consists of acclerations. Again in each
group, the responses at each grid are given at each time step are reported.

The DLOAD entry selects the corresponding TLOAD1 entry ( = 35) from the Bulk Data Section.
Thus the dynamic loading as defined by TLOAD1 entry is selected. The TSTEP entry selects the
time step for numerical integration. This selects the TSTEP ( = 45) entry from the Bulk Data
Section.

Bulk Data Section:

5.11
TLOAD1: Defines the dynamic loading. This entry selects the loading defined in Table with ID
40 multiplied by the magnitude defined in the DAREA entry 29 with a delay defined in DELAY
entry 31. The loading is in the form of a force and thus the TYPE column is equal to zero.

DAREA: Defines the magnitude and the location of the application of theforce.

DELAY: Defines the amount of and the location of the delay. Delay is applied at the same
location as defined in the DAREA card.

TABLED1: Defines the force-time relationship. The table must be ended with an ENDT column.
The relation is interpolated and extrapolated to obtain the forces at the unspecified time
instances.

TSTEP: Defines the step size of 0.005 seconds for 10 time steps.

V.2.3 Time Domain Transient Analysis using PATRAN

In this section, it will be demonstrated how to generate, using PATRAN, part of the bulk
data file (.bdf) for the transient analysis of a 1_D cantilever beam described in the example in
Section V.3 of the lab manual. The required additional data entries to be included in the .bdf file
for transient analysis of this example will then be provided. For detailed theoretical exposition of
Transient Analysis, refer to Section V.2. The step-by-step procedure to use P3/PATRAN for
Transient Analysis is described below.

Step 1. Geometry
Create points 1 and 2 at (x y z) of (0 0 0) and (72 0 0) respectively. Create a 2_Point straight
curve by connecting points 1 and 2.

Step 2. Mesh Generation

First create mesh seeds, four seeds for the curve. Then create a mesh to connect the mesh seeds.
Before completing this step, do not forget to Optimize and and also do Equivalencing.

Step 3. Material Selection

A material with 'Elastic Modulus' of 3.E7, a 'Poission Ratio' of .33 and a density of 0.28356 is
specified in the 'Input Properties...' form. You can specify any name for the material under

5.12
'Material Name'. The density is converted to slug/in3 by specifying a value of 0.002588 for the
Solution Parameter Wt.-Mass Conversion. This is done under Analysis.

Step 4. Element Properties

Proceed with element properties' specification. However in the 'Element Props Applicatin
Menu' form, 'Dimension' should be '1D' and 'Type' should be 'Beam'. Under 'Options' select
'General Section' The user is free to supply 'Property Set Name' of choice. , 'Input
Properties...' is clicked on and the appropriate form is displayed. The information to be used in
this form is shown below.
Material Name: m:steel (Note: Material name 'steel' is used for illustartive purpose)
Bar Orientation: <1 0 1> (Note: This is the orientation vector, V. See p. 2.25-2.26 of the
manual).
[Offset @ Node 1]: Not applicable, skip.
[Offset @ Node 2]: Not applicable, skip
[Pinned DOFs @ Node 1]: Not applicable, skip.
[Pinned DOFs @ Node 2]:
Area: 8.0 (Note: This is a real scalar)
[Inertia 1, 1]: 10.667 (Note: See p. 2.16 of the manual).
[Inertia 2, 2]: 2.667
[Inertia 2, 1]: Not applicable, skip.
Torsional Constant: 13.334 (Note: This is 'J'. See p. 2.16 of the manual)
[Shear Stiff, Y]: 0.8333 (Note: See p. 2.16 of the manual).
[Shear Stiff, Z]: 0.8333
[Nonstructural Mass]: Not applicable, skip.
[Y of Point C]: 2. (Note: See p. 2.16 of the manual).
[Z of Point C]: 1.
[Y of Point D]: 2.
[Z of Point D]: -1.
[Y of Point E]: -2.
[Z of Point E]: -1.
[Y of Point F]: -2.
[Z of Point F]: 1.
Station Distances: Not applicable, skip.

Step 5. Load Case

5.13
Create a load case (you can just retain the deafult load case). Select Time Dependent as the
Load Case Type.

Step 6. Displacement BCs

Constrain all six degrees of freedom for point 1, that is 'Translations' and 'Rotations' vectors
are both equal to <0 0 0>. Remember that boundary conditions are specified using the Load/BCs
radio button.

Step 7. Applied Transient Force

The first step here is to define a tabular field that provides information about the applied force as
a function of time. Then the specified field will then be applied at the desired point on the curve.
For the first part, the radio button Fields is used. By clicking on the Fields button, we have the
appropriate radio button menu form. In this form, select Create as Action, Non Spatial as
Object and Tabular Input as Method. Then, supply any name of choice under Field Name.
Under Table Definition, click on Time (t) as Active Independent Variable. Then click on
Input Data... to supply the data of applied force as a function of time. In the Input Data... form,
enter the following data for the column titled Time(t): -3.0, 0.0, 3.0. For the column titled Value,
enter the following data: 6.9, 5.6, and 5.6. Each data point is for each cell under the appropriate
column.. To enter a data into a cell, click on the reqiured cell, type in the required value in Input
Scalar Data which is now activated and then press Enter on the keyboard. This will place the
data enterd in the cell. After all the data is enetred, there will be three rows and two columns. The
first row will have -3.0 vs. 6.9, the second row 0.0 vs. 5.6 and finally the third row 3.0 vs. 5.6. A
plot of this data is shown in Figure 5.2.
Click on OK to accept the entered data. Go back to the Fields main menu form and click on
[Options...]. In the Options form, select Linear Extrapolation under Extrapolation Option
and then click on OK to accept the options. Go back to the Fields main menu form and click on
Apply to register the field just created. You will now see the name you chose for the Field under
Exixting Fields. This process completes the transient force data field creation.
To apply the force field just created, click on the Load/BCs radio button again. Go through the
process of Force Creation as usual. Do not forget to supply your name of choice under New Set
Name. Clicking on the Input Data... button, you will have the usual Input Data form. However
in this form, since our load case is time dependent, there are tow clolumns to enter our force data.
The first cloumn is for the force data which is space dependence. This column is titled Spatial
Dependence. Since we do not have any such force in this example, we just have to enter <0 0 0>
for the force entry under this column. The second column is for time dependent force data, hence
it has the title Time Dependence. This is where we need to enter our data. Below each column in

5.14
this form, you have another column that allows for the option to specify the applied force from a
field. Since we have created a time dependent field in the first part of this step, the name of the
created field now appears in the column titled Time Dependent Fields. To use this field as our
time dependent force, we first click in the Force cell under the Time Dependence column to
make the cell ready to accept data. We then go to the Time Dependent Fields column and click
on the name of the field we want to apply. This will place the created field in the Force data cell.
We then click on OK to conclude data input.
The force just created is applied at Node 5 of the our 1D curve. This node is specified by going
through Select Application Region... as usual. To conclude the whole process, click on Apply in
the Load/BCs main menu form. Doing this will place the name of the force just created under
Exixting Sets as usual.

Step 8. Creation of .bdf File

Since we need to modify the .bdf file before submittin our job to NASTRAN, instead of the
usual Full Run we normally select as Method in the Analysis main form, select Analysis Deck
in this case. This will only produce the .bdf file without proceeding to analysis automatically.
Accept the default Translation Parameters. In the Solution Type form, select TRANSIENT
RESPONSE since we want to carry out transient analysis. This will ensure the choice of the
Solution Sequence which in this case is sequence 109. In the Solution Parameters subordinate
form, the only parameter needed to be modified is Wt._Mass Conversion as said earlier (see
Material Selection section). The value of this parameter is changed to 0.002588. Next, go to
Subcase Create. Create a subacse as usual. You can just supply Default as the name of the
subcase. Clicking on the Subcase Parameters... button, you have a form that asks for Ending
Time and Number of Time Steps. Enter 0.05 as Ending Time and 5 as Number of Time Steps.
Then click on OK to accept these values. Go back to the Subcase Create form and click on
Output Requests... . In the Output Request form, select Displacements, Velocities and finally
Accelerations as the result entities to be output in the .op2 and .f06 result files. After all these,
go back to the Analysis main menu form and click on Apply. This will start the process of
creating the .bdf file.

Step 9. Modification of .bdf File

Modify the following entry in the .bdf file:
TLOAD1, 4, 5, , , 1
to
TLOAD1, 4, 5, 31, , 1
Add the following entries anywhere below BEGIN BULK in the .bdf file

5.15
DAREA, 5, 5, 3, _100.0
DELAY, 31, 5, 3, 0.2
After all the modifications, save the .bdf file. Then submit your job to NASTRAN simply by
typing at the system prompt:
nastran file.bdf
where file is the name you supplied for your job.

Step 10. Post Processing

Perform the post processing of your .op2 file the usual way. You will have 6 results, one for each
of the time levels 0.0, 0.01, 0.02, 0.03, 0.04 and 0.05
Plot the magnitudes of displacement, velocity and acceleration at time 0.05.
Step 11. Homework
Refer to the in_class lab. session 4 of section 4.6 on page 4.47 of the manual. Use the frame
structure for transient analysis in this homework. The transient force loads are (i) F = 100sin(2t)
lbs. at the point with coordinate (0 1 0) and (ii) F = 2t + 5 lbs. at the point with coordinate (1 1 1).
Both forces act in the negative x_direction (i.e. scaled by -1.0 in direction 1 in the entry
DAREA). Use 0 t 0.5 sec. at an interval of 0.1 sec. Plot the magnitudes of displacement,
velocity and acceleration at time 0.5.

V.3 FREQUENCY DOMAIN ANALYSIS

In the previous sections, the use of NASTRAN and PATRAN to analyze the responses
due to time dependent loading conditions has been demonstrated. In this section, the analyses of
structures (using NASTRAN) subjected to oscillatory excitation will be discussed. This
excitation is explicitly defined in the frequency domain; all of the applied forces are known at
each forcing frequency. The computed responses are complex numbers defined as magnitude and
phase (with respect to the forcing functions) or as real and imaginary components. The
commands are similar to the ones used for transient analysis, but different solution sequence and
the loading format.

Executive Control Section

SOL 26 (or 68 or 108)

SOL 26 is the older sequence, SOL 68 is the sequence which can handle super elements and SOL
108 is the structured sequence with advanced capabilities.

5.16
Case Control Section

DLOAD = n
where n corresponds to the SID of the RLOADi card in the Bulk Data Section. This entry defines
the oscillatory excitation.
FREQUENCY = n
where n corresponds to the SID of the FREQ card in Bulk Data Section.This entry defines the
frequency steps at which the responses arerequired.

Bulk Data Section: RLOAD1

Description: Defines the frequency dependent dynamic load of the form:

p( f ) = A(C ( f ) + iD( f ))e i ( 2 )

where A is the multiplication factor, C and D are the functions of frequencies, f is the frequency,
is the phase lead in degrees, and is the time lag in seconds.

Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
RLOAD1 SID DAREA DELAY DPHASE TC TD

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
TLOAD1 5 3 1

or RLOAD1,5,3,,,1

Fields Contents
SID Set identification number. (Integer > 0)
DAREA Identification number of DAREA entry set (or thermal load set in
Heat Transfer Analysis) which defines A. (Integer >0)
DELAY Identification number of the DELAY entry set that defines .

5.17
 (Integer 0 or blank)
DPHASE Identification number of the DPHASE entry set that defines
(Integer 0 )
TC Set identification number of the TABLEDi entry that gives
C(f). See Remark 2. (Integer 0)
TD Set identification number of the TABLEDi entry that gives
D(f). See Remark 2. (Integer 0)

Remarks
1) Dynamic load sets must be selected with the Case Control Command, DLOAD = SID.
2) If any of DELAY, DPHASE, TC or TD fields are blank or zero, the corresponding , , C(f),
D(f) will be zero. Either TC or TD field may be blank or zero, but not both.
3) SID must be unique for RLOAD1, TLOAD1, RLOAD2, and TLOAD2 entries.

Bulk Data Section: FREQ

Description: Defines the set of frequencies to be used in the solution of frequency response
problems by specification of a starting frequency, frequency increment, and the number of
increments desired.

Format:
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FREQ SID F1 F2 F3 F4 F5 F6 -etc-

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FREQ 3 2.98 3.05 17.9 21.3 25.6 28.8 31.2

or FREQ,3,2.98,3.05,17.9,21.3,25.6,28.8,31.2

Fields Contents
SID Set Identification number. (Integer > 0)
Fi Frequency values at which the force is applied. (Real > 0.0)

Remarks
1) The FREQ entries are selected with Case Control Command FREQUENCY = SID.

5.18
2) The units for Fi is cycles per unit time.
3) All FREQ, FREQ1, FREQ2 entries with the same frequency set identification numbers will
be used. Duplicate frequencies will be ignored. Frequencies, fN and fN-1 are considered
duplicated if
| fN - fN-1 | < 105 * | fMAX - fMIN |

Bulk Data Section: FREQ1

Description: Defines the set of frequencies to be used in the solution of frequency response
problems by specification of a starting frequency, frequency increment, and the number of
increments desired.

Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FREQ1 SID F1 DF NDF

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
FREQ1 6 2.9 0.5 13

or FREQ1,6,2.9,0.5,13
Fields Contents
SID Set Identification number. (Integer > 0)
F1 First Frequency value at which the force is applied. (Real > 0.0)
DF Frequency increment. (Real > 0.0)
NDF Number of frequency increments. (Integer > 0.0)

Remarks
1) The FREQ1 entries are selected with Case Control Command FREQUENCY = SID.
2) The units for F1 is cycles per unit time.
3) The frequencies defined by this entry are given by:
fi = F1 + DF * (i - 1)
where i = 1 to (NDF + 1).

5.19
4) All FREQ, FREQ1 and FREQ2 entries with the same frequency set identification numbers
will be used. Duplicate frequencies will be ignored. Frequencies, fN and fN-1 are considered
duplicated if
| fN - fN-1 | < 105 * | fMAX - fMIN |

Bulk Data Section: DPHASE

Description: Defines the phase lead term in the equation of the dynamic loading function.

Format

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DPHASE SID P1 C1 TH1 P2 C2 TH2

Example

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
DPHASE 4 21 6 2.1 8 6 7.2

or DPHASE,4,21,6,2.1,8,6,7.2

Fields Contents
SID Set Identification number. (Integer > 0)
Pi Grid, extra, or scalar point identification number. (Integer > 0)
Ci Component number. ( Integer 1 through 6 for grid points; zero or
blank for extra or scalar points.)
THi Phase angle in degrees. (Real)

Remarks
1) One or two dynamic load phase lead terms may be defined on a single entry.
2) SID must be referenced on a TLOAdi or RLOADi entry. Refer to RLOADi or RLOAD2
entry for the formulas that define hoe the phase lead anlge is used.
3) A DAREA entry must be defined for the same grid point and component.

Example

5.20
Perform the frequency domain analysis for the beam described in Sections V.3 and V.4 subjected
to the forcing function
p( f ) = 100. C ( f ) e i (20 0.4f )
0

where C(f) is defined by the graph shown in Fig. 5.3

5.6
C(f)

4.0

2.0 4.0 6.0

Frequency
Fig. 5.3: Forcing Function for Frequency Domain Analysis.

The .bdf file is given as follows:

ID MEM441, FREQUENCY
$STRUCTURED SOLUTION SEQUENCE FOR FREQ. RESPONSE ANALYSIS
SOL 108
TIME 100
CEND
$LOAD SET 35, FREQUENCY SET 45
DLOAD=35
FREQUENCY=45
TITLE=CANTILEVER BEAM USING BAR ELEMENTS,
$ FREQUENCY DOMANIN TRANSIENT ANALYSIS.
DISP=ALL
ACCELERATION=ALL
VELOCITY=ALL
SPC=100
BEGIN BULK
SPC1,100,123456,1
RLOAD1,35,29,31,32,40
$RLOAD1, DAREA, DELAY, DPHASE, TC, TD

5.21
DAREA,29,5,3,-100.0
DELAY,31,5,3,0.2
TABLED1,40, , , , , , , , 123
+23,2.0,4.0,4.0,5.6,6.0,5.6,ENDT
FREQ,45,2.98,3.05,17.9,21.3,25.6,28.8,31.2,3456
+456,29.2,22.4,19.3
DPHASE,32,5,3,20.0
GRID,1,,0.,0.,0.
GRID,2,,18.,0.,0.
GRID,3,,36.,0.,0.
GRID,4,,54.,0.,0.
GRID,5,,72.,0.,0.
CBAR,1,100,1,2,1.,0.,1.
CBAR,2,100,2,3,1.,0.,1.
CBAR,3,100,3,4,1.,0.,1.
CBAR,4,100,4,5,1.,0.,1.
PBAR,100,10,8.,10.667,2.667,13.334, , , 123
+23,2.,1.,2.,-1.,-2.,-1.,-2.,1.,456
+56,.8333,.8333
MAT1,10,3.+7,,0.33,0.28356
ENDDATA

Explanation

Executive Control Section:

The new thing in this section is the solution sequence 108, which is the solution sequence for
frequency domain analysis. The unstructured solution sequences 26 or 68 also can be used.

Case Control Section:

The FREQUENCY = n is the new card here. This card selects the FREQ or FREQ1 card from
the Bulk Data Section.

Bulk Data Section:

In this section, the three new entries introduced are the RLOAD1, FREQ and the DPHASE. The
RLOAD1 entry is similar to theTLOAD1 entry, but this entry selects the DAREA, DELAY,
DPHASE, TC and TD entry IDs. The DAREA and DELAY entries are exactly the same as

5.22
defined in the previous sections, the DPHASE entry defines the phase angle applied to the
forcing function. The TC and TD are the IDs of the TABLEDi entries giving the values of C and
D.

5.23
 VI. NASTRAN ANALYSIS OF LAMINATED COMPOSITE PANELS

VI. 1 Introduction

Laminated Composite Panels have been widely used in shipbuilding, automobiles,

airplanes for its light weight, yet adequate strength. Such panels are made of laminated plies
with long or chopped fibers aligned in different orientations and embedded in resins. Figure 1
shows a composite panel made of 8 plies. The angles in this figure indicate the fiber orientation
in each ply with reference to the material coordinate system of the composite panel. The angle
is measured from the X-axis of the material coordinate system to the fiber direction. The
lamniates are usually coded based upon the orientations of the plies. Starting from the botton of
the composite panel (i.e., -Z in the material coordinate system in NASTRAN), the plies in the
panel labeled as [0/-45/90/45/0/0/45/90/-45/0] in Fig. 20 have fibers oriented as 00, -450, 900,
450, 00, 00, -450, 900, 450 and 00 in sequence. Since the plies are usually laminated
symmetrically with respect to the centroid plane, the label can be shorted as[0/-45/90/45/0]S .

!
!
Figure 1. Composite Panels of 8 Plies
!

6.1$
 !
!
Figure 2. Labeling Orientations of Fibers in a Laminated Composite Panel

The stress-strain relation in each ply follows the rule for an orthotropic material which
is given by
- 1 *
+ E 12 0 (
' 1 $ + 1 E2 ( ' 1 $
! ! + 12 1 ! !
& 2 # = + 0 ( & 2 # (1)
E E (
!2 ! + 1 2 ! !
% 12 " 1 ( % 12 "
+ 0 0 (
,+ G12 )(
Note that the Poisson ratio 21 can be calculated from the rest of the material properties, 12 ,
E
E1 and E2 as 21 = v12 2 . In the composite laminated panel, the direction 1 is along the
E1
direction of fiber and 2 of resin. Thus, E1 is much stronger than E2 .

VI.2 MAT8 and PCOMP Cards

The laminated composite panel can be analyzed in the same way as the metal plates or
shells using CQUAD elements. However, since the laminated composite panel is a stacked of
plies each of which is modeled as an orthotropic materials, a new material card, MAT8 is used
here, instead of MAT1, to represent the relation given in Eq. (1). Furthermore, PCOMP is used,
instead of PSHELL, to detail the stacking sequence, the orientation and the fiber orientation of
each ply. The details of those two cards are explained below.

6.2$
Bulk Data Entry: MAT8 - Material Property Definition, Form 8
Description: Defines the material properties for linear, temperature-independent, orthotropic
materials
Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MAT8 MID E1 E2 NU12 G12 G1Z G2Z RHO
A1 A2 TREF Xt Xc Yt Yc S
GE F12 STRN

Example:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
MAT8 171 3.e+7 1.e+6 0.33 2.e+6 0.056
or
MAT1,171,3.e+7,1.e6,.33,2.e6,,,0.056

The Youngs moduli of the material is 3.e+7 units along fiber direction and 1.e+6 in the
direction normal to it. The in-plane Poissons ratio is 0.33, the in-plane shear modulus is 2.e+6
and the mass density 0.056. Since the allowable stresses and the failure modes are not the major
concern here, other input data in the second and the third rows can take default values.

Field Contents

MID Material identification number. Referenced on a PSHELL of PCOMP entry

only. (a positive integer )
E1 Modulus of elasticity in longitudinal direction, also defined as the fiber direction
or 1-direction. (Real, positive)
E2 Modulus of elasticity in lateral direction, also defined as the matrix direction or
2-direction. (Real, positive)
NU12 In-plane Poissons Ratio, v12. (Real, positive)
G12 In-plane Shear Modulus (Real, positive)
G1Z Transverse shear modulus for shear in 1-Z plane, where Z is the through
thickness direction (Real positive; Default implies infinite shear modulus)

6.3$
G2Z Transverse shear modulus for shear in 2-Z plane. (Real, positive; Default implies
infinite shear modulus.)
RHO Mass density (Real, positive)
Ai Thermal Expansion Coefficient in the ith-direction. (Real, positive)
TREF Reference temperature for the calculation of a) thermal loads, or b) a
Temperature-dependent thermal expansion coefficient.
Xt, Xc Allowable stresses or strains in tension and compression, respectively, in the
longitudinal direction. Required, if failure index is desired. Default value of Xc
is Xt. (Real, positive)
Yt, Yc Allowable stresses or strains in tension and compression, respectively, in the
lateral direction. Required of failure index is desired. Default value of Yc is Yt.
(Real, positive)
GE Structural damping coefficient. (Real, positive)
F12 Interaction term in the tensor polynomial theory of Tsai-Wu. Required if failure
index by Tsai-Wu theory is desired and if value of F12 is different from 0.0
(Real, positive).
STRN For the maximum strain theory only (see STRN in PCOM entry). Indicates
whether Xt, Xc, Yt, Yc and S are stress or strain allowable. (Real = 1.0 for strain
allowable; blank (default) for stress allowable).
Remarks:
1. If G1Z and G2Z values are specified as zero or blank, then transverse shear flexibility
calculations will not be performed, which is equivalent to zero shear flexibility (i.e., infinite
shear stiffness).
2. An approximate value for G1Z and G2Z is the in-plane shear modulus G12. If test data are
not available to accurately determine G1Z and G2Z for the material and transverse shear
calculations are deemed essential; the value of G12 may be supplied for G1Z and G2Z. In
solutions 66 and 106, linear and nonlinear elastic material properties in the residual
structure will be updated as prescribed in the TEMPERATURE Case Control command.
3. Xt, Xc, Yt, Yc and S are used for composite element failure calculations when requested in
the FT field of the PCOMP entry.
4. TREF and GE are ignored if this entry is referenced by a PCOMP entry
5. TREF used in two ways:
a. In nonlinear static analysis (SOLs 66 and 106), TREF is used only for the
calculation of a temperature-dependent thermal expansion coefficient. The reference
temperature for the calculation of thermal loads is obtained from the

6.4$
 TEMPERATURE (INITIAL) set selection. See Figure 1 in Remark 10 in the MAT1
description.
b. In all solutions except 66 and 106, TREF is used only as the reference temperature
for the calculation of thermal loads. TEMPERATURE (INITIAL) may be used for
this purpose, but TREF must then be blank.

Bulk Data Entry: PCOMP - Shell Element Property for Laminated Panels
Description: Defines the membrane, bending, transverse shear, and coupling properties of thin
shell elements.
Format:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PCOMP PID Z0 NSM SB FT TREF GE LAM
MID1 T1 THETA1 SOUT1 MID2 T2 THETA2 SOUT2
MID3 T3 THETA3 SOUT3 Etc.

Examples:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
PCOMP 181 -0.224 7.45 10000.0 HOFF
171 0.056 0.0 YES 45.0
-45.0 90.0

or
PCOMP,181,-.0224,7.45, 10000.0, HOFF,,,,ABC
+BC, 171,0.056,0.0, YES,,,45.,,,CDF
+DF,,,-45.,,,,90.

The PCOMP card, ID number 181 specifies the material properties. The distance between the
reference plane to the bottom surface is 0.224 in the negative Z direction. The non-structure
mass per unit area is 7.45, the allowable shear stress is 10000, while the failure is calculated
based upon the Hoffman theory. The value of the thickness is 0.056. The element is made of
four plies whose thickness and the material properties are the same. The laminated plate is not
symmetric and arranged as [0./45./-45./90].

6.5$
Field Contents
PID Property identification number (0<Integer <1000000)
Z0 Distance from the reference plane to the bottom surface. (Real;
Default = -1/2 the thickness of the element).
NSM Nonstructural mass per unit area. (Real)
SB Allowable shear stress of the bonding material (allowable
interlaminar shear stress). Required if failure index is desired.
(Real > 0.0).
FT Failure theory. The following theories are allowed (Character or
blank. If blank then no failure calculation will be performed):

HILL for the Hill theory

HOFF for the Hoffman theory
TSAI for the Tsai-Wu theory
STRN for the maximum strain theory
TREF Reference temperature. (Real).
GE Damping coefficient. (Real; Default = 0.0).
LAM Symmetric lamination option. If LAM = SYM, only plies on
one side of the element centerline are specified. (Character of
blank. If blank, all plies must be specified).
MIDi Material ID of the various plies. The plies are identified by
serially numbering them from 1 at the bottom layer. The MIDs
must refer to MAT1, MAT2, or MAT8 Bulk Data entries. See
Remark 1. (Integer>0 or blank except MID1 must be specified).
Ti Thicknesses of the various plies. (Real of blank except T1 must
be specified).
THETAi Orientation angle of the longitudinal direction of each ply with
the material axis of the element. (If the material angle on the
element connection entry is 0.0, the material axis and side 1-2 of
the element coincide). The plies are to be numbered serially
starting with 1 at the bottom layer. The bottom layer is defined as
the surface with the largest Z value in the element coordinate
system. (Real) .
SOUTi YES, requires the stress out for the associated ply. Blank or NO,
no stress output.

6.6$
Remarks:

1. The default for MID2 , , MIDn is the last defined MIDi. In the example above, MID1
is the default for MID2, MID3 and MID4. The same logic applies to Ti.
2. At least one of the four values (MIDi, Ti, THETAi, SOUTi) must be present for a ply to
exist. The minimum number of plies is one.
3. TREF given on the PCOMP entry will be used for all plies of the element; it will
override values supplied on the material entries for individual plies.
4. GE given on the PCOMP entry will be used for the element; it will override values
supplied on the material entries fir individual plies.
5. Stress and strain output for individual plies are available in all superelement static and
normal modes analysis. In SOL 24 or SOL 47, RF Alter RFXXS79 is required.
6. Strains for individual plies are output if PARAM, LSTRN,1 is input. IF PARAM,
NOCOMPS is set to -1, stress output for individual plies will be suppressed and the
homogeneous stress output will be printed.
7. ELFORCE and ELSTRESS requests must be present for all elements for which ply
stress or failure index output is desired.
8. A function of this entry is to derive equivalent internal PSHELL and MATi entries to
describe the composite element. Any sorted echo request will also cause printout and/or
punch of the derived PSHELL and MATi entries in User Information Message 4379 and
/or the punch file. (See the MSC/NASTRAN Users manual, Selection 1.13 for proper
interpretation of the output from User information Message 4379.). However, if these
equivalent PSHELL and MAT2 entries are input, then stress or strain output for
individual ply is not available and PARAM, NOCOMPS, -1 must be supplied.
9. The failure index for the boundary material is calculated as failure index = max(t1z,
t2z)/SB.
10. If the value specified for Z0 is not equal to the negative of one-half the thickness of the
element and PARAM, NOCOMPS, -1 is specified, the homogeneous element stresses
are incorrect, while lamina stresses and element forces and strains are correct. For the
correct homogeneous stresses, use ZOFFS on the corresponding connection entry.

6.7$
VI.3 Example Problem

A 4x4 laminated composite panel is loaded with a uniformly distributed load, 5 psi.
All edges of the panel are fully fixed. The thickness of the panel is 0.3 inches. The laminated
panel is made of three plies, [0/90/0]. The material properties are given as

E11 = 25e6 psi

E22 = 1e6 psi
G12 = 0.5e6 psi
G13 = G23 = 0.2e6 psi
Poissons ratio: 12 = 0.25
Density: = 0.09 lb/in3

Analyze the panel for deflection and stresses.

Figure 3. Mesh Plot of Laminated Panel with BC and Load

Figure 4. Ply Orientation of Laminated Panel

6.8$
The panel is discretized into 16 CQUAD elements. The bdf file of the model is listed below.
$ NASTRAN input file created by the Patran 2010.1.2 64-Bit (MD Enabled)
$ input file translator on November 10, 2014 at 11:16:37.
$ Direct Text Input for Nastran System Cell Section
$ Direct Text Input for File Management Section
$ Direct Text Input for Executive Control
$ Linear Static Analysis, Database
SOL 101
CEND
$ Direct Text Input for Global Case Control Data
TITLE = MD Nastran job created on 10-Nov-14 at 11:14:05
ECHO = NONE
SUBCASE 1
$ Subcase name : Default
SUBTITLE=Default
SPC = 2
LOAD = 2
DISPLACEMENT(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
STRESS(SORT1,REAL,VONMISES,BILIN)=ALL
BEGIN BULK
$ Direct Text Input for Bulk Data
PARAM POST 0
PARAM PRTMAXIM YES
$ Elements and Element Properties for region : composite_plate
$ Composite Property Record created from material record : laminate
$ Composite Material Description :
PCOMP 1
1 .1 0. YES 1 .1 90. YES
1 .1 0. YES
$ Pset: "composite_plate" will be imported as: "pcomp.1"
CQUAD4 1 1 1 2 7 6 0.
CQUAD4 2 1 2 3 8 7 0.
CQUAD4 3 1 3 4 9 8 0.
CQUAD4 4 1 4 5 10 9 0.
CQUAD4 5 1 6 7 12 11 0.
CQUAD4 6 1 7 8 13 12 0.
CQUAD4 7 1 8 9 14 13 0.
CQUAD4 8 1 9 10 15 14 0.
CQUAD4 9 1 11 12 17 16 0.
CQUAD4 10 1 12 13 18 17 0.
CQUAD4 11 1 13 14 19 18 0.
CQUAD4 12 1 14 15 20 19 0.
CQUAD4 13 1 16 17 22 21 0.
CQUAD4 14 1 17 18 23 22 0.
CQUAD4 15 1 18 19 24 23 0.
CQUAD4 16 1 19 20 25 24 0.
$ Referenced Material Records
$ Material Record : lamina
$ Description of Material : Date: 10-Nov-14 Time: 11:10:52
MAT8 1 2.5+7 1.+6 .25 500000. 200000. 200000. .09

6.9$
$ Nodes of the Entire Model
GRID 1 0. 0. 0.
GRID 2 1. 0. 0.
GRID 3 2. 0. 0.
GRID 4 3. 0. 0.
GRID 5 4. 0. 0.
GRID 6 0. 1. 0.
GRID 7 1. 1. 0.
GRID 8 2. 1. 0.
GRID 9 3. 1. 0.
GRID 10 4. 1. 0.
GRID 11 0. 2. 0.
GRID 12 1. 2. 0.
GRID 13 2. 2. 0.
GRID 14 3. 2. 0.
GRID 15 4. 2. 0.
GRID 16 0. 3. 0.
GRID 17 1. 3. 0.
GRID 18 2. 3. 0.
GRID 19 3. 3. 0.
GRID 20 4. 3. 0.
GRID 21 0. 4. 0.
GRID 22 1. 4. 0.
GRID 23 2. 4. 0.
GRID 24 3. 4. 0.
GRID 25 4. 4. 0.
$ Loads for Load Case : Default
SPCADD 2 1
LOAD 2 1. 1. 1
$ Displacement Constraints of Load Set : constraint_xy
SPC1 1 123456 1 2 3 4 5 6
10 11 15 16 20 21 22 23
24 25
$ Pressure Loads of Load Set : pressure_uniform
PLOAD4 1 1 -5. THRU 16
$ Referenced Coordinate Frames
ENDDATA b662e607

Remarks:
In comparison to the previous Patran procedure, the construction of a laminated plate for
composite panel involves the major change in the subsequent three steps.

Step 1: Define laminates material

In this step, we will define the material properties of the constitutions of the composite
material. The laminated composite, which is made of the same material but layered at various
orientations in this example, has its properties defined under the Material tab. In order to
define these material properties, one must choose Create under Action, 2d Orthotropic
under Object and Manual Input under Method. Also, one must enter the chosen material
name in the Material Name box; for this case, the name was chosen to be lat_mat. Later,

6.10$
click on the Input Properties to enter the properties of the composite such as Youngs
moduli, shear moduli, Poissons ratio and density. Lastly, we will have to click on OK and
Apply when done to register the laminated panel material.
The result should look like the figure 5 below.

Figure 5. Material Properties Tab

Step2: Define laminates properties

After defining the material of the laminated panel, its properties are described in this
step. The property of a laminated plate involves the thickness and the orientation of each of its
layers. In order to build up the laminated plate from each of the individual layer, one would
open the Material tab; after that, choose Create under Action, Composite under Object
and Laminate under Method. One also must enter the name of the laminated plate under
Material Name, in this case the laminated plate was name lat.
The sub window will appear prompting the user to input each of the layer thickness and
orientation. In this example, the composite material has three 0.1 inch thick layers which are

6.11$
positioned at 0 deg, 90 deg and 0 deg respectively. The Input Data box will be used to enter
the name of the laminate material, for this particular example, the name of each layer is
lat_mat not lat. This box is also used to enter the thickness and orientation for each layer.
The complete table should look similar to the figure 6 below.

Figure 6. Laminated Composite Properties Tab

Step 3: Set up and run the analysis

In order to run the analysis for the laminated panel, one would follow the succeeding
steps, select Analyze, Entire Model and Full Run under Action, Object and Type
correspondingly. Later, one will have to look down to the bottom of the Analysis tab and click
on Subcases. On the Subcases sub tab, select Default and click on Output Requests to
open the next form.
On the Output Requests form, select Advanced on the Form Type box. Also, select
Element Stresses in Select Result Type box and Stresses in Output Requests box. Click
OK and Apply to activate the choice. You should be done for now and ready to run the
analysis test on the model.
The figure 7 below presents data needed to fill in the sub tabs Output Requests.

6.12$
 Figure 7. Analysis Tab

The deformation of the panel as reported by the associated f06 file is shown in Fig. 8.
The part of the stress output is reported in Table 1. Note that the stress is reported for each ply
at the center of the CQUAD element on the mid-plane of the ply. As examples, the stress
contour plot of x for Ply 1 is shown in Fig. 9. Furthermore, the stress contour plots of von
Mises stresses for each of three plies are shown in Fig. 10 to 12, respectively.

6.13$
 !
Figure 8. Deformation Contour Plot

Figure 9. x in Ply 1.

6.14$
 Figure 10. von Mises Stress in Ply 1

Figure 11. von Mises Stress in Ply 2

Figure 12. von Mises Stress in Ply 3

6.15$

$
 Table 1. Stresses in Laminated Composite Elements

S T R E S S E S I N L A Y E R E D C O M P O S I T E E L E M E N T S ( Q U A D 4 )
ELEMENT PLY STRESSES IN FIBER AND MATRIX DIRECTIONS INTER-LAMINAR STRESSES PRINCIPAL STRESSES (ZERO SHEAR) MAX
ID ID NORMAL-1 NORMAL-2 SHEAR-12 SHEAR XZ-MAT SHEAR YZ-MAT ANGLE MAJOR MINOR SHEAR
0 1 1 -4.33995E+01 -5.53926E+00 -3.39293E+00 -1.28798E+01 -1.96389E+00 -84.92 -5.23760E+00 -4.37011E+01 1.92318E+01
0 1 2 -4.78226E-06 -1.10429E-07 1.26396E-07 -1.28798E+01 -1.96389E+00 88.45 -1.07012E-07 -4.78568E-06 2.33933E-06
0 1 3 4.33995E+01 5.53926E+00 3.39293E+00 -1.79479E-14 -1.28281E-14 5.08 4.37011E+01 5.23760E+00 1.92318E+01
0 2 1 3.90727E+01 -1.17678E+01 -1.81428E+00 -3.73910E+00 -4.66679E+00 -2.04 3.91374E+01 -1.18324E+01 2.54849E+01
0 2 2 -1.13362E-05 -5.07550E-08 6.75870E-08 -3.73910E+00 -4.66679E+00 89.66 -5.03503E-08 -1.13366E-05 5.64312E-06
0 2 3 -3.90727E+01 1.17678E+01 1.81428E+00 -5.21041E-15 -3.04834E-14 87.96 1.18324E+01 -3.91374E+01 2.54849E+01
0 3 1 3.90727E+01 -1.17678E+01 1.81428E+00 3.73910E+00 -4.66679E+00 2.04 3.91374E+01 -1.18324E+01 2.54849E+01
0 3 2 -1.13362E-05 -5.07550E-08 -6.75870E-08 3.73910E+00 -4.66679E+00 -89.66 -5.03503E-08 -1.13366E-05 5.64312E-06
0 3 3 -3.90727E+01 1.17678E+01 -1.81428E+00 5.21041E-15 -3.04834E-14 -87.96 1.18324E+01 -3.91374E+01 2.54849E+01
0 4 1 -4.33995E+01 -5.53926E+00 3.39293E+00 1.28798E+01 -1.96389E+00 84.92 -5.23760E+00 -4.37011E+01 1.92318E+01
0 4 2 -4.78226E-06 -1.10429E-07 -1.26396E-07 1.28798E+01 -1.96389E+00 -88.45 -1.07012E-07 -4.78568E-06 2.33933E-06
0 4 3 4.33995E+01 5.53926E+00 -3.39293E+00 1.79479E-14 -1.28281E-14 -5.08 4.37011E+01 5.23760E+00 1.92318E+01
0 5 1 -9.90257E+01 4.11501E+00 -2.87314E+00 -3.10029E+01 -8.61216E-02 -88.41 4.19499E+00 -9.91057E+01 5.16503E+01
0 5 2 4.72920E-06 -1.01801E-07 1.07033E-07 -3.10029E+01 -8.61216E-02 1.27 4.73157E-06 -1.04171E-07 2.41787E-06
0 5 3 9.90257E+01 -4.11501E+00 2.87314E+00 -4.32022E-14 -5.62546E-16 1.59 9.91057E+01 -4.19498E+00 5.16503E+01
0 6 1 1.03352E+02 1.31920E+01 -1.29449E+00 -9.78424E+00 -3.62654E-01 -0.82 1.03371E+02 1.31734E+01 4.50988E+01
0 6 2 1.13892E-05 2.62985E-07 4.82233E-08 -9.78424E+00 -3.62654E-01 0.25 1.13894E-05 2.62776E-07 5.56333E-06
0 6 3 -1.03352E+02 -1.31920E+01 1.29449E+00 -1.36343E-14 -2.36885E-15 89.18 -1.31734E+01 -1.03371E+02 4.50988E+01
0 7 1 1.03352E+02 1.31920E+01 1.29449E+00 9.78424E+00 -3.62654E-01 0.82 1.03371E+02 1.31734E+01 4.50988E+01
0 7 2 1.13892E-05 2.62985E-07 -4.82233E-08 9.78424E+00 -3.62654E-01 -0.25 1.13894E-05 2.62776E-07 5.56333E-06
0 7 3 -1.03352E+02 -1.31920E+01 -1.29449E+00 1.36343E-14 -2.36885E-15 -89.18 -1.31734E+01 -1.03371E+02 4.50988E+01
0 8 1 -9.90257E+01 4.11501E+00 2.87314E+00 3.10029E+01 -8.61217E-02 88.41 4.19499E+00 -9.91057E+01 5.16503E+01
0 8 2 4.72920E-06 -1.01801E-07 -1.07033E-07 3.10029E+01 -8.61217E-02 -1.27 4.73157E-06 -1.04171E-07 2.41787E-06
0 8 3 9.90257E+01 -4.11501E+00 -2.87314E+00 4.32022E-14 -5.62546E-16 -1.59 9.91057E+01 -4.19499E+00 5.16503E+01
0 9 1 -9.90257E+01 4.11501E+00 2.87314E+00 -3.10029E+01 8.61217E-02 88.41 4.19499E+00 -9.91057E+01 5.16503E+01

6.16$

$
VI.4 Homework Assignment:

Assume that the cantilever plate given in Example 2 on Page 2.35 is now made of laminated
composite panel with four plies. The total thickness is maintained the same as 0.6 inches. Each
ply is then with a thickness of 0.15 inches. [0/45/-450/0]. The material properties are given as

E11 = 30e6 psi

E22 = 2e6 psi
G12 = 0.4e6 psi
G13 = G23 = 0.2e6 psi
Poissons ratio: 12 = 0.3
Density: = 0.09 lb/in3

Analyze the cantilever beam and plot out the deflection and the von Mises stress on the top ply.
( See Section II. 7 to get PATRAN plots from a bdf file )

z
z
y
2 3 5 7 9 0.6"

x y
6"
1 2 3 4

1 18" 4 18" 6 18" 8 18" 10

6.17%