re A First Course . =9 Cee gd Fi enn on Cae DARYL L. LOGAN NOILIG3 GNOOD3S WMret PWS SERIES IN ENGINEERING a Second Edition png of Motes Sor rca Maho fr Eng mA bngnering an nrc ee aeepacencr Spsene Desi, second Eaton Ce er ronson enering Sze EON a Per Comecniel Engrg, Seo Ei tan Pomp of So Dynamics fat Cnc f Teil Grape Pada, Power Electromics Fancher, Engineering Economy: Copial Allocation Theory eee: Tasodheako, Mechanics of Materials Third Edition Cher and Sarma, Power Sytems Analysis and Design nana fevducion 1 Fld Mechanix, Second Esition Jeno ad Karim, Digital Design anim, Elec Opel Devices and Systems outs fnrndaction 1 CAD Using CADKEY, Second Edition Kanwhe The Pinte Flement Method in Machine Design ce Fen Cowen the Fine Element Method, Second Edition Mth ond King. Enyinering Mechanic: Sates, Second Edition McC and King, Enpnering Mechanics An Introduction to Dynamics, Sond Edition McC und King. Engineering Mechanic: Statics and An Introduction to Dynamics, Second Fition Polaris and Sel, Signals and Systems, Second Edition Poslanas and Sely Element of Signals and Systems Rent Hul/Abbarchian, Physical Metllaryy Principle, Third Edition Reyaoids, Ut Operations and Processes in Enironmental Engineering Sack Maar Srctral Analy ‘Schemas and Wong. Fundamentals of Surveying, Third Edition Sep, Fendamenalt of Structural Stel Design ‘Shes and Kong. Applied Blectromagnetiom. Second Edition See Mamtarreny Fates Wenman. VAX FORTRAN, Scood Edition ‘eseman. FORTRAN for Scientists and Engincers fove, FORTRAN Second Edition SECOND EDITIO ——__ 559s ee A First Course oe in the Finite Element Method DARYL L. LOGAN Rose-Hulman Institute of Technology PWS Publishing Company BOSTONMING ah va, Boston, MA 02116-4924 20 Park Pi TE Ty rae Sy ay mee, tare echnical: pane. Tiara pa en porno oe pba, PS Png Comper (Pe nang Company dine of Wade ery Comer Catenin Pablation Data Ag Day count element method Day. Logan — inde 7m Iaten behographical references and inde, IsENesiesabe8 MOT 1 Pte demented 1 Te TAurhstaa it p \ Sra01s1959—4e f sens orp Se Contents ee 1 Introduction Prologue 1 11 Brief History 2 1.2 Introduction to Matrix Notation 4 13 Role ofthe Computer 6 14 General Steps of the Finite Element Method 6 15 Applications ofthe Finite Element Method 13 1.6 Advantages of the Finite Element Method 20 1.7 Computer Programs for the Finite Element Method 20 References 22 Problems 25 2 Introduction to the Stiffness (Displacement) Method a Introduction 27 21 Definition ofthe Stiffness Matrix 28 22 Derivation of the Stiffness Matrix for a Spring Element 28 23 Example ofa Spring Assemblage 32 24 Assembling the Total Stiffness Matrix by Superposition (Direct Stiffness Method) 34 25 Boundary Conditions 37 26 Potential Energy Approach to Derive Spring Element Equations 47Development of Truss Equations Iearoducnon 60 1 Derivation of the Stiflness Matrix for a Bar Element 12 Selecting Approximation Functions for Displacements 13 Trandformation of Vectors in Two Dimensions 68 9 Seifiness Matrix 71 15 Computation of Stress fora Bar in the x-y Plane 75 Xb Solution ofa Plane Truss 77 1 Transformation Matrix and Stiffness Matrix for a Bar in Three-Dimensional Space #3 Inclined oF Skewed Supports 88 9 Potential Energy Approach to Derive Bar Element Equations 93 5.10 Comparison of Finite Element Solution to Exact Solution 104 w 66 M1 Galerkin’s Residual Method and Its Application to a One-Dimensional Rar 108 References 111 Problems 112 ‘Symmetry, Bandwidth, and a Computer Program for Truss Analysis 128 Ietroduction 128 41 Use of Symmetry in Structure and Partitioning for Solution of Bguations 128 42 Banded-Symmetric Matrices and Bandwidth 132 £2 Flowchart for the Solution of a Truss Problem by the Direct Saline Method 138 44 Description of a Computer Program for Truss Analysis 139 Retecences 142 Problems 148 5 Development of Beam Equations Contents a Introduction 151, a 5.1 Beam Stiffess 152 a 52 Example of Assemblage of Beam Stilfness Matrices 157 5.3 Examples of Beam Analysis Using the Direct Stiffness Method 139 54 Distributed Loading 169 55 Beam Element with Nodal Hinge 178 56 Potcaial Energy Approach to Derive Beam Element Equations 182 5.7 Galetkin's Method to Derive Beam Element Equations 185 References 187 Problems 187 6 Plane Frame and Grid Equations 195 Introduction 195 6.1 Two-Dimensional Arbitrarily Oriented Beam Element 195 62 Rigid Plane Frame Examples 199 6.3 Inclined or Skewed Supports—Frame Element 216 64 Grid Equations 217 6.5 Beam Element Arbitrarily Oriented in Space 234 66 Concept of Substructure Analysis 239 617 Description of a Computer Program for Plane Frame and Grid Analysis 245 References 252 Problems 252 7 Development of the Plane Stress and Plane Strain Stiffness Equations Introduction 270 iL 7.1 General Steps in the Formulation of the Plane Triangular Element Equations 271 ‘. 7.2 Derivation of the Constant-Strain 1 Matrix and Equations 278viii Contents 73 Treatment of Body and Surface Forces 290 7.4 Explicit Expression for the Constant-Strain Triangle Stiffness Matrix 294 75 Finite Element Solution of a Plane Stress Problem 296 References 305 Problems 306 8 Practical Considerations in Modeli Interpreting Results; and a Computer Program for Plane Stress/Strain Analysi 31 Introduction 311 4 Finite Element Modeling 312 42 Equilibeium and Compatibility of Finite Element Results 322 83 Convergence of Solution 324 4 Interpretation of Stresses 325 85 Static Condensation 326 86 Flowchart of a CST Stiffness Program 330 8.7 Description of a Computer Program for Plane Stress/Strain Analysis 330 References 338 Problems 339 9 Development of the Linear-Strain Triangle Equations 349 Introduetion 349 9.1 Derivation ofthe Linear-Strain Triangular Element Stiffoess Matrix and Equations 349 92 Example LST Stiffness Determination 354 93 Comparison of Elements 356 References 399 Problems 360 10 Axisymmetric Elements 362 Introduction 362 Contents - 10.1 Derivation of the Stiffness Matrix 362 102 Solution of an Axisymmetric Pressure Vessel 372 103 Applications of Axisymmetric Elements 379 References 342 Problems 383 11 \soparametric Formulation m8 Introduction 388 11.1 Isoparametric Formulation ofthe Bar Element Stiffness Matrix 389° 112 Rectangular Plane Stress Element 392 113 Isoparametric Formulation of the Plane Element Stiffness Matrix 395 11.4 Gaussian Quadrature (Numerical Integration) 403 115 Evaluation ofthe Stiffness Matrix by Gaussian Quadrature 406 116 Higher-Order Shape Functions 410 References 414 Problems 414 12 Three-Dimensional Stress Analysis as | Introduction 418 | 121 Three-Dimensional Stress and Strain 418 122 Tetrahedral Element 420 123 Isoparametric Formulation 427 References 431 Problems 432 13 Heat Transfer and Mass Transport 434 Introduction 434 13.1 Derivation of the Basic Differential Equation 435 132 Heat Transfer with Convection 438 133 Typical Units; Thermal Conductivities, K;and Heat-Transier Coafficients,h 439 134 Finite Element Formulation Using « Variational Method 440i“ is 16 Contents “ 5 One-Dimensional Finite Element Formulation 135 One-Di 7 136 Two-Dimensional Finite Element Formulation 13.7 Line or Point Sources 463 138 One-Dimensional Heat Transfer with Mass Transport 466 139 Finite Element Formulation of Heat Transfer with Mass Transport by Galerkin’s Method 467 13.10 Flowchart of a Heat-Transfer Program — 471 an 13.11 Description of a Computer Program for Heat Transfer References 475, Problems 477 Fluid Flow Introduction 485, 141. Derivation of the Basic Differential Equations 486 142 One-Dimensional Finite Element Formulation 490) 143 Two-Dimensional Finite Element Formulation 498 M44 Flowchart ofa Fluid-Flow Program $02 145 Description of a Computer Program for Fluid Flow $03 References $05 Problems $08 Thermal Stress Introduetion $12 151 Formulation of the Thermal Stress Problem and Examples Problems $33 Seroctirel Dynamics and Time-Dependent Introduction 538 16.1 Dynamics of a Spring-Mass System $39, 162 Direct Derivation ofthe Bar Element Equations $40 163 Numerical Integration in Time 544 164 Natural Frequencies ofa One-Dimensional Bar 555 512 512 Comms 165 Time-Dependent One-Dimensional Har Anslysis 8 1646 Beam Element Mass Matrices and Natural Frequencies Si} 167 Teuss and Plane Frame Analysis $67 168 Time-Dependent Heat Transfer $75 References $82 Problems 583 Appendix A Matrix Algebra Introduction $88 ALL Definition of a Matrix $88 ‘A2 Matrix Operations $89 A} Cofactor or Adjoint Method to Determine the Inverse ofa Matrix 595 A4 Inverse of a Matrix by Row Reduction 597 References 598 Problems $99 Appendix B_ Methods for Solution of Simultaneous Linear Equations or Introduction 601 B.1 General Form of the Equations 601 B2 Uniqueness, Nonuniqueness, and Nonexistence of Solution S02 B3 Methods for Solving Linear Algebraic Equations 603 References 614 Problems 614 Appendix C Equations from Elasticity Theory ae Introduction 616 C.1 Differential Equations of Equilibrium — 616 C2 Strain/Displacement and Compatibility Equations 618 3 StresyStrain Relationships 620 Reference 623Appendix D Equivalent Nodal Forces Problems 624 Appendix E Principle of Virtual Work References 630 Answers to Selected Problems 624 627 653 Preface ‘The purpose ofthis second edition of the book is again to provide & simple, basic approach to the finite element method that cam be understood by beth ‘undergraduate and graduate students without the usual prerequmstes (wach a= structural analysis) required by most available texts in this area. The book is ‘written primarily as a basic learning tool for the undergraduate student in civil and mechanical engineering whose main interest sim stress analysis and ‘heat transfer. However, the concepts are presented in sulicietly simple form so that the book serves as a valuable learning aid for students of other backgrounds, as well as for practicing engineers. The text is geared toward ‘those who want to apply the finite element method to solve practical physical problems. ‘General principles are presented for each topic, followed by traditions ‘applications of these principles, which are in turn followed by computer applications where possible. This approach is taken to illustrate concepts used for computer analysis of large-scale problems. ‘The book proceeds from basic to advanced topics and can be suitably used in a two-course sequence. Topics include basic treatments of (1) simple springs and bars, leading to two- and three-dimensional truss analysis: (2) ‘beam bending, leading to plane frame and grid analysis, and space frame analysis; (3) elementary plane stress/strain elements, leading to more ‘advanced plane stress/strain elements, (4) axisymetric stres: (5) ssoparamettic formulation ofthe finite element method: (6) three-dimensional stress (7) heat transfer and fluid mass transport; (8) basic uid mechanics, (@) thermal stress and (10) time-dependent stress and heat transfer. New topics/features include: how to handle inclined or skewed supports, beam element with nodal hinge, beam element arbitrarily located in space, the concept of substructure analysis, a completely new chapter on Mid ‘mechanics, and a diskette including the source codes of six basic programs sed in the text,Preface ‘The direct approach, the principle of minimum potential eeray, nd Gains edt method areinrodueed at various stages a8 required, fo level uations needed for analysis, seve ences incl (1) basic matrix algebra used throughout the est, (2) solution methods for simultancous equations, (3) basic theory of elasticity, ani ‘of vietual work. SE re ad prblens apes throughout te text Most ofthe examples are solved “longhand” to illustrate the concepts; many of them are Solved by digital computer to illustrate the use of the computer programs provided on the diskette enclosed in the back of the book. More than 300 Pid-of-chapter problems, including a number of new ones, are provided to reinforce concepts. Answers to many problems are included in the back of the book. Those end-oF-chapter problems to be solved using a computer program are marked with a computer symbol emptor programs ae incorporated direst relevant pls inthe text to create a natural extension from basi principles to longhand examples land then to computer program examples. The programs are written specii- cally for instructional purposes and source codes written in FORTRAN language are included on the accompanying diskette, Each program solves @ specific clas or type of problem, These programs are easy for students to use. A single lecture is sufficient ro explain how to use most of them. ‘To run any of the six special-purpose programs, you must create a "EXE file of the program. This is done using a FORTRAN compiler program such as the MS-DOS FORTRAN compiler. Following is an outline of suggested topics for a first course (approxi- ‘mately 40 lectures, $0 minutes each) using this textbook. Tope ember of Lecares 1 1 2 : (Chapter 3, Stone 31-210 5 cum! 1 Chapters 4 Chapter 5 Secon 51-56 4 ape Seon 61-63, 67 ‘ Capi? ‘ Chapter A ape 8, Seon 1-138 1810-181 $ sm i 1 use this outline in a one-quarter course for ssocaisinclend nevinsbaleagenntap Ofte Ss deared.Chaper 1 can be tepaced forse cmp Mlinois University: John O. Dow, University of Colorada: Joueph J Renae Worcester Polytech ; . bas iytechnic Institute; and Stephen R. Swanson. University of Tam grateful to Ted Belytschko for his excellent teaching dkenen tho ich men wing tree t the many students who teat. Speci thants to Row Cenfac, Buy Dena, Remmssuse Kariotos, Howard Koswara, Hidajat Harintho, Han Salempinesn. Joe Kesari, Yanping Lu, and Khailan Zhang for checking and solving problems in the text. Thanks also to Peter Soller, David Danner, £4 Koy. and ‘Yanping Lu for “cleaning up” computer programs, and » special thank you to Professor Joseph Rencis and Chili Bao for getting computer program DFRAME running properly. Finally, a very special thank you to my family, Diane, Kath, Daryl Je. sd Paul or the many series ducing te Svlopment of th eed ition. Dery! L Logan—_ = See Notation ENGLISH SYMBOLS ma 8 pane 5 a D E £ f L SI ISS z «, ra (coefficients used to express displacement in ‘cross-sectional area ‘matrix relating strains to nodal displacements or relating temperature gradient to nodal temperatures specific heat of a material ‘matrix relating stresses to nodal displacements direction cosine in two dimensions direction cosines in three dimensions element and structure nodal displacement matrix, both in global coordinates ocal-coordinate clement nodal displacement matrix ‘matrix relating stresses to strains exponential function ‘modulus of elasticity global-coordinate nodal force matrix local-coordinate element nodal force matrix body force matrix heat transfer force matrix heat flux force matrix heat source force matrix surface force matrix slobal-coordinate structure force matrix, ‘condensed force matrix ‘global nodal forcesPoP equivalent force matrix temperature gradient matrix or hydraulic gradient matrix heat-ransfer (or convection) coefficient nodes of a triangular element principal moment of inertia Jacobian matrix spring stiffness wobal-coordinate clement stiffness or conduction matrix condensed stiffness matris, and conduction part ofthe stiffness matrix in heat-transfer problems local-coordinate element stiffness matrix convective part of the stiffness matrix in heat-transfer problems lobal-coordinate structure stiffness matrix thermal conductivities (or permeabilities for fluid mechanics) in the cand y directions, respectively Jength of « bar or beam element ‘maximum difference in node numbers in an clement ‘general moment expression focal mass matrix local nodal moments ‘elobal mass matrix ‘matrix used to relate displacements to generalized coordinates for a linear-steain triangle formulation matrix used (o relate strains to generalized coordinates fora linear- strain triangle formulation bandwidth of a structure ‘number of degrees of freedom per node shape (interpolation or basis) funetion matrix shape functions ‘surface pressure (or nodal heads in fuid mechanics) radial and axial (longitudinal) pressures, respectively concentrated load ‘concentrated local force matrix beat flow (ux) per unit area rate of heat ow beat flow per unit area on a boundary surface ‘heat source generated per unit volume ot internal luid source line oF point heat source radial, circumferential and axial coordinates, respectively residual in Galerkin’ integral Notation body force in the radial direction ‘nodal reactions in x and y directions, respectively natural coordinates stacked osopurametn clement thickness of a plane element ‘nodal temperatures ofa triangular element temperature funetion freesstream temperature displacement, force, and stiffness transformation matrix surface traction matrix in thei direction displacement functions in the x, y, and = directions, respectively strain energy ‘change in stored energy velocity of fuid flow shear force in a beam distributed loading on a beam or along an edge of a plane clement work ‘nodal coordinates in the x, y, and = directions, respectively local element coordinate axes structure global or reference coordinate axes body force matrix body forces in the x and y directions, respectively body force in longitudinal direction (axisymmetric case) or inthe = direction (three-dimensional case) GREEK SYMBOLS Fs Boe TB 6 & 4 coeficient of thermal expansion used to express the shape functions defined by Eq. (72.10) and Egs(1225)-(1228) spring or bar deformation normal teat thermal strain matrix Poisson's ratio. ‘nodal angle of rotation or slope ina beam element total potential energy functional fr heat-transfer problem, ‘mass density of a material weight density of a materialxx Novation o angular velocity a potential energy of forces . fluid head or potential « normal stress on thermal stress matrix r shear stress 6 angle between the x-axis and the local & axis for two-dimensional problems principal angle 4.8% angles between the global x,y, and z axes and the local 2 axis, respectively ¥ ‘general displacement function matrix OTHER SYMBOLS a) dx a time differential o the dot aver a variable denotes that the variable is being differential with respect to time denotes a rectangular oF a square matrix denotes a column matrix the underline of a variable denotes a matrix the hat over a variable denotes that the variable is being described ina local coordinate system 1 denotes the inverse of a matrix ry denotes the transpose of a matrix derivative of a variable with respect to x partial derivative with respect to x partial derivative with respect to each variable in {d} . ‘denotes the end of the solution of an example problem — 2 oe eee Introduction Nt aia mth PROLOGUE ‘The Mite clement metho iss numerical method for sl “ éninecring and mathematical phys Typical problem arent of tera engineering and mathematial physics that are solvable by ee of te Re clemcat method include structural analysis, heat transi fad ow, sane transport and electromagnetic potent For problems involving complicated geometries, oadimg, and materia broperies, itis genealy not possible to obtain anata! mathematical ‘olution. Analytica solutions are those given bya mathomaticalexpemon that yields the values ofthe desired unkown quanites a any locasom mt body thee toa stucture or pial stem of interest) anda than nab for an infinite number of locations inthe body. These aaivical soasons teoerally require the solution of orinaty or partial diferent suatons, which, beause ofthe complicated geometric loadings, and mateal prope tesa not usualy obtainable Hence we ned tore on tumoral mete suchas theft clement method for acceptable solutions Tee Rie clememt formulation of the problem resus in a spsem of smaltancous algeaee equation for solution, ater ha reiting the solution of derctal sea tions. These mumerica methods el approsimate vlc he vows at discrete numbers of pints the continuum. Hence. hs proce of moding a body by dividing tito an equivalent spsem of salir Boies ots Gnte clement) interconnected at points common to two oe more ements (nodal points or nodes) and/or Boundary lines andor saaces i called diseeriaton. nthe nite clement metho, stead of alvin he peblem for the entre body in one operation, oe formulates the equations or each inte element and combines them o obtain the solution of the wha boy. rie, the solution for structural problems typaly rele to deer mining the dplacenents a each node and the strecs within each element ‘aking up the sractore this subjected to appli loads tn onscstura problems, the nodal unknowns may, for instance, be temperatures ot id Pressures du o thermal or uid utes “This chapter ft presets bret history of the development ofthe te element method, You will see from this historical account that the methodVd problems in the past has only become a practical oe for solving engineetnk POT Scere 35 years (paralleling the developments associated Wi us hist ‘account is followed by an intro- ern ial somes Th pe nn for mais etd (ade ractica by the development ofthe moder digital computer) in forms pea for solution is described. This section discusses bot ee tl coat get Spe te at ror ted on te ine eet method. Net ger seein of fe spe aylved to obtain cation 2 problem 5 P Se i ccoatearaat oc ca This desta ation, Various reprentaive applications are pene to atte he apy ofthe meth to sve problems suck ree et sconce ven eet cae Mts uae Fay, Chapter | he sme ofthe aang ofthe ire vais, Fin, CO om ofeagecing a mate physics Brief History This section presents a brit history ofthe finite clement method as applied to both structural and nonstructural areas of engineering and to mathemat- ical physics. Referenes cited here are intended to augment this short intro- duction t the historia background The modern development ofthe finite clement method began in the 1940 iin the field of structural engineering with the work by Hrennikoff [1] in 1941 find McHenry [2] in 1983, who used a latice of line (one-dimensional) elements (bars and beams) for the solution of stresses in continuous solids. In «paper published in 1943, but not widely recognized for many years, Courant [3] proposed setting up the solution of stresses ina variational form, Then he introduced piecewise interpolation (or shape functions over triangular subre- sions making up the whole region asa method o obtain approximate numer- ‘cal solutions. tn 1947, Levy [4] developed the flexibility or force method, and, in 1953, his work (S] suggested that another method (the stiffness oF displacement method) could bea promising alternative for use in analyzing statically redundant aircraft structures However, hisequations were cutmber- some to solve by hand, and thus the method became popular only with the advent ofthe high-speed digital computer. kn 1954 Argytis and Kelsey [6, 7) developed matrix structural analysis ‘methods using energy principle. This development illustrated the important role that energy principles would play in the finite element method. ‘The first treatment of t¥o-gimensional clements was by Turner, Clough, Martin, and Topp [8] in 1956. They detived stiffness matrices for truss clements, beam cements, and two-dimensional triangular and recangulst elements in plane stress and outlined the procedure commonly known as the Martin [12] in 1961, by Gallagh a Melosh 16} 1965. Aaah, Padlg. and Baad (13) in 1962 ant by Argyris (15) in 1964, The special case of axisymmets bo Clough and Rashi 16] and Wison [17 in 1968 ee ost ite element work up to saminda en a However, large deflection and thermal analysis were considered by Taree Dill Martin, and Melosh [18] in 1960 and material nonlinearities by Gallagher, Padlog, and Bijlaard [13] in 1962, whereas ‘were initially treated by Gallagher and Padlog [19] in 1963. Extension of the ‘method to visco-elasticty problems was done by Zienkiewice. Watson. and King [20] in 1968, 4In 1965, Archer [21] considered dynamic analysis in the developmeet of the consistent-mass matrix, which is applicable t0 analysis of distmbutcd- ‘mass systems such as bars and beams in structural analysis. ‘With Melosh’s [14] realization in 1963 that the finite clement method Could be set up in terms of a variational formulation it began to be wsed 89 solve nonstructural applications. Field problems, such as the determination of the torsion of a shaft, uid flow, and heat conduction, were solved Sy Zienkiewicz and Cheung [22] in 1965, Martin (23) in 1968, and Wilson and Nickel [24] in 1966, Further extension ofthe method was made possible by the adaptation of Weighted residual methods, first to derive the previously known elasticity {equations used in structural analysis by Szabo and Lee (25} in 1969 and then for transient field problems by Zienkiewicz and Parekh (26) in 1970. It was. then recognized that when direct formulations and variational formulations are difficult or not possible to use, the method of weighted residuals may at times be appropriate. For example, in 1977 Lyness, Owen, and Zsenkiewice (27) applied the method of weighted residuals to the determination of mage neti field Recently, problems associated with large-displacement noalinear dynam- ‘behavior, and improved numerical techniques forthe solution of the result ing systems of equations have been considered by Belytschko (28,29) in 1976, AA relatively new field of application of the finite element method is that1 Introduction of bioengineering [30, 31). This fel is still troubled by all the difficulties such as nonlinear materials, geometric nonlinearities, and other complexities still being discovered From the eutly 1980s tothe present, enormous advances have been made in the application of the finite element method to solve complicated engin- ccering problems. Engincers, applied mathematicians, and other scientists will undoubtedly continue to develop new applications. For an extensive bibliog- raphy on the finite element method, consult the work of Whiteman [32] or Norrie and de Vries [33]. Introduction to Matrix Notation Matrix methods are a necessary tool used in the finite element method for purposes of simplifying the formulation of the element stiffness equations, for purposes of longhand solutions of various problems, and, most important, for Use in programming the methods for high-speed electronic digital computers Hence, matrix notation represents a simple and easy-to-use notation for writing and solving sets of simultaneous algebraic equations. ‘Appendix A presents a discussion of the significant matrix concepts used throughout the text. We will present here only a brief summary of the nota- tion used in this text. ‘A matrix isa rectangular array of quantities arranged in rows and columns that is often used to aid in expressing and solving a system of algebraic equa- tions. As examples of matrices that will be described in subsequent chapters, the force components (Fix, FiysFies Fas Fays Faas os Fac Foy Fg) acting. at the various nodes of points (1, 2,...,) On a structure and the corresponding ‘set of nodal displacements (dy, diy, dye, aes ayy daar ---> dns ys yg) 0 both be expressed as fa diy A, 4, Ay 4, A 4, - Fy , 4, Wake] h dada dee ay) 4 by f os ‘The subscripts to the right of F and d identify the node and the direction of force or displacement, respectively. For instance, F,, denotes the force at node | applied in the x direction. The matrices in Eqs. (1.2.1) are called column 12 Inteoduetion to Matrix Motation, matrices The brake notation | } wil be vd rouge ‘can mat. The wet fae ig icon array is shmply represented by (F) oF {d}.-A more {hroushout this text to represent any sectangular array i the underlining of (oosibly column mattes or rectangular maar the topes hanes lari the content ofthe dicusionssocuted withthe nea ‘structure stiffness matrices [k) and [J a represented by square matrices given as Kay his Ls pywpe| te bas : ay Ka where, in structural theory, the elements ky and Ky are often relerret to as -suffness influence coefficients, ‘You will leam that the global nodal forces F and the global nodal displacements d are elated through use ofthe global stiffness matrix K by P=ké (12a Equation (1.24) is called the global stiffness equation Its the basi equation formulated in the stiffness or displacement method of analysis. Using the ‘compact notation of underlining the variables, as in Eg, (1.24 showld mot ‘cause you any difficulties in determining which matrices are column oF rec- tangular arrays, Subsequent chapters will discuss the clement stiffness matrices & for various element types, such as bars, beams. and plane stress. They wil also cover the procedure for obtaining the global stiflnes: matnees K for var ‘ous structures and for solving Eq, (1.24) forthe unknown displacements i matrix Using matrix concepts and operations will become rowtine with practi: {ey will be valuable tools for solving small problems loaghand. However, mateix methods are crucial to the use of digital computers-mecexsary for the Solution of complicated problems with thet associated large number of ‘ultaneous equations.13 14 Role of the Computer ma feta i a mrt te sc eg aan though the finite clement me ae tran ti tee tat ce rent cn iar a inmomeo The development of the computer resulted in computation: pe = ‘velopment. Numerous special-purpose and general-purpose Progran ae were rn Yaa ee eth fuer tenet emi an ome finite element method, ‘Tone the computer, the analyst, baving defined the finite element model, Inputs the information into the computer. This formation may include the soe toa the clement nodal coordinates, the manner in which elements are Connected together, the material properties of the elements, the applied loads, boundary conditions, oF constraints, and the kind of analysis to be performed. The computer then uses this information to generate and solve the equations necessary to-carry out the analysis. Gonoral Steps of the Finite Element Method This section presents the general steps inctuded in a finite element method formulation and solution to an engineering problem. We will use these steps as our guide in developing solutions for structural and nonstructural prob- lems in subsequent chapters For simpliciy's sake forthe presentation of the steps to follow, we will consider only the structural problem. The nonstructural heat-transfer and fluid mechanics problems and their analogies to the structural problem are considered in Chapters 13 and 14 Typically, forthe structural stressanalyss problem, the engineer seeks 0 determine displacements and stress throughout the structure, which is in cxuilibrium and is subjected to applied loads. For many structures, itis dificult to determine the distnbution of deformation using conventional methods and thus te finite element method is necessarily used. There are two general approaches associated with the finite clement ietbod. One approach alle he fc, fle, method, we inter fores asthe unknowns of the problem. To obtain the governing equations, fe te muoen te roe, oot ie erin are found by introducing compatibility equations. The result is a set of algebraic equations for determining the redundant of unknown forces. Step 1 14 General Sisps of the Finite Element Method 7 ‘The second approach sums the displacenest of tee ‘eating forces to displacements eae eet ni a unknowns (lors, of come mass cadre eee Faricmore sr phe ow seca ars hey oncana ons displacement formlation for solving struct prob Spree tae Ine Conzqmt, ah apne sah a Ti aie wesLannaiee eel anenee imcronmene! dena ce eter k tei eee soeciacd gina Saar oa a seep scl ny ohare nee ee iol eee c Lae ee re ae been Mawar ee ete eee ee beets were yew opel energ peyetis erapone cesar monet aerate describing the behavior of each node results in a series: of algebrasc equations: wore oes We oeric cia een oan cai seeseteiee cates te eeee GuTipesananie naa peace teeaeees Neieaipoceuiuie sare tnceatamnee manta: Parra aee ete oor cally for springs, bars, trusses, beams, plane frames, plane stress. axisymmetric ee aren te nana aah SEaL neers ate an eee eae eee a ican ie Scans eee Srirettt ad oti sen rs ee oehie sees SS eee Se ee ee ee Discretize and Select Element Types Step 1 involves dividing the body into an equivalent system of inite elements, With associated nodes and choosing the most appropriate element type. The {otal number of elements used and their variation in size and type within a1 tmrodution nt The elements must ters ofengincening nough to reduce sven body are primarily matters Tr cuts and yet large enoug! ‘made small enough to give ust ‘higher-order elements) are : canon mon em marge elements can be used where results fines in ater chapter particularly wr enen erated with mesh sn gnc The creed body oF me soften rai) Wl Eaton por peoneor ogame sTUANG US 5, The choo cements nd 8 ile ee a pial makeup ofthe oy gant the results ob. Hudementconcera- tothe actu ea cme, two, or tree-imensional ealizations is ing the appropriateness of oe, Ho OT open clement for a partic acide ar protiem sone of the aor as hat must ecu Oo PY Oe Soe edn hs ent—are shown in Figure Il ote primary tine elements. Figure 1-1(a consist of har (or truss) and ee area but are usally represented tan eens Thay ve onscoma eeak l ee Segments tn general the erost-ectinal area within the ars be Sondre o be constant throughout this text. These “Snot aren ed to mode rss nd rae suctures oe Figure 1-2 on pa 1 for stance) The simplest ie cent (alle nea lane) tus two nodes one a each nd although higher-order elements having three nodes or more (called quadratic, cube, ele. element) also exist. The line Clemeats are the simplest of elements to consider and will be discussed in Chapters 2,3 5 and 6 to llistate many ofthe basi concepts ofthe inte element method. "The bsic to dimensional (or lane) elements, Figure 1-1(6), ae loaded ty forces in their own plane (plane stressor plane strain conditions). They are trangular or quadrilateral elements. The simplest two-dimensional elements have comer node only near elements with straight sides or boundaries (6 Chapter 7, atbough there ate also higher-order ements, typically ith midside nodes (called quadratic elements) and curved sides (see Chapters 9 and 1) The elements ean have variable thicknesses throughout or Be 6 stant. They arc ften used to mode! a wide range ofengincering problems (se© Figures 1-3 and {on page 10, The most common thee- Introduction othe Stiffness (Displacement) Method Fy = (20090005) = —10KN Fe, = a0) ~ 2004001) = 0 f= (2090005) + 400/001) — 20N}0015)=0 (2534) ; 20U)(0.,01) + (400(0.015) — (200)(0.02) ~ 0 2100015) + 20002) = 1OKN tof the nodal result of gp. (254) yl the reaction F., opposite that of Pe Se nace nde 3 by 8200 mm. Ths es vriex ulibgum ofthe whole spring sembage i (d) Next, we make use of local clement Eq, (2.2.17) to obtain the forces im each element. Fa ia Element 1 Hal [0 2 lfawest (2335) a5” | =200 200] }o.00s} Simplifying Eq, (2.535) yields f= -10KN fy, = LOK (25.36) Element 2 {ela [oee cae eal (2537) Simplifying Eq (2.537) yields fu=-1KN f= kN (2538) Element 3 Jaw 200-200] foor te ~|-200 200} lors AR Simplifying Eq. (2.539, we have f= -1KN f= tk (2840) Element 4 fs) _[ 200 ~200) (ors ti}-[-ao “malfeee} asa Simplifying Eq. (2.841), we obtain Sag= VEN fay KN (2542) EXAMPLE 23 28 Boundary Conditions 5 Finally to review the major concepts presente in hi chap, we sve the following example problem. ro > 4 (0) Using the ideas presented in Seton 23 forthe syste far sine sorng shown in Figure 11, expe the boundary cones, te coma bay or continuity condition similar oy (239) ad these ogee condions similar to Bs (234)(28) Then formulate the gos stn Ia and ution slain othe unknown goa opacones ex. The ping constants he cemens ately eta, Por force at node 2. — (b) Using the diet stifaese method, formulate the same bal ifs imatrt and eauation asin Par a. a \ veliecs ae SAH ye nate EEN FIGURE 2-41 Spring assemblage for solution {@) The boundary conditions ure =O d= 0 dyad ion at node is ayy = ag ‘The nodal equilibrium conditions are A= ft Pm f+ Fy, = 2 Fay = fo The compatibility con 12! = da, 2844 where the sign convention for positive element noxtal forces given by Figure 2-1 was used in writing Eqs. (2.548), Figure 2-12 shows the element aod nodal {force free-body diagrams. Using the local stiffness matrix Eq, (2.2.17) applied to each element, and ‘compatibility condition Eg, (2.544), we obtain the total or global equilibriumFIGURE 2.12 _Foe-bogy dagams cf elanans and ode of sing amamiege of Figure 2-7 ‘equations as Fist hide —jhadae - aa —Hadae Pm hydatid + idan — Bada tal ie Poem haat had Fag = —Kadag | bydae tn mates form, we expres Eqs. (25.46) as Fis hy ak 0 oO dye P| _|-h kthth —h by) [de ih thtk he eit wif] 0 che ees Walt ateme fa) Lo * Sige = “oeiitael| te ‘Therefor, the lob stiffness mati isthe square, symmetric matrix on the ‘right side of Eq, (2.5.47), Making use of the boundary conditions, Eqs. (2.5.43), and then considering the second equation of Eqs. (2546 or (2.547), we solve fords, a 2 ot kat hy We could have obtained this same result by deleting rows 1, 3, and 4 in the F and d matrices and rows and columns 1, 3, and 4 in K, corresponding to zero: displacement, as previously described in Section 2.4, and then solving for d,,. Using Eqs (2546), we now solve forthe plobal forces as Fem —hidie Fare —hadae Fae hyde (2549) ‘The forces given by Eqs. (2.549) cam be interpreted as the global reactions in this example. The negative signs in front of these forces indicate they are directed to the lft (opposite the x axis) (b) Using the direc stiffness method, we formulate the global stiffness ‘matrix. First, using Eq. (2.2.18), we express each clement stiffness matrix as dy. de dn de dag dae ef HK] ye f tb) yf es x [2 t] € [2 “| w-[p tl (2550) Ps (2548) 2.6 2.0 Potential Energy Avproneh to Derive Spring Hemant tauetione AP here the deyrees of freedom associated with each temens are Sed im the columns above each matrix. Using the dlvectsifiness method as ovtimed i Section 2:4, we add terms from exch element stfiness matin nto te appre Driate corresponding row and column inthe gobal wifes marin to oman a a ae 4, hy o 6 why ket hy + ky hy es are “by ly 0 i eh ‘We observe that each element stiffness matrix & has been added seo the location in the global K corresponding to the identical degree of Soedomn associated with the element & For instance, element 3 amociaied wah degrees of freedom d;, and dy; hence, ts contributions to K wee i the 2-2 2-4,4-2,and 4-4 locations of K as indicated in Bq. (2S. 51)y the seem Having assembled the global K by the direct stiffeess method. the giobal equations are then formulated in the usual manner by making wae of the general Eq, (23.10), F = Kd These equations have been previowsly cheated by Eq. (25.47) and therefore are not repeated . Potential Energy Approach to Derive Spring Element Equations (One of the alternative methods oftcn used to derive the clement equation and the stiffness matrix for an clement is based on the principe of minimum rotential energy. (The use of this principle in structural mechanic & Sally described in Reference [4]) This method has the advantage of beng more general than the method given in Section 22 savoiving nodal and ckemest equilibrium equations long with the sres strain law foc the element. Then, the principle of minimum potential energy ss more adaptable for the deter rmination of clement equations for complicated clements (howe wth large numbers of degrees of freedom) such as the plane stressstram element ‘Again, we state that the principle of virtual work (sce Appendix E) applicable for any material behavior, whereas the principle of ‘minima potential energy is applicable only for elastic materials. However. Noe pew ‘ples yield the same element equations for linear-slastic materials, whch are the only kind considered in this text. Moreover, the principle of minimum Potential energy. being included inthe general category of sarsatomal methods (as is the principle of virtual work), leads to other variational fanctiows (or functionals) similar to potential energy that can tbe formelated for other lasses of problems, primarily of the nonstructural type. These other pros lems are generally classified as field problems and include, among others,12 berduction othe Stiiess ( Diplacement) Method torsion ofa bat, heat transer (sce Chapter 13) fui ow (Chapter 14), and clecric potential Toe asses of problems, for which a variational formulation isnot clearly definable, can be formulated by weighted residual methods, of which qaitLinsmethod isthe most often used. We will describe Galerkin’ method oases ication to a har element in Section 3.11, (For more information SrA ted residual methods, alo conslt References [5], [6]. and (73) se recat the pancple of minimum potential cDergy 28 used to erie th bat cement equations. We will illustrate this seemingly compli- cere concep complicated possibly because ofthe lack of physical insight aa ee ath this approach) applied to the simplest of elements in hope that ‘tie become a more comfortable concept when applied out of necesity to hove complicated element types in subsequent chapters. ‘The total potential energy x, of a structure is expressed in terms of displacements. tothe ite element formulation, these will generally be nodal foplacemente such that y= yl, dyna) When is minimized with eanoct to these displacements, equilibrium equations reault For the spring ement, we will show that the same nodal equilibrium equations hi ~ f fest as previously derived in Section 22. ‘We fit state the principle of minimum potential energy as follows: Ofailthe displacements that satisfy the given boundary conditions ofa structure, {hove that satay the equations of equilibrium are distinguishable bya stationary ‘lve ofthe potential energy Ifthe saionary valve i a minimum, the equilib ‘um state stable ‘To explain this principle, we must frst explain the concepts of potential energy and of a stationary value ofa function. We will now discuss these two concepts. "Total potential energy is defined as the sum of the internal strain energy U and the potential energy of the external forces 9 that i, 1,2U+0 261) Strnin energy is the capacity of internal forces (or stresses) to do work through ‘deformations (strains) inthe structure; isthe capacity of frees such as body forces, surface traction forces, and applied nodal forces to do work through deformation of the structure. Recall that a linear spring has force related to deformation by F = kx, sch eng conta i a a 13) ‘The diflerential internal work (or strain energy) dU in the spring is the {internal force multiplied by the change in displacement through which the force moves, given by dU = Fdx 262) Now we express F as Fok (263) Using Eq. (26.3) in Eq, (262, the differential strain energy becomes aU = kxds (264) YS FIGURE 249. Foce-delometioncuve for Sree aeng FIGURE 2.16 Staionay vie of afuncion “The total strain energy is then given by ua [eae 2a Upon explicit integration of Fg (265), we oain vate (ass Using Bq. (262) in Bq, 266), we have U = Vader aa Equation (26.7) indicates that the strain energy is the area under the forse deformation curve, “The potential energy of the external force, being opposite in sgn from the external work expression because the potential energy of the external force lost when the work is done by the external force, i given by Q=—Fx easy ‘Therefore substituting Eqs. (2.66) and (268) into (26.1), the total potential energy becomes ny = phx? = Be 269) ‘The concept of a stationary value of a function G (used it the definition (of the principle of minimum potential energy) is shown in Figure 2-14 Here G is expressed as a function of the variable x. The stationary value can be a2 Intredection to the Stifes (Displacement) Method maximum, a minimum, of « neutral point of G(x) To finda value of x yielding, sn atv of sh we wse erential calcu to diferente G with respect to x and set the expression equal to 2er0, as follows: ee 26.1 eo (26.10) ‘An analogous process will subsequently be used to replace G with x, and x with dere vals (nodal displacements. With an understanding of ‘ariational calculus (see Reference (8), the first variation of x, (denoted by {,) could be used to minimize z,. However, we will avoid the details of ‘variational calculus and show that we can really use the familiar differential Calculus to perform the minimization of x. To apply the principle of mini- ‘mum potential energy —that i, to minimize x,—we take the variation of t, ‘defined in general as Oe gh ap ae + pgtbha t+ tbe (2641) “The principle states that equilibrium exits when the d, define a structure state ‘such that dx, = Ofor arbitrary admissible variations 6 from the equilibrium state, An admissible variation is one in which the displacement field still satisfies the boundary conditions and interelement continuity. Figure 2-15 shows the hypothetical actual axial displacement and an admissible one for a bar with specified boundary displacements 2, and dy. Here 6@ represents the variation ind In the general finite element formulation, 6a would be replaced bby ddj, This implies that any of the 5d, might be nonzero. Hence, to satisfy ‘x, = 0, all coefficients associated with the dd, must be zero independently. Thos, on, on, a Wah 2dooM) or Spe0 (2612) where m equations must be solved for the n values of d, that define the static ‘equilibrium state of the structure. Equation (2.6.12) shows that for our pur- ‘poses throughout this text, we can interpret the eariation of n, as a compact, Adm placement fantom, + 8 FIGURE 2-15 Actual and admissible displacement functions 24 Potential toeray ABprosch 16 Derive Boring Hhement tauenions == OY ‘notation equivalent o differentiation of, with reepect so pore ABT 1-115, a. oe as reior[_| ~j] asa The asvembled plobalsifiess matrix is 4 2x} —1 2-1 (394) ° 1 ‘The assembled global equations are then 1-1 0) fa. 1500 2x10} —1 2 1/44; ={ 9000 (3945) o 1 1flacaol ay 750 where the boundary condition ds, = 0 has been substituted into Eq, (3.945) Now solving Eqs. | and 2 of Eg. (19.45), we obtain dy, = =0.006 in cama ,, = ~0.00825 in. ‘The element stresses are as follows:NOE Development of Tras Ewrions _ 1] f 4, = -0.006 -e{ 30,30] las, ~~ 0.00525} a6 (3947) 2 — 11] dss ae «=4[-3 =| = 5250 psi (T) Coy 4.10 Comparison of Finite Element Solution to Exact Solution ‘We will now compare the finite element solutions for Example 3.10 using one, ‘wo. our and eight elements to model the bar element and the exact solution, The exact solution for displacement is obtained by solving the following equation: att aE sere, using the following free-body digram, we have 1100» x o: -!ooominoo%) +4000 =0 For other kinds of elements (other than beams}, this adjustment is ignored in practice The adjustment is less important for plane and solid elements than for beams. Also, these adjustments are more difficult to formulate for an sdement of general shape. Gelerkin’s Residual Method and Its Application ‘to # One-Dimensional Bar General Formulation ‘We have developed the bar finite element equations by the direct method in Section 3.1 and by the potential energy method (one of a number of varia ‘somal method) i Section 39. In Gelds other than structural/solid mechanics, son gone probable that a variational principle, analogous tothe principle of smememem potential energy, for instance, may not be Known or even exist. In 211 Gastar Reta men | teamed ns per ne a met sana tt eat eet residuals app dey To te ie equation can be wed to develop teint oman egos Ine ne ‘edie Galt esal metd pn sep cknent This development pve he ses pons ae Kis method tote beam conn Cape Sante nose pensar cent Geil he omen cob conde ton, cometon ad mas notes dc Chap On tothe mar penne hears fou mt roe (cern wile tobe he Ga meted cesar applied to develop the finite clement equations. i“ - hte ate number fer el thts Among tesco ton ast squares and eas gure collin For moron ie mos sce Rett []) However se Gastint method more wel Knows Ta weed eda thd lor approve uncon cose to aprons he lndpete e sed iecmmstorsmper Satins prob et ys etal euaon Tal con il notin nr sty th govern itera equation Ths the sr ton ofthe tl fnton ins he seta gut relia eae the whol repo the problem a flows: jffnor mien any segs Oe fice Decne wom cn ifm : seepoation function, such as Using Gaterki’s method, we choos the interpola fg. (21 rin terms ofN shape factions forthe independent variable inthe lifcrenel equation, In general, this substitution yields the residual R #0. By the Gelesin cterin, the shape functions N are chosea to play the role of the weighting functions W Thus, foreach we have Equation (111.3) resus a total of equations. Equation (11.3) applies to. raat i ein ofa bd) without ference o Poundary condos aa appt Yoad o placements. To obtain Doundary cond such as pci ration by parts t0 E.G 3h which ye integrals Spica fr the region and its boundary Gunz)Ulm 2 Development of Tras Exwatons ‘er Erement Formulation now etate Galerkin’ method 0 Form SSL Weep wt base iret eat TS tnd Seton 31 i a)= Gila, (40%) -0 5, # ua Ris now defined to were constants Aand fate now assumed, The resi DEST eg Gacrins ction, 0.11.3, 1 Ea. 11.4, we have late the bar element stiffness jon, without distributed ¥d (gp ; ie-0 (i= 1,2) Gus) (a(eg)se ‘ We now apply integration by parts to Eq, (3.11.5), Integration by parts is piven in general by ela ans) where wand o ate simply variables in the general equation. Letting wan dun Nas a 117) & ‘ Bg (411.5) and integrating by parts according to Eq, (3.11.6), Eq. (3.11.5) becomes a da’ da (setae vane da\\* ("| da dN, ™ - Tap dt = (waeit)f [taettt¥aeeo ana) ‘where the integration by parts introduces the boundary conditions. Recall that, because d = [NV] {d}, we have Fo ta a, ug) irae ear dt af 4-[- 84 sau ‘Using Eq. (3.11.10) in Eq. 3.11.8), we then express Eq, (3.11.8) as “aN 1 1], fa a af alt tft} ~ (sacs) Equation (3.11.11) & really two equations (one for Nj= N, and one for N= N,} Fe, wing the weighting lncion Nc Mee hand @=12) GaLID m ae (“Mt 1 fa Ne sates Hotel eT [=a -t He Betan, cosy where fe EA(AQ/d8) because any Evaluating Eq. 1113) yids | | "0 8M =O at xm AE. Ten~ deed hie Gans) Similarly, using N, = Ny, we obtain “Crypt ty ga, ayy AE E he) CUE Deli -(oeg an Simplifying Eq. (3.11.15) yields AE Tn where fy, = EA(di/d) because N= Lat x= Land N= Oat x= 0. Equa- tions 3.11.14) and (3.11.16 are then sen to be the same as Eqs. (3113) and (3927) derived, respectively, by the iret and variational methods 2) = Sas (4116) REFERENCES [1] Turner, M. J, Clough, RW, Martin, H.C, and Topp, LJ, “Stiivess and Defletion Analysis of Complex Structures" Journal of the Aeronantial Sciences, Vol. 23, No.9, Sept. 1986, pp. 805-824 [2] Martin, H.C, “Plane Elasticity Problems and the Direct Stiffness Method,” The Trend in Engineering, Vol. 13, Jan. 1961, pp. 5-19. [3] Melosh, RJ, “Bass for Derivation of Matrices fr the Direct Stifbess Method.” Journal of the American Insitute of Aeronautics and Astronatics, Vol. 1, No.7, Tuly 1963, pp. 1631-1637 [4] Oden, J. T, and Ripperger, E. A, Mechanics of Blaste Structures, 2nd ed, [5] Finlayson, B. A, The Method of Weighted Residuals and Variational Principles, ‘Academic Press, New York, 1972 [6] Zienkiewice, O.C, The Finite Element Method, Srd od, McGraw-Hill, London, 197, (7] Cook, R. D, Malkus, B.S. and Plesha, M. E, Concepts and Applications of Fine Element Analysis, 31d ed, Wiley, New York, 1989. 8) Forray, M. Jy Variational Calculus i Science and Engineering, McGraw-Hill, New York, 1968.Um 12> Dermat TEs PROBLEMS snese matin K ofthe assemblage shown in Figure Copa cng the ins matics ofthe individual ars. Note Fa ee fy An dy Ev Eo yb La and La Hee 4 I sould be in er Symbols sed for erosesectonal art, modulus of Seny, and eth, respecte RI Compote the total st 1 FIGURE P31 A Now let Ay = As = Ay = A, Ey = Ba = If nodes 1 and 4 are fixed and a force P acts isection. find expressions for the displacement of nodes 2 and 3 in terms of 4, EL, and P. «Now lt A= 1 in, E= 10 x 10° psi, L = 10 in, and P = 1000 1b. 4. Determine the numerical values of the displacements of nodes 2 and 3, ii. Determine the numerical values of the reactions at nodes 1 and 4 4d Determine the stresses in elements 1, 2, and 3. 42-311 Forthe bar assemblages shown in Figures P3-2-P3-1!, determine the nodal displacements, the forces in each element, and the reactions. Use the direct stiffness method for these problems. Bw 2, 3,4 10N £= 21060 Te Te Azax io 20a — 0 foe +0 Ficune rss @ 2 wm 190m 5 < == O48 e- 0x ws in —+-—$ 9 Lon A= som! FIGURE P34 ae ee FIGURE P35 IGURE P37 set 1GURE Ps. 1@ win 2 FIGURE Ps. FIGURE P3.10 yon. wn — nS : el Te Bai 6 We 70Gr Peo Sw Nwan a3 se as = 2106P. FIGURE PII Seivefor hail dplacement and srs inthe tapered ba shown in Figure Pe St ane and then two constancarea clement. Evalate the aren at Teco fcach clement length: Use that area foreach clement. Let Ay = 2 PESTS ins = 10 10" psi and P= 1000 Tb. Compare your fie iemetvltons wt th exact solution. FIGURE P3.12 Determine the stiffness matrix for the bar element with end nodes and mid Jength node shown in Figure P3-13. Let axial displacement u = a, + 43x + yx", (This is a higher-order element in that strain now varies linearly ‘rough the element) =a FIGURE P3143 ‘Gives the following displacement function forthe two-noded bar element + bx? 1s thes valid displacement function? Discuss why or why not For cach of the bar clements shown in Figure P3-15, evaluate the global x-) linens matecn B= 30 x 10 pa £ = 21007 tase FIGURE Pats For the bar elements shown in Figure P3.16, the global displacements have been determined to be d;, = 0.5 in, d;, = 00, dy, = 025 in. and dy, ~ 0.75 in, Determine the local $ displacements at each end of the bars. Let E = 12 x 10° psi, A = OS in, and L = 60 in for each element ay ty FIGURE P3168 Ie316 Dewkapmen of Tra Exweions AA? Forth ar ements shown i Fig 217th loa placentae Ee ay vined to = 00, d,,= 25mm, ds, mm, and ds, = 3.9 bee determined 10 ef virions theendsofeach bat. Let B= 21 | BR eo toes mand L = 3m foreach lement % | 2 rieine Fer N, ss ‘Wg he td ein 35 dtm ial tn Maureen eee peta re P avec 30) Tor ieee rs seiko gs > ene dag ‘ment of node 2 using the stiffness method. Also determine the stress in element 1. Let d= $in3, B= 1x 108 ps, and £ = 100 in s voKip < 1 a: 2 fr Hom, t-te 4 = 100mm 4-00 FIGURE P3.20 FIGURE P32 = = 3.21 Find the horizontal and vertical displacements of node | forthe truss shown | in Figure P3-21, Assume AB is the same for each element.meee 5 Deven of Tram Eewerions 1p3.22 solve for the horizontal and verticg ‘node 1 and determine the stress in each jode 1. Allelements have A, = Lint ia Figure 22 For the truss shown components of displacement at ‘Gement Also verify force equilibrium at n Sad E = 10 = 10° psi. Let = 100in. FIGURE P3.22 3.23 For the truss shown in Figure P3-23, solve for the horizontal and vertical ‘components of displacement at node 1. Also determine the stress in element L Let A= Lim}, E = 100 x 10* psi, and L = 100 in, 1000 32 P= 1000 +n FIGURE P3.26 424 Determine the nodal displacements and the element forces for the truss shown ‘= Figure P3-24. Assume all elements have the same AE. 428 Now remove the element connecting nodes 2 and 4 in Figure P3-24, Thea ‘éescrmive the nodal displacements and element forces. FIGURE Poza ne 4126 Now remove both cros element nodal displacement? Ifnot wy?" iM P3-24 Can you determine the 427 Determine the dpccmen cong posers at the plane truss shown in Fiute P32) Lar pe and the clement forces for for all elements. Verity force equilib = 3in? and E30 «10° rium at node 3. ? FIGURE paz7 3.28 Show that for the transformation matrix T of Eq, (3415), T7 = T~"and hence, Eq, (34.21) is indeed correct, thus also illustrating that k = T"ETis the ‘expression forthe global stiffess matrix for an element 4.29-3.30 For the plane trusses shown in Figures P3-29 and P3-30, determine the horizontal and vertical displacements of node ! and the stresses in each clement. All elements have E = 210 GPa and A = 40 x 10°* m* own aww FIGURE P3.29 FIGURE P3220 431 Remove element 1 from Figure P3-30 and resolve the problem. Compare the displacements and streses to the results for Problem 3.30.plane . ine the nodal displace. shown in Figure P3-32, determine As eaten foes andsteer and th spport eations Allelemens SESE DO Gpa and 4 = 30 x 10-* m*, Verify force equilibrium at nodes Pad 4 Use symmetry in your model FIGURE P3.32 For the plane truss supported by the spring at node 1 in Figure P3-33, determine the nodal displacements and the stresses in euch element. Let E = 210GPa and A = 5.0 x 10°* m? for both truss elements. 333 FIGURE 3.33 a at FIGURE P3.38 444 For the plane truss shown in Figure P3-34, node 2 settles an amount 8 = 0.05 ‘= Determine the forces and stresses in each element due to this settlement. Let E = 30 x 10° pai and A= 2 in? for each element. To 7 335 For the symmetic plane truss FIGURE 3.25 336-3.37 | For the space truss elements shown in Figures P3-36 and P3.37, the global displacements at node 1 have been determined to be d,, = 0.1 ity dy, = 02 in, and d,, = 0.15 in. Determine the displacement along the local £ axis at ‘node 1 of the elements. The coordinates, in inches, are shown in the figures. J, 1,4 FIGURE P3.36 FIGURE P37————————rt— iC 1225 Devotee of Trams Eqnations sae pace truss elements shown in Figures P3-38 and P3-39, the global ae eres at node 2 have been determined 10 be ds, = 5 mm, d3,— Seinen and d, = 15 mm. Determine the displacement along the local axis sowie Sos ihe cements. The coordinates, in meters, are shown in the figures, 0.0.0 «2. - FIGURE p34 meee: cia uner ree 342 For the space truss subjected to 1000-5 load in the x direction, as shown in Figure P3-42, determine the displacement of node $. Also determine the stresses in each element. Let A = 4in? and E = 30 x 10* ps forall elements. ‘The coordinates of each node, in inches, are shown inthe figure. Nodes 1,2 4.40-3.41 For the space trusses shown in Figures P3-40 and P3-41, determine the ‘nodal displacements and the stresses in each element. Let E = 210 GPa and See ero maps 2) ek cocoa ae ent ee A= 10 x 10°* m? for all elements, Verify force equilibrium at node 1. The coordinates ofeach node, in meters, are shown in the figure. All supports are ball-and-socket joints FIGURE P3-42REF Development of Truss Equaions Probleme 0.18 4 | 1235 ‘0008 p- x : ha FIGURE P347 FIGURE Paae 3.49 Use the principle of minimum potential energy develope in Section 38 to solve the bar problems shown in Figure P349, that i plot the total potential ‘encray for variations in the displacement of the free end ofthe bat to dete. mine the minimum potential energy. Observe that the displacement that yields the minimum potential energy also yields the stable equilibrium pos me Seis tion. Use displacement increments of 002 in, beginning with x = 006. Let E = 30 x 10° psi and A = 2 in? forthe bars 443 For the space truss subjected to the 4000 Ib load acting as shown in Figure 3.43, determine the displacement of node 4. Also determine the stresses in cach clement. Let A= 6 in? and E = 30 x 10° psi for all elements. The 200% coordinates ofeach node in inches, are shown inthe figure. Nodes 1, 2, and | re supported by ball-and-socket joints (fixed supports) 344. Veily £4. (379) fork by fst expanding T°, given by Ea, 377,108 6x 6 wm $—————— nan ‘square matrix in a manner similar to that done in Section 3.4 for the two- 2 Cry r dimensional case. Then expand £ to a 6 « 6 matrix by adding appropriate rows and columns of zeros (forthe d, terms) to Eq. (24.17), and finally, x perform the matrix triple product & = TT e : 345 Derive Eq. (3.721) for stress in space truss elements by a process similar to that used to derive Eq. (3.5.6) for stress in a plane truss element, PLS Ee Pee 346-348 For the plane trusses with inclined supports shown in Figures P3-46-3-48, solve for the nodal displacements and element stresses in the bars. Let A= 2 a9 in?, E = 30 x 10° pei and L = 30 in. for each truss, Derive the stiffness matrix for the nonprismatic bar shown in Figure P3-S0 Using the principle of minimum potential energy. Let £ be constant. FIGURE 2-46 Aa) = Ay + Aol FIGURE P3-50ee ‘equal-length elements and (b) four equal-length elements. Let A = 2in.? ang = 30 x 10° psi Compare the finite clement solution with an exact solution, 555 FIGURE P3.s1 382 For the bar subjected to the uniform line load in the axial direction shown in. Figure P3-52, determine the nodal displacements and axial stress distribution using (a) two equallength elements and (b) four equal-length elements. Com- ppare the finite element results with an exact solution. Let A= 2 in? and 30 x 10° psi 1, = 100 tn Floune pase FIGURE P3-53 4.53 For the bar fixed at both ends and subjected to the uniformly distributed foading shown in Figure P3-53, determine the displacement at the middle of ‘the bar and the stress in the bar. Let A = 2 in? and E = 30 x 10° psi. is For the bar hanging under its own weight shown in Figure P3-S4, determine the nodal displacements using (a) two equal-length elements and (b) four ‘equal-iength elements. Let A = 2 in, £ = 30 x 10° psi, and weight density 2. = 0283 Ibjin. (Hint: The internal force is a function of x. Use the potential energy approach) Ulli, FIGURE pase Probleme ro Determine the energy equivalent shown acting on the bat element odal forces forthe ial distributed loadin sin Figure P3-55, _— FIGURE P3.5541 GWA P THER 4 Symmetry, Bandwidth, anda Computer Program for Truss Analysis an INTRODUCTION lin this chapter, we wll first introduce the use of symmetry to reduce the size t's problem, and thus facilitate its solution. We will use an example truss problem to illustrate the concept. We will solve the resulting set of stiffness ‘equations by a process of partitioning the matrix equations. ‘Next, we will introduce the concepts of banded-symmetric matrices and bandwidth because we will implement these concepts in a computer program 10 analyze trusses. Because the longhand solution of large problems is not feasible, we conclude this chapter with a description of a computer program implementa~ ‘on of the matrix sifness method. We will give a general computer flowchart for solution of both two- and three-dimensional truss structures and discuss the manner in which data is input into the computer program listed on the accompanying disk Use of Symmetry in Structure and Partitioning for Solution of Equations ln many instances, we can use symmetry to facilitate the solution of a prob Jem. Symmetry means correspondence in size, shape, and position of loads: smaterial properties; and boundary conditions that are on opposite sides of ® riding line or plane. Use of symmetry allows us to consider a red problem instead of the actual problem. Thus, the order of the total stifinss ‘airs and total set of stiffness equations can be reduced. Longhand solutio® EXAMPLE 41 ee 41 Une of Symmetry in Structure and Pattoning 129 time is then reduced and com substantially decreased," “lution ime for large-scale problems is Solve the plane truss problem sh ‘eight elements and five modes node 4. Nodes 1 and 5 ure pin stiffnesses of /2AE and bars again, A and E represent the a bar. own in Figure 41. The tras i composed of a shown, A vertical loud of 2 applied a sepports. Bar ements 12.7. and ¥ hve 3,4,5, and 6 have axial slfnes of AE Here, &os-etional area and modulus oes of In this problem, we will use a plane of symmetry. The vertical pla verpendculir tothe plat ts psig through de tt te plane of symmetry because identical geometry, material, loading. and bound. ary conditions occur atthe corresponding locations on opposite sides of this plane. For loads such as 2P, occurring in the plane of symmetry, one-half of the total load must be applied to the reduced tructure For elssenty mesure fing in the plane of symmetry, one-half of the cross-sectional area must be used in the reduced structure, Furthermore, for nodes in the plane of symme- try, the displacement components normal to the plane of symmetry must be set to zero in the reduced structure; that is, we set dy, =O, dy, =0, and dg, = 0, Figure 4-2 shows the reduced structure to be used to analyze the plane truss of Figure 4-1 ‘We begin the solution of the problem by determining the angles @ for ‘each bar element. For instance, for clement I, assuming 2 1 be directed {tom node 1 to node 2, we obtain = 45°. Table 41 is used in determining each clement stiffness matrix. ‘There are a total of eight nodal components of displacement for the truss before boundary constraints are imposed. Therefore, K must be of order 8 x 8, For element 1, using Eq, (34.23) along with Table 41 for the direction FIGURE 42 Trussot FIGURE 41 Figure 4.1 reduced by symmetrym0 Semmes, emt, end a Competer Program for Trass Analysis ats for the tse of Figure 4-2 4 " 14 i 4 (4a i i ) = 4 Sioa, for elements 2. 3,4, and 5, we obtain de dy dae doy poee|-4 $b (412) Vus-$ 4 4-4 die dy das day i oot pr tt] 0 0 (413) |i 00 “ ° ponte] 0 4 (14) “Elo 0 ° o = i tie dyy ° 00 cael aware} (41s) “T/o oo o o-b0 4 wher, Eqs (4.11)-(4.1.5) the column labels indicate the degrees of freedom ‘eocauted with each clement. Also, because elements 4 and 5 lie in the plane eymmetry. one hail of their original areas have been used in Eqs. (4.1.4) and sembly sulting set of stiffness equations is 7 ae] 9 Aft] _ fe H{et=1 ob carey ry Slay” Le) Solution by Partitioning (Condensing) of Matrices We will solve Eq. (4.1.6) by separating the matrices into submatrices by drawing dashed horizontal and vertical lies as shown in Eq. (41) The Purpose ofthis partitioning is to expedite further calculations into which the ‘matrix enters. We will use partitioning here to make one part of a matrix all zeros. (In Section 2.5, we used partitioning to separate unknown displace- ‘ments from known ones in solving for d.) It is most appropriate to begin with the general st of stiffness equations partitioned as follows: Ky Ka These equations can then be applied to the specific example, Eq. (4.16, Equation (41.7) can be written as separate matrix equations as follows 417) (41s) (419) Solving Eq (4:18) for dy, we obtain dy =-Kib (41.10) Using Eq, (4.1.10) in Eq, (4.19) yields Ka(-Kiik E (aun Simplifying Eq, (4.1.11), we obtain a-K. (4412) Def (413) we can express Eq (4.1.12) a5 %o. ai kas Multiplying both sides of Ea. (4.1.14) by (&.)"' we obtain