Mar 102

MA*2003*Z*Z*Z*SM-MA102-NT1 1
MSC.Marc Advanced Course

2 MSC.Marc Advanced Course
Copyright 2003 MSC.Software Corporation
All rights reserved. Printed in U.S.A.
Corporate Europe
MSC.Software Corporation MSC.Software GmbH
2 MacArthur Place Am Moosfeld
Santa Ana, CA 92707 81829 Mnchen, GERMANY
Telephone: (714) 540-8900 Telephone: (49) (89) 431 987 0
Fax: (714) 784-4056 Fax: (49) (89) 436 1716
Asia Pacific Worldwide Web
MSC Japan Ltd. www.mscsoftware.com
Entsuji-Gadelius Building
2-39, Akasaka 5-chome
Minato-ku, Tokyo 107-0052, JAPAN
Telephone: (81) (3) 3505 0266
Fax: (81) (3) 3505 0914
Part Number: MA*2003*Z*Z*Z*SM-MA102-NT1
Disclaimer
THE CONCEPTS, METHODS, AND EXAMPLES PRESENTED IN THE DOCUMENTATION ARE FOR ILLUSTRATIVE AND
EDUCATIONAL PURPOSES ONLY, AND ARE NOT INTENDED TO BE EXHAUSTIVE OR TO APPLY TO ANY PARTICULAR
ENGINEERING PROBLEM OR DESIGN. USER ASSUMES ALL RISKS AND LIABILITY FOR RESULTS OBTAINED BY THE USE OF THE
COMPUTER PROGRAMS DESCRIBED HEREIN. IN NO EVENT SHALL MSC.SOFTWARE CORPORATION BE LIABLE TO ANYONE FOR
ANY SPECIAL, COLLATERAL, INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES ARISING OUT OF, RESULTING FROM, OR
IN CONNECTION WITH USE OF THE CONTENTS OR INFORMATION IN THE DOCUMENTATION.
MSC.SOFTWARE CORPORATION ASSUMES NO LIABILITY OR RESPONSIBILITY FOR ANY ERRORS THAT MAY APPEAR IN THE
DOCUMENTATION. THE DOCUMENTATION IS PROVIDED ON AN AS-IS BASIS AND ALL EXPRESS AND IMPLIED CONDITIONS,
REPRESENTATIONS AND WARRANTIES, INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY OR FITNESS FOR A
PARTICULAR PURPOSE, ARE DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD TO BE LEGALLY
INVALID.
MSC.SOFTWARE CORPORATION RESERVES THE RIGHT TO MAKE CHANGES IN SPECIFICATIONS AND OTHER INFORMATION
CONTAINED IN THE DOCUMENTATION WITHOUT PRIOR NOTICE.
Trademarks
MSC, Dytran, MARC, and Patran are registered trademarks of MSC.Software Corporation or its subsidiaries in the United States and/or other
countries. MSC., MSC.Dytran, MSC.Marc, and MSC.Patran are trademarks of MSC.Software Corporation.
NASTRAN is a registered trademark of the National Aeronautics and Space Administration. MSC.Nastran is an enhanced proprietary version
developed and maintained by MSC.Software Corporation. All other trademarks are the property of their respective owners.
Third Party Software Program Credits
METIS is copyrighted by the regents of the University of Minnesota.
NT-MPICH is developed by Lehrstuhl fr Betriebssysteme der RWTH Aachen. Copyright 1992-2003 Lehrstuhl fr
Betriebssysteme der RWTH Aachen.
Government Use
Use, duplication, or disclosure by the U.S. Government is subject to restrictions as set forth in FAR 12.212 (Commercial Computer Software) and
DFARS 227.7202 (Commercial Computer Software and Commercial Computer Software Documentation), as applicable.
MSC.Marc Advanced Course 3
Contents
MSC.Marc Advanced Course ....................................................... 1
Table of Contents ............................................................................. 3
CHAPTER 1 Introduction to MSC.Marc .................................... 9
About MSC.Marc 10
MSC.Marc Solutions Aerospace 11
MSC.Marc Solutions Automotive 13
MSC.Marc Solutions Manufacturing 14
MSC.Marc Solutions Biomed 15
MSC.Marc Solutions Electronics 16
Course Objective 17
Documentation 18
MSC.Marc/Mentat File Road Map 19
CHAPTER 2 Review of The Finite Element Method ................ 21
Introduction to the Finite ELEMENT Method 22
The Choice of the Element Type(s) 31
Nonlinear FEA and Iterative Solution 34
Direct Incremental-iterative Solution Procedures 37
Iterative Solution Procedures 39
Convergence Checking 40
Equation Solvers 44
CHAPTER 3 Geometrically Nonlinear Analysis ....................... 45
Updating the Geometry of a 2 Bar Truss 46
Linear Strain-displacement relation 47
Nonlinear Strain-displacement Relation 48
Follower Forces 50
Load Incrementation 51
Bucking Analyses 54
Table of Contents
Contents
CHAPTER 4 Nonlinear Material Behavior ............................... 57
Types of Nonlinear Material Behavior 58
What about Linear Materials? How is E Measured? 59
Plasticity: Constitutive Relation, Uniaxial Tensile Test 60
Every Integration Point Tracks the Constitutive Relation 62
Rigid-Plastic Flow 67
Creep 69
Viscoplasticity 73
Metal Forming Methods 74
Neo-Hookean Material Model 75
2-Constant Mooney Extensional Deformation 82
Other Mooney-Rivlin Models 84
Ogden Models 86
Foam Models 87
General Formulation of Elastomers 88
Finite Element Formulation 91
Large Strain Viscoelasticity 93
Illustration of Large Strain Viscoelastic Behavior 98
CHAPTER 5 Contact Analysis.................................................. 101
Definition of Contact Bodies 102
Control of Rigid Bodies 104
Contact Procedure 105
Bias Factor 106
Contact Flowchart 109
Symmetry Body 110
Rigid with Heat Transfer 111
Contact Table 113
Exclude Segments during Contact Detection 116
Contents
Friction Model Types 119
Stick-Slip Friction Model 120
Coulomb (Sliding) Friction Model 121
Shear (Sliding) Friction Model 122
Glued Contact 124
Release Option 126
Forces on Rigid Bodies 127
CHAPTER 6 Heat Transfer and Thermal Stresses.................. 129
Mathmetical Formulation 130
Boundary conditions 132
Analogy between Heat Transfer and Stress Analysis 134
Spatial Finite Element Discretization 135
Linear Heat Transfer Analysis 136
Nonlinear Heat Transfer Analysis 137
Control Values Heat Transfer Analysis 139
Initial Time Step Estimate 140
Surface to Surface Radiation 141
Laminar Flow and Heat Transfer 147
Coupled Thermo-mechanical Analysis 154
Heat Generated from Mechanical Sources 155
Coefficient of Thermal Expansion 156
CHAPTER 7 Dynamic Analysis................................................ 157
Dynamic Analysis Methods 158
One Degree of Freedom System 159
Multiple Degree of Freedom System 164
Free Vibration Analysis 165
Modal Superposition Method 166
Harmonic Analysis 167
Contents
Transient Analysis 168
CHAPTER 8 User Subroutines................................................. 169
When to use Subroutines or Special Routines 170
Classification of User Subroutines 171
Programing Conventions 172
Geometry Modifications Subroutines 173
Output Quantities Subroutines 174
Loads and Boundary Conditions Subroutines 174
Constitutive Relations Subroutines 177
Viscoplasticity Subroutines 179
Viscoelasticity Subroutines 179
Hydrodynamic Lubrication Subroutines 180
Special Routine PLDUMP 180
ELMVAR Utility Routine 181
Submit MSC.Marc with User Subroutines 183
MSC.MARC User Subroutine Example 184
CHAPTER 9 Automatic Mesh Refinement .............................. 193
Adaptive Meshing 194
Example: Clamped Cantilever Beam 195
Error Criteria for Linear Analyses 198
Error Estimates Mentioned in Literature 199
Error Criteria Available in MSC.Marc 199
Strain Energy Criterion 201
Zienkiewicz-Zhu (Stress) 202
Zienkiewicz-Zhu (Strain Energy) 206
Equivalent Stress/Strain 208
Location within Box 209
Solution Gradient (Heat Transfer Analysis) 210
Contents
User Criterion 211
Remeshing Strategy for Continuum Elements 212
Possible Subdivisions for Local Mesh Refinement 213
Multipoint Constraint Equations 214
For a Linear Adaptive Analysis 217
Important MSC.Marc Mentat Options 218
Improved Geometry Description 219
Attaching Nodes to Geometrical Entities 221
Example: Cylinder-cylinder Intersection 223
Modified Remeshing Strategy for Shell Elements 225
General 3D Situation 227
Nonlinear Adaptive Analyses 228
Error Criteria for Nonlinear Analyses 229
Node in Contact 230
Equivalent Plastic Strain 231
Rezoning Required 232
Use of MSC.Marc for Nonlinear Adaptive Analyses 235
CHAPTER 10 Workshop Problems ............................................ 237
Some MSC.Marc Mentat Hints and Shortcuts 238
Mechanics of Mentat 240
Comprehensive Sample Session 248
Statics S1 Tensile Specimen 297
Statics S2 Cantilever Beam 316
Statics S3 Elastomeric Curve Fit and Seal 328
Statics S4 Break Forming 345
Statics S5 Super Plastic Forming 357
Statics S6 Creep of a Tube 377
Statics S7 Interference Fit 390
Statics S8 Hertz Contact Problem 397
Contents
Statics S9 Elastomeric Arch 403
Heat Transfer H1 Transient Cooling Fin 409
Heat Transfer H2 Steady State Cooling Fin 415
Heat Transfer H3 Coupled Transient Cooling Fin 418
Heat Transfer H4 Dynamics with Friction Heating 423
Heat Transfer H5 Radiation with Viewfactors 432
Dynamics D1 Cantilever Beam Modal Analysis 438
Dynamics D2 Cantilever Beam Harmonic Analysis 441
Dynamics D3 Cantilever Beam Transient Analysis 444
CHAPTER 11 Notes and Course Critique.................................. 453
Class Notes 454
Class Critique 459
CHAPTER 1 Introduction to MSC.Marc
This course offers complete workshop-paced
training based upon real world problems like
those shown in this first chapter.
Attendees are involved with both the
MSC.Marc and MSC.Marc Mentat programs
by lectures and hands on modeling of
nonlinear mechanics problems.
Tell Me and Ill Forget
Show Me and Ill Remember
Involve Me and Ill Understand
Introduction to MSC.Marc About MSC.Marc
About MSC.Marc
MSC.Software Corporation, the world leader in nonlinear and
coupled physics simulation, introduces you to MSC.Marc, an
integrated member of the MSC.VisualNastran Enterprise family.
MSC.Marc is a general-purpose finite element program for advanced
engineering analysis.
Since 1971, it has been known for its versatility in helping market
leaders in various industries, design better products and solve simple
to complex, real-world engineering applications. MSC.Marc allows the
user to perform a wide variety of structural, fluid, and coupled
analyses using the finite element method. These procedures provide
solutions for simple to complex linear and nonlinear engineering
problems. Analysts can graphically access all features via the
MSC.Marc Mentat or the MSC.Patran interfaces. Also included in
MSC.Marc is the parallel processing of large problems using Domain
Decomposition.
Today, MSC.Marc has evolved and matured through constant
development and worldwide use by thousands of engineers. Please
read on to discover how its many capabilities will enable you to solve
the most challenging nonlinear and coupled physics problems. For
more information please visit our web site at www.marc.com.
MSC.Marc Solutions Aerospace Introduction to MSC.Marc
MSC.Marc Solutions Aerospace
Simulation of the stretch forming of
aircraft panels. As a result of these
simulations, the panels can be formed
properly the first time and their final shape
matches the design without reworking.
Introduction to MSC.Marc MSC.Marc Solutions Aerospace
MSC.Marc Solutions Aerospace
Super Plastic Forming (SPF), is a
process that forms shapes through the
application of high temperature and
controlled pressure to certain kinds
of materials.
MSC.Marc Solutions Automotive Introduction to MSC.Marc
MSC.Marc Solutions Automotive
This simulation illustrates the very large
displacements due to the contact
between the shaft and the boot.
Self-contact also occurs between the
bellows of the boot.
Introduction to MSC.Marc MSC.Marc Solutions Manufacturing
MSC.Marc Solutions Manufacturing
Here, a set of rollers deforms a flat sheet
into a channel shape. This simulation
shows half of the environment, which, in
the end, is used to form a U-shaped
channel.
MSC.Marc Solutions Biomed Introduction to MSC.Marc
MSC.Marc Solutions Biomed
This biomedical pump is composed of a
polymer material, and the milking
motion was simulated using defined
positions versus time of each of the
rigid fingers.
Introduction to MSC.Marc MSC.Marc Solutions Electronics
MSC.Marc Solutions Electronics
Uniform heating of the paper traveling
in a printer is an important quality
issue, since thermal gradients will cause
distortion in the text as the ink drys on
the paper.
Course Objective Introduction to MSC.Marc
Course Objective
50% Short overview of background theory
50% MSC.Marc program options with special emphasis on
using MSC.Marc Mentat with hands on modeling of
engineering problems:
* Linear elastic analysis
* Geometrically nonlinear analysis
* Material nonlinear analysis
* Contact analysis
* Dynamic analysis
* Heat transfer and thermal stress analysis
MSC.Marc QA Procedures
Always referring to standard version
MSC.Marc office distributes list with
known program errors and possible
work-arounds
Customer may get fix on request
Introduction to MSC.Marc Documentation
Documentation
MSC.Marc:
Volume A: User Information
Volume B: Element Library
Volume C: Program Input
Volume D: User Subroutines
see ~marc/user
(all user sub templates)
Volume E: Demonstration Problems
see ~marc/demo for over 450 demos
MSC.Marc Mentat:
New Features
Users Guide
Training Notes
On-line Help
MSC.Marc/Mentat File Road Map Introduction to MSC.Marc
MSC.Marc/Mentat File Road Map
INPUT SOLVER
OUTPUT
MARC
MENTAT
model1.mud
file=model1_job1
file.dat run_marc -j file -u sub file.log
file.out
file.t08
file.cnt
sub.f
file.t09
file.t19
file.t16
Submit Post_Open
Reset/Kill
Monitor_Job
Introduction to MSC.Marc MSC.Marc/Mentat File Road Map
CHAPTER 2 Review of The Finite Element
Method
The purpose of this chapter is to review
the finite element method for better
understanding of how any finite
element program works. Special
attention is given to the subject of
solving nonlinear problems.
Review of The Finite Element Method Introduction to the Finite ELEMENT Method
Introduction to the Finite ELEMENT Method
The finite element method has the following six steps. The success of
this method is due in part to the simplicity of these steps, and how
finite element programs such as MSC.Marc implement these steps.
Step 1 Shape Functions: The FEM expresses the unknown
field, , in terms of the nodal point unknowns, , by using
the shape functions, , over the domain of the element, ,
as,
Step 2 Material Loop: The FEM expresses the dependent flux
fields, , such as the strain (stress) or heat flux in terms of
the nodal point unknowns as,
Step 3 Element Matrices: The FEM equilibrates each element
with its environment which can be expressed as,
where the element matrices, and have consistently
lumped all physical significance of the element at its nodes,
u x ( ) a
N x ( )
e
u x ( ) N x ( )a
e
=
x ( )
x ( ) L u x ( ) [ ] Ba = = ( ) D x ( ) = = ;
K
e
a
e
f
e
0 = +
K
e
f
e
Introduction to the Finite ELEMENT Method Review of The Finite Element Method
where:
are the element matrices. represents physical properties
such as stiffness or conductivity, and , represents loads
experienced by the element, . These loads may be: body
loads, , such as weight or internal heat generation in volume
; surface loads, , such as pressure or convection on surface
; or concentrated loads, .
Step 4 Assembly: The FEM assembles all elements to form a
complete structure in such a manner to equilibrate the structure
with its environment which requires,
and to insure continuity of nodal point unknowns, .
Therefore, a finite number of system equations results as,
K
e
B
T
DB
V d
=
f
e
N x ( )
T
b V d
N x ( )
T
t S d
F + + =
K
e
f
e
e
b
e
t
e
F
K K
e
f f
e
e
= ;
e
=
a a
e
=
Ka f 0 = +
Step 5 Solve Equations: The FEM specifies the boundary
conditions, namely, the nodal point values on the boundary and
the system equations are partitioned as,
where: are the unknown nodal values; are the specified
nodal values; are the applied nodal loads; are the nodal
point reactions. Hence the solution becomes,
Step 6 Recover: The FEM recovers the stresses by substituting
the unknown nodal values found in Step 5 back into Step 2 to
find the dependent flux fields such as strain, stress and
heat flux.
K
uu
K
us
K
su
K
ss
a
u
a
s
f
a
f
r
=
a
u
a
s
f
a
f
r
a
u
K
uu
1
= f
a
K
us
a
s
+ ( )
f
r
K
su
a
u
K
ss
a
s
+ ( ) =
Notice the cpu costs of each of the steps from a typical run.
s t a r t o f i n c r e m e n t 1
start of assembly cycle number is 0 time = 249.72
start of matrix solution time = 326.78
end of matrix solution time = 372.77
......
start of assembly cycle number is 12 time = 2301.75
start of matrix solution time = 2380.45
end of matrix solution time = 2426.46
e n d o f i n c r e m e n t 1 time = 2504.60
The last recycle took 80% of the total time in Assembly and
Recovery, with the remainder in the Solver. Furthermore, this
increment took 12 recycles for a total time of 2500 seconds. A
linear analysis would only take 372 seconds! Assembly and
Recovery times have to do with performing the integrations
shown in Step 3. Reducing the amount of integration points
will greatly improve performance.
What about CPU time (cost)?
Step Time points (sec)
Time
(sec)
% of
Total
Assembly 2380.45 - 2301.75 78.70 39
Solve 2426.46 - 2380.45 46.01 22
Recover 2504.60 - 2426.46 78.14 39
Total 2504.60 - 2301.75 202.85 100
Linear and Nonlinear FEA differ greatly in the amount of time
spent on each of the three solution steps.
In linear FEA, step 5, Solve Equations dominates (or has in the
past dominated) the overall cpu time.
In nonlinear FEA, the costs of each step are more equal. With
recent advances in solver technology, the time spent in
assembly and recovery now exceed that spent in the solver.
CPU time used in %
Step Linear Analysis Nonlinear Analysis
Assembly 15 30
Solve 70 30
Recover 15 30
Total 100 100
The finite element method can convert any differential
equation to a set of algebraic equations. It assumes the shape of
the solution in the domain of the element and satisfies
equilibrium. For instance, in heat transfer let, ,
then the system equations become:
where becomes the capacitance matrix, and , becomes the
conductivity matrix. In dynamics, let , and the
system equations become:
The time integration of the above system equations is usually
done by the finite difference method, generating completely
algebraic equations that can are solved by computers.
Efficient algebraic equation solvers become important. If these
six steps have any nonlinearities, the resulting set of algebraic
equations are also nonlinear and iterative solution procedures
must be used. Before considering these iterative solution pro-
cedures and equation solvers, lets examine these six steps
more closely.
b C
p
u
+ =
Ma
Ka f 0 = + +
M K
b u
+ =
Ma
Ka f 0 = + +
Lets choose a specific element
type and use the six steps above to
solve a simple problem. Choosing
the shape function in step 1
determines the type of element to
be used. For simplicity, a bar
element is chosen.
Step 1: The shape function:
Step 2: The strain and stress:
Step 3: The element matrices:
With the element chosen, lets solve a simple problem.
u
1
u
2
x 0 = x l =
x
u x ( )
x
u x ( ) 1
x
l
--
x
l
--
u
1
u
2
=
x ( )
1
l
---
1
l
---
u
1
u
2
= E = ;
K
e
B
T
DB
V
AE
l
-------
\ .
| |
1 1
1 1
= d
=
The problem to solve is a bar
subjected to its own self
weight. The bar is hung from
a support, we want to know
the deformations and the
stresses. Lets start with a
one element solution.
Step 4 and 5: Assemble the elements and solve.
Step 6: Recover Stresses
The stress is constant over the element and is 50% of the
maximum value. The displacement at node 2 is exact.
x 0 =
x L =
E A , ,
u
1
u
2
AE
L
-------
\ .
| |
1 1
1 1
0
u
2
r
1
AL
2
-------------
=
u
2
AL
2E
----------
\ .
| |
= r
1
AL
2
-------------
= ;
e L
2
------
=
Lets look carefully at the solution by plotting the
displacements and stress versus the axial position in the bar.
The FEM approximation to the stress is a piecewise constant
representation to a line, and the displacement approximation is
a piecewise linear representation to a parabola. If a parabolic
shape function would have been selected, the finite element
solution would have agree exactly with the theoretical solution.
Knowing the element type is very important.
The Choice of the Element Type(s) Review of The Finite Element Method
The Choice of the Element Type(s)
For any element chosen, element matrices like:
must be integrated efficiently. A popular choice is to use
Gaussian Quadrature. Where the integration the domain of the
element is transformed to a unit cube, square, or line. Hence:
where is the determinant of the Jacobian operator, and
then the element matrices are numerically integrated by
Gaussian Quadrature as:
then,
K
e
B
T
DB
V d
=
V d dxdydz det J ddd = =
det J J
------
J

x
-----
=
K
e
B
T
DB
V
F
ddd
F
B
T
DB det J ( ) = ;
1
1
1
1
1
1
= d
=
Fddd
1
1
1
1
1
1
k
F
i

j

k
, , ( )
k
Review of The Finite Element Method The Choice of the Element Type(s)
where the Gaussian points (called integration or Gauss points),
, and the weights, , are found to make the
integration of polynomials exact. For example, to integrate a
polynomial of order 2n-1 exactly, we have the following
2n equations:
These simultaneous equations can be solved explicitly in terms
of Legendre polynomials. The table below shows the Gauss
points and weights for n=1,3:
MSC.Marc reports the element variables at the integration
points in the output and post files. MSC.Marc Mentat will
extrapolate the post file results to the nodes and nodal average
the element variables. Reduced integration elements are
elements specifically designed to use fewer integration points,
hence great savings in assembly and recovery. Though the
reduced elements are economic, they are not advised for large
amounts of plasticity. If you start using them to get quick
results, remember to switch back to full integration elements
on the last run.
Gauss points/weights
n
1 0 2.0000
2 .57735 1.0000
3 .77459
.00000
.55555
.88888
i

j

k
, , ( )
i

j

k
, , ( )
P x ( ) x d
1
1

i
P x
i
( )
i 1 =
n
=
x
i
The Choice of the Element Type(s) Review of The Finite Element Method
Accurate results from the finite ELEMENT method depend on
the choice of the element, how many are used and where the
nodes are located.
Remember the ELEMENT is the middle name of the finite
ELEMENT Method. There are over 150 types to choose from
in MSC.Marc. Virtually all of the elements in MSC.Marc are
available for nonlinear analysis.
MSC.Marc Mentat assists you in picking elements based upon
the analysis type you are performing and the topology of the
elements used in meshing.
Element types fall into three basic categories:
Continuum Elements are: bricks, plane stress, plane
strain, generalized plane strain, and axisymmetric.
Shell Elements are: beams, plates, shells, trusses, and
shells of revolution.
Special Elements are: gaps, pipe bend, shear panel,
semi-infinite, incompressible, and reduced integration.
Mixing different element types is easy if the number of degrees
of freedom is the same for all types. Otherwise, constraints
must be used to insure proper compatibility.
Review of The Finite Element Method Nonlinear FEA and Iterative Solution
Nonlinear FEA and Iterative Solution
For the solution step, we must solve the equations:
For the linear case, Gauss elimination is applied directly.
However, for nonlinear equations both the stiffness and
external forces may be functions of the nodal displacements,
Newton-Raphson Method
To solve such a nonlinear set of equations we apply the
Newton-Raphson method. This is an iterative method. Given a
general nonlinear equation, , and a known point ,
we calculate a correction, as follows:
with
K [ ] a { } F { } = or I F 0 =
I a ( ) F a ( ) 0 =
f a ( ) 0 = a
i
a
i 1 +
a
i 1 +
f a
i
( )
f a
i
( )
-------------
=
a
i 1 +
a
i
a
i 1 +
+ =
Nonlinear FEA and Iterative Solution Review of The Finite Element Method
by defining the Tangent Stiffness as,
and the Residual as,
we can rewrite the N-R equation in a more familiar form:
And use Gauss elimination techniques (or something similar)
to solve this set of equations for .
With each iteration, the residual should decrease. If it does, the
method converges to the correct solution.
If the extent of nonlinearity is too great (loads are too large),
the method may diverge, or simply not converge. For this
reason, we load the structure gradually or incrementally.
f a
i
( ) K
i
T
a
i
( )
u
I a
i
( ) F a
i
( ) ( ) =
f a
i
( ) R a
i
( ) I a
i
( ) F a
i
( ) =
K
i
T
a
i
( ) a
i 1 +
R a
i
( ) =
a
i 1 +
Review of The Finite Element Method Nonlinear FEA and Iterative Solution
The following simple example is shown for a single
nonlinear equation:
where,
0 0.25 0.753 0.969 0.777 1.027 0.144
1 1.027 0.144 0.517 0.278 1.305 0.035
2 1.305 0.035 0.263 0.133 1.438 0.009
3 1.438 0.009 0.132 0.068 1.506 0.002
f x ( ) x ( ) sin 1 0 =
f x ( ) x ( ) cos =
x
0
0.25 = and u
i 1 +
f a
i
( )
f a
i
( )
-------------
=
i
x
i
f
i
f
i
x
i 1 +
x
i 1 +
Residual
Direct Incremental-iterative Solution Procedures Review of The Finite Element Method
Direct Incremental-iterative Solution Procedures
Full Newton-Raphson (default)
Assembly and decomposition of stiffness matrix during
every iteration or recycle
Good convergence
Expensive for large systems
u
2
Residual
u
1
Review of The Finite Element Method Direct Incremental-iterative Solution Procedures
Modified Newton-Raphson:
Assembly and decomposition of stiffness matrix only at
start of increment
Slow convergence behavior
Effective for mildly nonlinear problems without
material nonlinearities
u
2
Residual
u
1
Iterative Solution Procedures Review of The Finite Element Method
Iterative Solution Procedures
To contrast direct and iterative solvers, lets suppose the system
equation is, f(x) = sqrt(x) -1 = 0. The direct solver would
require the construction of the inverse of the stiffness matrix.
An iterative approach would not. A Direct Solver must get an
inverse, for a newton method we have,
where as the iterative solver does not. Here since sqrt(x)=1,
then the equation can be rewritten as x=sqrt(x) and hence,
The convergence of the two are plotted below:
x
i 1 +
x
i f x
i
( )
f' x
i
( )
------------
=
Inverse
x
i 1 +
sqrt x
i
( ) =
No Inverse
Review of The Finite Element Method Convergence Checking
Convergence Checking
Termination of iterative procedure when convergence ratio
is less than (default ). Types
include:
Residual checking, possible in one cycle:
Relative: and/or
Absolute: and/or
= maximum residual force
= maximum reaction force
Displacement checking, not possible in one cycle:
Relative: and/or
Absolute: and/or
with: , , , etc.
= maximum displacement change
= maximum displacement increment
Tol
1
Tol
1
Tol
2
0.1 = =
F
residual
max
F
reaction
max
----------------------------------
Tol
1
<
M
residual
max
M
reaction
max
-----------------------------------
Tol
2
<
F
residual
max
Tol
1
< M
residual
max
Tol
2
<
F
residual
max
F
reaction
max
u
max
u d
max
-------------------
Tol
1
<

max
d
max
-------------------
Tol
2
<
u
max
Tol
1
<
max
Tol
2
<
u
1
d u
1
= u
1
u
1
= u
2
u
2
u
1
=
u
max
du
max
Convergence Checking Review of The Finite Element Method
Residual Checking, the Residuals and Reactions
Note that the residuals are the difference between the external
forces and the internal forces at each node, namely:
The nodal reactions are from the system equations, namely:
The maximum residuals and reactions occur at different
degrees-of freedom that have the largest magnitude, namely:
and
Finally, residual convergence checking can be switched from
relative to absolute. When this feature is selected, the switch
will occur when the Minimum Reaction Cutoff (user specified)
is reached. When the switch occurs, the Maximum Residual
will be less that the Maximum Absolute Residual that is
specified. Both the reactions and residuals are placed on the
post file as a single vector for postprocessing.
F
residual
F
external
B
T
V d
=
F
reaction
f
r
K
su
a
u
K
ss
a
s
+ ( ) = =
F
residual
max
Max
F
i
residual
( ) i 1 maxdof , = ; =
F
reaction
max
Max F
i
reaction
( ) i 1 maxdof , = ; =
Review of The Finite Element Method Convergence Checking
Displacement Checking, the Maximum Displacement
Change and Maximum Displacement Increment
More that one recycle is necessary because the maximum displacement
change is the maximum displacement increment on the first recycle.
For relative displacement checking, the convergence ratio becomes:
Graphically this is shown as:
conv
ratio
u
max
u d
max
-------------------
u
i 1 +
u
i
max
u
i
max
-------------------------------------------
= =
Residual
u
i
u
i 1 +
u
i

Convergence Checking Review of The Finite Element Method
Convergence Ratio versus Recycle Number:
There are those cases where convergence does not happen,
and things can get interesting like the Mandelbrot set.
Review of The Finite Element Method Equation Solvers
Equation Solvers
Equation solvers are used during the solution phase as
shown in step 5. Below is a list of current MSC.Marc
solvers. the multifrontal sparse solver is the default.
CHAPTER 3 Geometrically Nonlinear Analysis
The purpose of this chapter is to focus
on the nonlinearities associated with the
geometry of large displacements that
can arise because the original and
deformed geometry is no longer
assumed to be the same.
Geometrically Nonlinear Analysis Updating the Geometry of a 2 Bar Truss
Updating the Geometry of a 2 Bar Truss
Lets examine the two bar truss below, and update the deformation as
the truss deforms because of the load applied.
We see that the displacement
under the load P depends upon
the current configuration.
Iterating for the final
displacement, d, for various
applied loads shows the affect
of updating the geometry.
Note that accounting for large
displacement shows a stiffer
structure.
2a
a
f
P
d
U
L
AE
-------
\ .
| |
f
2
=
d
P
U PL
AE
-------
g ( ) = =
d
a
-- -
g ( ) ;
1 1 + ( )
2
+ [ ]
2 1 + ( )
2
--------------------------------- -
= =
f
L
Linear Strain-displacement relation Geometrically Nonlinear Analysis
Linear Strain-displacement relation
Small rigid body rotation:
,
hence:
Arbitrary rigid body rotation:
,
hence:
L
u
1
u
2
v
1
v
2
1 2
u
x
-----
1
L
------
0
1
L
---
0
u
1
v
1
u
2
v
2
u
2
L
-----
= = =
u
2
0 v
2
L
0
u
2
L cos 1 ( ) = v
2
L sin =
cos 1 0 =
Geometrically Nonlinear Analysis Nonlinear Strain-displacement Relation
Nonlinear Strain-displacement Relation
2D truss element (Green-Lagrange strain definition):
with:
and
Arbitrary rigid body rotation:
Geometrically nonlinear strain-displacement relation 3D
continuum:
..........
u
x
-----
1
2
---
u
x
-----
\ .
| |
2
1
2
---
v
x
-----
\ .
| |
2
+ + =
u
x
-----
u
2
L
-----
=
v
x
-----
v
2
L
-----
=
cos 1
1
2
---
cos 1 ( )
2 1
2
---
sin ( )
2
+ + 0 = =
xx
u
x
-----
1
2
---
u
x
-----
\ .
| |
2
1
2
---
v
x
-----
\ .
| |
2
1
2
---
w
x
------
\ .
| |
2
+ + + =
xy
u
y
-----
v
x
-----
u
x
-----
u
y
-----
v
x
-----
v
y
-----
w
x
------
w
y
------
+ + + + =
Nonlinear Strain-displacement Relation Geometrically Nonlinear Analysis
Green-Lagrange Strain Definition is Activated using the
LARGE DISP Parameter Option
The stress tensor corresponding to the Green-Lagrange
strain tensor is the Second Piola-Kirchhoff stress tensor:
Geometrically linear and nonlinear response 2D
truss assembly:
Parameter option
Stress tensor Strain tensor
large disp
Green-Lagrange
2nd Piola-
Engineering Engineering
Kirchhoff
F
v
geometrically
linear
geometrically
nonlinear
F
v
Geometrically Nonlinear Analysis Follower Forces
Follower Forces
Distributed loads are taken into account by means of
equivalent nodal loads; changes in direction and area can
be taken into account using the parameter option
FOLLOW FOR
Total Lagrange versus Updated Lagrange
Updated Lagrange is especially useful for beam and shell
structures with large rotations and for large strain plasticity
problems. Updated Lagrange is activated using the UPDATE
parameter option
0
1
2
3
0
1
2
3
Total
Updated
Lagrange
Lagrange
Load Incrementation Geometrically Nonlinear Analysis
Load Incrementation
F
u
F
u
F
u
constant or
variable
load
increments
variable
load
increments
snap-through
snap-back
Geometrically Nonlinear Analysis Load Incrementation
Constant Load Increments
with:
: incremental load during increment
: total load to be applied
: total number of increments
Variable load increments (equilibrium):
with:
: factor depending on equilibrium during increment
; ,
Initial load increment must be defined by user
F
i F
N
inc
----------
=
F
i
i
F
N
inc
F
i
F
i 1
=
i 1 i 2 0 >
Load Incrementation Geometrically Nonlinear Analysis
Variable Load Increments (Arc Length Method):
with:
: factor following from requirement that the
length of the incremental displacement vector
(arc length) remains constant during one step;
may be positive or negative
Initial load increment must be defined by user
The arc length may be changed based on number of
iterations (increase if necessary number is smaller
than desired number, decrease if necessary number is
larger than desired number)
Be careful with initially nearly linear response
(limit maximum arc length multiplier):
F
i
F =
F
u
Geometrically Nonlinear Analysis Bucking Analyses
Bucking Analyses
Linear pre-buckling state:
The buckling load is given by:
where is the smallest value following from the
eigenvalue problem:
with based on the stresses corresponding to
Perform buckling analysis immediately after
increment 0
u
F
F
cr
F
1
F
u
e.g. buckling of
Euler column
F
cr
F
cr
1
F
1
=
1
det K
L
K
+ ( ) 0 =
K
F
1
Bucking Analyses Geometrically Nonlinear Analysis
Nonlinear Pre-buckling State:
The buckling loads are given by:
, etc.
where is the smallest value following from an
eigenvalue problem similar to the linear buckling
analysis, but with the complete tangent matrix
taken into account
Perform buckling analysis after increment 1, etc.
Usually, a complete incremental analysis is
preferable
F
1
cr
F
2
cr
u
F
F
cr
F
1
F
2
F
1
cr
1
F
1
=
F
2
cr
F
1

2
F
2
F
1
( ) + =
2
K
L
K
NL
K
+ +
Geometrically Nonlinear Analysis Bucking Analyses
CHAPTER 4 Nonlinear Material Behavior
The purpose of this chapter is to focus on
those nonlinearities caused by the
material behavior.
Nonlinear Material Behavior Types of Nonlinear Material Behavior
Types of Nonlinear Material Behavior
Plasticity
Creep
Nonlinear elasticity
(incompressible Mooney)
Nonlinear elasticity
(incompressible/compressible)
Visco-elasticity
Visco-plasticity
Damage
Cracking
Etc.
Introductory
course
Special topic
courses
What about Linear Materials? How is E Measured? Nonlinear Material Behavior
What about Linear Materials? How is E Measured?
Tension/Compression
Torsion
Bending
Wave Speed
Do you expect all of these Es to be the same for
the same material?
E
P A
L ( ) L
-------------------
\ .
| |
=
E 2 1 + ( )
Tc J
-------------
\ .
| |
=
E
PL
3
3I
---------
=
E c
2
=
T ,
P ,
P L ,
Nonlinear Material Behavior Plasticity: Constitutive Relation, Uniaxial Tensile Test
Plasticity: Constitutive Relation, Uniaxial Tensile Test
Uniaxial data is transferred into a multi-axial stress
situation using a yield criterion: , is the
true (Cauchy) stress tensor e.g. von Mises yield criterion:
where,
is the true (Cauchy) stress tensor, whereas, is the
yield stress from a uniaxial test.
p
y
F
ij
( )
y
0 =
ij
F
ij
( ) { }
2
xx
2
yy
2
zz
2
xx
yy

yy
zz
+ + =
zz
xx
3
xy
2
yz
2
zx
2
+ + ( ) +
ij

y
Plasticity: Constitutive Relation, Uniaxial Tensile Test Nonlinear Material Behavior
Plasticity: von Mises Yield Surface is just a Cylinder
Multi-axial stress converted to a single equivalent stress
to compare with uniaxial strength using a yield criterion:
, is the equivalent von
Mises Stress.
The von Mises yield criterion in terms of principal stresses,
, becomes:
ij
vm
F
ij
( )
y
0
vm

y
= =
vm
i
i 1 3 , = ( ) ,
2 F
ij
( ) { }
2
1

3
( )
2
1

2
( )
2
2

3
( )
2
+ + =
Nonlinear Material Behavior Every Integration Point Tracks the Constitutive Relation
Every Integration Point Tracks the Constitutive Relation
Plasticity: Other Yield Criteria:
Every Integration Point Tracks the Constitutive Relation Nonlinear Material Behavior
Hardening Rules
Ideally plastic:
Isotropic hardening:
Kinematic hardening:
Combined:
F
ij
( )
y
0 =
F
ij
( )
y

p
( ) 0 =
F
ij

ij
( )
y
0 =
F
ij

ij
( )
y

p
( ) 0 =
ideally plastic
isotropic hardening
kinematic hardening
Stress and Strain Measures
Engineering stress/strain: ,
Cauchy stress/Log strain: ,
The strain corresponding to the true stress is the true or
logarithmic strain, . It is called the
Total Equivalent Strain. The Cauchy stresses and
logarithmic strains are activated using the parameter
option UPDATE. The Cauchy stress can be estimated as:
, assuming that .
For large strain plasticity problems, include also the
parameter option FINITE, or Large Strain NRM.
A
0
L
0
,
F F
A L
0
L + , L =
F
F
F
A
0
------
=
L
L
0
-------
=
F
A
---
= ln 1
+ ( ) =

d
0
t
ln 1
+ ( ) = =

( ) = AL A
0
L
0
=
Every Integration Point Tracks the Constitutive Relation Nonlinear Material Behavior
Plasticity: Entering Material Workhardening Data
Usual material data for plasticity is contained in the stress-
strain diagram. Where the strain is the total strain.
MSC.Marc expects the work hardening to be from the
stress-plastic strain diagram.
Usual material data

MSC.Marc expects this
Slope =
p

y
E =
E
Temperature Dependent Plasticity
Usual material data for plasticity is contained in the stress-
strain diagram for several temperatures.
In MSC.Marc analysis, it is better to use subroutine WKSLP.
Usual material data

WKSLP Interpolates over all Temperatures
Rigid-Plastic Flow Nonlinear Material Behavior
Rigid-Plastic Flow
The rigid-plastic flow analysis is an approach to large deformation
analysis which can be used for metal forming problems. Two
formulations are available: an Eulerian (steady state) and Lagrangian
(transient) approach. The effects of elasticity are not included. If these
effects are important, this option should not be used.
Steady State Analysis
The steady state R-P flow formulation is based on an Eulerian
reference system. For problems in which a steady-state
solution is not appropriate, an alternative method is available to
update the coordinates. User subroutine UPNOD is used to
update the nodal coordinates at the end of a step according to
the relationship, where n refers to the step
number,
n
is the nodal velocity components, and t is an
arbitrary time step. t is selected in such a way as to allow only
a reasonable change in mesh shape while ensuring stability
with each step. Updating the mesh requires judicious selection
of a time step. This requires some knowledge of the magnitude
of the nodal velocities that will be encountered. The time step
should be selected such that the strain increment is never more
than one percent for any given increment. The quantities under
the title of ENGSTN in the printouts actually refer to the strain
rate at an element integration point. The reaction forces output
by the program gives the limit loads on the structure. For
steady state analysis, user subroutine UNEWTN can be used to
define a viscosity.
x
i
n
x
i
n 1
v
n
t + =
Nonlinear Material Behavior Rigid-Plastic Flow
Transient Analysis
In the transient procedure, there is an automatic updating of the mesh
at the end of each increment. During the analysis, the updated mesh
can exhibit severe distortion and the solution might be unable to
converge. Mesh rezoning can be used to overcome this difficulty.
Technical Background
The rigid-plastic flow capability is based on iteration for the velocity
field in an incompressible, non-Newtonian fluid. The normal flow
condition for a nonzero strain rate can be expressed as:
where is the equivalent strain
rate, is the yield stress (which may be rate-dependent) and
gives the deviatoric stress. The effective viscosity is
evaluated as: . Note that as . A cutoff value of
strain rate is used in the program to avoid this difficulty. An initial
value for is necessary to start the iterations. These values can be
specified in the PARAMETERS option. The default cut-off value is
10
-12
, and the default initial strain rate value is 10
-4
. User subroutine
URPFLO can be used to define the flow stress.
ij
2
3
---
---
\ .
|
| |

ij

( )
ij
= =
2
3
---
ij
ij
=
ij

ij
1
3
---
ij
kk
=
2
3
---

---
=
0 ,

Creep Nonlinear Material Behavior
Creep
Constitutive Relation, uniaxial creep test
Multi-axial stress converted to a single equivalent stress
to compare with uniaxial strength using the von Mises
yield criterion:
Every integration point must track the Constitutive Relation.
c
y
Slope = E
2 F
ij
( ) { }
2
1

3
( )
2
1

2
( )
2
2

3
( )
2
+ + =
Nonlinear Material Behavior Creep
Creep Material Input
Material behavior is defined by describing the creep strain
rate. MSC.Marc integrates this wrt time to calculate the
creep strain in any increment.
Creep strain rates are, in general, a function of applied
stress, the current creep strain, the temperature, and the
current time.
In MSC.Marc, two methods exist for describing these
dependencies. Each dependency can be given by a piece-
wise linear function, or by a power law.
Piecewise linear:
Power Law:
A combined elastic-plastic-creep model developed at Oak
Ridge National Lab is also available in MSC.Marc.
More general creep strain rate equations may be specified
using the MSC.Marc usersubroutine CRPLAW.
c
A f ( ) g
c
( ) h T ( )
dk t ( )
dt
------------
=
c
A
m
c
( )
n
T
p
qt
q 1
( ) =
Creep Nonlinear Material Behavior
Creep Hardening Rules at Stress Changes
a) Time Hardening (Time fixed for creep increment)
b) Strain Hardening (Strain fixed for creep increment)
b
a
a
b
Nonlinear Material Behavior Creep
Stress and Temperature Dependent Creep
Creep test data are for several stress and temperature
levels. In the analysis, it is better to use subroutine
CRPLAW for interpolations and hardening rules.
c
Typical material data at one temperature
Viscoplasticity Nonlinear Material Behavior
Viscoplasticity
There are two procedures in MSC.Marc for viscoplastic analysis:
explicit and implicit. A brief description of each procedure follows:
Explicit Method
The elasto-viscoplasticity model in MSC.Marc is a modified
creep model to which a plastic element is added. The plastic
element is inactive when the stress is less than the yield
stress of the material. You can use the elasto-viscoplasticity
model to solve time-dependent plasticity and creep as well
as plasticity problems with a nonassociated flow law. The
requirements for solving the viscoplastic problem are:
CREEP parameter and creep controls load incrementation
immediately followed by a series of creep increments
specified by AUTO CREEP, and the use of user subroutine
CRPLAW and/or user subroutine NASSOC. Since the
viscoplasticity model in MSC.Marc is a modified creep
model, you should familiarize yourself with the creep
analysis procedure.
Implicit Method
A general viscoplastic material law can be implemented
through user subroutine UVSCPL. When using this
method, you are responsible for defining the inelastic strain
increment and the current stress.
Nonlinear Material Behavior Metal Forming Methods
Metal Forming Methods
Superplastic Forming
Hydro Forming
Stretch Forming
Bulk Forming
Powder Compaction
Creep Forming
Overview stress/strain parameter options:
Parameter option Stress tensor
Strain tensor
large disp Green-Lagrange
2nd Piola-
Engineering Engineering
Kirchhoff
update
finite
(large disp)
Logarithmic Cauchy
Neo-Hookean Material Model Nonlinear Material Behavior
Neo-Hookean Material Model
It took nearly 300 years to add the new part to Hookes Law and a
few more years for finite element analysis to deal with incompressible
materials by Herrmann.
Hookean (1660)
Neo-Hookean (1940)
2 1 + ( )G = G =
G 1 + ( ) 1 + ( )
2
[ ] = G =
Nonlinear Material Behavior Neo-Hookean Material Model
Definitions, Stretch Ratios, Engineering Strain:

Incompressibility:
From Thermodynamics and statistical mechanics,
First-order approximation (neo-Hookean):
i
L
i
L
i
+
L
i
--------------------
1
i
+ = = eng. strain,
i
L
i
L
i
( ) =
t
1
t
1
t
2
t
2
t
3
t
3
1
L
1
2
L
2
3
L
3
L
1
L
2
L
3
3
1 =
W
1
2
---
G
1
2
2
2
3
2
3 + + ( ) =
(Nearly) incompressible material:
, hence
Ordinary solid (e.g. steel): and are the same order of
magnitude. Whereas, in rubber the ratio of to is of the
order ; hence the response to a stress is effectively
determined solely by the shear modulus .
Bulk Modulus K
Shear Modulus G
------------------------------------------
2 1 + ( )
3 1 2 ( )
------------------------
=
1
2
---
K
G
----

G K
G K
10
4
G
Neo-Hookean Material Model
Experimental Verification using Simple Extension
Hence:
Engineering Stress:
True Stress:
Simple, one parameter material model.
Positive G guarantees material model stability.
1
=
2

3
1 = =
W
1
2
---
G
2 2
---
3 +
\ .
| |
=
dW d G
1
2
-----
\ .
| |
= = =
G 1
1
1 + ( )
2
-------------------
+
\ .
| |
=
t

1
----------
G
2 1
---
\ .
| |
= = =
Neo-Hookean Material Extension Deformation
Theory versus experiments:
1 2 3 4 5 6 7 8
Extension ratio
0.0
2.0
4.0
6.0
E
n
g
i
n
e
e
r
i
n
g

s
t
r
e
s
s

(
N
/
m
m
2
)
theory
experiment
Neo-Hookean Material Shear Deformation
Experimental Verification using Simple Shear:
If , then and
Equivalent shear strain :
Strain energy function:
Shear stress depends linearly on shear strain
X
Y
atan
1
=
2
1
---
=
3
1 =
---
=
W
1
2
---
G
2 1
2
-----
2 +
\ .
| |
1
2
---
G
2
= =

W d
d
------- -
G = =
Neo-Hookean Material Shear Deformation
Theory versus experiments:
0 1 2 3 4 5
Shear strain
0.0
0.4
0.8
1.2
1.6
S
h
e
a
r

s
t
r
e
s
s

(
N
/
m
m
2
)
theory
experiment
Nonlinear Material Behavior 2-Constant Mooney Extensional Deformation
2-Constant Mooney Extensional Deformation
Basic assumptions:
(1) The rubber is incompressible and isotropic
(2) Hookes law is obeyed in simple shear
Strain energy function with two constants:
Simple shear:
Hence
W C
1

1
2
2
2
3
2
3 + + ( ) C
2
1
1
2
-----
1
2
2
-----
+
1
3
2
-----
3 +
\ .
|
| |
+ =
W C
1
C
2
+ ( )
1
2 1
1
2
-----
2 +
\ .
|
| |
C
1
C
2
+ ( )
2
= =
dW d 2 C
1
C
2
+ ( ) = =
G 2 C
1
C
2
+ ( ) =
2-Constant Mooney Extensional Deformation Nonlinear Material Behavior
2-Constant Mooney Extensional Deformation
or
Theory versus experiments
2
1
2
-----
\ .
| |
C
1
C
2
------
+
\ .
| |
=

2 1
2
( )
------------------------------
C
1
C
2
------
+ =
A
B
C
D
E
F
G
0.5 0.6 0.7 0.8 0.9 1.0
1/
0.1
0.2
0.3
0.4
/
2
(
1
/
2
)

(
N
/
m
m
2
)
Nonlinear Material Behavior Other Mooney-Rivlin Models
Other Mooney-Rivlin Models
Basic assumptions:
(1) The rubber is incompressible and isotropic in the
unstrained state.
(2) The strain energy function must depend on even
powers of .
The three simplest possible even-powered functions
(invariants):
Incompressibility implies that , so that:
i
I
1

1
2
2
2
3
2
+ + =
I
2

1
2
2
2
2
2
3
2
3
2
1
2
+ + =
I
3

1
2
2
2
3
2
=
I
3
1 =
W W I
1
I
2
, ( ) =
Other Mooney-Rivlin Models Nonlinear Material Behavior
Other Mooney-Rivlin Models
Mooney material in terms of invariants:
(Mooneys original notation)
(Mooney-Rivlin notation)
Some other proposed energy functions:
The Signiorini form:
The Yeoh form:
Third-order Deformation Form
(James, Green and Simpson):
W C
1
I
1
3 ( ) C
2
I
2
3 ( ) + =
W C
10
I
1
3 ( ) C
01
I
2
3 ( ) + =
W C
10
I
1
3 ( ) C
01
I
2
3 ( ) C
20
I
1
3 ( )
2
+ + =
W C
10
I
1
3 ( ) C
20
I
1
3 ( )
2
C
30
I
1
3 ( )
3
+ + =
W C
10
I
1
3 ( ) C
01
I
2
3 ( ) C
11
I
1
3 ( ) I
2
3 ( ) + + + =
C
20
I
1
3 ( )
2
C +
30
I
1
3 ( )
3
Nonlinear Material Behavior Ogden Models
Ogden Models
Slightly compressible rubber:
and are material constants,
is the initial bulk modulus, and
is the volumetric ratio, defined by
The order of magnitude of the volumetric changes per unit
volume should be 0.01
Usually, the number of terms taken into account in
the Ogden models is or .
W

n
n
------
J
3
--------
n
+ + ( ) 3 4.5K J
1
3
---
1
\ .
|
| |
2
+
n 1 =
N
n

n
K
J
J
1
3
=
N 2 = N 3 =
Foam Models Nonlinear Material Behavior
Foam Models
Elastomer foams:
, and are material constants
W

n
n
------

1
n
+ + 3 [ ]

n
n
------
1 J
n
( )
n 1 =
N
+
n 1 =
N
n

n

n
Nonlinear Material Behavior General Formulation of Elastomers
General Formulation of Elastomers
Material points in undeformed configuration: ; material
points in deformed configuration:
Lagrange description:

is the deformation gradient tensor
Green-Lagrange strain tensor:
Right Cauchy-Green strain tensor:
Some additional relations:
X
i
x
i
x
i
x
i
X
j
( ) =
dx
i
F
ij
dX
j
with F
ij
x
i
X
j
--------
= =
F
ij
dx ( )
2
dX ( )
2
2E
ij
dX
i
dX
j
=
dx ( )
2
C
ij
dX
i
dX
j
=
C
ij

ij
2E
ij
+ =
C
ij
x
k
X
i
--------
x
k
X
j
--------
F
ki
F
kj
= =
General Formulation of Elastomers Nonlinear Material Behavior
Introduce displacement vector :
With respect to principal directions:
Invariants of :
E
ij
1
2
---
x
k
X
i
--------
x
k
X
j
--------

ij
1
2
---
F
ki
F
kj

ij
[ ] = =
u
i
x
i
X
i
u
i
+ =
E
ij
1
2
---
u
i j ,
u
j i ,
u
k i ,
u
k j ,
+ + ( ) =
C
ij

ki
u
k i ,
+ ( )
kj
u
k j ,
+ ( ) =
C
i' j'
1
2
0 0
0
2
2
0
0 0
3
2
=
C
ij
I
1
C
ii
=
I
2
1
2
---
C
ii
C
jj
C
ij
C
ij
( ) =
I
3
det C
ij
=
Nonlinear Material Behavior General Formulation of Elastomers
Strain energy function:
Second-Piola Kirchhoff stresses:
True or Cauchy stresses:
Zero deformation:
hence:
so that the stresses can be expressed in terms of
displacements and the hydrostatic pressure
W
*
W I
1
I
2
, ( ) h I
3
1 ( ) + =
S
ij
2
W
I
1
--------
ij
2
W
I
2
--------

ij
C
kk
C
ij
[ ] 2h
I
3
C
ij
----------
+ + =
t
ij
0
-----

ik
u
i k ,
+ ( )S
kl

jl
u
j l ,
+ ( ) =
S
ij
0
2
W
I
1
--------
4
W
I
2
--------
+
0
2h +
0
\ .
| |
ij
=
p 2
W
I
1
--------
0
4
W
I
2
--------
0
2h =
Finite Element Formulation Nonlinear Material Behavior
Finite Element Formulation
Modified virtual work equation:
In addition to the displacements, within an element we
need to interpolate the pressure:
The incremental stresses are related to the linear strain
increment by
The incremental set of equations to be solved reads:
S
ij
V
E
ij
dV Q
i
u
i
V d
V
T
i
u
i
A d
A
-
+
I
3
1 ( ) V d
V
+ 0 =
u
i
X
i
( ) N
X
i
( )u
i
=
p X
i
( ) h
X
i
( )p
=
S
ij
D
ijkl
n
E
kl
-
p C
ij
n
( )
1
=
K
0 ( )
K
1 ( )
+ [ ] H [ ]
H [ ]
T
0 [ ]
u
p
P R
g
=
Nonlinear Material Behavior Finite Element Formulation
with:
: the linear stiffness matrix
: the geometric stiffness matrix
: the nodal pressure coupling matrix
: nodal load vector
: internal stress vector
: vector quantity representing the
incompressibility constraint
K
0 ( )
[ ]
K
1 ( )
[ ]
H [ ]
P
R
g
Large Strain Viscoelasticity Nonlinear Material Behavior
Large Strain Viscoelasticity
The behavior of rubber is in most cases considered to be time
independent elastic. This approximation is no longer valid, if specific
hysteresis effects need to be taken into account. The theory of linear
viscoelasticity cannot be applied directly since there is no linear
relation between the applied strain and the resulting stress. Various
forms are proposed in literature to describe nonlinear viscoelasticity.
In MSC.Marc, a rather simple form, based on an extension of the
elastic energy function as proposed by Simo is used. The model is
based in the observation that for short time loading more energy is
required then in a long term loading. Also if one loads at a high rate
and keeps the deformation constant for a specific period of time, part
of the elastic energy is released.
Large Strain Viscoelasticity based on Energy
For an elastomeric time independent material the constitutive
equation is expressed in terms of an energy function W. For a large
strain viscoelastic material Simo generalized the small strain visco-
elasticity material behavior to a large strain viscoelastic material using
the energy function. The energy functional is now a time dependent
function and is written in the following form:
W E
ij
,Q
ij
n
( ) W
0
E
ij
( ) Q
ij
n
E
ij
n 1 =
N
=
Nonlinear Material Behavior Large Strain Viscoelasticity
where are the components of the Green-Lagrange straintensor,
internal variables and the elastic strain energy density for
instantaneous deformation. In MSC.Marc, it is assumed that is the
energy density for instantaneous deformations is given by the third-
order James, Green and Simpson form, or the energy function as
defined by Ogden.
The components of the second Piola-Kirchoff stress then follow from:
The energy function can also be written in terms of the long term
moduli resulting in a different set of internal variables :
where is the elastic strain energy for long term deformations.
Using this energy definition the stresses are obtained from:
Observing the similarity with the equations for small strain visco-
elasticity the internal variables can be obtained from a convolution
expression:
E
ij
Q
ij
n
W
0
W
0
S
ij
W
E
ij
----------
W
0
E
ij
----------
Q
ij
n
n 1 =
N
= =
T
ij
n
W E
ij
T
ij
n
, ( ) W
E
ij
( ) T
ij
n
E
ij
n 1 =
N
+ =
W
S
ij
W
E ( )
E
ij
--------------------
T
ij
n
n 1 =
N
+ =
where are internal stresses following from the time dependent part
of the energy functions.
Let the total strain energy be expressed as a Prony series expansion:
Observing the difficulty in finding accurate expressions for the
multiaxial aspect of the elastic energy in time independent rubber a
further simplification is used. We assume that the energy expression
for each term is of similar form to the short time elastic energy and
only different by a scalar multiplier . This equation can
now be rewritten as:
where is a scalar multiplier for the energy function based on the
short term values.The stress strain relation is now given by:
T
ij
n
S
ij
n
( ) t ( )
n
[ ]
.
d exp
0
t
=
S
ij
n
S
ij
n W
n
E
ij
----------
=
W W
W
n
t
n
( ) exp
n 1 =
N
+ =
W
0
W
n
W
0
=
W W
n
W
0
t
n
( ) exp
n 1 =
N
+ =
n
S
ij
t ( ) S
ij
t ( ) T
ij
n
t ( )
n 1 =
N
+ =
Nonlinear Material Behavior Large Strain Viscoelasticity
Analogue to the derivation for small strain visco-elasticity a recurrent
relation can be derived expressing the stress increment as a function of
the strain increment and the internal stresses at the start of the
increment:
The functions and are a function of the time step h in the time
interval :
S
ij
W
E
ij
-----------
1
n
n 1 =
N
\ .
|
|
| |
W
0
E
ij
----------
= =
T
ij
n
n
S
ij
0
t ( ) t ( )
n
[ ]
.
d exp
0
t
=
S
ij
t
m
( ) S
ij
t
m
( ) S
ij
n
t
m
( )
n 1 =
N
+ =
S
ij
t
m
( ) S
ij
t
m
( ) S
ij
t
m
( ) =
S
ij
n
t
m
( )
n
h ( ) S
ij
n
t
m
( ) S
ij
n
t
m
h ( ) [ ]
n
h ( )S
ij
n
t
m
h ( )
n 1 =
N
=

t
m
1 t
m
, [ ]
n
h ( ) 1 h ( )
n
exp =
n
h ( )
n
h ( )

n
h
-----
=
The equations above are based on the long term moduli. Since in the
MSC.Marc program always the instantaneous values of the energy
function are given on the MOONEY option the equations are
reformulated in terms of the short time values of the energy function:
It is assumed that the viscoelastic behavior in MSC.Marc acts only on
the deviatoric behavior. The incompressible behavior is taken into
account using special Herrmann elements.
S
ij
t
m
( ) 1 1
n
h ( ) [ ]
n
n 1 =
N
S
ij
0
t
m
( ) S
ij
0
t
m
h ( ) =
n
S
ij
n
t
m
h ( )
n 1 =
N
S
ij
n
t
m
( )
n
h ( )
n
S
ij
0
t
m
( ) S
ij
n
t
m
h ( ) [ ] =
n
h ( )S
ij
n
t
m
h ( )
Nonlinear Material Behavior Illustration of Large Strain Viscoelastic Behavior
Illustration of Large Strain Viscoelastic Behavior
A large strain viscoelastic material is characterized by the following
time dependent elastic energy function:
where is the energy function for very slow processes. is an
extra amount of energy necessary for time dependent processes. To
each amount , a characteristic time is associated. At time zero (or
for time processes: ) the elastic energy reduces to:
If we assume that the energy function for each time dependent part is
different only by a scalar constant:
the equations reduce to:
or
W t ( ) W
W
n
t
n
( ) exp
n 1 =
N
+ =
W
W
n
W
n
t
n
<
W 0 ( ) W
0
W
W
n
n 1 =
N
+ = =
W
n
n
W
0
=
W
0
W
W
0
n
n 1 =
N
+ = W
1
n
n 1 =
N
\ .
|
|
| |
W
0
=
Illustration of Large Strain Viscoelastic Behavior Nonlinear Material Behavior
The time dependent energy is then given by:
If we restrict ourselves for simplicity of the discussion to the case N = 1
we have:
W t ( ) W
0
W
0
n
W
0
n
t
n
( ) exp
n 1 =
N
+
n 1 =
N
=
W
0
1
n
1 t
n
( ) exp ( )
n 1 =
N
=
W
1 ( )W
0
=
W t ( ) W
0
1 1 t
n
( ) exp ( ) [ ] =
Nonlinear Material Behavior Illustration of Large Strain Viscoelastic Behavior
CHAPTER 5 Contact Analysis
This features allows for the automated solution
of problems where contact occurs between
deformable and rigid bodies. It does not require
special elements to be placed at the points of
contact. This contact algorithm automatically
detects nodes entering contact and generates the
appropriate constraints to insure no penetration
occurs and maintains compatibility of
displacements across touching surfaces.
Contact Analysis Definition of Contact Bodies
Definition of Contact Bodies
Contact Body Any group of
elements or geometric entities
that may contact themselves
or others.
Types of Contact Bodies:
Deformable Collection of
elements. Rigid Collection
of geometric entities or heat
transfer elements
Add elements to contact body,
here 90 elements are added to
contact body, cbody1.
Analytic contact may be used
to smooth facets of element
edges or faces.
By default Rigid bodies are
controlled with displacement,
unless specified here.
Geometric curves/surfaces
have to be properly oriented.
Definition of Contact Bodies Contact Analysis
Definition of Contact Bodies
Contact Body Any group of
elements or geometric
entities that may contact
themselves or others.
Types of Contact Bodies:
Deformable Collection of
elements. Rigid Collection
of geometric entities or heat
transfer elements
Add elements to contact
body, here 1 curve is added
to body, cbody2.
This bodies position is
controlled by a table.
Contact Analysis Control of Rigid Bodies
Control of Rigid Bodies
Rigid bodies can be controlled by their velocity, position, or load.
q Prescribe translational and/or rotational velocity as a
function of time using a time table.
q Prescribe position/rotation as a function of time.
q Prescribe force on rigid body as a function of time:
define force on additional node
connect node to rigid contact body
motion of rigid contact body is in direction of
applied force; motion in perpendicular direction
is constrained
Contact Procedure Contact Analysis
Contact Procedure
Deformable to Rigid Body Contact
with:
: incremental displacement vector of node
: unit normal vector with proper orientation
: contact distance (Default = h/20 or t/4)
: separation force (Default = Maximum Residual)
Case 1: Node A does not touch, no constraint applied.
Case 2: Node A is near rigid body within tolerance, con-
tact constraint pulls node to contact surface if .
Case 3: Node A penetrates within tolerance, contact con-
strain pushes node to contact surface.
Case 4: Node A penetrates out of tolerance and increment
gets split (loads reduced) until no penetration.
D D A
n
u
A
Rigid Body
Deformable Body
3 2
Case 1
4
(set of elements)
(set of curves or
surfaces)
u
A
n d D
d
Cases 2,3: Contact Detected When
Case 1: Contact Not Detected When
u
A
n D d <
Case 4: Penetration Detected When
u
A
n D d + >
u
A
A
n
D
F
s
F F
s
<
Contact Analysis Bias Factor
Bias Factor
By default, the contact tolerance is equally applied to both sides of a
segment; this can be changed by introducing a bias factor .
( ):
Choosing may be useful to
reduce increment splitting, since the distance to cause
penetration is increased
improve accuracy, since the distance below which a
node comes into contact is reduced
:
B
0 B 1
D
contact
D
contact
D
contact
1 B + ( )
D
contact
1 B ( )
default
with bias factor
B 0 >
default with bias factor
Bias Factor Contact Analysis
Deformable to Deformable Contact
Discrete deformable contact (default) is based on piecewise
linear geometry description of either 2-node edges in 2
dimensions or 4-node faces in 3 dimensions on the outer
surface of all contacting meshes.
Then the contact constraint:
defines tying relation for displacement component of
contacting node in local -direction
applys correction on position in local -direction
actual geometry
finite element approximation
contacted body
contacting body
contact tolerance
y
x
A
y
y
Contact Analysis Bias Factor
Potential Errors due to Piecewise Linear Description:
Tying relation may be not completely correct due to the
assumption that the normal direction is constant for a
complete segment.
If contacting node slides from one segment to another,
a discontinuity in the normal direction may occur.
The correction on the position of the contacting node
may be not completely correct.
Analytical Deformable Contact Bodies:
Replace 2-node linear edges by cubic splines (2D)
or 4-node bi-linear patches by bi-cubic Coons
surfaces (3D).
You must take care of nodes (2D) and edges (3D)
where the outer normal vector is discontinuous.
You may wish to use extended precision.
Advantages are smoother contact where in 2D,
-continuity is obtained, and in 3D, at least pointwise
-continuity is obtained. Analytical deformable contact
must be turned on, whereas, rigid bodies default to analytic
contact where the curves or surfaces are represented as
NURBS during contact.
C
1
C
1
Contact Flowchart Contact Analysis
Contact Flowchart
Input
Initial set up of contact bodies
Incremental data input
Check on contact
Set up of contact constraints
Apply distributed loads
Assemble stiffness matrix; include friction
Apply contact constraints
Solve set of equations
Recover stresses
Converged solution?
Separation?
Penetration?
Last increment?
Stop
No
No
Yes
Yes
Yes
Change
contact
constraints
Split
increment
Update contact constraints
No
No
Yes
b
e
g
i
n

i
n
c
r
e
m
e
n
t
b
e
g
i
n

i
t
e
r
a
t
i
o
n
Contact Analysis Symmetry Body
Symmetry Body
Symmetry bodies often provide an easy way to impose symmetry
conditions; they may be used instead of the TRANSFORMATION and
SERVO LINK options that WILL CAUSE PROBLEMS in contact as
these nodes come into contact. A symmetry plane is characterized by a
very high separation force, so that only a movement tangential to the
contact segment is possible The symmetry plane option can only be
invoked for rigid surfaces
deformable_body
symmetry_plane_1
symmetry_plane_2
none
Rigid with Heat Transfer Contact Analysis
Rigid with Heat Transfer
Model 1: deformable-rigid (stress or coupled analysis)
billet
channel
35
20
o
4
4.75
50
R = 6
25
20
billet
channel
none
deformable-rigid (stress or coupled analysis)
geometrical entities
(straight lines and a
circular arc)
MARC element 10
Contact Analysis Rigid with Heat Transfer
Model 2: deformable-rigid (coupled analysis)
Model 3: deformable-deformable (stress or
coupled analysis)
billet
channel
none
deformable-rigid (coupled analysis)
MARC element 40
MARC element 10
Rigid w Heat Transfer
billet
channel
none
deformable-deformable (stress or coupled analysis)
MARC element 10
MARC element 10
Contact Table Contact Analysis
Contact Table
Contact Table Properties:
Single-sided contact:
Only body 2 may contact itself
1
2
3
4
Contact Analysis Contact Table
Contact Table
Very useful for specifying parameters between
contacting bodies
Contact tables must be turned on initially in contact
control, or during any loadcase to become active. With no
contact tables active, all bodies can come into contact
including self contact
Contact Table Contact Analysis
Contact Areas
Very useful for defining certain nodes of a body that may
enter contact.
Like contact tables, contact areas must be turned on
initially in contact control, or during any loadcase to
become active. With no contact areas active, all nodes of
all bodies can come into contact.
Both contact table and contact areas can reduce the
amount of node to segment checking and can save
compute time.
Contact Analysis Exclude Segments during Contact Detection
Exclude Segments during Contact Detection
Exclude segment will influence the searching done for nodes detected
in the contact zone during self contact.
Options to influence search for contact include:
q Contact table: define which bodies can potentially come
into contact (defined per loadcase)
q Contact node: define which nodes of a body can poten-
tially come into contact (defined per loadcase)
q Single sided contact: searching for contact is not done
with respect to bodies with a lower body number
(defined for the whole analysis)
q Exclude: define which segments of a body can never be
contacted (defined per loadcase)
Contact table, contact node and
exclude affect the initial search
for contact; once a node is in
contact, this is not undone by
these options.
Exclude Segments during Contact Detection Contact Analysis
Effect of Exclude Option:
Standard contact
excluded segments
With exclude option
Contact Analysis Exclude Segments during Contact Detection
Contacting Nodes and Contacted Segments
For 3D continua, an automatic check on the direction of the
normal vectors is included:
Contact will not be accepted if
Shell Thickness is taken into Account According to:
q 2D: one fourth of the shell thickness only if the body is
contacted.
q 3D: one fourth of the shell thickness for both the
contacting and the contacted body.
Contacting body nodes Contacted body patches
n
node
n
patch
0.05 >
Friction Model Types Contact Analysis
Friction Model Types
Friction coefficient is specified in contact body or contact
table. Although, the coefficient is entered a specific friction
model type must be selected for friction to be active.
Contact Analysis Stick-Slip Friction Model
Stick-Slip Friction Model
Discovered by Leonardo da Vinci in the 15th century and
verified by experiments by Charles A. Coulomb in the 18th
century, this stick-slip friction model uses a penalty method
to describe the step function of Columbs Law.
with:
: incremental tangential displacement
: slip to stick transition region (default
: coefficient multiplier (default 1.05)
: friction force tolerance (default 0.05)
: small constant, so that (fixed at )
u
t
F
t
2
2
F
n
F
n
F
t
F
n
static F
t
F
n
, kinetic
u
t
1
6
10
e
0 1
6
10
Coulomb (Sliding) Friction Model Contact Analysis
Coulomb (Sliding) Friction Model
Implementation of this friction model has been done using
nonlinear dashpots whose stiffness depend on the relative
sliding velocity as:
MSC.Marc approximation:
with:
: relative sliding velocity below which sticking is
simulated (Default = 1.0!)
slip
slip
MARC approximation
F
t
v
r
stick
C
F
t
F
n
2
---
v
r
C
----
\ .
| |
atan
C
Contact Analysis Shear (Sliding) Friction Model
Shear (Sliding) Friction Model
This friction model has been done using nonlinear dashpots
whose stiffness depend on the relative sliding velocity as:
However, the friction force depends upon a fraction of the
shear strength of the material, not the normal force:
with:
: relative sliding velocity below which sticking is
simulated (Default = 1.0!)
slip
slip
MARC approximation
t
v
r
stick
C
t

Y
3
-------
\ .
| |
2
---
v
r
C
----
\ .
| |
atan
C
Shear (Sliding) Friction Model Contact Analysis
Coulomb Sliding Friction Model use Stresses or Forces
Nodal stresses: (Applies to continuum elements)
define distributed load in opposite direction of
relative sliding velocity: ,where, , friction
coefficient and, , normal stress in contact point
Nodal forces: (Always used for shell elements)
define nodal load in opposite direction of relative
sliding velocity, , where, , friction coefficient
and, , normal force in contact point
p
t
p
t

n
=
n
p
t
v
r
F
t
F
t
F
n
=
F
n
F
t
v
r
F
t
Contact Analysis Glued Contact
Glued Contact
Sometimes a complex body can be split up into parts which can be
meshed relatively easy:
define each part as a contact body
invoke the glue option (CONTACT TABLE) to obtain
tying equations not only normal but also tangential to
contact segments
enter a large separation force
Z
body 1
body 2
Glued Contact Contact Analysis
Glued Contact
Gluing rigid to deformable bodies can help simulate testing because
testing of materials generally involves measuring the force and
displacement of the rigid grips. Here is an example of a planar tension
(pure shear) rubber specimen being pulled by two grips. The grip force
versus displacement curve is directly available on the post file and can
be compared directly to the force and displacement measured.
Contact Analysis Release Option
Release Option
The release option provides the possibility to deactivate a contact
body: upon entering a body to be released, all nodes being in contact
with this body will be released. Using the release option, e.g. a spring-
back effect can be simulated. Releasing nodes occurs at the beginning
of an increment. Make sure that the released body moves away to
avoid recontacting
Interference Check / Interference Closure Amount
By means of the interference check, an initial overlap will be removed
at the beginning of increment 1
By means of an interference closure amount an overlap or a gap
between contacting bodies can be defined per increment:
positive value: overlap
negative value: gap
Forces on Rigid Bodies Contact Analysis
Forces on Rigid Bodies
During the analysis rigid bodies have all forces and moments resolved
to a single point which is the centroid shown below.

This makes rigid bodies useful to monitor the force versus
displacement behavior as shown at the right.
Body 3 Force Y
Contact Analysis Forces on Rigid Bodies
Forces on Rigid Bodies
Vector plotting External Force will show the forces at each node
resulting from the contact constraints.
CHAPTER 6 Heat Transfer and Thermal Stresses
The purpose of this chapter is to focus how
MSC.Marc performs heat transfer analysis
and its impact to thermal stress analysis.
Heat Transfer and Thermal Stresses Mathmetical Formulation
Mathmetical Formulation
energy conservation law:
with:
: mass density
: specific heat
: gradient operator
T x
t , ( )
Q x
t , ( )
source distribution
temperature
distribution
CONTINUUM
v
t , ( )
q
t , ( )
material flow velocity
energy flow
density
X
Y
Z
x
x y z , , ( ) =
T
n
C
p
T
t
------

T
q
+ Q =
C
p
, ,
\ .
| |
T
Mathmetical Formulation Heat Transfer and Thermal Stresses
Mathematical Formulation
Energy flow density is given by a diffusion and convection
part:
where is the conductivity matrix. Assume that the
continuum is incompressible and that there is no spatial
variation of and :
Without convection this reduces to:
where is the conductivity
q
T C
p
Tv
+ =
C
p
C
p
T
t
------
C
p
v
T ( ) + Q =
C
p
T
t
------

2
T Q =
Heat Transfer and Thermal Stresses Boundary conditions

Boundary conditions
Description Corresponding
Mentat BC type
1) prescribed temperature: fixed temperature
2) prescribed energy flow edge/face flux
density normal to surface:
3) convective heat transfer edge/face film
to the environment:
: film coefficient
: environmental
temperature
T T x t , ( ) =

T
n
------
q =

T
n
------
h T T
( ) =
h
T

Boundary conditions Heat Transfer and Thermal Stresses
Boundary Conditions
4) radiative heat transfer
to the environment:
: Stefan-Boltzmann
constant
: emissivity
5) heat source or sink:
localized: point flux
distributed: volume flux
Initial condition in transient analysis:
T
n
------
T
4
T
4
( ) =
Q x
t , ( ) Q t ( ) x
0
( ) =
Q x
t , ( ) Q x
t , ( ) =
T x
0 , ( ) T
i
x
( ) =
Heat Transfer and Thermal Stresses Analogy between Heat Transfer and Stress Analysis
Analogy between Heat Transfer and Stress Analysis
Heat transfer analysis Stress analysis
temperature T displacement vector u
Fourier law stress-strain law
(distributed) heat flux (distributed) force
(scalar) (vector)
film (convective boundary spring
condition)
radiative boundary condition nonlinear spring
heat flux density vector stress tensor
transient dynamic
(1st order) (2nd order)
heat capacity mass
steady state static
F
k u
0
( ) =
q h T T
( ) =
q T
4
T
4
( ) =
Spatial Finite Element Discretization Heat Transfer and Thermal Stresses
Spatial Finite Element Discretization
Approximate temperatures for a set of discrete points
(nodes) in spaceby expressing temperature in terms
of locally definedshape functions and nodal values
where are the shape functions and are the
nodal temperatures. Using the Galerkin method, the heat
transfer problem can be written as a coupled set of first
order ordinary differential equations:
with:
: heat capacity matrix
: conductivity and convection matrix
: contribution from convective boundary condition
: vector of nodal fluxes
In case of convection:
Upwinding (SUPG method)
Nonsymmetric system matrix
T x t , ( )
T x
t , ( )
N
1
x
( ) N
2
x
( ) ... N
m
x
( )
T
1
t ( )
T
2
t ( )
.
T
m
t ( )
N
T
T
= =
N
i
x
( ) T
i
t ( )
CT
K F + [ ]T
+ Q
=
C
K
F
Q
K
Heat Transfer and Thermal Stresses Linear Heat Transfer Analysis
Linear Heat Transfer Analysis
steady state: solution can be obtained by a single matrix
inversion:
transient: time discretization by means of finite differences:
Approximate nodal temperature at discrete points in time
MSC.Marc uses a backward difference scheme is used to
approximate time derivative as:
resulting finite difference scheme:
T
K F + ( )
1
Q
=
T
n
T
t
0
nt + ( ) =
T
n T
n
T
n 1
t
------------------------
C
t
-----
K F + + T
n C
t
-----
T
n 1
Q
+ =
Nonlinear Heat Transfer Analysis Heat Transfer and Thermal Stresses
Nonlinear Heat Transfer Analysis
Nonlinearity can be caused by:
temperature dependent conductivity or specific heat
radiation boundary condition
temperature dependent film coefficient or heat flux
Consequently, heat capacity matrix, conductivity matrix,
film matrix and equivalent nodal flux vector may be
temperature dependent.
steady state: applicable in case of mild nonlinearity; e.g. if
conductivity is slightly temperature dependent
Solution can be obtained iteratively:
start with:
next approximations are obtained by successive
substitution:
continue until:
KT
0
Q
=
T
n 1 +
K T
n
( ) [ ]
1
Q
n
( ) =
T
n 1 +
T
n
max
T
tol3
Heat Transfer and Thermal Stresses Nonlinear Heat Transfer Analysis

Nonlinear Heat Transfer Analysis
transient analysis: necessary in case of severe nonlinearity;
e.g. radiation boundary condition
Solution can be obtained using nonlinear backward
difference scheme:
where is the average nodal temperature vector in
time increment first iteration within an increment ,
is taken as an extrapolated value of the previous
two increments:
for the next iterations , follows from:
iterations are stopped if
C T
*
( )
t
---------------
K T
*
( ) F T
*
( ) + + T
n C T
*
( )
t
---------------
T
n 1
Q
*
( ) + =
T
*
t n
T
*
T
1
* 1
2
---
3T
n 1
T
n 2
( ) =
i T
*
T
i
* 1
2
---
T
n 1
T
i 1
n
+ ( ) =
T
i 1
*
T
i
*
max
T
tol3

Control Values Heat Transfer Analysis Heat Transfer and Thermal Stresses
Control Values Heat Transfer Analysis
: maximum incremental temperature change at node
(def = 20); if the automatic time stepping scheme is selected,
the time step will be automatically increased or decreased if
necessary
: maximum nodal temperature change before
properties are reevaluated and matrices reassembled
(def = 100)
: maximum error in estimated nodal temperature;
used for property reevaluation (def = 0, test isbypassed)
T
tol1
T
tol2
T
tol3
Heat Transfer and Thermal Stresses Initial Time Step Estimate
Initial Time Step Estimate
Sometimes oscillatory behaviour if time step is too small;
e.g. heat penetration in an originally iso-thermal block
Better approximation can be obtained if:
time step is INCREASED
mesh is refined
heat capacity matrix is lumped (linear elements)
Further MSC.Marc capabilities:
user subroutines for nonlinear boundary conditions tyings
and heat transfer shell element with parabolic distribution
in thickness direction phase transitions
: analytical solution
: FEM approximation
T
T
p
T
0
x
Surface to Surface Radiation Heat Transfer and Thermal Stresses
Surface to Surface Radiation
In many analyses, the radiative transfer of heat between
surfaces plays a significant role. To properly model this
effect, it is necessary to compute the proportion of one
surface which is visible from a second surface known as the
viewfactor. The viewfactor, defined by a fourth-order
integral, presents many difficulties in its computation.
Primary among these is the large amount of computing
power needed, especially when shadowing effects are
included. The radiative flow of hear from surface 1 to
surface 2 is given by:
in which, F
12
is the viewfactor and is calculated as:
In MSC.Marc there are two computational approaches,
namely, the Direct Adaptive Integration, and the Monte
Carlo Method method. The Direct Adaptive Integration is
only for axisymmetric models where as the method
discussed herein is the Monte Carlo Method method and is
applicable to any surface to surface radiation.
q
12
F
12
T
1
4
T
2
4
( ) =
F
12
1
A
1
------

1

2
cos cos
r
2
-----------------------------
A
2
d A
1
d
A
2
A
1
=
Heat Transfer and Thermal Stresses Surface to Surface Radiation
Monte Carlo Method for Surface Radiation
In this method, the idea is to randomly emit rays from the surface in
question. The percentage of these rays which hit another surface is the
viewfactor between the surfaces. The Monte Carlo method computes
N viewfactors at one time, providing linear scaling. In fact, the larger
the number of surfaces, the faster the viewfactors are computed
compared to the Direct Adaptive Integration and Adaptive Contour
Integration. Hence, this method is adopted for the viewfactor
calculations in MSC.Marc Mentat. Some of the notable features of the
viewfactor calculation in MSC.Marc Mentat and MSC.Marc are:
1.) You are not required to specify blocking elements. This is
embedded into the algorithm completely and, hence, done
automatically. This is specially useful in three-dimensional
analysis since, for complex geometries, it is impractical to
predict what surfaces are blocking other surfaces.
2.) The cost of calculation is nearly linearly proportional to the
number of elements which means that, for big problems, the
cost does not increase significantly.
3.) The methodology in MSC.Marc guarantees that the sum of
viewfactors is always one.
4.) Shadowing effects (due to two surfaces being hidden from one
another by other surfaces) can be modeled.
Monte Carlo Method for Surface Radiation
Ray Emission: The surface should be able to randomly emit
a ray from its surface.
Ray Intersection: Given an origin, direction, and length of
a ray, the surface should be able to determine whether it is
hit by that ray, and if hit, the length of the ray at the point
of intersection.
Surface 1
Surface 2
Point Light
Eye Ray
Light Source Rays
Shadow
Monte Carlo Method Example
1.) Pick all edges/faces that can radiate. Here, the outer
surface of the inner sphere and the inner surface of the
outer sphere has edge radiation turned on.
Notice that there are 36 objects (edges of elements) that can
view each other.
2.) In jobs, select the type of radiation (Axisymmetric),
storage file name (e5x15b.vfs) and compute the
viewfactors by clicking start.
3.) In jobs, under analysis options, identify the viewfactor
file and then submit to MSC.Marc.
4.) The results of this problem (Chapter 5 problem 15 in
Volume E) are show below. Radiation is the only
mechanism for heat transfer across the vacuum between
these two concentric spheres.
Laminar Flow and Heat Transfer Heat Transfer and Thermal Stresses
Laminar Flow and Heat Transfer
Note that a fluid flowing over a surface can exchange heat
with the surface according to:
where the second term convects heat to or away from the
surface. To illustrate how this would work consider the
following Problem 17 in Chapter 5 of Volume E.
This problem demonstrates the air cooling of an electronic
chip at room temperature. The comparison of the no-
inclusion of heat convection, e5x17a.dat, and the inclusion
of the contribution of heat convection, e5x17b.dat, by air is
made. The nonsymmetric solver is turned on automatically
when heat convection is included.
This problem is modeled using the two techniques
summarized below.
Element
Element type 39 is used for both the air region and the chip body.
The model is shown in Figure 5.17-1.
Data Set
Element
Type(s)
Number of
Elements
Number of
Nodes
Differentiating
Features
e5x17a 39 360 399 Exclude
convection
e5x17b 39 360 399 Include
convection
C
p
T
t
------

f
C
p
f
v
T ( ) + Q =
Heat Transfer and Thermal Stresses Laminar Flow and Heat Transfer
Material properties
Room temperature thermal properties for air are used. The
specific heat is 5.8e-5 kJ/kg.C, the density is 0.24l kg/m
3
, and
thermal conductivity is 0.129e-2 W/m.C. Thermal properties for
pure copper are used for the chip. The specific heat is 0.93 kJ/
kg.C, the density is 0.0914 kg/m
3
, and thermal conductivity is
8.94 W/m.C. Assume the variation of properties with
temperature is negligible.
Initial Conditions
The initial nodal temperature for chips is 40C and for air is 10C
throughout.
Boundary Conditions
The temperature of the air far away from chips is fixed at 10C
and velocity of the air is kept at a constant 100 cm/second. The
velocity of the chips is zero.
Transient Nonauto
A fixed, large time step is used to simulate the cooling process
near steady-state condition.
Results
The temperature distributions shown in Figure 5.17-3 and
Figure 5.17-4 which indicate the effect of heat convection on the
cooling of the chips. The chips have cooled down faster on the left
side because, as heat convection of the air is included, more heat is
carried away by the air. The effect of the boundary layer between
the air and the surface of the chips is neglected. Because the
Courant number is too large, which indicates the time step is also
too large, numerical dispersion occurs at the air region far away
from the chips.
Parameters, Options, and Subroutines Summary
Example e5x17a.dat:
Parameters Model Definition Options History Definition Options
ALL POINTS CONNECTIVITY CONTINUE
COMMENT CONTROL TRANSIENT
DIST LOADS COORDINATE
END DEFINE
HEAT END OPTION
PRINT FIXED TEMP
SETNAME INITIAL TEMP
SIZING ISOTROPIC
TITLE NO PRINT
POST
VELOCITY
Example e5x17b.dat:
Parameters Model Definition Options History Definition Options
ALL POINTS CONNECTIVITY CONTINUE
COMMENT CONTROL
DIST LOADS COORDINATE
END DEFINE TRANSIENT
HEAT END OPTION
PRINT FIXED TEMP
SETNAME INITIAL TEMP
SIZING ISOTROPIC
TITLE NO PRINT
POST
VELOCITY
Figure 5.17-1 Complete Finite Element Mesh
Figure 5.17-2 Finite Element Mesh of Chips and Board
2.0 cm
1
.
5

c
m
Y
Z X
0
.
1
2
5
0
.
0
7
5
.2 .5 .3
Y
Z X
Figure 5.17-3 Temperature Distribution Excluding Convection
Y
Z X
INC
SUB
TIME
FREQ
:
:
:
:
50
0
5.000e+01
0.000e+00
prob e5.17a cooling of chips: heat convection
Temperature
3.689e+01
3.420e+01
3.151e+01
2.882e+01
2.613e+01
2.344e+01
2.076e+01
1.807e+01
1.538e+01
1.269e+01
1.000e+01
Figure 5.17-4 Temperature Distribution Including Convection
Y
Z X
INC
SUB
TIME
FREQ
:
:
:
:
50
0
5.000e+01
0.000e+00
prob e5.17b cooling of chips: heat convection
Temperature
2.281e+01
2.139e+01
1.997e+01
1.856e+01
1.714e+01
1.572e+01
1.430e+01
1.288e+01
1.146e+01
1.005e+01
8.628e+00
Heat Transfer and Thermal Stresses Coupled Thermo-mechanical Analysis
Coupled Thermo-mechanical Analysis
Overview physical effects attributing to coupling between
heat transfer part and mechanical part:
Heat transfer and mechanical analysis can be done:
separately (temperature in mechanical analysis can be
directly defined or read in from heat transfer post file)
fully coupled (include all coupling effects; staggered
algorithm)
heat generated
due to plastic
deformation
changing geometry
due to large
deformations
changing contact
conditions
heat transfer
analysis
mechanical
analysis
thermal
expansion
temperature
dependent material
properties
temperature
dependent bound-
ary conditions
Heat Generated from Mechanical Sources Heat Transfer and Thermal Stresses
Heat Generated from Mechanical Sources
To determine the heat generated from mechanical sources
such as the work done plastic deformation (volumetric heat
generation) or friction heating (surface heat generation), it
is necessary to specify how much mechanical energy is
converted into thermal energy. In MSC.Marc, this is
specified with the heat generation conversion factor
specified in the Jobs Parameter section in MSC.Marc
Mentat. This conversion factor is composed of two factors.
The first factor called Fraction, is the amount of
mechanical energy converted to thermal energy, and is
between zero and one. Work by Farren and Taylor suggest
a Fraction of .9 is appropriate for metal plasticity. The
second factor called Units must account for consistent units
of mechanical and thermal energy. For example, if the
mechanical properties are such that mechanical energy is in
Lbf-in, and the thermal energy is in BTUs then the value of
Units becomes:
The default conversion factor is 1 where 100% of the
mechanical energy is converted to heat and the units of
mechanical and thermal energy are the same.
C
convfactor
Fraction ( ) Units ( ) =
Units
BTU
777.9 12 ( )Lbf in
--------------------------------------------
=
Heat Transfer and Thermal Stresses Coefficient of Thermal Expansion
Coefficient of Thermal Expansion
The MSC.Marc program always uses an instantaneous
thermal expansion coefficient defined as:
or for isotropic behavior:
In most cases, the thermal expansion data is given with
respect to a reference temperature , as:
Hence,
and the conversion of the expansion data to the
instantaneous thermal expansion coefficient becomes:
d
th
ij

ij
dT =
d
th
ij

ij
dT =
T
o
th
T T
o
( ) =
d
th
d
dT
-------
T T
o
( ) + dT =

d
dT
-------
T T
o
( ) + =
CHAPTER 7 Dynamic Analysis
The purpose of this chapter is to
focus on how MSC.Marc performs
dynamic analysis.
Dynamic Analysis Dynamic Analysis Methods
Dynamic Analysis Methods
Eigenvalue extractions linear with pre-stress
Lanczos method
Power Sweep
Harmonic response linear with pre-stress
Real (no Damping)
Imaginary (Damping)
Transient analysis linear and nonlinear
Explicit
Implicit
Contact
One Degree of Freedom System Dynamic Analysis
One Degree of Freedom System
Undamped free vibrations
Equation of motion:
with non-trivial solution:

;
hence:
k
m
u
mu
ku + 0 =
u Re ue
st
[ ] =
s
2
2
+ [ ]u 0 = k m =
s i =
u A t + ( ) cos =
Dynamic Analysis One Degree of Freedom System
One Degree of Freedom System: damped free vibrations
Equation of motion:
with harmonic solution:

;
hence:
critical damping:
k
m
u
c
mu
cu
ku + + 0 =
u Re ue
st
[ ] =
s
2 c
m
----
s
2
+ + u 0 = k m =
s
c
2m
-------
c
2m
-------
\ .
| |
2
2
=
c
crit
2m =
One Degree of Freedom System
Define damping factor:
: below critical damping oscillating behavior
: over critical damping exponential decay
c
c
crit
----------
=
1 <
1
time
d
i
s
p
l
a
c
e
m
e
n
t
1 <
time
d
i
s
p
l
a
c
e
m
e
n
t
1
Dynamic Analysis One Degree of Freedom System
One Degree of Freedom System: harmonic oscillations
Equation of motion:
then define:
;
static solution is:
k
m
u
c
pe
it
mu
cu
ku + + pe
it
=
0
k
m
---- =

0
------
=
u
s
p
k
---
=
One Degree of Freedom System; harmonic oscillations
Dynamic magnification factor:
with a phase angle of:
D
u
u
s
-----
1
1
2
( )
2
2 ( )
2
+
--------------------------------------------------
= =
2
1
2
--------------
\ .
| |
atan =
0.0 1.0 2.0 3.0
0.0
1.0
2.0
3.0
4.0
0 =
0.2 =
0.5 =
0.7 =
1 =
D
0.0 1.0 2.0 3.0

0.0
90.0
180.0
0 =
0.2 =
0.7 =
1 =
0.5 =
Dynamic Analysis Multiple Degree of Freedom System
Multiple Degree of Freedom System
interpolation functions:
basic equations:
with:
stiffness matrix
mass matrix
consistent or lumped
damping matrix
Rayleigh
u x t , ( ) N x ( )u
t ( ) =
Mu
Cu
Ku
+ + F
=
K B
T
SB V d
V
K
i
+ =
M N
T
N V d
V
M
i
+ =
C M
t
--------
+
\ .
| |
K + =
Free Vibration Analysis Dynamic Analysis
Free Vibration Analysis
General solution
solve eigenvalue problem:
Power sweep method:
initial shift frequency
number of modes
auto shift parameter
Lanczos method:
frequency range
number of modes
Mu
Ku
+ 0 =
u
e
it
=
M
1
2
------
K
=
Dynamic Analysis Modal Superposition Method
Modal Superposition Method
Consider linear undamped system
with N eigenvectors ;
resulting set of uncoupled equations:
where
; ;
User selects number of participating modes modal damping
User specifies modal damping factors
Mu
Ku
+ P
t ( ) =
i
u
i
y
i
i 1 =
N
=
m
i
y
i
k
i
y
i
+ f
i
t ( ) =
m
i

i
T
M
i
= k
i

i
T
K
i
= f
i

i
T
P t ( ) =
m
i
y
i
c
i
y
i
k
i
y
i
+ + f
i
t ( ) =
c
i
2m
i
i
c
crit i ,

i
= =
Harmonic Analysis Dynamic Analysis
Harmonic Analysis
Equation of motion:
without damping:
with damping:
User must specify:
amplitude
frequency (range)
complex damping
Mu
Cu
Ku
+ + P
e
it
=
K
2
M [ ]u
=
K iC
2
M + [ ]
u
r
u
i
P
r
P
i
=
C C
el
C
d
K M 2K + + ( )
+ + =
Dynamic Analysis Transient Analysis
Transient Analysis
Equation of motion:
Finite Difference operators are use to directly integrate in
time. This leads to two basic apporaches, implicit and
explicit.
The Newmark-Beta (Implicit) operator matirx
is unconditionally stable but requires matrix inversion.
The Central Difference & Lumped mass (Explicit)
operator matrix
is fast but conditionally stable and depends on the Courant
stability limit.
Mu
Cu
Ku
+ + P
t ( ) =
4
t
2
--------
M
2
t
-----
C K + + u
n 1 +
history u
n
( ) =
M
t
2
--------
u
n 1 +
history u
n
( ) =
CHAPTER 8 User Subroutines
The user subroutine feature constitutes
one of the real strengths of the program,
allowing you to substitute your own
subroutines for several existing in the
program. This feature provides you with
a wide latitude for solving nonstandard
problems. These routines are easily
inserted into the program. When such
a routine is supplied, you are simply
replacing the one which exists in the
program file using appropriate
control setup.
User Subroutines When to use Subroutines or Special Routines
When to use Subroutines or Special Routines
User subroutines may be advantageous or even necessary to complete
your specific problem. User subroutines can be called during any stage
of the solution process as shown below:
Input
Assemble
Solve
Recover
Converge?
Output
Done?
More Input
No
Stop
Yes
Classification of User Subroutines User Subroutines
Modification of the input or output is very simple and useful. If you
require something special such as:
Very shell thickness by means of an analytical formula.
Special constitutive relations.
Automatically adjust loads depending upon the material
response, such as superplastic forming.
Shopping? See Volume E for 35 examples of:
...... Whatever, it is all in Volume D.
Classification of User Subroutines
1.0 Modify Input Geometry: UFXORD, UFCONN,
USHELL, UTHICK...
2.0 Modify Output Results: PLOTV, UPOSTV,
IMPD, ELEVAR....
3.0 Define Loads & BC: FORCEM, FLUX,
FORCDT, FILM....
4.0 Define Material: WKSLP, HOOKLW, HYPLA,
ANKOND...
5.0 General Routines: UBGINC, UEDINC,
USDATA, UELOOP, UELEM....
anelas ankond anplas crede crplaw film flow flux
forcdt forcem gapu hooklw hypela2 impd motion orient
plotv rebar sstran ubeam ubear ufconn uforms ufour
ufxord ugroov uinstr urpflo ushell ussd uthick utrans
uveloc vswell wkslp
User Subroutines Programing Conventions
Programing Conventions
Start with the user stubs in ~marc/user directory e.g. ~marc/user/
wkslp.f (or .for on NT) looks like:
subroutine wkslp(m,nn,kc,mats,slope,ebarp,eqrate,stryt,dt,ifirst)
implicit real*8 (a-h,o-z)
c* * * * * *
c stryt current value of yield strength
c slope work hardening slope
c ebarp equivalent plastic strain
c eqrate equivalent plastic strain rate
c dt temperature
c ifirst flag distinguishing tenth cycle properties for
c* * * * * *
return
end
The implicit real*8 is very important as well as the
variables in the call list of the subroutine.
It is a good idea to test the program with your own main routine before
running it with MSC.Marc to verify that your programing is doing
what is required. For wkslp the test would look like:
do 1 i=1,10
ebarp=float(i)/1.0d0
call wkslp(1,1,1,1,slope,ebarp,0.0d0,stryt,0.0d0,1)
write(6,*) ebarp,stryt,slope
1 continue
end
This will supply your wkslp with values of the plastic strain
and print the return variables.
Geometry Modifications Subroutines User Subroutines
Geometry Modifications Subroutines
User
Subroutine
Required Parameters or
Model Definition Options
Purpose
MAP2D MESH2D
MAPPER
Define the coordinates of key boundary nodes for
mesh generation.
REBAR ELEMENTS
(rebar element types)
Define the orientation and effective thickness of
the elements.
SSTRAN SUBSTRUCTURE Define the transformation matrix applied to
a superelement.
UACTIVE Activate or deactivate elements.
UACTUAT GEOMETRY Define the length of the actuator element.
UADAP ADAPTIVE (parameter)
ADAPTIVE (model definition
option)
Define a user-defined error criteria for
adaptive meshing.
UCOORD ADAPTIVE (parameter)
ADAPTIVE (model definition
option)
Describe of the location of newly created nodes.
UFCONN UFCONN Modify the connectivity of an element.
UFRORD REZONING
REZONE
UFRORD
Modify the coordinates of a node during rezoning.
UFXORD UFXORD Modify the initial nodal coordinates.
UPNOD R-P FLOW Update the nodal coordinates in a rigid plastic
analysis using the Eularian procedure.
URCONN UFCONN Modify the connectivity of an element
during rezoning.
USHELL Define the integration point thickness for
shell elements.
UTHICK NODAL THICKNESS Define the initial thickness at the nodes for
shell elements.
UTRANS UTRANFORM Define a transformation to be applied to the degrees
of freedom at a node.
Popular
User Subroutines Output Quantities Subroutines
Output Quantities Subroutines
Loads and Boundary Conditions Subroutines
User
Subroutine
Purpose
ELEVAR UDUMP Allows postprocessing of element results.
ELEVEC UDUMP Allows postprocessing of element results in
harmonic analysis.
INTCRD Makes available integration point coordinates.
IMPD UDUMP Allows postprocessing of nodal vector results.
PLOTV POST Defines element quantity to be written to post file.
UBGINC Dummy routine available at the beginning of
each increment.
UBGITR Dummy routine available at the beginning of
each iteration.
UEDINC Dummy routine available at the end of
each increment.
UELOOP Dummy routine available during major
element loops.
UPOSTV POST Defines nodal vecors that are to be written to a
post file.
User
Subroutine
Purpose
CREDE THERMAL LOADS Definition of state variable including temperature.
CUPLFX COUPLE
DIST FLUXES (flux type 101)
Heat generated due to inelastic behavior in
coupled analysis.
DIGEOM CONTACT (2D)
CONTACT (3D)
Definition of rigid surface.
FILM HEAT or COUPLE
FILMS (Model Definition)
FILMS (History Definition)
Definition of convective heat transfer coefficient and
sink temperature.
FLOW HEAT
CHANNEL
Definition of mass flow rate.
FLUX HEAT or COUPLE
DIST FLUXES (Model Definition)
DIST FLUXES (History Definition)
Definition of distributed flux.
FORCDF FORCDT
FIXED DISP or
DISP CHANGE
Definition of point load or kinematic boundary
condition in a harmonic analysis.
Loads and Boundary Conditions Subroutines User Subroutines
FORCDT FORCDT
FIXED DISP or DISP CHANGE
FIXED TEMPERATURE or
TEMP CHANGE
Definition of point load or prescribed displacement in
stress analysis. Definition of point flux or prescribed
temperature in heat transfer analysis.
FORCEM DIST LOADS (Model Definition)
DIST LOADS (History Definition)
Definition of distributed load.
GAPT HEAT
CONRAD GAP
Definition of thermal contact gap temperature.
INITPL INITIAL PLASTIC STRAIN Definition of initial plastic strain.
INITPO PORE
INITIAL PORE
Definition of initial pore pressure in a uncoupled
soil analysis.
INITSV INITIAL STATE Definition of initial values of state variables.
MOTION CONTACT (2D)
CONTACT (3D)
UMOTION
MOTION CHANGE (History
Definition)
Definition of velocity of rigid surfaces.
NEWPO PORE
CHANGE PORE (Model Definition)
CHANGE PORE (History
Definition)
Change pore pressure in an uncoupled soil analysis.
NEWSV CHANGE STATE (Model
Definition)
CHANGE STATE (History
Definition)
Change value of the state variable.
SEPFOR CONTACT (2D)
CONTACT (3D)
Definition of force required for separation.
SEPSTR CONTACT (2D)
CONTACT (3D)
Definition of stress required for separation.
UCONTACT CONTACT (2D)
CONTACT (3D)
UCONTACT
Allow user-defined contact procedure.
UFORMS TYING Definition of user-defined constraint matrices.
UFOUR FOURIER Definition of function giving nonuniform variation
about the circumference in Fourier analysis.
UFRIC CONTACT (2D)
CONTACT (3D)
UFRICTION
Definition of friction coefficient.
UHTCOE CONTACT (2D)
CONTACT (3D)
UHTCOEF
Definition of heat transfer coefficient to environment
for coupled contact analysis.
UHTCON CONTACT (2D)
CONTACT (3D)
UHTCON
Definition of heat transfer coefficient between bodies
in contact in coupled analysis.
User
Subroutine
Purpose
User Subroutines Loads and Boundary Conditions Subroutines
UINSTR ISTRESS Definition of initial stress.
UNORST CONTACT (2D)
CONTACT (3D)
USER
Definition of normal stress for user elements
in contact.
USDATA USDATA Definition of user-definer constants.
USINC INITIAL DISP
INITIAL VEL
INITIAL TEMP
Definition of initial displacement, initial velocity, or
temperature.
USSD DYNAMIC CHANGE
RESPONSE SPECTRUM
Definition of spectrum displacement
density function.
UVELOC HEAT Definition of convective velocities.
User
Subroutine
Purpose
Constitutive Relations Subroutines User Subroutines
Constitutive Relations Subroutines
User
Subroutine
Purpose
ANELAS ORTHOTROPIC or
ANISOTROPIC
Definition of factors to scale elastic stress
strain law.
ANEXP ORTHOTROPIC or
ANISOTROPIC
Definition of thermal strain increment.
ANKOND ORTHOTROPIC or
ANISOTROPIC
Definition of thermal conductivity or electrical
resistance in Joule heating.
ANPLAS ORTHOTROPIC or
ANISOTROPIC
Definition of parameters for Hill yield criteria
CRPLAW CREEP Definition of function to describe creep strain rate.
GAPU GAP DATA Definition of contact gap closure distance
GENSTR SHELL SECT Definition of generalized stress-strain law for
shells.
HOOKLW ORTHOTROPIC or
ANISOTROPIC
Definition of elastic stress-strain or
compliance relation.
HYPELA HYPOELASTIC Definition of nonlinear stress-strain relationship.
HYPELA2 HYPOELASTIC Definition of nonlinear stress-strain relationship.
ORIENT ORIENTATION Definition of preferred material orientation for
orthotropic or anisotropic behavior.
TENSOF ISOTROPIC
CRACK DATA
Definition of tension softening modulus.
UBEAM HYPOELASTIC Definition of nonlinear generalized stress-strain
law for element types 52 or 98.
UCOMPL HARMONIC Definition of stress-strain rate relationship for
harmonic analysis.
UCRACK ISOTROPIC
CRACK DATA
Definition of ultimate stress for cracking analysis.
UELDAM OGDEN
DAMAGE
Definition of damage parameters for Ogden
rubber model.
UENERG MOONEY Definition of strain energy function.
UEPS ELECTRO or
EL-MA
ORTHOTROPIC
Definition of anisotropic electrical permittivity.
UFAIL FAIL DATA Definition of composite failure criteria.
UMOONEY MOONEY Definition of temperature dependent
Mooney-Rivlin constants.
UMU MAGNETO or
EL-MA
ORTHOTROPIC
Definition of anisotropic magnetic permeability.
UNEWTN R-P FLOW or
FLUID
Definition of material viscosity.
User Subroutines Constitutive Relations Subroutines
UOGDEN OGDEN Definition of Ogden material parameters.
UPERM PORE Definition of soil permeability.
UPHI HARMONIC
MOONEY
PHI-COEFFICIENTS
Definition of phi coefficients for rubber-viscoelastic
harmonic analysis.
UPOWDR POWDER Definition of powder material data.
UPSTRECH ODGEN Definition of generalized principal stretch based
elasticity models.
URPFLO R-P FLOW Definition of yield surface for rigid plastic flow.
USELEM USER Definition of consistent nodal loads,
mass matrix, stiffness matrix, and residuals
for user-defined element.
USHRET ISOTROPIC
CRACK DATA
Definition of shear retention factor for elements
that have cracks.
USIGMA EL-MA Definition of anisotropic electrical conductivity.
USPCHT HEAT or
COUPLE or
FLUID
Definition of specific heat.
USPRNG SPRINGS or
FOUNDATION
Definition of nonlinear spring or
foundation stiffness.
UVOID DAMAGE Definition of initial void fraction for Gurson
damage model.
UVOIDN DAMAGE Definition of void nucleation for Gurson
damage model.
VSWELL CREEP Definition of volumetric swelling.
WKSLP ISOTROPIC or
ORTHOTROPIC or
ANISOTROPIC
WORK HARD
Definition of work hardening or strain
hardening data.
User
Subroutine
Purpose
Viscoplasticity Subroutines User Subroutines
Viscoplasticity Subroutines
Viscoelasticity Subroutines
User
Subroutine
Purpose
ASSOC ISOTROPIC GEN-PLAST Definition of the direction of incremental plastic
strain in generalized plasticity model.
CRPLAW CREEP Definition of inelastic strain rate for explicit
viscoplasticity model.
NASSOC CREEP Definition of direction of incremental viscoplastic
strain for explicit viscoplasticity model.
SINCER ISOTROPIC GEN-PLAST Definition of fraction of increment which is elastic for
generalized plasticity model.
UVSCPL CREEP
ISOTROPIC VISCO-PLASTIC
Definition of inelastic strain rate for implicit
viscoplastic model.
YIEL ISOTROPIC,
ORTHOTROPIC or
ANISOTROPIC
Definition of yield stress.
ZERO ISOTROPIC,
ORTHOTROPIC or
ANISOTROPIC
Definition of equivalent stress.
User
Subroutine
Purpose
CRPVIS VISCO ELAS Definition of generalized Kelvin model using explicit
procedure.
HOOKVI VISCELORTH Definition of anisotropic viscoelastic material law for
a particular relaxation time.
TRSFAC VISCELPROP or
VISCELORTH or
VISCELMOON or
VISCELOGDEN and
SHIFT FUNCTION
Definition of shift function for thermo-rheologically
simple material.
User Subroutines Hydrodynamic Lubrication Subroutines
Hydrodynamic Lubrication Subroutines
Special Routine PLDUMP
This routine is a stand-alone program that provides examination of the
postprocessing file created by the POST option. This allows you to
perform additional calculations based upon results calculated in
MSC.Marc, and to create a post file. These results can then be viewed
with MSC.Marc Mentat.
User
Subroutine
Purpose
UBEAR BEARING Define the orientation of the film surface.
UGROOV BEARING Define the groove depth.
URESTR BEARING
RESTRICTOR
Define the nonuniform restrictor coefficient and
pump pressures.
UTHICK BEARING
NODAL THICKNESS
THICKNS CHANGE
Define the lubricant thickness.
UVELOC BEARING
VELOCITY
Define the nodal velocity of bearing surface.
ELMVAR Utility Routine User Subroutines
ELMVAR Utility Routine
To facilitate extraction of solution results, it is possible to use
subroutine ELMVAR. This routine can be called from any user
subroutine that is within an element loop. This routine is used in
conjunction with the MSC.Marc post element post codes to return the
calculated values to you.
This routine is called with the following header:
CALL ELMVAR (ICODE,M,NN,KC,VAR)
where:
This subroutine can be called from user subroutines:
ICODE is the post code.
M is your element number.
NN is the integration point number.
KC is the layer number.
VAR is the current value(s) of the items requested.
ANELAS FILM NEWSV UCRACK UNEWTN USIGMA
ANEXP FLUX ORIENT UELDAM UOGDEN USPCHT
ANKOND FORCEM PLOTV UELOOP UPERM UVOIDN
ANPLAS GENSTR REBAR UENERG UPOWDR UVSCPL
ASSOC HOOKLW SINCER UEPS UPSTRECH VSWELL
CRPLAW HOOKVI TENSOF UFAIL URESTR WKSLP
CRPVIS HYPELA TRSFAC UHTCOE URPFLO YIEL
CUPFLX HYPELA2 UACTIVE UHTCON USELEM ZERO
ELEVAR INTCRD UADAP UMOONY USHELL
ELEVEC NASSOC UCOMPL UMU USHRET
User Subroutines ELMVAR Utility Routine
Example
Suppose you would like the plastic strain tensor from within user
subroutine UADAPT for a user-defined adaptive meshing criteria. In
this example, there are no shell elements, so KC=1 and the number of
integration points per element = 4, so INTEL=4. The plastic strain
tensor is code 321. The plastic strains are stored in a local array
EPTEN. You could create the following routine:
SUBROUTINE UADAP
(M,XORD,DSXT,NCRDMX,NDEGMX,LM,NNODE,USER)
IMPLICIT REAL *8 (A-H, O-Z)
DIMENSION XORD(NCRDMX, *),DSXT(NDEGMX, *),LM(*)
DIMENSION EPTEN (6,28)
KC=1
INTEL=4
ICODE=321
DO NN=1,INTEL
CALL ELMVAR(ICODE,M,NN,KC,EPTEN(1,NN))
ENDDO
USER CODE TO DEFINE TASK
RETURN
END

Submit MSC.Marc with User Subroutines User Subroutines
Submit MSC.Marc with User Subroutines
All subroutines will go into one file, say model1.f and upon proper
submission the subprograms will be automatically compiled and linked
to create your own executable called model1.marc. Lets suppose that
one wants to use subroutine WKSLP to prescribe the work hardening
slope for plastic behavior.
In MSC.Mentat select the
user subroutine from the
materials menu as shown.
This will write the work
hardening option to tell
MSC.Marc to call WKSLP
when MSC.Marc runs.
Then identify the file
model1.f containing the
subroutine WKSLP. When
MSC.Mentat submits the
job you can select one of
three modes below,
choosing either 1 or 2
before 3.
From Mentat From UNIX
1.) Compile/No Save run_marc -user model1 ...
2.) Compile/Save run_marc -user model1 -save y
3.) Run Saved Executable run_marc -prog model1 ...
User Subroutines MSC.MARC User Subroutine Example
MSC.MARC User Subroutine Example
This example will use a user subroutine that will model a temperature
dependent plastic material behavior. Lets assume that the
temperature dependent yield surface can be expressed as:
Of course, an actual example this surface would usually be a two-
dimensional table of values experimentally determined and subroutine
WKSLP would use the actual data, interpolating for the proper value
of the current yield strength based upon the current values of the state
variables, , and . Example features include:
1.) The user subroutine WKSLP.
2.) Testing the routine with a main program that will write
a MSC.Marc Mentat procedure file to draw the yield
surface. Because the yield surface is two-dimensional
this is a good idea, and it shows how to take advantage of
MSC.Marc Mentat to help.
3.) Running a one element model to check overall response.
This is a very common procedure when modeling
nonlinear material properties and many times this
simple checkout model is the actual specimen that was
used to observe the material data.
y
H
p
T
0
, ( )R T ( )
2 20T 10 ( ) atan
2 10 ( ) atan
----------------------------------------------
1
p
( ) log
10
+ ( ) = =
p
T
MSC.MARC User Subroutine Example User Subroutines
User Subroutine WKSLP
First, lets write the routine by getting the stub in the
~marc/user directory, then adding to it.
c* * * * * *
c dt temperature
c* * * * * *
parameter (e=2.718282d0)
z (x,y)=(2.0d0-atan(20.0d0*x-10.0d0))/(2.0d0-atan(-10.0d0))*
1 (1.0d0+log10(1.0d0+4*y))
zp(x,y)=(2.0d0-atan(20.0d0*x-10.0d0))/(2.0d0-atan(-10.0d0))*
1 (log10(e))*4.0d0/(1.0d0+4*y)
c* * * * * *
x=dt
y=ebarp
stryt=z(x,y)
slope=zp(x,y)
if(m.eq.1) write(6,*) dt,ebarp,stryt,slope,dt,ebarp,stryt,slope
return
end
It is very important to note that the return variable slope is
the rate of change of the yield surface with respect to plastic
strain NOT total strain.
Also, writing out results to the output file is useful to verify
that the routine has been called and performs accordingly.
Test Routine with a Main Program
Now to test the routine. Lets construct a main program to
call WKSLP with the temperature and plastic strain
changing over the expected range, in this case, both range
from 0 to 1. The test program will write a MSC.Marc
Mentat procedure file called example.proc that will plot
the surface defined by WKSLP.
idiv=21
open (unit=16,file=example.proc)
write(16,*) *set_curve_type polyline
do 1 i = 1,idiv
write(16,*) *add_curves
x=float(i-1)/float(idiv-1)
do 2 j = 1,idiv
y=float(j-1)/float(idiv-1)
call wkslp(1,1,1,1,slope,y,0.0d0,z,x,1)
write(16,100) x,y,z
100 format( point(1e15.6,1e15.6,1e15.6))
2 continue
write(16,*) # | End of List
1 continue
do 3 k = 1,2
y=k-1
do 4 j = 1,idiv
x=float(j-1)/float(idiv-1)
call wkslp(1,1,1,1,slope,y,0.0d0,z,x,1)
write(16,100) x,y,z
4 continue
3 continue
WKSLP is called two times to establish construction curves
upon which to build the yield surface.
The remaining portion of the test program trims to the
coordinate planes (x=y=z=0) and builds all surfaces from
the curves.
z=0
write(16,*) *set_curve_type line
write(16,100) 0,0,z
write(16,100) 1,0,z
write(16,100) 0,0,z
write(16,100) 0,1,z
write(16,100) 0,1,z
write(16,100) 1,1,z
write(16,100) 1,0,z
write(16,100) 1,1,z
write(16,*) *set_surface_type ruled
do 6 k=1,idiv-1
write(16,*) *add_surfaces
write(16,*) k, k+1
6 continue
write(16,*) *set_surface_type ruled
write(16,*) 1, idiv+4
write(16,*) idiv+1, idiv+3
write(16,*) idiv, idiv+6
end
The surface can now be displayed in MSC.Marc Mentat
by executing the procedure file example.proc. The
lighter or blue surface is the yield surface where the
z-direction is the strength, the x-direction is the
temperature and the y-direction is the plastic strain.
Run a One Element Model
The one element test case will be an axisymmetric disc with both ends
axially constrained that is heated and cooled. The disc will have steel
properties and the plasticity will be defined by subroutine WKSLP.
This yield surface will be the same shape but magnified by the actual
initial yield of 36ksi, and the actual temperature range will be from 0
to 2000
o
F. The model1.f file containing WKSLP is show below.
c* * * * * *
c dt temperature
c* * * * * *
parameter (e=2.718282d0)
z (x,y)=(2.0d0-atan(20.0d0*x-10.0d0))/(2.0d0-atan(-10.0d0))*
1 (1.0d0+log10(1.0d0+4*y))
zp(x,y)=(2.0d0-atan(20.0d0*x-10.0d0))/(2.0d0-atan(-10.0d0))*
1 (log10(e))*4.0d0/(1.0d0+4*y)
c* * * * * *
x=dt/2000.0d0
y=ebarp
yinit=36.0d3
stryt=z(x,y)*yinit
slope=zp(x,y)*yinit
if(m.eq.1) write(6,*) dt,ebarp,stryt,slope,dt,ebarp,stryt,slope
return
end
Upon completing the analysis, we will want to track the response and
show it on the yield surface.
Results from One Element Model
The results from the axisymmetric disc show that all is working. Here,
the Equivalent von Mises Stress is plotted against the temperature.
There is an elastic loading until yield, plastic flow until the end of the
heating cycle, then reverse yielding upon cooling leaving a large tensile
residual stress at the end of cooling.
Results from One Element Model
Here, the Equivalent von Mises Stress is plotted against the total
equivalent plastic strain. These two plots show how the response tracks
the yield surface as projected on the x=y=0 planes.
Elastic
Nothing is like 3D
The stress-strain-temperature response of the disc is shown as a red
(black) on the yellow (gray) yield surface.
Every point MUST TRACK the CONSTITUTIVE relation.
What would happen if we increase the thermal coefficient
of expansion or ran more heating/cooling cycles?
CHAPTER 9 Automatic Mesh Refinement
Because the results of FEA depend upon
the mesh, one always wants to generate a
better one. With this in mind this chapter
will cover local refinement, called
adaptive meshing.
Automatic Mesh Refinement Adaptive Meshing
Adaptive Meshing
Adaptive meshing is a automatic meshing strategy to improve results
by adding elements in areas that meet a specific error criteria. This
strategy can improve results.
Many practical problems:
Engineering
problem
F.E.M. solution
Exact solution
Differences in:
* displacements
* stresses
* temperatures
* etc.
Mesh 1
Mesh 2
Mesh 3
(finer than 1)
(finer than 2)
Solution 1
Solution 2
Solution 3
2
(<
1
)
etc.
3
(<
2
)
Example: Clamped Cantilever Beam Automatic Mesh Refinement
Example: Clamped Cantilever Beam
Geometry: , ,
Material: ,
Load:
Solutions according to beam theory:
, with ,
hence
hence
A
B
F
L
t
h
x
y
L 100 = h 10 = t 2 =
E 2 10
5
= 0 =
F 100 =
w
A
exact FL
3
3EI
---------
6FL
5Ght
-------------
+ = I
th
3
12
-------
= G
E
2 1 + ( )
---------------------
=
w
A
exact
1.006 =
xx
B
exact FLh
2I
----------
=
xx
B
exact
300 =
Automatic Mesh Refinement Example: Clamped Cantilever Beam
F.E. solution (Marc element type 3 with assumed strain formulation,
load applied via distributed shear load) uniform refinement
mesh 1: 4x1 elements, 20 d.o.f.
,
mesh 2: 8x2 elements, 54 d.o.f..
,
mesh 3: 16x4 elems, 170 d.o.f..
,
Define errors: ,
w
A
FE
0.989 =
xx
B
FE
262.188 =
w
A
FE
1.001 =
xx
B
FE
281.084 =
w
A
FE
1.005 =
xx
B
FE
292.113 =
w
w
A
exact
w
A
FE
w
A
exact
-------------------------------
=
xx
B
exact
xx
B
FE
xx
B
exact
------------------------------
=
10.0 100.0
degrees of freedom
0.00
0.05
0.10
0.15
e
r
r
o
r

displacement
stress
Example: Clamped Cantilever Beam Automatic Mesh Refinement
F.E. solution (Marc element type 3 with assumed strain formulation,
load applied via distributed shear load) locally modified mesh
mesh 1: 4x1 elements, 20 d.o.f.
,
mesh 2: 9 elements, 36 d.o.f..
,
mesh 3: 15 elems, 54 d.o.f..
,
Define errors: ,
w
A
FE
0.989 =
xx
B
FE
262.188 =
w
A
FE
0.995 =
xx
B
FE
281.084 =
w
A
FE
0.996 =
xx
B
FE
292.451 =
w
w
A
exact
w
A
FE
w
A
exact
-------------------------------
=
xx
B
exact
xx
B
FE
xx
B
exact
------------------------------
=
10.0 100.0
degrees of freedom
0.00
0.05
0.10
0.15
e
r
r
o
r

w (regular mesh)
(regular mesh)
w (modified mesh)
(modified mesh)
Automatic Mesh Refinement Error Criteria for Linear Analyses
Error Criteria for Linear Analyses
A Finite Element mesh is called sufficiently fine when an error
criterion is satisfied (or when multiple error criteria are satisfied)
Linear adaptive stress analysis:
sufficiently
Input
Set up
Ku
=
Determine
u
K
1
f
=
Calculate
strains and
stresses
Mesh
fine?
Stop
Modify
FE mesh
No
Yes
Error Estimates Mentioned in Literature Automatic Mesh Refinement
Error Estimates Mentioned in Literature
estimates based on residuals
estimates according to Zienkiewicz and Zhu
estimates based on strain energy
estimates based on geometrical considerations
estimates based on invariants of stress and strain tensors
MSC.Marc Requirements
suitable for various material laws
suitable for linear and nonlinear analyses
suitable for various element types
relatively easy to implement
Error Criteria Available in MSC.Marc
(for linear analyses and lower-order continuum and shell elements):
Automatic Mesh Refinement Error Criteria Available in MSC.Marc
Classical:
Strain energy
Zienkiewicz and Zhu (slightly different from most
applications in recent literature)
Equivalent stress or strain
Heuristic:
Location within box (not very useful within a
linear analysis)
Solution gradient (heat transfer analysis)
User criterion
Strain Energy Criterion Automatic Mesh Refinement
Strain Energy Criterion
Determine total strain energy by summing element contributions:
with: : vector with stress components : vector with strain
components (shear terms equal twice the tensor
components) : domain of element , : number of
elements for which the current criterion is active. The
average strain energy is evaluated according to:
For element the error criterion is not fulfilled if:
with: : user-defined value
Remarks:
Typical value for :
An FE mesh is considered to be optimal if all elements
have the same amount of strain energy (this implies that
for a constant stress patch test some elements may not
fulfil the error criterion)
E
total
E
i
i 1 =
N
adapt
1
2
---
)
`

i 1 =
N
adapt
= =
i
i N
adapt
E
average
E
total
N
adapt
---------------
=
i
E
i
f
1
E
average
> f
1
f
1
1.5 f
1
2 < <
Automatic Mesh Refinement Zienkiewicz-Zhu (Stress)
Zienkiewicz-Zhu (Stress)
A measure for the error in the stresses is given by:
with:
: exact stress solution
: FE stress solution
Pointwise specification of errors is difficult. Hence, an
integral measure is preferable. In MSC.Marc, the norm
has been implemented:

Since is generally unknown, instead of an
improvement of the FE solution is used:
and
e
exact
exact
L
2
e
2
e
T
e
= e
i
2
d
)
`

i 1 =
N
adapt
exact
exact
= e
*
2
e
*T
e
*
d
=
Zienkiewicz-Zhu (Stress) Automatic Mesh Refinement
A projection technique is applied to calculate :
Interpolate by the same shape function as the
displacements:
Introduce a weighted residual requirement:
hence:
where is evaluated in a lumped form
*
N
*
=
N
T
( )d
0 =
*
N
T
Nd
)
`

1
N
T
=
N
T
Nd
Automatic Mesh Refinement Zienkiewicz-Zhu (Stress)

Example: Stress Projection (MSC.Marc Element 6)
1 2
3 4
1
4
2
I
II

xx
I
100 =
xx
II
50 =
x
y
thickness = 1
N
T
Nd
1
12
----- -
6 1 3 2
1 2 1 0
3 1 6 2
2 0 2 4
1
12
----- -
12 0 0 0
0 4 0 0
0 0 12 0
0 0 0 8
=
N
T
1
3
---
200
100
200
100
=
*
,
1 0 0 0
0 3 0 0
0 0 1 0
0 0 0 3 2
1
3
---
200
100
200
100
200 3
100
200 3
50
= =
Zienkiewicz-Zhu (Stress) Automatic Mesh Refinement
Global Error Norm
Now for element the error criterion is not fulfilled
if:
and
with: , , :
user-defined values
Remarks:
Typical value for :
Default values:
Mesh is considered to be optimal if all elements have the
same contribution to the global error
Factor can be used to emphasize the global error, for
which is a measure (typically )
*
2
i 1 =
N
adapt
*
2
+
------------------------------------------------------ =
i
f
1
>
e
*
i
2
e
*
2
N
adapt
---------------
f
2
f
3
f
1
----
+
\ .
| |
> f
1
f
2
f
3
f
1
0.05 f
1
0.20 < <
f
2
1 = f
3
0 =
f
3
f
1
f
2
f
3
+ 1
Automatic Mesh Refinement Zienkiewicz-Zhu (Strain Energy)
Zienkiewicz-Zhu (Strain Energy)
Similar to previous stress based criterion, but with strain energy
instead of stress. Hence:
with:
: exact strain energy
: FE strain energy
Integral measure:
Approximate by :
and
Projection technique:
with
e
E
E
exact
E =
E
exact
E
e
E
2
e
E
2
d
=
E
exact
E
*
e
E
*
E
*
E = e
E
*
2
e
E
*2
d
=
E
*
NE
*
= E
*
N
T
Nd
)
`

1
N
T
Ed
=
Zienkiewicz-Zhu (Strain Energy) Automatic Mesh Refinement
Global error norm:
For element , the error criterion is not fulfilled if:
and
with:
, , : user-defined values
Remarks similar to previous stress criterion with
instead of and instead of .
e
E
*
2
E
2
d
i 1 =
N
adapt
e
E
*
2
+
------------------------------------------------- =
i f
1
>
e
E
*
i
2
e
E
*
2
N
adapt
---------------
f
4
f
5
f
1
----
+
\ .
| |
>
f
1
f
4
f
5
f
4
f
2
f
5
f
3
Automatic Mesh Refinement Equivalent Stress/Strain
Equivalent Stress/Strain
Determine maximum equivalent stress/strain:
(relative) or: (absolute)
or:
with , , and : user-defined values
Remarks:
Typical values for to are problem dependent
An optimal mesh does not exist
Criterion may be useful for problems with a single
stress/strain concentration
max
eq
max
max
i
eq
( i 1 N
adapt
, ) = =
max
eq
max
max
i
eq
( i 1 N
adapt
, ) = =
i
max
i
eq
f
1
max
eq
>
max
i
eq
f
2
>
max
i
eq
f
3
max
eq
>
max
i
eq
f
4
>
f
1
f
2
f
3
f
4
f
1
f
4
Location within Box Automatic Mesh Refinement
Location within Box
For element , the error criterion is not fulfilled if one of its
nodes in the deformed configuration lies within a user-
specified box, defined with respect to the global Cartesian
-, -and -coordinates:
with:
: original coordinates
: total displacements
: any node of element
to : user-defined values of box boundaries
Remarks:
Criterion may be used to manually obtain a locally
refined FE mesh
i
X Y Z
f
1
X U + ( )
i
node
j
f
2
< <
f
3
Y V + ( )
i
node
j
f
4
< <
f
5
Z W + ( )
i
node
j
f
6
< <
X Y Z , ,
U V W , ,
node
j
i
f
1
f
6
Automatic Mesh Refinement Solution Gradient (Heat Transfer Analysis)
Solution Gradient (Heat Transfer Analysis)
Determine maximum gradient of temperature :
(relative) or: (relative)
or:
or:
(absolute) or: (absolute)
with: , : user-defined values
Remarks:
Typical value for :
An optimal mesh only exists for problems with (nearly)
homogeneous temperature gradients
T
gr
max
max
T
X
------
i
T
Y
------
i
T
Z
------
i
, ,
\
|
i 1 N
adapt
,
.
|
= =
i
T
X
------
i
f
1
gr
max
>
T
Y
------
i
f
1
gr
max
>
T
Z
------
i
f
1
gr
max
>
T
X
------
i
f
2
>
T
Y
------
i
f
2
>
T
Z
------
i
f
2
>
f
1
f
2
f
1
f
1
0.75 =
User Criterion Automatic Mesh Refinement
User Criterion
Implemented via user subroutine UADAP
User selects a variable which maximum is calculated
by MSC.Marc:
with: , : user-defined values
Remark:
and are entered via input file
V
user
V
max
user
max V
i
user
( i 1 N
adapt
, ) = =
i
V
i
user
f
1
V
max
user
> V
i
user
f
2
>
f
1
f
2
f
1
f
2
Automatic Mesh Refinement Remeshing Strategy for Continuum Elements
Remeshing Strategy for Continuum Elements
If an error criterion is not fulfilled, an FE mesh can be adapted
according to:
r-method: change the location of nodal points
p-method: change the order of the element shape functions
h-method: change the element size (global remeshing or
local mesh refinement)
Possible Subdivisions for Local Mesh Refinement Automatic Mesh Refinement
Possible Subdivisions for Local Mesh Refinement
+: no conflicts with adjacent elements
-: aspect ratio and skewness worsen
+:aspect ratio and skewness do not change
-: multipoint constraint (tying) equations must be set
up at interfaces of elements with different
refinement level
Available
in
MSC.Marc
Automatic Mesh Refinement Multipoint Constraint Equations
Multipoint Constraint Equations
2D:

with: : vector with nodal degrees of freedom
3D:
, ,
, ,
1
2
3
u
3
1
2
---
u
1
u
2
+ ( ) =
u
1
2
3
4
5
6
7
8
9
u
5
1
2
---
u
1
u
2
+ ( ) = u
6
1
2
---
u
2
u
3
+ ( ) =
u
7
1
2
---
u
3
u
4
+ ( ) = u
8
1
2
---
u
4
u
1
+ ( ) =
u
9
1
4
---
u
1
u
2
u
3
u
4
+ + + ( ) =
Multipoint Constraint Equations Automatic Mesh Refinement
In order to obtain some smoothness of the FE solution, the
difference between refinement levels of neighboring elements will
never be larger than one (one-level-rule):
The one-level-rule will also be applied to elements for
which no adaptive criterion has been activated. Increase
of number of elements:
Hence: maximum number of refinement levels must be
chosen carefully
level QUAD4 TRIA3 HEX8 TET4
1 4 4 8 8
2 16 16 64 64
3 64 64 512 512
4 256 256 4096 4096
refined due to
one-level-rule
Automatic Mesh Refinement Multipoint Constraint Equations
Boundary Conditions
Transformations and tyings on master nodes are not
applied to newly created nodes.
Element sets are automatically extended and available for
postprocessing.
For a Linear Adaptive Analysis Automatic Mesh Refinement
For a Linear Adaptive Analysis
estimate in final mesh number of nodes, number of
elements and number of boundary conditions
especially for error criteria where an optimal mesh does not
exist, limit the maximum number of refinement levels
choose reasonable SIZING to make in-core solution
possible
activate the ELASTIC parameter option for a stress
analysis; perform a STEADY STATE analysis in case of a
heat transfer analysis
error criteria can only be activated for linear continuum
elements (plane stress, plane strain, axisymmetric, 3D) and
shell element 75
J-integrals, modal dynamics and element-by-element
iterative solver are not supported
Automatic Mesh Refinement Important MSC.Marc Mentat Options
Important MSC.Marc Mentat Options
for linear adaptive analysis:
JOBS
ADAPTIVE MESHING
CRITERIA
e.g.
MEAN STRAIN
ENERGY
Enter VALUE,
MAX # LEVELS
Add ELEMENTS
JOBS
JOB
PARAMETERS
ADAPTIVE
MESHING
e.g.
MECHANICAL
Enter upper bounds for ELEMENTS,
NODES, FIXED DOFs
Improved Geometry Description Automatic Mesh Refinement
Improved Geometry Description
Steps in an FE analysis:
Especially for coarse meshes, the actual geometry may be
poorly represented.
Engineering
problem
Problem
idealization
FE
discretization
Automatic Mesh Refinement Improved Geometry Description
By default, in an adaptive analysis, the position of newly created nodes
is straightforwardly derived from the position of the master nodes:
initial bad representation of geometry is not improved
Improvement can be obtained by taking into
account geometrical information: attach nodes to
geometrical entities
master
nodes
master nodes
geometry
attached to
geometry
Attaching Nodes to Geometrical Entities Automatic Mesh Refinement
Attaching Nodes to Geometrical Entities
2D Entities Available:
polyline
circular arc
circle
NURB curve (full description)
NURB curve (internally generated)
3D Entities Available:
plane
sphere
cylinder/cone
NURB surface (full description)
NURB surface (internally generated)
Note: geometrical entities created by MSC.Marc Mentat
are always written as NURB entities with full description
Automatic Mesh Refinement Attaching Nodes to Geometrical Entities
Attaching Newly Created Nodes
If the master nodes of an element edge have been attached
to a geometrical entity, the newly created edge node will
also be attached to this geometrical entity.
If, in a 3D analysis, all nodes of shell element 75 or all nodes
of the face of a tet4 or hex8 element have been attached to a
geometrical entity, all newly created nodes corresponding
to the shell element or the tet4 or hex8 face will be attached
to this geometrical entity.
If, in a 3D analysis, the master nodes of an element edge
have been attached to two geometrical entities, the newly
created edge node will be positioned on the intersection of
the geometrical entities.
Example: Cylinder-cylinder Intersection Automatic Mesh Refinement
Example: Cylinder-cylinder Intersection
Initial FE mesh:
X Y
Z
4
nodes attached
to small cylinder
nodes attached
to big cylinder
nodes attached
to both cylinders
elements to
be refined
Automatic Mesh Refinement Example: Cylinder-cylinder Intersection
FE mesh after 3 refinement levels:
X Y
Z
cylinder-cylinder intersection
INC : 0
SUB : 3
TIME : 0.000e+00
FREQ : 0.000e+00
4
gaps due to attaching
nodes to the cylinder
Modified Remeshing Strategy for Shell Elements Automatic Mesh Refinement
Modified Remeshing Strategy for Shell Elements
Homogeneous bending deformation with 4-noded plate/shell
elements (MARC element 75):
problem description:
deformed configuration (enlarged):
0 =
all degrees
of freedom
suppressed
FE mesh I
FE mesh II
x
y
x
z
analytical
FE mesh II
FE mesh I
Automatic Mesh Refinement Modified Remeshing Strategy for Shell Elements
What to do at interfaces of elements with different
refinement levels?
in-plane displacements must be tied linearly
transverse displacement cannot be tied linearly and are
left free
all rotations are tied linearly
The following patch test is now passed for constant
in-plane stresses, pure bending and pure torsion:
??
6 degrees of
freedom per
node
x
z
y
x
z
y
General 3D Situation Automatic Mesh Refinement
General 3D Situation
The displacement of node in direction is left free,
where is determined by:
*
if nodes 1 and 2 are not attached to a surface
*
if nodes 1 and 2 are attached to a surface ( is the unit
normal vector to this surface at node 3)
Remark:
use the ORIENTATION option when the Zienkiewicz-Zhu
stress criterion is used, due to local direction of stress
components
x
z
y
1
2
3
E
L3
E
L2
E
L1
4
3 E
L3
E
L3
E
L3
E
L1
E
L2
E
L1
E
L2
-----------------------------
=
E
L3
n
surface
=
n
surface
Automatic Mesh Refinement Nonlinear Adaptive Analyses
Nonlinear Adaptive Analyses
Basic flow scheme nonlinear adaptive stress analysis:
Input
Set up
Ku
=
Determine
u
K
1
f
=
Calculate strains
Next
Modify
FE mesh
No
Yes
increment
and stresses
sufficiently
Mesh
fine?
Converged
solution?
Next
iteration
No
Yes
Error Criteria for Nonlinear Analyses Automatic Mesh Refinement
Error Criteria for Nonlinear Analyses
All the criteria previously mentioned for linear analyses
can be used in a nonlinear analysis.
Note: in a nonlinear analysis the strain energy for element
is calculated according to:
with:
: current increment number
: stress vector at beginning of increment
: incremental stress vector for increment
: incremental strain vector for increment
In addition, the following criteria may be used:
node in contact
equivalent plastic strain
i
E
i

n
T 1
2
---
n
T
+
\ .
| |
n
d
)
`

n 1 =
n
current
=
n
current
n
n
n
n
n
n
Automatic Mesh Refinement Node in Contact
Node in Contact
at least one of its nodes comes into contact with either a
deformable or a rigid body
or:
at least one of its nodes belongs to a segment which is
contacted by another node
Remarks:
Default contact tolerance will automatically
be adapted
Be careful with an initially too coarse FE mesh
to avoid penetration:
i
Equivalent Plastic Strain Automatic Mesh Refinement
Equivalent Plastic Strain
Determine maximum equivalent plastic strain:
(relative)
or:
(absolute)
Remarks:
Typical values for and are problem
dependent
max
eq pl
max
max
i
eq pl
( i 1 N
adapt
, ) = =
i
max
i
eq pl
f
5
max
eq pl
>
max
i
eq pl
f
6
>
f
5
f
6
Automatic Mesh Refinement Rezoning Required
Rezoning Required
Compared to a linear analysis, modifying an FE mesh now also
involves local rezoning, since nodal and integration point values must
be transferred to newly created nodes and elements:
nodal point values for the newly created nodes follow from
the average values of:
two master nodes (new node on an element edge)
three or four master nodes (new node on an
element face)
four or eight master nodes (new node in the
interior of a 3D continuum element)
integration point values of the newly created elements are
obtained according to the following schemes:
quad4 elements with reduced integration and
hourglass control:
; ; ;
A
1
A
4
A
5
A
3
A
2
v
A
2
v
A
1
= v
A
3
v
A
1
= v
A
4
v
A
1
= v
A
5
v
A
1
=
Rezoning Required Automatic Mesh Refinement
fully integrated quad4 elements:
;
;
Step 1: determine values in original nodal points:
1 2
3 4
E
A
1
B
1
C
1
D
1
v
E
0.25 v
A
1
v
B
1
v
C
1
v
D
1
+ + + ( ) =
v
1
v
E
1.733 v
E
v
A
1
( ) + =
v
2
v
E
1.733 v
E
v
B
1
( ) + =
v
3
v
E
1.733 v
E
v
D
1
( ) + =
v
4
v
E
1.733 v
E
v
C
1
( ) + =
Step 2: determine values in newly created nodal points:
1 2
3 4
5
6
7
8 9
v
5
0.5 v
1
v
2
+ ( ) = v
6
0.5 v
2
v
3
+ ( ) =
v
7
0.5 v
3
v
4
+ ( ) = v
8
0.5 v
4
v
1
+ ( ) =
v
9
0.25 v
1
v
2
v
3
v
4
+ + + ( ) =
Automatic Mesh Refinement Rezoning Required
;
;
etc. where is the shape function of the element
for tria3, tet4, and hex8 elements the procedure is
similar (notice that both tria3 and tet4 elements have
only 1 integration point)
Step 3: determine values in newly created integration points
5
6
7
8 9
1 2
3 4
A
2
B
2
C
2
D
2
A
3
B
3
C
3
D
3
A
4
B
4
C
4
D
4
A
5
B
5
C
5
D
5
v
A
2
N A
2
( )
v
1
v
5
v
9
v
8
T
= v
B
2
N B
2
( )
v
1
v
5
v
9
v
8
T
=
v
C
2
N C
2
( )
v
1
v
5
v
9
v
8
T
= v
D
2
N D
2
( )
v
1
v
5
v
9
v
8
T
=
N
Use of MSC.Marc for Nonlinear Adaptive Analyses Automatic Mesh Refinement
Use of MSC.Marc for Nonlinear Adaptive Analyses
if applicable, attach nodes to geometrical entities
for a contact analysis, estimate in final mesh total number
of nodes on the outer boundary of deformable bodies
Important MSC.Marc Mentat options with respect to a
nonlinear adaptive analysis:
MESH
e.g.
ATTACH NODE
GENERATION TO SURFACE
ATTACH
JOBS
JOB
PARAMETERS
ADAPTIVE
MESHING
e.g.
MECHANICAL
Enter upper bounds for CONTACT
SEGMENTS and NODES; switch
on ATTACHED NODES button
Automatic Mesh Refinement Use of MSC.Marc for Nonlinear Adaptive Analyses
CHAPTER 10 Workshop Problems
The purpose of this chapter is to provide
the hands on problems associated with the
previous chapters.
Workshop Problems Some MSC.Marc Mentat Hints and Shortcuts
Some MSC.Marc Mentat Hints and Shortcuts
1.) Enter MENTAT to begin, Quit to stop
2.) Mouse in Graphics: Left to pick, Right to accept pick
3.) Mouse in Menu: Left to pick another menu or function,
Middle for help, Right to return to previous menu. <cr> means
keyboard return.
4.) Save your work frequently. Go to FILES and select SAVE AS
and specify a file name. Use SAVE from then on. This will save
the current MENTAT database to disk.
5.) Dialog region at the lower left of screen displays current
activity and prompts for input. Check this region frequently to
see if input is required.
6.) Dynamic Viewing can be used to position the model in the
graphics area. When activated, the mouse buttons, Left
translates the model, Right zooms in/out, Middle rotates in 3D.
Use RESET VIEW and FILL to return to original view. Be
sure to turn off DYNAMIC VIEW before picking in the
graphics area.
7.) CTRL P/N recall previous/next commands entered.
Some MSC.Marc Mentat Hints and Shortcuts Workshop Problems
8.) All of the workshop problems have Mentat procedure files.
They are located in a training directory under Mentats main
directory. The directory/file structure looks like:
~mentat/examples/training/
s1/ s2/ s3/ s4/ s5/ s6/ s7/ s8/ s9/ h1/ h2/ h3/ h4/ h5/ d1/ d2/ d3/
Where, say in directory s4, there is a
procedure file called s4.proc. It will
automatically run Mentat to build, run
MARC, and process the results. These
directories can be copied to your local
disk area to work on during the
workshop. Furthermore, all training
notes can be accessed on-line from the
menu below.
Check out the HELP menu.
s4
notes
Workshop Problems Mechanics of Mentat
Mechanics of Mentat
Before you get started with Mentat, you need to know how to
communicate with the program. The goal of this section is to give you
an overview of how Mentat works and to provide you with the basic
information to interact comfortably with the program. Upon
completion of this chapter, you should have a clearer understanding of
the following areas:
The basic window layout
How Mentat communicates with you
How you communicate with Mentat
The menu system
Mechanics of Mentat Workshop Problems
MSC.Marc/Mentat Window Layout
The starting point for all communication with Mentat is the window
shown below that appears at the start of the program.
The Mentat window is divided into three major areas:
The Graphics area is used to display the current state of the database.
When you start Mentat, the graphics area is blank to indicate
that the database is empty.
The Menu area is reserved to show the selectable menu-items and is
divided into two sub-areas, the Static and Dynamic menus. The
contents of the dynamic menu area change as the menu-items are
selected. In contrast, the static menu is always present and con-
tains items that are applicable and selectable at all times.
The Dialogue area is a scrollable area of about five visible lines where
all program prompts, warnings, and responses appear, and
where the user can input data or commands. Within the dialogue
area is the status area which is reserved to communicate the state
of the program to the user. Either working or ready appears
in the status area to reflect the current state of the program.
Graphics Dynamic
Menu
Static Menu
Dialogue Status
How MSC.Marc Mentat Communicates with You
Mentat communicates with you via prompts and messages and
other visual queues. Mentat's prompts urge you to take action
through the input of data or commands. These prompts have 3
types of trailing punctuation marks to indicate the required
type of input:
: enter numeric data, e.g. .283/384;
> enter a character string, typically a command, file
name or set name;
? enter a YES or NO answer.
If you misspell a keyword or enter an incorrect response,
Mentat warns you through a message posted in the dialogue
area. Mentat does not require that you complete every action
you initiate. For example, if you are prompted for a filename,
and you change your mind, entering a <CR> instead of typing in
the filename will tell Mentat to abort the action. If the program
is waiting for a list of items to operate on, and instead you
enter a command that also requires a list of items or any
additional data, Mentat will ignore your original request and
process the command. If the command you enter does not
request additional data, you are returned to the original data
request from before the interrupt.
The program assumes at all times that you want to repeat the
previous operation on a new set of items and will prompt you
for a new list to operate on. This process repeats itself until
you indicate otherwise, typically by entering a new command
or a <CR>.
How You Communicate with MSC.Marc Mentat
All interaction with Mentat is done through the mouse, keyboard or a
combination of both. This section first discusses the usage of the
mouse, followed by a discussion on how to use the keyboard as a means
to enter commands and data.
The Mouse
The mouse is used to select items from the menu area or to point at
items in the graphics area. It is important to make a distinction
between using the mouse in the menu area versus the graphics area
because the three mouse buttons have very different functions in each
area. Below is a graphical representation of the mouse, mouse buttons,
and corresponding cursor.
The Mouse, Mouse Buttons, and Corresponding Cursor
The left button is represented by <ML>, the middle button by <MM>, and
the right button by <MR>. For a two button mouse <MM> = <ML> + <MR>
depressed at the same time. Click refers to a quick single depress-
release action.
Cursor, <^>
<MM>
<MR>
<ML>
Using the Mouse to Select a Menu Item
To select a menu item with the mouse, move the over the item that
you want to select and click the <ML>. To return to the previous menu,
move the over the menu area, and click the <MR>. Alternatively, you
can click on the RETURN button in the menu area using <ML>. Clicking on
the MAIN button takes you to the main menu.
On-line Help
Each menu item has a help panel with a short description and
explanation of the function of that menu item. To activate the help
feature, position the over the menu item on which you require help,
followed by a click of the <MM>. The help panel disappears the moment
you select another menu item.
Using the Mouse in the Menu Area
< >
< >
< >
<ML>
Use this button,
for command selection
<MR>
for RETURN
Use this button,
<MM>
for on-line help
Use this button,
Using the Mouse to Point
The mouse is used in two ways to operate in the graphics area: to point
to, or pick, existing items as well as to point to, or pick, the location of
yet to be created items.
1. To pick the mouse is used for this by moving the over the item to be
identified followed by a click of the <ML>. Henceforth called by clicking
on an item. You can undo that action by clicking the <MM> anywhere in
the graphics area. At times, you will need to identify more than a single
item. A list of items must be terminated by a click of the <MR> with the
positioned anywhere in the graphics area. Alternatively you can
click on the END LIST button in the menu area using <ML>.
2. To locate a position in Mentat, it is possible to define a grid that is
positioned in space and where the grid consists of points that can be
pointed to. If you click in the vicinity of a grid point, the coordinates of
the item that you created will be snapped to that grid point. In addition,
you can also pick an existing node, point, or surface-grid-point to
specify a location.
Using the Mouse in the Graphics Area
< >
< >
<ML>
Use this button,
to pick
<MR>
for end of item list
Use this button,
<MM>
to undo last pick
Use this button,
Keyboard Input
Not all data can be entered through the mouse; numerical and literal
data must be entered via the keyboard. The program mode prescribes
the specific requirements for proper entry of each type of data. The
program can be in data mode or in command/literal data mode and is
described under the following two headings.
Numerical Data
You must use the keyboard for numerical data entry. The program
interprets the data entry according to the context in which it is used. If
the program expects a real number and you enter an integer, Mentat will
automatically convert the number to its floating point value. Conversely,
if a floating point format number is entered where an integer is
expected, the program will convert the real number to an integer.
Scientific notation for real numbers is allowed in the following formats:
.12345e01
.12345e01
-0.12345e-01
The interpreter does not allow imbedded blanks in the format.
Whenever the program encounters an illegal format, the message bad
float! will appear in the dialogue area. The prompt for numerical
data is a colon (:).
Literal Data
Literal data is used for file, set and macro names. A literal data string
may not be abbreviated. Commands as introduced in the beginning of
are considered string data (as opposed to literal string data) and can be
abbreviated as long as the character string is unique within the Mentat
command library. For example, *add_elements cannot be
abbreviated to *add because of the other commands that start with the
same characters such as *add_nodes and *add_curves. The
program checks the input for validity against the internal library of
valid responses. For example, if you enter an ambiguous or misspelled
command, Mentat responds by listing all the valid entries that start with
the same first letter of the command. The prompt for literal data is a
greater-than symbol (>).
If the program is in data mode which is identified by the : prompt, you
must enter a command preceded by an * (asterisk) to instruct the
program that you are entering a command.
For example: Enter node (1): *add_nodes
Workshop Problems Comprehensive Sample Session
Comprehensive Sample Session
In this hands-on session, you will create a simple 3D mesh and add all
appropriate boundary conditions, material properties, etc. You will
run the analysis and view the results.
A linear elastic analysis of the following 3D structure will be
performed:
Boundary conditions:
face 1: clamped
face 2: loaded by a uniformly distributed
shear load
(force per unit area), magnitude 40, direction
Material properties:
Youngs modulus
Poissons ratio
20 20
20
30
200
60
face 1
face 2
x
y
z
0 1 1
T
E 4
5
10 =
0.3 =
Comprehensive Sample Session Workshop Problems
Start up window Mentat:
Static menu
Graphic area
Dynamic menu
Status area Dialogue area
Mouse buttons:
ML MM
MR
menu area
graphic area
select
command
help on
command
return to
previous menu
pick
entity
undo last
pick
end of list
single pick
box pick
polygon pick
(CTRL)
Mesh generation: top menu
mesh entities
geometric entities
mesh entities
geometric entities
Mesh generation (continued): set and display
grid for easy input of coordinates and fill view
COORDINATE SYSTEM: SET GRID ON
U DOMAIN 0 200 <cr>
U SPACING 20 <cr>
V DOMAIN 0 80 <cr>
V SPACING 10 <cr>
FILL
RETURN
Mesh generation (continued): create points
(geometric entities), switch off grid and fill view
POINTS: ADD (Add the following points with mouse clicks)
(0,80,0)
(20,80,0)
(40,80,0)
(0,60,0)
(20,60,0)
(40,60,0)
(20,0,0)
(40,0,0)
(200,80,0)
(200,60,0)
COORDINATE SYSTEM: SET GRID OFF
FILL
Mesh generation (continued): create quad
surfaces (geometric entities)
SURFACES: ADD
Pick corner points for quad surfaces with mouse clicks to
obtain four surfaces as shown. A half-arrowhead is used to
indicate the first side of the surface.
surface type to be created
first side of surface
Mesh generation (continued): convert surfaces
to elements (mesh entities)
CONVERT
DIVISIONS 6 2 <cr>
BIAS FACTORS -0.3 0 <cr>
SURFACES TO ELEMENTS
(Pick the rightmost surface)
DIVISIONS 2 2 <cr>
BIAS FACTORS 0 0
(Pick the two small surfaces)
DIVISIONS 2 3 <cr>
(Pick the lower surface)
RETURN
first edge of element
and node numbering
direction
Mesh generation (continued): modify sweep
tolerance and use sweep option to merge
coincident nodes
SWEEP
TOLERANCE 0.001 <cr>
SWEEP: NODES
ALL:EXIST.
RETURN
Mesh generation (continued): use renumber
option to obtain consecutive numbering
RENUMBER: ALL
RETURN
Mesh generation (continued): use expand option
to expand the mesh in z-direction
EXPAND
TRANSLATIONS 0 0 15 <cr>
REPETITIONS 2 <cr>
MODE: REMOVE (no action required, this is the default)
ELEMENTS
ALL: EXISTING
RETURN
original elements will be removed
Mesh generation (continued): remove unused
nodes and repeat renumber command
SWEEP
REMOVE UNUSED: NODES
SWEEP: ALL
RETURN
RENUMBER: ALL
RETURN
Mesh generation (continued): show view 4 and
fill view
VIEW
SHOW VIEW: 4
FILL
Mesh generation (continued): define increment
of rotation for the model
VIEW SETTINGS
MODEL INCREMENTS: ROTATE 90 <cr>
RETURN
Mesh generation (continued): rotate model in
positive direction around model x- and y-axis
and fill view
MANIPULATE MODEL
ROTATE IN MODEL SPACE: X+
ROTATE IN MODEL SPACE: Y+
FILL
Mesh generation (continued): plot elements in
solid mode, switch off plotting geometric entities
PLOT
turn off POINTS and SURFACES
ELEMENTS: SOLID
REDRAW
SAVE
MAIN
Boundary conditions: top menu
BOUNDARY CONDITIONS
MECHANICAL
Boundary conditions (continued): mechanical
boundary conditions, fixed displacements
NEW
NAME clamped <cr>
FIXED DISPLACEMENT
ON: X DISPLACE
ON: Y DISPLACE
ON: Z DISPLACE
Boundary conditions (continued): switch to
view 1 and select appropriate nodes
VIEW
SHOW VIEW 1
RETURN
NODES: ADD
Add as shown by the box pick method
END LIST
(for end list use button or use right mouse click in graphics area)
box pick method
Boundary conditions (continued): switch to
view 4, define mechanical boundary conditions,
face loads
VIEW
SHOW VIEW 4
RETURN
NEW
NAME shear <cr>
FACE LOAD
U SHEAR 28.2843 <cr>
V SHEAR 28.2843 <cr>
OK
Boundary conditions (continued): zoom in
locally and select appropriate element faces
ZOOM
Zoom in on the right end of structure
FACES: ADD
Add appropriate element faces with mouse
END LIST
RETURN
Boundary conditions (continued): overview of
boundary conditions
ID BOUNDARY CONDITIONS
FILL
SAVE
MAIN
Material properties: top menu
MATERIAL PROPERTIES
Material properties (continued): mechanical
material type, isotropic properties, apply to
all elements
NEW
NAME linear_elastic
ISOTROPIC
E=400000 <cr>
NU=0.3 <cr>
OK
ELEMENTS: ADD
ALL: EXISTING
SAVE
MAIN
Geometric properties: top menu
GEOMETRIC PROPERTIES
Geometric properties (continued): select assumed
strain formulation for all existing elements to
improve bending behavior
NEW
NAME assumed_strain <cr>
3-D
SOLID
ASSUMED STRAIN
OK
ELEMENTS: ADD
ALL: EXISTING
SAVE
MAIN
Jobs: define mechanical analysis; for a single
linear analysis no loadcases are necessary and
the default analysis options can be used
JOBS
NEW
NAME example_3d <cr>
MECHANICAL
Jobs (continued): select post file quantities
JOB RESULTS
TENSORS: STRESS
OK
Jobs (continued): check if boundary conditions
are selected as initial loads
INITIAL LOADS
OK
OK
Jobs (continued): select mechanical 3D solid
element type 7, save model
ELEMENT TYPES
3-D SOLID
Select element type 7
OK
ALL: EXISTING
SAVE
RETURN
Jobs (continued): save Mentat database and
submit job model1_example_3d
RUN
SUBMIT 1
Submitting a job:
submit1: Mentat-directory/bin/submit1
marc2003 -j model1_example _3d -q b -v no
* use model1_example_3d.dat as data file
* run job in background
* dont wait for confirmation of correct input
outputfile: model1_example_3d.out
log file: model1_example_3d.log
post file: model1_example_3d.t19 (formatted)
model1_example_3d.t16 (binary)
MSC.Marc data file:
Parameter
options
Model
definition
options
History
definition
options
sufficient
for a
single
linear
analysis
title
sizing
elements
...
...
end
connectivity
coordinates
isotropic
geometry
fixed disp
dist loads
point load
...
...
end option
...
continue
...
...
continue
dist loads
point load
disp change
...
Jobs (continued): use monitor to observe
current status
MONITOR
OK
RETURN
MAIN
MSC.Marc post file:
Header
Increment 0
coordinates
connectivity
nodal
quantities;
element
quantities
if selected
Increment 1
nodal
quantities;
element
quantities
if selected
open post file
next increment;
skip to increment 0
next increment;
skip to increment 1
Postprocessing: use open default option
RESULTS
OPEN DEFAULT
Postprocessing (continued): skip to increment 0
and select equivalent von Mises stress
to be displayed
NEXT INC
SCALAR
EQUIVALENT STRESS
OK
Postprocessing (continued): plot deformed and
undeformed structure for increment 0 using
contour bands
DEFORMED AND ORIGINAL
CONTOUR BANDS
Postprocessing (continued): deformed shape
settings. You may magnify the displacements.
DEFORMED SHAPE: SETTINGS
MANUAL FACTOR: 4
MANUAL FACTOR: 1
RETURN
manual deformation scaling
scale factor 1 by default
Postprocessing (continued): define plotting
style settings for cutting planes
SCALAR PLOT: SETTINGS
POINT 20 -10 0 <cr>
NORMAL 1 1 0 <cr>
PLANES 8 <cr>
SPACING 25 <cr>
RETURN
Postprocessing (continued): select cutting planes
to visualize the equivalent von Mises stress
CUTTING PLANES
Postprocessing (continued): switch back to
contour bands plotting and define a node path
for a path plot
CONTOUR BANDS
PATH PLOT
NODE PATH
Pick the nodes shown to define the path
END LIST
Postprocessing (continued): add path plot
curve and scale the plot axes
VARIABLES
ADD CURVE
ARC LENGTH (X variable)
EQUIVALENT STRESS (Y variable)
FIT
RETURN
RETURN
Postprocessing (continued): vector plot
of displacements
MORE
VECTOR PLOT: ON
Workshop tasks:
* Perform the discussed 3D analysis and store the
Mentat commands in a procedure file, which
can be created in the UTILS menu
* Analyze the same 3D structure, but now subjected
to a distributed shear load with a magnitude 40 and
a direction of (bending load)
* Analyze the structure subjected to the bending
load using 4-node plane strain elements (select
Marc element type 11) and compare the results
with the 3D solution
0 1 0
T
Additional workshop: linear elastic analysis of an
infinitely long pressurized thick-walled cylinder
Dimensions: , ,
Apply fixed displacements in axial direction
Internal pressure:
Material: ,
p
r R
L
section to
be considered
L 4 = r 5 = R 12 =
p 15 =
E 2.1
5
10 = 0.3 =
Workshop tasks:
Determine the radial stress as a function of the
radial coordinate using:
A: axisymmetric element 10:
B: plane strain element 11 (model one quarter of
the cross section):
C: brick element 7 (model one quarter of the section
to be considered):
Apply the correct boundary conditions and compare
the results
4.0 6.0 8.0 10.0 12.0
radial coordinate
-15.0
-10.0
-5.0
0.0
r
a
d
i
a
l

s
t
r
e
s
s
Infinitely long pressurized thick-walled cylinder
Analytical solution
10 axisymmetric elements; radial bias -0.5
80 plane strain element; radial bias -0.5
80 brick elements; radial bias -0.5
Statics S1 Tensile Specimen
Overview: This example session
describes the simulation of the loading of
a dog-bone tensile specimen. This session
builds the geometry, exports an iges file
and demonstrates different types of
meshing strategies including: overlay,
advancing front, and mapped meshing.
Using the mapped mesh, the tensile
specimen is subjected to an axial load and
submitted to MSC.Marc. MSC.Marc
Mentat then post process the results of
the tensile specimen.
After the first run, the specimens gage
section is changed and re-run to compare
with the original specimen.
Finally, the material is changed from an
isotropic to an orthotropic material. The
material direction does not line up with
the pull direction, and the deformed
shape becomes skewed.
Overlay Mesh
Adv. Front Mesh
Loads
Mapped Mesh
Skewed Orthotropic
Begin at the main menu
MESH GENERATION
COORDINATE SYSTEM SET
GRID ON
U DOMAIN -1.5 1.5 <cr>
V DOMAIN -1.5 1.5 <cr>
FILL
RETURN
CURVE TYPE
Select Arc CENTER/POINT/ANGLE
RETURN
CURVES ADD
0 1.5 0 <cr>
0 -1.5 0 <cr>
-21 <cr> (degrees)
MOVE
TRANSLATIONS
0 1.75 0 <cr>
CURVES
use left mouse to pick curve, right will END LIST
RETURN
SYMMETRY
NORMAL type in 0 1 0
CURVES select the arc,
END LIST
NORMAL type in 1 0 0
CURVES
ALL: EXISTING
RETURN
COORDINATE SYSTEM: SET GRID OFF
RETURN
DUPLICATE
TRANSLATIONS type in .425 0 0
POINTS, select two right most points
END LIST
TRANSLATIONS type in -.425 0 0
POINTS, select two left most points
END LIST
RETURN
CURVE TYPE
LINE
RETURN
CURVES ADD
Select pairs of points beginning at the upper left of the top arc and
move CCW to complete the boundary of the model.
The following to save this geometry in an IGES file.
FILES
EXPORT IGES ten.spec.iges OK RETURN
MAIN
SAVE
The next section shows how to mesh the geometry
several ways. Lets look at the Overlay Technique
MESH GENERATION
AUTOMESH
2D PLANAR MESH-
ING
DIVISIONS
type in 20 20 <cr>
OVERLAY QUAD
ALL: EXISTING
UNDO
DIVISIONS
type in 40 40 <cr>
OVERLAY QUAD MESH
ALL: EXISTING
RETURN
UNDO this will undo your last command
Advancing Front Technique
AUTOMESH
2D PLANAR MESHING
QUAD MESH (Advancing Front)
ALL: EXISTING
UNDO
RETURN
CURVE DIVISIONS
FIXED AVG LENGTH
FORCE EVEN DIV
APPLY CURVE DIVISIONS
ALL: EXISTING
RETURN
2D PLANAR MESHING
QUAD MESH (Advancing Front)
ALL: EXISTING
UNDO
RETURN
Mapped Meshing Technique
CURVE DIVISIONS, CLEAR CURVE DIVISIONS
ALL: EXISTING, RETURN (twice)
SURFACE TYPE
RULED RETURN
ADD SURFACE Pick top left/bottom arcs
ADD SURFACE Pick top right/bottom arcs
CHECK
FLIP CURVES Pick top right/top left lines, RETURN
ADD SURFACE Pick right top/bottom line
ADD SURFACE Pick left top/bottom line
CONVERT
Pick left and right curved surfaces
DIVISIONS 5 10
Pick left and right rectangular surfaces, RETURN
SWEEP
ALL & RETURN
RENUMBER
ALL & RETURN
MAIN
BOUNDARY CONDITIONS
MECHANICAL, FIXED DISPLACEMENT
ON X DISPLCEMENT, OK
NODES ADD
Select all nodes on left edge
END LIST
NEW
FIXED DISPLACEMENT
ON Y DISPLACEMENT, OK
NODES ADD
Select center node on left edge
END LIST
NEW
EDGE LOAD
ON PRESSURE -30000, OK
EDGES ADD
Select all edges on right edge
END LIST
MAIN
MATERIAL PROP.
NEW
ISOTROPIC
E = 1E7
= .3, OK
ELEMENTS ADD
ALL: EXISTING
RETURN
PLANAR
PLANE STRESS
THICKNESS = 0.25
ASSUMED STRAIN
This improves the elements behavior in bending.
OK
ELEMENTS ADD
ALL: EXISTING
MAIN

JOBS
MECHANICAL
PLANE STRESS
ANALYSIS OPTIONS
LARGE DISPLACEMENT
OK
JOB RESULTS
TENSORS STRESS, OK (twice)
SAVE
RUN
SUBMIT1
MONITOR (some element upside/down)
OK
MAIN
MESH GENERATION
CHECK UPSIDE DOWN
FLIP ELEMENTS
ALL: SELECTED
UPSIDE DOWN
Number of upside/down elements: 0
RETURN (twice)
Go back to RUN and resubmit.
JOBS
SAVE
RUN
SUBMIT1
MONITOR
OK
Is the job complete?
Did it do what I expect?
MAIN
RESULTS
OPEN DEFAULT
SCALAR
COMP 11 OF STRESS, OK
CONTOUR BANDS
RESULTS
MORE
VECTOR Pick Reaction Force, OK
VECTOR PLOT ON
VECTOR Pick External Force, OK
RESULTS
PATH PLOT
NODE PATH N1 N2
path from N1 to N2
END LIST
VARIABLES
ADD CURVE
ARC LENGTH, COMP 11 OF STRESS
FIT
RETURN
YMIN 0
N1
N2
RESULTS
PATH PLOT
TABLES COPY TO
> 1
TABLES
FIT
XSTEP 100
YSTEP 100
INTEGRATE
FIT
Where: p = 30000, W=0.898518, t=0.25
N1
N2
W
11
t y d
N1
N2
26926t 26955t pWt 6739 = = =

Statics S1 Tensile Specimen Uniform Gage Section
The previous stress analysis shows that the stress field is not uniform in
the gage section. Redesign the specimen such that it has a 1" constant
gage section at the center.
MAIN
RESULTS
CLOSE, MAIN
FILES
SAVE AS model2
RESET PROGRAM
RETURN
MESH GENERATION
ATTACH
DETACH NODES, ALL: EXISTING
DETACH ELEMENTS, ALL: EXISTING
SELECT ELEMENTS
Pick all elements to the right of the net section
END LIST
ELEMENTS STORE
right, ALL:SELECTED, RETURN (twice)
SUBDIVIDE
DIVISIONS 1 1 1
ELEMENTS, ALL:SELECTED, RETURN
MOVE
TRANSLATIONS 1 0 0
ELEMENTS
right <cr>, RETURN
SWEEP
REMOVE UNUSED NODES
ALL, FILL, RETURN
PLOT
CURVES OFF
SURFACES OFF
POINTS OFF
REGEN
RETURN to mesh generation
ELEMENT ADD N1, N2, N3, N4
SUBDIVIDE, DIVISIONS 10 10 1
ELEMENT pick element just added then
RETURN
SWEEP, ALL, RETURN
RENUMBER, ALL, MAIN
MATERIAL PROPERTIES
ELEMENTS ADD, ALL: EXISTING, RETURN
N1 N2
N3 N4
PLANAR
PLANE STRESS
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
JOBS
RUN
SUBMIT(1)
RESULTS
OPEN DEFAULT
SCALAR Comp 11 Of Stress, OK
CONTOUR BANDS
Statics S1 Tensile Specimen Composite Material
What about composites? Suppose we want to analyze an orthotropic
material whose material axis does not line up with the structures
geometric axis.
FILES
OPEN model1
SAVE AS model3
RETURN
MATERIAL PROPERTIES
ORTHOTROPIC
E11 = 3E7, E22=E33=1E6
ALL s = .3
ALL GS = 5E5
OK
ORIENTATION
NEW
EDGE41 ON
ANGLE 45
ADD ELEMENTS
ALL: EXISTING
SAVE

Re-run and check results, deformed shape is skewed.
Statics S2 Cantilever Beam
Overview: This example session
describes the simulation of loading a
cantilever beam with a tip load. This
model will be used latter for dynamics
and will be saved.
The linear elastic solution is found. The
bending stresses and tip displacements
are then compared to theory.
The material properties are changed to
include plasticity with workhardening.
The beam is then loaded with a larger
load of 1500 pounds with 50 equal load
steps. We will see how every integration
point must track the materials
constitutive relation.
10" X 1" X 1"
500 #
Here is a cantilever beam. Later, we will also look at its dynamic
behavior and the static results here will be needed later.
FILES
NEW, OK
SAVE AS beam1,OK
RETURN
MESH GENERATION
NODE ADD
0 0 0 <cr>
10 0 0 <cr>
10 1 0 <cr>
0 1 0 <cr>
FILL
ELEMENT ADD (Pick above nodes in CCW)
SUBDIVIDE
DIVISIONS 10 4 1 <cr>
ELEMENTS
ALL: EXISTING, RETURN
SWEEP
ALL & RETURN
RENUMBER
ALL & RETURN
MAIN
10" X 1" X 1"
500 #
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISPLACEMENT
ON X DISPLCEMENT, OK
NODES ADD, Select all nodes on left edge
END LIST
NEW
FIXED DISPLACEMENT
ON Y DISPLACEMENT, OK
NODES ADD, Select bottom node on left edge
END LIST
NEW
POINT LOAD
ON Y FORCE -500, OK
NODES ADD, Select top right node
END LIST
RETURN
ID BOUNDARY CONDITIONS
MAIN
MATERIAL PROP.
NEW
ISOTROPIC
E = 3E7
= .3
= .283/386
OK
ELEMENTS ADD
ALL: EXISTING
RETURN
GEOMETRIC PROP.
PLANAR
PLANE STRESS
THICKNESS = 1 <cr>
ASSUMED STRAIN
OK
ELEMENTS ADD
ALL: EXISTING, MAIN
JOBS
MECHANICAL
PLANE STRESS
JOB RESULTS
TENSORS STRESS, OK (twice)
SAVE
RUN
SUBMIT1
MONITOR, OK

MAIN
RESULTS
OPEN DEFAULT
NEXT
SCALAR
COMP 11 OF STRESS
CONTOUR BANDS
SCALAR
DISPLACEMENT Y, OK
Complete Modeling: Check Load
Peak Bending Stress +/- 29Ksi, Max Disp 6.7e-2.
How does this compare to beam theory?
What can improve the results?
Statics S2 Cantilever Beam Add Plasticity
Here is a cantilever beam.
Lets convert it to an elastic-
plastic model.
10" X 1" X 1"
750 #
FILES
OPEN beam1
SAVE AS beam1p, OK
RETURN
MATERIAL PROPERTIES
TABLES
NEW
TABLE TYPE: eq_plastic_strain, OK
POINT ADD
0.000 20E3 <cr>
0.109 25E3 <cr>
0.305 30E3 <cr>
FIT
> XY
MAIN
MATERIAL PROPERTIES
TABLE
NEW
TABLE TYPE: TIME, OK
FORMULA, ENTER
1.5*V1 (will ramp load from 0 to 750# in one second)
FIT
SHOW MODEL
RETURN
ISOTROPIC
ELASTIC-PLASTIC
INITIAL YIELD STRESS = 1.0
TABLE1 = table1
OK
OK
RETURN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply3 (point load), OK
POINT LOAD
Y FORCE (pick table2, time), OK
MAIN
LOADCASES
MECHANICAL
STATIC
OK
RETURN (twice)
JOBS
MECHANICAL
SELECT lcase1
ANALYSIS OPTIONS
LARGE DISPLACEMENT
LARGE STRAIN ADDITIVE
OK
JOB RESULTS
EQUIVALENT VON MISES STRESS
TOTAL EQUIVALENT PLASTIC STRAIN
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR, OK, MAIN
RESULTS
OPEN DEFAULT
NEXT
DEF & ORIG
SCALAR Total Equivalent Plastic Strain
LAST
CONTOUR BANDS
RESULTS
HISTORY PLOT
SET NODES (pick top left node), END LIST
COLLECT GLOBAL DATA
NODE/VARIABLES
ADD VARIABLE
Total Equivalent Plastic Strain
Equivalent Von Mises Stress
FIT
RETURN
> XY
UTILS
GENERALIED XY PLOT
FIT
This will overlay the history plot of the stress strain response of this
node with the stress-strain material behavior. Remember that
continuum mechanics requires that the continuum be in equilibrium
and that every point must track the constitutive relation.
Stress strain response(+) tracking the constitutive
relation.
Statics S3 Elastomeric Curve Fit and Seal
Overview: A hollow elastomeric cylinder
will be squeezed to closure and released
using the contact option. Symmetry is
used and only one half of the cylinder
is modeled.
The experimental material data is fitted to
an Ogden elastomeric material.
In the first run, the cylinder is squeezed
and released with out friction. Then
friction between the rigid bodies and the
tube is added using the stick-slip friction
option. A friction coefficient of 0.2 is used.
The results of the two runs are compared.
The friction case shows higher squeeze
loads and some hysteresis.
In the last run, the tube has a closed air
cavity and its results are compared to
frictionless tube with out a closed cavity.
HYSTERSIS ON
FORCE X
HIGHER
LOAD
Here is a hollow elastomeric
cylinder. It will be squeezed and
released. The material properties
will be fit to an Ogden model.
FILES
NEW, OK
SAVE AS elasto1
RETURN
MESH GENERATION
COORDINATE SYSTEM SET: GRID ON
U DOMAIN -1.1 1 <cr>
V DOMAIN 0 1.0 <cr>
FILL
RETURN
CURVES ADD (pick indicated points from grid)
POINT( 1.0, 0.0, 0.0)
POINT( -1.1, 0.0, 0.0)
POINT( -1.1, 1.0, 0.0)
POINT( 1.0, 1.0, 0.0)
CURVE TYPE
Select Arc CENTER/POINT/POINT
RETURN
CURVES ADD
(-1.1 .5 0 ) (-1.0 0.0 0) (-1.0 1.0 0)
(-1.1 .5 0 ) (-1.0 0.1 0) (-1.0 0.9 0)
SURFACE TYPE
RULED
RETURN
SURFACES ADD
Pick inner then
outer curve
CONVERT
DIVISIONS (30 3)
ALL: EXISTING
RETURN
SWEEP
ALL, RETURN
RENUMBER
ALL, RETURN
COORDINATE SYS: SET GRID OFF, MAIN
BNDRY. CONDITIONS
MECHANICAL
FIXED DISP
X=0, OK
ADD NODES
(pick nodes along x=0)
END LIST
MAIN
MATERIAL PROPERTIES
EXPERIMENTAL DATA FITING
TABLES
NEW
TABLE TYPE experimental_data
ADD POINTS
0.0 0.0 <cr>
0.9 100.0 <cr>
1.6 250.0 <cr>
1.9 300.0 <cr>
2.2 500.0 <cr>
2.4 600.0 <cr>
2.6 700.0 <cr>
2.9 1000.0 <cr>
FIT
NAME tension <cr>
RETURN
UNIAXIAL (pick table tension)
ELASTOMERS
OGDEN
UNIAXIAL
# TERMS = 2
POS. COEFF
MATH CHECKS
COMPUTE
APPLY
OK
RETURN (twice)
ELEMENTS ADD
ALL: EXISTING
MAIN
CONTACT
CONTACT BODIES
DEFORMABLE, OK
ELEMENTS ADD
ALL: EXISTING
TABLES, NEW, NAME position <cr>
TABLE TYPE TIME, OK
ADD POINT
0 0
.5 1
1 0
SHOW MODEL
RETURN
NEW
RIGID, DISCRETE
POSTION PARAMETERS
Y = -.4, TABLE1 position
OK (twice)
NAME, top <cr>
CURVES ADD
top curve, END LIST
ID CONTACT
NEW
RIGID,
DISCRETE
POSTION
PARAMETERS
Y = +.4 <cr>
TABLE1 pos.
OK (twice)
NAME, bottom
CURVES ADD
bottom curve
END LIST
MAIN
LOADCASES
MECHANICAL
STATIC
TIME = .5
STEPS = 50
SOLUTION CONTROL
# RECYCLES = 30
NON-POSITIVE
DEVIATIORIC STRESS, OK
CONVERGENCE CHECK DISPLACEMENTS
OK
OK
COPY (this copies lcase1 into lcase2)
MAIN
JOBS
MECHANICAL
Select lcase1, lcase2
JOB RESULTS
EQUIVALENT CAUCHY STRESS
OK (twice)
ELEMENT TYPES
PLANE STRAIN SOLID
80
OK
ALL: EXISTING
RETURN (twice)
RUN
SUBMIT1
MONITOR (oops elems inside-out)
OK, RETURN
MESH GENERATION
CHECK
UPSIDE DOWN
FLIP ELEMENTS
ALL: SELECTED
RETURN (twice)
JOBS
RUN
SUBMIT1
MONITOR, OK
SAVE
POSTPROCESS
OPEN DEFAULT
NEXT INC
DEF ONLY
SCALAR
EQUIVALENT CAUCHY STRESS
OK
CONTOUR BAND
SKIP TO INC 50
POSTPROCESS
HISTORY
SET NODES
1, #END LIST
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Pos Y top
Force Y bottom
FIT
Add Friction to the surfaces
FILES
NEW, OK
OPEN elasto1
SAVE AS elasto1f
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE
FRICTION COEFF = .2, OK
NEXT
NEXT
MAIN
JOBS
MECHANICAL
CONTACT CONTROL
STICK-SLIP
OK (twice)
RUN
SAVE
SUBMIT1
MONITOR
POSTPROCESS
HISTORY
SET NODES
1, #END LIST
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD GLOBAL CRV, Pos Y top, Force Y top
ADD GLOBAL CRV, Pos Y top, Force X top
FIT
Normal Force Y on top contact body
Tangential Force X
on top contact body
hystersis
Another interesting change to this model will be to also simulate the
compression of the air inside the tube, assuming that the tube is closed
and the air cannot escape. Lets start with the original model and add a
closed cavity representing the air inside the tube.
FILES
NEW, OK
OPEN elasto1
SAVE AS elasto1c
RETURN
FILL
MESH GENERATION
ELEMENT CLASS
LINE (2)
RETURN
ADD ELEMENT
(pick nodes N1 and N3 indicated)
CAVITY
NEW
SELECT
METHOD PATH, RETURN
EDGES (pick N1 N2 N3), END LIST
RETURN
EDGES ADD, ALL: SELECTED
REF. PRESSURE = 1
REF. TEMPERATURE = 1
N1
N3
N2
REF. DENSITY = 1.8E-5
MAIN
BOUNDARY CONDITIONS
NEW
MECHANICAL
MORE
CAVITY MASS LOAD
MASS
CLOSED CAVITY, OK
CAVITIES ADD
cavity1, END LIST
MAIN
LOADCASE
MECHANICAL
STATIC
LOADS
apply2 (on)
OK (twice)
NEXT
STATIC
LOADS
apply2 (on)
OK (twice)
MAIN
JOBS
MECHANICAL
INITIAL LOADS
apply2 (on)
OK
JOB PARAMETERS
CAVITY PARAMETERS
AMBIENT PRESSURE = 1
OK (twice)
ELEMENT TYPES
MECHANICAL
MISCELLANEOUS
171, OK
(pick element previously added)
END LIST
RETURN (twice)
SAVE, RUN, SUBMIT (1)
OPEN POST FILE (RESULTS MENU)
HISTORY PLOT
COLLECT GLOBAL DATA
NODE/VARIABLES
ADD GLOBAL CURV
Pos Y top, Force Y top
Volume Cavity 1, Pressure Cavity 1
FIT
The force to crush the tube is considerably larger(5x) than before
and nearly equal to the cavity pressure since the area of contact is
about 1 inch.
SHOW MODEL, RETURN
DEF ONLY
SKIP TO INC 50
The deformed shape at increment 50 shows the inner walls have
yet to close, whereas without the closed air cavity, the inner walls
will touch.
Statics S4 Break Forming
Overview: A flat sheet is formed into an
angled bracket by punching it though a hole
in a table using the contact option.
The cylindrical punch drives the sheet down
into the hole of the table to a total stroke of
0.3. The punch then returns to its original
position. The material is elastic plastic with
work hardening.
At the bottom of the stroke, the total plastic
strain is nearly 45%. The vertical punch force
is plotted versus its vertical position. This
force rises quickly, hardens though about half
of the stroke, then softens near the end of the
stroke. Upon lifting the punch, the punch
force drops rapidly and the sheet has very
little springback.
The stress-plastic strain response of a point in
the sheet under the punch is plotted and
shown to overlay the material data. This
workshop problem exemplifies how every
point in the sheet must follow the materials
constitutive behavior as well as being in
equilibrium throughout the deformation.
The vertical line in the history plot to the
right is the elastic unloading of this point in
the sheet.
This is a break forming problem where a punch indents a sheet over a
table to make an bracket. The problem geometry is shown below:
MESH GENERATION
COORDINATE SYSTEM SET: GRID ON
V DOMAIN -.7 .4 <cr>
FILL, RETURN
CURVES ADD (pick indicated points on grid)
POINT (1,0,0), POINT(.3,0,0)
POINT(.3,0,0), POINT(.3,-.6,0)
POINT(.3,-.6,0), POINT(-.3,-.6,0)
POINT(-.3,-.6,0), POINT(-.3,0,0)
POINT(-.3,0,0), POINT(-1,0,0)
Grid spacing 0.1" X 0.1"
CURVE TYPE
FILLET
RETURN
CURVES ADD
(right horizontal curve, right vertical curve)
radius = 0.1 <cr>
(left horizontal curve, left vertical curve)
radius = 0.1 <cr>
CURVE TYPE
CIRCLES: CENTER/RADIUS
RETURN
CURVES ADD
0 .2 0
.1 <cr>
ELEMENTS ADD (pick points on grid)
POINT (-.9,0,0) , POINT(.9,0,0)
POINT(.9,.1,0), POINT(-.9,.1.0)
SUBDIVIDE
DIVISIONS 30, 3, 1
ELEMENTS
ALL:EXISTING, RETURN
SWEEP, ALL, RETURN
RENUMBER, ALL, RETURN
COORDINATE SYS: SET GRID OFF
RETURN
RETURN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISP
X=0
OK
NODES:ADD
(pick nodes along x=0, except node touching circle)
MAIN
MATERIAL PROPERTIES
ISOTROPIC
E = 3E7
= .3, OK
ELEMENTS ADD
ALL: EXISTING
TABLES
NEW, 1 IND. VARIABLE
TABLE TYPE
eq_plastic_strain, OK
FORMULA, ENTER
5E4*(1+V1^.6) <cr>
FIT
NEW, 1 IND. VARIABLE
TABLE TYPE time, OK
ADD POINT
0, 0, .5, -.3, 1, 0 <cr>

FIT
SHOW MODEL
RETURN
ISOTROPIC
ELASTIC-PLASTIC
INITIAL YIELD STRESS = 1
TABLE1 = table1 (eq_plastic_strain)
OK (twice), MAIN
CONTACT
CONTACT BODIES
DEFORMABLE OK
ELEMENTS ADD ALL:EXISTING
NEW
RIGID
POSITION PARAMS: Y=1
TABLE=table2 (time), OK (twice)
CURVE ADD, pick cylinder, END LIST
ID CONTACT
NEW
CONTACT BODY TYPE RIGID OK
CURVES ADD, pick all remaining curves
END LIST, MAIN
LOADCASES
MECHANICAL
STATIC
LOADCASE TIME = .5
# OF STEPS 50
CONVERGENCE TESTING
DISPLACEMENT
OK (twice)
NEW
STATIC
LOADCASE TIME = .5
# OF STEPS 20
CONVERGENCE TESTING
DISPLACEMENT
RELATIVE/ABSOLUTE
MIN. DISP. CUTOFF = 1E-5
MAX ABS. DISP = 1E-5
OK
SOLUTION CONTROL
NON-POSITIVE DEFINITE
OK (twice), MAIN
JOBS
MECHANICAL
ANALYSIS OPTIONS
LARGE DISPLACEMENT
ADV. OPT., CONSTANT DILATATION,OK
LARGE STRAIN ADDITIVE
OK
lcase1
lcase2
ANALYSIS DIMENSION: PLANE STRAIN
JOB RESULTS
EQUIVALENT VON MISES STRESS
OK
CONTACT CONTROL
ADVANCED CONTACT CONTROL
SEPARATION FORCE = .1 <cr>, OK
OK (twice)
SAVE
RUN
SUBMIT
MONITOR
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
CONTOURS
SKIP TO INCREMENT 50
RESULTS
SKIP TO INCREMENT 70
RESULTS
HISTORY PLOT
SET NODES (pick bottom middle node)
COLLECT GLOBAL DATA
NODES/VARIABLES
ADD VARIABLE
Pos Y cbody2
Force Y cbody2
FIT
RESULTS
HISTORY PLOT
CLEAR CURVES
COLLECT DATA 1 11111 1
NODES/VARIABLES
ADD VARIABLE
Total Equivalent Plastic Strain
FIT
Statics S5 Super Plastic Forming
Overview: A flat sheet is formed into a
rigid die by pressure. The die is three-
dimensional and represents a corner of a
pan. This fine-grained material is assumed
to be a rigid-plastic material with no
elasticity and the flow stress is only a
function of the strain rate.
As the sheet contacts the die, friction
causes the thickness of the sheet to vary. In
addition, the pressure must be adjusted to
keep this strain rate sensitive material
within a certain target range. This is
necessary to maintain the proper flow of
the super plastic material.
Prediction of thinning of the sheet is very
important since the sheet may become too
thin for its application.
A user subroutine is used to adjust the
pressure on the sheet to keep within the
target strain rate. Furthermore, the
maximum pressure is limited by the
capacity of the rig used to form the sheet.
This problem will do a Super
Plastic Forming of a corner. It
requires a Fortran compiler
present on your machine. Also,
you will copy user subroutines
from the MSC.Marc
demonstration directory.
To proceed, lets copy the user subroutines to your working directory.
Enter the command:
find / -name u3x32b.f -print
it should return something like, ~marc/demo/u3x32b.f
then copy it to your working directory,
cp ~marc/demo/u3x32b.f .
FILES
SAVE AS spf
RETURN
MESH GENERATION
COORDINATE SYSTEM SET:
GRID ON
U DOMAIN -7 7<cr>
U SPACING = .5<cr>
V DOMAIN 0 5 <cr>
V SPACING = .5<cr>
FILL
RETURN
CURVES: ADD
point( 7.0, 4.5, 0.0)
point( 4.0, 4.5, 0.0)
point( 4.0, 4.5, 0.0)
point( 3.5, 0.0, 0.0)
point( 3.5, 0.0, 0.0)
point( 0.0, 0.0, 0.0)
point( 0.0, 0.0, 0.0)
point( -4.0, 0.0, 0.0)
CURVE TYPE
FILLET, RETURN
CURVES: ADD
1
2
.5
8
3
.5
VIEW
SHOW VIEW 2
FILL
RETURN
EXPAND
SHIFT
TRANSLATIONS
0 0 3.5 <cr>
CURVES
ALL: EXISTING
RESET
SHIFT
POINT
0 0 3.5 <cr>
ROTATIONS
0 -90/10 0 <cr>
REPETITIONS
10 <cr>
CURVES
(pick curves shown)
PICK
THESE
FILL
RESET
SHIFT
POINT
0 0 3.5 <cr>
TRANSLATIONS
-4 0 0 <cr>
REPETITIONS
1 <cr>
CURVES
pick curves shown, END LIST
RETURN
CURVES REMOVE
ALL: EXISTING
Now we will add nodes on the grid that will form the surfaces that will
contain the mesh.
SELECT
SURFACES
ALL: EXISTING
MAKE INVISIBLE
RETURN
CURVE TYPE
LINE
RETURN
CURVES: ADD
POINT( -3.5, 5.0, 0.5)
POINT( 0.0, 5.0, 0.5)
PICK
THESE
POINT( 0.0, 5.0, 0.5)
POINT( 6.5, 5.0, 0.5)
EXPAND
RESET
SHIFT
TRANSLATIONS
0 0 3.0 <cr>
CURVES
ALL: EXISTING
RESET
SHIFT
POINT
0 0 3.5 <cr>
ROTATIONS
0 -90/10 0 <cr>
REPETITIONS
10 <cr>
CURVES
(pick curve shown)
RESET
SHIFT
TRANSLATIONS
-3.5 0 0 <cr>
REPETITIONS
1 <cr>
PICK
PICK
CURVES
pick curve shown
END LIST
RETURN
CONVERT
DIVISIONS
10 1 <cr>
SURF. TO ELEMS
pick those shown
DIVISIONS
10 10 <cr>
SURF. TO ELEMS
pick remaining
END LIST
RETURN
SWEEP
ALL
RETURN
RENUMBER
ALL
RETURN
RETURN
INITIAL CONDITIONS
MECHANICAL
STRESS
USER SUB. UINSTR
PICK
ON STRESS1, OK
ELEMENTS ADD
ALL: EXISTING
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
SELECT
ELEMENTS
ALL: EXISTING
MAKE VISIBLE
RETURN
NEW
FIX X,Y,Z = 0
SELECT
METHOD PATH
NODES
(pick 1st middle and last node of outer path)
END LIST
RETURN
NODES: ADD
ALL: SELECTED
NEW
FIX X = 0, OK
NODES: ADD
(along x=0)
END LIST
NEW
FIX Z = 0
NODES: ADD
(along z=0)
END LIST
NEW
FACE LOAD
USER SUB. (it uses
sub FORCEM)
ON PRESSURE, OK
FACES: ADD
ALL: EXISTING
MAIN
MATERIAL PROPERTIES
NEW
ISOTROPIC
PLASTICITY
RIGID-PLASTIC
USER SUB URPFLO
INITIAL YIELD = 50 <cr>
OK
OK
ELEMENTS ADD
ALL: EXISTING
RETURN
3-D
MEMBRANE
THICKNESS = .080 <cr>
OK
ELEMENTS ADD
ALL: EXISTING
SELECT
MAKE INVISIBLE
MAIN
VIEW
SHOW VIEW 1, RETURN
CONTACT
CONTACT BODIES
NEW, NAME workpiece <cr>
DEFORMABLE
FRIC. COEFF = .3 <cr>
OK
ELEMENTS ADD
ALL: EXISTING
NEW, NAME die <cr>
RIGID
VELOCITY PARAMETERS
INIT. VEL Y = 1 <cr>
OK
FRIC. COEFF = .3 <cr>
OK
SURFACES ADD (pick surfaces forming die)
ID BACKFACES (flip die surfaces until gold
color will touch the workpiece)
MAIN
LOADCASES
MECHANICAL
RIGID PLASTIC
SOLUTION CONTROL
MAX # RECYCLES = 30 <cr>
NON-POSITIVE DEFINITE
OK
TOTAL LOADCASE TIME = 3000 <cr>
MECH. LOAD (EQUILIB)
MAX INC = 500 <cr>
# RECYCLES = 10 <cr>
INIT FRAC = 1e-4 <cr>
MAX FRAC = 5e-3 <cr>
MIN MULT. = 2e-2 <cr>
MAX MULT. = 1.05 <cr>
OK
OK
MAIN
JOBS
MECHANICAL
lcase1
ANALYSIS OPTIONS
LARGE DISPLACEMENT
FOLLOWER FORCE
OK
JOB RESULTS
EQUIVALENT PLASTIC STRAIN RATE
THICKNESS OF ELEMENT
user1 (this will be pressure in sub PLOTV)
OK
CONTACT CONTROL
BIAS = .9 <cr>
COULOMB
REL. SLIDING VEL. = 2E-5 <cr>
SEPARATION FORCE = 1E6 <cr>
OK (twice)
ELEMENT TYPES, MECHANICAL
3-D MEM/SHELL
18, OK
SAVE, OK (twice)
RUN
USER SUB. FILE
u3x32b.f, OK
COMPILE AND SAVE
SUBMIT1
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
NEXT INC
DEF ONLY
CONTOUR BAND
SCALAR
THICKNESS
SCAN
SKIP TO INC
last increment
Node B
Node A
PATH PLOT
SET NODES
(Node A) (Node B)
END LIST
VARIABLES
ADD CURVE
ARC LENGTH
THICKNESS
FIT
RESULTS
HISTORY PLOT
SET NODE
1
END LIST
COLLECT DATA
0 111111 1 <cr>
NODES/VARIABLES
ADD VARIABLE
TIME
USER1 (pressure)
FIT
Because of the popularity of using MSC.Marc for spf, this
feature has now been integrated into the product and no longer
requires user subroutines to perform the analysis. This would
allow us to take advantage of other features like adaptive
meshing which would be difficult with the approach above.
To this end, lets use the above model turning on the newer
features.
FILES
OPEN spf
SAVE AS spf_new
RETURN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply4, OK
FACE LOAD
SUPERPLASTICITY CONTROL
ON PRESSURE NEGATIVE, OK
MAIN
INITIAL CONDITIONS
REM icond1, OK
MATERIAL PROPERTIES
ISOTROPIC
PLASTICITY: RIGID-PLASTIC
SUPERPLASTIC CONTROL
fill out as shown below
OK (twice)
MAIN
B
N
=
LOADCASES
MECHANICAL
STATIC
SUPERPLASTIC CONTROL
PRESSURE
fill out as shown below
OK (twice)
MAIN
JOBS
RUN
USER SUBROUTINE CLEAR
SUBMIT1
MONITOR
OK
SAVE
On postprocessing.... similar results
Other things to try.... local remeshing
Statics S6 Creep of a Tube
Overview: A stainless steel oval tube is
pressurized at high temperature and over
time will creep. Only half of the tube is
modeled due to symmetry.
The material constitutive behavior has
the creep strain rate dependent upon the
stress level. The material data has been
fitted with a power relation where the
creep strain rate becomes: ,
where a and b are material constants.
The oval tube will bulge and become a
completely circular tube over time.
The tube will finally rupture due to the
large strains.
Plotting the displacement of the bulge
versus time shows a quick growth
followed by a slower growth, because
the stresses drop with time.
A more complex constitutive relation
may be easily modeled with the user
subroutine, CRPLAW.
c
a
b
=
FILES
NEW,OK
SAVE AS creep
RETURN
MESH GENERATION
COORDINATE SYSTEM:
SET, GRID ON
U DOMAIN 0 1<cr>
U SPACING 0.065 <cr>
V DOMAIN -1 1<cr>
V SPACING 0.065 <cr>
FILL
RETURN
CURVE TYPE ARC
CENTER/PT/PT
RETURN
CURVES: ADD
(arcs shown)
CURVE TYPE
LINE
RETURN
CURVES: ADD
(lines shown)
SURFACE TYPE
RULED, RETURN
SURFACES ADD
(pick interior and opposite
exterior arcs continue for lines)
CONVERT
DIVISONS
15 4 <cr>
SURF. TO ELEMS
pick largest surface
DIVISIONS
10 4 <cr>
SURF. TO ELEMS
pick smallest surface
RETURN
SYMMETRY
NORMAL 0 1 0
ELEMENTS
CHECK
UPSIDE DOWN
FLIP
ALL: SELECTED, RETURN
SWEEP
ALL, RETURN
RENUMBER
ALL, MAIN
INITIAL CONDITIONS
MECHANICAL
NODAL TEMPERATURE
VALUE = 1660 <cr>
OK
NODES ADD
ALL: EXISTING
RETURN (twice)
BOUNDARY CONDITIONS
MECHANICAL
FIX DISPLACEMENT
FIX X=0
NODES: ADD
all on x=0 axis, END LIST
NEW
FIX Y=0
NODES: ADD
at line of symmetry y=0
NEW, MORE
NODAL TEMPERATURE
TEMPERATURE = 1600, OK
NODES ADD, ALL: EXISTING
NEW
EDGE LOAD
PRESSURE = 66 <cr>, OK
SELECT
METHOD PATH, OK
EDGES
pick node path on interior
RETURN
EDGES: ADD
ALL: SELECTED
MAIN
MATERIAL PROPERTIES
ISOTROPIC
E = 21.4E6 <cr>
= .3 <cr>
CREEP
COEFICIENT = 4E-24 <cr>
STRESS EXPONENT = 4.51 <cr>
OK (twice)
ELEMENTS ADD
ALL: EXISTING
RETURN

PLANAR
PLANE STRAIN
THICKNESS = 1 <cr>
CONSTANT DILATATION
ASSUMED STRAIN, OK
ELEMENTS ADD
ALL: EXISTING
MAIN
LOADCASES
MECHANICAL
CREEP
TOTAL LOADCASE TIME = 3.47E6 <cr>
CREEP STRAIN/STRESS PARAMETERS
INITIAL TIME STEP = 1 <cr>
MAX. # INCS = 2000 <cr>
STRESS CHANGE TOL = 1 <cr>
OK (twice)
RETURN (twice)
JOBS
MECHANICAL
lcase1
PLANE STRAIN
ANALYSIS OPTIONS
LARGE DISPLACEMENT
ADVANCED OPTIONS
UPDATE LAGRANGE
OK
FOLLOW FORCE ON
OK
JOB RESULTS
Total Equivalent Creep Strain
Temperature (Integration Point)
OK (twice)
SAVE
RUN, SUBMIT1, MONITOR, OK, RETURN
RESULTS
OPEN DEFAULT
DEF & ORIG
CONTOUR BANDS
SCALAR
Total Equiv. Creep Strain
LAST
HISTORY PLOT
SET NODES, 80
END LIST
COLLECT DATA
0 11111 1 <cr>
NODES/VARIABLES
ADD VARIABLE
Time
Displacement X
FIT
POSTPROCESS
HISTORY PLOT
CLEAR CURVES
NODES/VARIABLES
ADD VARIABLE
Equiv. Von Mises Stress
FIT
NODE
80
Clearly as the tube creeps, the volume inside the tube increases. The
increase in volume decreases the internal pressure and the creep
deformation is reduced. To simulate this effect, we can model the
cavity of air inside the tube. This cavity will monitor the volume and
adjust the pressure according to the ideal gas law.
FILES
OPEN creep OK
SAVE AS creep2 OK
RETURN
MESH GENERATION
ELEM. CLASS LINE(2)
ELEMS ADD
pick interior nodes N1, N3
CAVITIES
NEW
SELECT
METHOD PATH
EDGES
pick interior nodes N1, N2, N3
END LIST
RETURN
EDGES ADD, ALL: SELECTED
REF. PRESSURE = 15
REF. TEMPERATURE = 1660
REF. DENSITY = 1.8E-5
MAIN
N1
N3
N2
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply4
MORE
CAVITY PRESSURE LOAD
PRESSURE = 66, OK
CAVITIES ADD
ALL EXISTING
NEW
CAVITY MASS LOAD
MASS
CLOSED CAVITY
CAVITIES ADD
ALL EXISTING
MAIN
CHECK
FLIP ELEMENTS
(pick element added to
close
the cavity properly)
MAIN
LOADCASES
MECHANICAL
CREEP
apply4 (off)
apply5 (on)
OK (twice)
JOBS
ELEMENT TYPES
MECHANICAL
MISCELLANEOUS
171
pick element added
OK (thrice)
SAVE, RUN
SUBMIT(1), MONITOR, OK
MAIN
RESULTS
OPEN DEFAULT
DEF & ORIG
CONTOUR BANDS
SCALAR
LAST
HISTORY PLOT
SET NODES, 80
NODE
80
END LIST
COLLECT DATA
0 11111 1 <cr>
NODES/VARIABLES
ADD VARIABLE
Time
Displacement X
FIT
GEN. XY PLOT COPY TO
SHOW HISTORY
RETURN
CLEAR CURVES
NODES/VARIABLES
ADD VARIABLE
Equiv. Von Mises Stress
FIT
MAIN
RESULTS
OPEN creep_job1.t16 OK
HISTORY PLOT
SET NODES, 80
END LIST
COLLECT DATA
0 11111 1 <cr>
NODES/VARIABLES
ADD VARIABLE
NODE
80
NODE
80
Time
Displacement X
FIT
GEN. XY PLOT COPY TO
FIT
Clearly the reduction in pressure due to the increase in volume
reduced the creep deformation of the tube.
Statics S7 Interference Fit
Overview: Two concentric cylinders are
fitted together with an interference fit
using the contact option and rigid
bodies of symmetry. Each cylinder is
modeled using axisymmetric elements.
Since the inner cylinder is slightly
bigger that the hole in the outer
cylinder, stresses will be generated as
the fit is finished. The hoop stress of the
outer cylinder will go into tension, and
the hoop stress of the inner cylinder will
go into compression.
The contour plot shows the strength
ratio, namely the ration of the
equivalent stress to the strength of the
material. This ratio is largest in the
outer cylinder where it touches the
inner cylinder.
Plotting the radial and hoop
components along the radius. It is seen
that the radial stress is continuous
across the interface, and the hoop stress
switches from compression in the inner
cylinder to tension in the outer cylinder.
FILES
NEW, OK
SAVE AS interf
RETURN
MESH GENERATION
COORDINATE SYS: SET GRID
ON
U DOMAIN 0 1.1<cr>
U SPACING 0.1 <cr>
V DOMAIN 0 3.1<cr>
V SPACING 0.1 <cr>
FILL
RETURN
CURVES: ADD
POINT(0.0,0.0,0.0)
POINT(0.0,3.1,0.0)
ELEMENTS: ADD
NODE(0.0, 1.0,0.0)
NODE(1.1, 1.0,0.0)
NODE(1.1, 2.0,0.0)
NODE(0.0, 2.0,0.0)
NODE(0.0, 2.1,0.0)
NODE(1.0, 2.1,0.0)
NODE(1.0, 3.1,0.0)
NODE(0.0, 3.1,0.0)
SUBDIVIDE
DIVISIONS 15 15 1 <cr>
ELEMENTS
ALL: EXISTING
RETURN
SWEEP
REMOVE UNUSED: NODES
ALL
RETURN
RENUMBER
NODES DIRECTED
0.0001 1 0 <cr>
RETURN
MOVE
TRANSLATIONS
0 -0.1 0 <cr>
ELEMENTS
(pick top cylinder), END LIST
MAIN
MATERIAL PROPERTIES
ISOTROPIC
E = 3E7 <cr>
= .3 <cr>
PLASTICITY
INITIAL YIELD=5E4 <cr>
OK (twice)
ELEMENT ADD
ALL: EXISTING
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE, OK
ELEMENTS: ADD
(pick inner cylinder)
NEW
DEFORMABLE, OK
ELEMENTS ADD
(pick outer cylinder)
NEW
SYMMETRY
DISCRETE, OK
CURVES ADD
(pick symmetry curve)
ID CONTACT
RETURN

CONTACT TABLES
NEW
PROPERTIES
TOUCH ALL
TOUCHING BODIES cbody1, cbody2
INTERFERENCE CLOSURE=4E-3 <cr>
OK (twice)
MAIN
LOADCASES
MECHANICAL, STATIC
CONTACT, CONTACT TABLE, ctable1, OK
# STEPS = 1
OK
MAIN
JOBS
MECHANICAL
lcase1
AXISYMMETRIC
JOB RESULTS
EQ. VON MISES
STR/YIELD RATIO
TENSORS STRESS
OK (twice)
ELEMENT TYPES
MECHANICAL
AXISYM. SOLID
116, OK
ALL: EXISTING
RETURN (twice)
SAVE, RUN, SUBMIT(1), MONITOR, OK
RESULTS
OPEN DEFAULT
LAST
SCALAR
EQ. STRESS/YIELD, OK, CONTOUR BANDS
RESULTS
PATH PLOT
NODE PATH
1 241 257 497, END LIST
VARIABLES
ADD CURVE
ARC LENGTH
Comp 22 Of Stress
ADD CURVE
ARC LENGTH
Comp 33 Of Stress
FIT
Component 22 of stress is the radial stress. It is in compression
and is continuous across the interface between the two cylinders.
Also the radial stress vanishes on the free surfaces of the
cylinders. Component 33 of stress is the hoop stress, with the
inner cylinder being compressed and the outer cylinder being
expanded. The Equivalent Stress/Yield Strength ratio in the
contour plot show that the outer cylinder at 94% of yield.
Statics S8 Hertz Contact Problem
Overview: In this example problem, a
steel cylinder with a radius of 5 is
pressed against a 2 deep aluminum base.
The problem is linear except for contact.
The problem is similar to the Hertz
contact problem (see Timoskenko and
Goodier, 1951).
In this problem, you will modify and
existing model and add quadratic
elements with contact. The steel material
properties have an Elastic Modulus of
30E6 and a Poissons ratio of 0.30 and the
aluminum properties have an Elastic
Modulus of 10E6 and a Poissons ratio
of 0.33. The cyclinder and base plate
are pressed together with a load of
100,000 lb/ inch.
FILES
OPEN
hertzbase.mud
OK
MAIN
JOBS
RUN
SUBMIT(1)
MONITOR
OK
MAIN
RESULTS
OPEN DEFAULT
DEF ONLY
SCALAR
Comp 22 of Stress
OK
CONTOUR BANDS
LAST
Here, we see that the peak stress using linear elements is around 141
Ksi in compression. We suspect that this is low due to the fact that
linear elements cant capture stress concentration as well as
quadratic elements. Therefore, we will change the element type and
rerun the problem.
CLOSE
FILES
SAVE AS hertzbasequad.mud
MAIN
First we will: move the aluminum sheet down one inch, attach edges to
the arc, change element types, sweep and move the aluminum sheet
back to its original position.
MESH GENERATION
MOVE
TRANSLATIONS
0 -1 0
ELEMENTS
(select the aluminum elements)
END LIST
SELECT
METHOD = PATH, OK
EDGES
(pick nodes N1, N2, N3)
END LIST
OK (twice)
ATTACH
EDGES TO CURVES
(select circular arc)
ALL: SELECTED EDGES
RETURN
CHANGE CLASS
QUAD(8)
ELEMENTS
ALL: EXISTING
RETURN
SWEEP, ALL, RETURN
RENUMBER, ALL, RETURN
N1
N2
N3
MOVE
TRANSLATIONS
0 1 0
ELEMENTS
(select aluminum elements)
END LIST
MAIN
JOBS
ELEMENT TYPES
MECHANICAL
PLANE STRAIN SOLID
27, OK
ALL: EXISTING
RETURN (twice)
MECHANICAL
CONTACT CONTROL
QUAD. SEGMENTS GENUINE, OK (thrice)
SAVE
RUN
SUBMIT(1)
MONITOR, OK
MAIN
RESULTS
OPEN DEFAULT
LAST
DEF ONLY
SCALAR
(Comp 22 of Stress)
CONTOUR BANDS
SELECT
SELECT CONTACT BODY ENTITIES
(steel), OK
MAKE VISIBLE
FILL
FEA versus Theoretical Solutions
From the 6th Edition of
Roarkss Formulas for Stress
and Strain (by Warren C.
Young, 1989, pg 651) we can
derive the contact patch, b,
and the maximum
compressive stress, .
The contact width for model depicted is given by .
Where , and the contact area for the half
model becomes, . The maximum stress
becomes
Theory versus FEA
Theory Linear Ele-
ments
Quadratic
Elements
230.9 131.8 229.8
Error
(%)
43.0 0.5
Max
b 1.60 pK
D
C
E
=
C
E
1
1
2
E
1
-----------------
1
2
2
E
2
-----------------
+ =
b
2
---
0.80 pK
D
C
E
0.276 = =
Max 0.798
p
K
D
C
E
--------------- 230.9Ksi = =
Max ksi [ ]
Statics S9 Elastomeric Arch
Overview: An elastomeric arch has a
center load applied and the objective
of the analysis is to determine the snap
though in the force displacement
response.
An adaptive load stepping method
called arc-length (modified Riks-Ramm)
is used.
Force
F
o
r
c
e
Displacement
MESH GENERATION
COORDINATE SYSTEM SET
CYLINDRICAL
SET: GRID ON
RETURN
CURVE TYPE ARC CPP
RETURN
CURVES ADD
0 0 0
.7 30 0
.7 150 0
0 0 0
.8 30 0
.8 150 0
SURFACE TYPE RULED
RETURN
SURFACE: ADD
2 1 <cr>
CONVERT
DIVISONS
20 3 <cr>
ALL: EXISTING
RETURN
GRID OFF
FILL
MAIN
BOUNDARY CONDITIONS
MECHANICAL
FIXED DISP
X=0
Y=0
NODES ADD
(nodes at both ends)
NEW
POINT LOAD
Y FORCE
-0.03 OK
TABLES
DATA POINTS ADD
0 0 1 1 2 0
TABLE TYPE TIME
SHOW MODEL, RETURN
NODES ADD
(top center node)
POINT LOAD
(attach table to y force)
MAIN
MATERIAL PROPERTIES
MORE
MOONEY
C10 = 1, OK
ELEMENTS ADD
ALL EXISTING
MAIN
LOADCASES
MECHANICAL
STATIC
ARC LENGTH PARAMETERS
INITIAL FRACTION = 0.1
OK (twice)
COPY
MAIN
JOBS: MECHANICAL
lcase1
lcase2
PLANE STRAIN
JOB RESULTS
CAUCHY STRESS TENSOR
OK (twice)
ELEMENT TYPES
MECHANICAL
PLANE STRAIN
80 OK
ALL EXISTING
RETURN (twice)
RUN
SUBMIT1
MONITOR
OK (twice)
SAVE
RESULTS
OPEN DEFAULT
SKIP TO INC 29, OK
DEF ONLY
CONTOUR BAND
SCALAR
EQ. CAUCHY STRESS
HISTORY PLOT
SET NODE
11
# END LIST
(pick top center node)
GLOBAL COLLECT DATA
NODES/VARIABLES
ADD VARIABLE
Displacement Y
External Force Y
FIT
Heat Transfer H1 Transient Cooling Fin
Overview: A planar slab of material is
subjected to heat loads and the
resulting transient response is
determined. The slab has convection
boundary conditions on the left and
right surfaces as shown. The top and
bottom horizontal surfaces are
adiabatic. The slab is at an initial
temperature of 70
o
F.
The left surface is exposed to a hot
environment whereas the right surface
is exposed to cooling conditions. The
purpose of the fin on the right side is to
create more surface area for cooling
and improve the cooling effectiveness of
the slab.
The contour of temperature is at a time
of 6 seconds, and the temperature
history of the two points shown are
plotted. These plots show that the slab
has yet to reach steady state.
Pick
FILES
SAVE AS heat1, OK
MAIN
MESH GENERATION
COORDINATE SYSTEM SET:
GRID ON
U SPACING 0.1 <cr>
U DOMAIN 0 1<cr>
V SPACING 0.1 <cr>
V DOMAIN -1 1<cr>
FILL, RETURN
ELEMENTS: ADD
NODE( 0.0, -1.0, 0.0)
NODE( 1.0, -1.0, 0.0)
NODE( 1.0, 1.0, 0.0)
NODE( 0.0, 1.0, 0.0)
SUBDIVIDE
DIVISIONS
8 9 1 <cr>
ELEMENTS
ALL: EXISTING
RETURN
ELEMENTS REMOVE
(pick those shown)
END LIST
COORDINATE SYS: SET GRID OFF
CHECK
UPSIDE DOWN
FLIP
ALL: SELECTED
RETURN
SWEEP
REMOVE UNUSED
NODES
ALL
RETURN
RENUMBER
ALL
MAIN
BOUNDARY CONDTIONS
THERMAL
EDGE FILM (TOP)
H=800/(3600*144) <cr>
Tinf=2500 <cr>
EDGES: ADD
pick edges on left vertical surface
NEW
EDGE FILM (TOP)
H=600/(3600*144) <cr>
Tinf=1000 <cr>
EDGES: ADD
pick right surfaces as shown, END LIST
MAIN
INITIAL CONDITIONS
THERMAL
TEMPERATURE = 70 <cr>
OK
NODES: ADD
ALL: EXISTING
RETURN (twice)
MATERIAL PROPERTIES
HEAT TRANSFER
CONDUCTIVITY
6E-4 <cr> (BTU/s/in/F)
SPECIFIC HEAT
.146 <cr> (BTU/LBF-F)
MASS DENSITY
.283 <cr> (LBF/in^3)
OK
ELEMENTS: ADD
ALL: EXISTING
RETURN
LOADCASES
HEAT TRANSFER
TRANSIENT
TOTAL LOADCASE TIME=6 <cr>
ADAPTIVE LOADING TEMPERATURE
MAX # INCREMENTS = 200 <cr>
INITIAL TIME STEP = 1 <cr>
OK (twice)
RETURN (twice)
JOBS
HEAT TRANSFER
lcase1
PLANAR
ANALYSIS OPTIONS
LUMP MASS
OK (twice)
SAVE
RUN
SUBMIT1
MONITOR
RETURN (twice)
RESULTS
OPEN DEFAULT
CONTOUR BANDS
SKIP TO INC
(last increment)
HISTORY PLOT
SET NODES
(pick those shown)
END LIST
COLLECT DATA
1 11111 1
NODES/VARS
ADD VARIABLE
Time
Temperature
FIT, RETURN (twice)
Pick
PATH PLOT
SHOW MODEL
NODE PATH
(pick two nodes shown)
END LIST
VARIABLES
ADD CURVE
Arc Length
Temperature
FIT
RETURN
YMIN = 70 <cr>
REWIND
MONITOR
Notice how the heat flows
into the fin, with the
interior slower to respond.
Heat Transfer H2 Steady State Cooling Fin
Overview: The planar slab of the
previous problem is subjected to
the same heat loads and after the 6
second transient, a steady state
loadcase follows. Following a
transient loadcase with a steady
state case can help determine if the
transient has completed.
Notice the jump to steady state in
the temperature history plot.
Increasing the time period of the
transient loadcase would then
show the complete transient
response from initial conditions
to steady state.
{
{
Jump to Steady State
Pick
Because we are not sure if the transient in the previous problem
reaches steady state, lets include another loadcase.
FILES
OPEN heat1
SAVE AS heat2, OK
RETURN
LOADCASES
HEAT TRANSFER
NEW
STEADY STATE
OK
RETURN (twice)
JOBS
HEAT TRANSFER
lcase2
OK
SAVE
RUN
SUBMIT1
MONITOR
RETURN
MAIN
Pick
RESULTS
OPEN DEFAULT
CONTOUR BANDS
SKIP TO INC
(last increment)
HISTORY PLOT
SET NODES
(pick those shown)
END LIST
COLLECT DATA
0 11111 1 <cr>
NODES/VARS
ADD VARIABLE
Time
Temperature
FIT
It is a good modeling practice to follow a transient with a steady state
loadcase to get the proper time constant whereby the transient runs
long enough to achieve steady state. Extra Credit #1 What is the
cooling efficiency, , with and without the cooling fin present? Where
the cooling efficiency is defined as:
{
{
Jump to Steady State
1
T
avg
fin
T
avg
no fin
--------------------------
=
Heat Transfer H3 Coupled Transient Cooling Fin
Overview: The transient thermal planar
slab of the previous problem is now
coupled with a thermal stress analysis.
Mechanical boundary conditions are
added to the previous model. Here, the
bottom horizontal surface is constrained
not to displace in the vertical and the left
vertical surface is constrained not to
displace in the horizontal.
Mechanical properties are also added to
the model including the thermal
coefficient of expansion.
The transient loadcase is changed to a
quasi-static coupled loadcase. Finally, the
element types are changed to plane stress
and the job is submitted.
Stresses are generated in the slab because
of thermal growth that is constrained by
the mechanical boundary conditions. By
plotting the stress at the points shown, we
see that the stress on the hot side occurs
well before steady state.
Pick
Even though the thermal efficiency may be better with the cooling fin,
the structural response may not. Lets see how to take the previous
model and convert into a couple heat/stress problem.
FILES
OPEN heat1
SAVE AS heat1s, OK
RETURN
BOUNDARY CONDITIONS
MECHANICAL
NEW
FIX X = 0
NODES ADD
all nodes on x=0
NEW
FIX Y = 0
NODES ADD
all nodes on y = -1
RETURN (twice)
MAT. PROPERTIES
ISOTROPIC
E = 3E7
= .3
THERMAL EXP=10E-6
OK (twice)
RETURN

LOADCASES
COUPLED
QUASI-STATIC
LOADS
Pick new bcs
CONV. TESTING
DISPLACEMENTS, OK
TOTAL LOADCASE TIME
60, OK
RETURN (twice)
JOBS
COUPLED
JOB RESULTS
EQUIVALENT VON MISES STRESS, OK
INITIAL LOADS
Select new bcs, OK
ELEMENT TYPES
COUPLED
PLANE STRESS
3
OK
ALL: EXISTING
RETURN (twice)
SAVE
RUN
SUBMIT1
MONITOR
RESULTS
OPEN DEFAULT
CONTOUR BANDS
LAST
SCALAR
EQ. VON MISES
HISTORY PLOT
SET NODES
(pick nodes shown)
END LIST
COLLECT DATA
0 1111 1 <cr>
NODES/VARIABLES
ADD VARIABLE
Temperature
Eq. Von Mises Stress
Notice how the stress peaks well
before steady state because of the
nonuniform temperatures during the transient.
Pick
Plane stress was used in this example. If plane strain elements (types
11, 27, etc.) were used, the out-of-plane strain for these elements is
zero. This generates a large out-of-plane stress since for plane strain
we have:
and the last term in the equation will dominate for large changes in
temperature. If there is no out-of-plane constraint to the thermal
growth physically, plane stress should be used. If the out-of-plane
thermal growth is restricted, such as plane remaining plane,
generalized plane strain elements (types 19, 29, etc.) should be used.
You may wish to try these elements and observe what happens.
zz
E
1 + ( ) 1 2 ( )
--------------------------------------

xx

yy
1 + ( )T + [ ] =
Heat Transfer H4 Dynamics with Friction Heating
Overview: A block has an initial velocity
and will slide over a rigid table. Due to
the weight of the block and friction
between the block and the table, the
block will slow down and heat up
because of friction.
Mechanical boundary conditions keep
the block moving in a straight line. Initial
conditions set the initial velocity and
temperature.
The coupled loadcase selected is a
dynamic transient with a time period
long enough to allow the block to come to
rest.
The temperature contours show how the
leading edge of the block touching the
table heat up faster than other portions
of the block.
A history plot of the velocity and
acceleration of the node shown show how
the block comes to a stop with the
velocity and acceleration becoming zero
at 1.4 seconds.
This is a problem of a block subjected
to its own weight that is sliding on a
table with an initial velocity. Friction
between the block and table generate
heat and reduce the speed.
FILES
NEW, OK
SAVE AS block
RETURN
MESH GENERATION
VIEW SHOW VIEW 4, OK
ADD ELEMENTS
NODE( -1.0, -1.0, 0.0)
NODE( 1.0, -1.0, 0.0)
NODE( 1.0, 1.0, 0.0)
NODE( -1.0, 1.0, 0.0)
ADD SURFACES
POINT( 1.0, -1.0, 0.0)
POINT( -1.0, -1.0, 0.0)
POINT( -1.0, 1.0, 0.0)
POINT( 1.0, 1.0, 0.0)
MOVE
SCALE 4 2 1 <cr>
SURFACES
ALL: EXISTING
MOVE
RESET
TRANSLATIONS
1.8 0 0 <cr>
SURFACES
ALL: EXISTING
RETURN
SUBDIVIDE
ELEMENTS
ALL: EXISTING
RETURN
EXPAND
TRANSLATIONS
0 0 1/2 <cr>
REPETITIONS
2 <cr>
ELEMENTS
ALL: EXISTING
RETURN
FILL
SWEEP
REMOVE UNUSED
NODES
ALL, RETURN
RENUMBER
ALL, RETURN
BNDRY. CONDITIONS
MECHANICAL
FIXED DISP Y = 0 <cr>
OK
NODES ADD
ALL: EXISTING
NEW
GRAVITY LOAD
ON Z ACCEL = -9.81 <cr> m/s
2
OK
ELEMENTS ADD
ALL: EXISTING
MAIN
INITIAL CONDITIONS
THERMAL
TEMP. = 0 <cr>
o
K
OK
NODES ADD
ALL: EXISTING
RETURN
NEW
MECHANICAL
VELOCITY
VEL X = 4.905 <cr> m/s
OK
NODES ADD
ALL: EXISTING
MAIN
MATERIAL PROP.
NEW
ISOTROPIC
E = 210E9 <cr> (N/m
2
)
= .3 <cr>
= 7854 <cr> (Kg/m
3
)
DAMPING
NUM. MULT
0.3, OK (twice)
HEAT TRANSFER
CONDUCTIVITY
60.5 <cr> (W/m
o
K)
SPECIFIC HEAT
434 <cr> (J/Kg
o
K)
MASS DENSITY
7854 <cr> (Kg/m
3
)
OK
ELEMENTS ADD
ALL: EXISTING
RETURN

CONTACT
CONTACT BODIES
DEFORMABLE
= .5 <cr>, OK
ELEMENTS ADD
ALL: EXISTING
CONTACT
CONTACT BODIES
NEW
RIGID
= .5 <cr>, OK
SURFACES ADD
ALL: EXISTING
MAIN
LOADCASES
COUPLED
DYNAMIC TRANSIENT
SOL. CONTROL
NON-POSITIVE DEF
OK
CONV. TEST. DISP
OK
TOTAT LOADCASE TIME
2 <cr>
FIXED # STEPS = 50 <cr>
OK
MAIN

JOBS
COUPLED
lcase1
ANALYSIS OPTIONS
LARGE DISPLACEMENT
LUMPED MASS
OK
CONTACT CONTROL
COULOMB
SLIDING VEL = 0.1 <cr>
SEP. FORCE = 1E11 <cr> (keep block on surface)
OK (twice)
JOB RESULTS
EQUIVALENT VM STRESS
TEMPERATURE
OK
JOB PARAMETERS
HEAT GEN: CONV FAC = 1E3 <cr> (should
be 1, but want larger temps for show)
OK
OK
SAVE
RUN
SUBMIT1
MONITOR
OK
RETURN
RESULTS
OPEN DEFAULT
CONTOUR BAND
DEF ON
SCALAR Temp.
SKIP TO 50
RESULTS
HISTORY PLOT
SET NODES
(pick node shown)
COLLECT DATA
0 11111 1
NODES/VARS
ADD VARIABLE
Time
Velocity x
ADD VARIABLE
Time
Acceleration x
FIT
Pick Node
Notice that the effect of friction was not 100% since the block should
come to a stop at 1 sec. This was due to the ever slipping friction model.
Rigid body dynamics gives:
where the initial velocity was selected as . Where is the
stopping time or 1 second.
Also from the friction heating, the friction force moves through a
distance and this mechanical energy is converted to thermal energy.
This thermal energy is input to the heat transfer portion of the
solution. The average rise in temperature for a block that comes to rest
from an initial velocity of , becomes:
In this case, the rise in temperature is 27.27
o
K.
How does this compare with the MSC.Marc predictions? (28
o
K)
Why is the block hotter at the leading bottom edge? What would you
do to improve the results?
u
g = u
gt u
0
+ = ; u g
t
2
2
----
u
0
t u
0
+ + = ;
u
0
gt
s
= t
s
u
0
T conv
factor
u
0
2
c
p
--------
\ .
|
| |
=
Heat Transfer H5 Radiation with Viewfactors
Overview: Two concentric spheres
have their inner and outer most
surfaces held at a fixed
temperature. They exchange heat
flow via radiation.
Thermal boundary conditions keep
the inner and outer most surfaces
fixed at 400 and 500 degrees C.
Another thermal boundary
conditions identifies that the outer
surface of the inner sphere and the
inner surface of the outer sphere
can radiate.
The heat transfer loadcase selected
is a steady state that will allow the
sphere to exchange heat flow via
radiation.
The temperature contours shows
this flow and the path plot shows
the radial change in temperature.
a
b
c
d
This will be an axisymmetric model and we can use cylindrical
coordinates to define the spheres.
MESH GENERATION
COORDINATE SYS.
CYLINDRICAL (on)
CURVE TYPE
CENTER POINT POINT
RETURN
CURVES ADD
0,0,0, 8,0,0, 8,180,0 <cr>
0,0,0, 10,0,0, 10,180,0 <cr>
0,0,0, 12,0,0, 12,180,0 <cr>
0,0,0, 14,0,0, 14,180,0 <cr>
SURFACE TYPE
RULLED, OK
SURFACE ADD
1, 2 <cr>
3, 4 <cr>
CONVERT
DIVISIONS 12 2 <CR>
ALL: EXISTING
RETURN
SWEEP
ALL
RETURN
CHECK ELEMENTS
UPSIDE DOWN
FLIP ELEMENTS
ALL SELECTED
UPSIDE DOWN
RETURN
RENUMBER
ALL
MAIN
BOUNDARY CONDITIONS
THERMAL
FIXED TEMP = 400, OK
(add all nodes for r = 8)
NEW
FIXED TEMP = 500, OK
(add all nodes for r = 14)
NEW
EDGE RADIATION
ON, OK
(add all edges r = 10 & 12)
Try using the path select
option to pick the nodes
on r=8, 14 and the edges
on r=10, 12. You only
need to pick a beginning
middle and ending node
for path select.
r=14 12 10 8
COMPUTE RADIATION VIEWFACTORS
TYPE AX
VIEWFACTOR FILE = model1.vfs, OK
START, OK
MAIN
MATERIAL PROPERTIES
HEAT TRANSFER
CONDUCTIVITY = 1E-4
EMISSIVITY = 0.4, OK
ELEMENTS ADD, ALL EXISTING, MAIN
LOADCASES
HEAT TRANSFER
STEADY STATE
SOLUTION CONTROL
MIN # RECYCLES
10, OK (twice)
MAIN
Here 1000 rays are
randomly cast from each
of the 24 edges to compute
the view factors. The
view-factors will be stored
in the file model1.vfs.
If the geometry changes
this would need to be
done again.
JOBS
HEAT TRANSFER
lcase1
AXISYMMETRIC
ANALYSIS OPTIONS
RADIATION
VIEWFACTOR FILE = model1.vfs, OK
LINEARIZE CALCULATION (off), OK
JOB PARAMETERS
UNITS AND CONSTANTS
TEMPERATURE IN CELSIUS (on)
STEFAN-BOLTZMANN = 5.67E-14, OK
OK, (thrice)
RUN, SUBMIT1, MONITOR, OK, SAVE
MAIN
RESULTS
OPEN DEFAULT
LAST
CONTOUR BAND
PATH PLOT
SET NODES
(pick node shown)
(a,b,c,d) #END LIST
a
b
c
d
VARIABLES
ADD CURVE
Arc Length
Temperature
FIT
Dynamics D1 Cantilever Beam Modal Analysis
Overview: A modal analysis of a
cantilever beam will be done to
determine the natural frequencies of
the cantilever beam shown.
The end load is turned off prior to the
modal analysis. The effects of pre-
stress change the natural frequencies,
like the tension in a guitar string.
However, it is not modeled here. Also,
the bending stresses due to the tip load
very slightly change the frequency.
The first ten lowest natural
frequencies and corresponding mode
shapes were requested. Here, the mode
shape of the lowest natural frequency
of 325 Hz is shown. As expected, it
shows easy wise bending.
10" X 1" X 1"
500 #
Here is a cantilever beam from
before. We will use it now.
FILES
OPEN d1
SAVE AS d12
OK
LOADCASES
MECHANICAL
DYNAMIC MODAL
OK
MAIN
JOBS
MECHANICAL
SELECT lcase1
INITIAL LOADS
turn off point load,
OK
OK
RUN
SUBMIT1, OK
SAVE
MAIN
10" X 1" X 1"
500 #
RESULTS
OPEN DEFAULT
NEXT
DEFORMED SHAPE SETTINGS AUTOMATIC
DEF & ORIG
SCAN
1st Natural frequency at 3.251E+02 cycles per time
Dynamics D2 Cantilever Beam Harmonic Analysis
Overview: A harmonic analysis of a
cantilever beam will be done to
determine the dynamic response of
the cantilever beam shown to an
oscillating tip load of 500 pounds.
The end load is turned on in the
harmonic loadcase, and the range of
excitation frequencies is 0 to 400
Hz, in 10 steps of 40 Hz.
Plotting the tip displacement
magnitude along the frequency
range, shows the static solution at 0
Hz, and the resonance around the
first natural frequency of 325 Hz,
ending with a phase reversal above
325 Hz.
10" X 1" X 1"
500 #
Static
Resonance
at 325 cps
Solution
FILES
OPEN d12
SAVE AS d13, OK
RETURN
BOUNDARY CONDITIONS
MECHANICAL
EDIT apply3
HARMONIC BCS
POINT LOAD, OK, MAIN
LOADCASES
MECHANICAL
DYNAMIC HARMONIC
LOADS
(pick point load), OK
LOWEST FREQ = 0 <cr>
HIGHEST FREQ = 400 <cr>
# OF FREQS = 40 <cr>
OK
MAIN
JOBS
MECHANICAL
SELECT lcase1 OK
RUN, SUBMIT1, SAVE
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODE (Pick the one with point load)
COLLECT DATA 0:0 0:40 1 <cr>
NODE/VARIABLES
ADD VARIABLE
FREQUENCY
DISPLACEMENT y
FIT
Static Solution
Resonance at
325 cps
Dynamics D3 Cantilever Beam Transient Analysis
Overview: A transient analysis of the
previous cantilever beam will be
done to determine the transient
dynamic response of the cantilever
beam shown to a suddenly appearing
tip load of 500 pounds.
The dynamic transient loadcase time
period is set to 3/(325 Hz) to get 3
cycles of response.
Plotting the tip displacement along
the time axis shows the tip oscillating
about the static solution.
The second run includes damping
and the tip displacement along the
time axis plot shows the tip
oscillating about the static solution
with the oscillations diminishing
with time.
The last run includes contact with a
bumper below the beam at mid span.
The beam contacts the bumper only
on the way down and separates from
the bumper when displacing
upward.
10" X 1" X 1"
500 #
Static Solution
Period
Static Solution
Period
10" X 1" X 1"
500 #
0.03"
Static Solution ?
Period ?
0.03"
Here is a cantilever beam from
before, we will use it now. The
beam is at rest and the load is
placed on the end at time t=0.
FILES
OPEN d1
SAVE AS d14, OK
RETURN
LOADCASE
MECHANICAL
DYNAMIC TRANSIENT
TIME = 3/325 <cr> (remember 1st natural
frequency)
STEPS = 150 <cr>
OK
MAIN
JOBS
MECHANICAL
SELECT lcase1
OK
SAVE, RUN, SUBMIT1, MONITOR, OK
MAIN
10" X 1" X 1"
500 #
RESULTS
OPEN DEFAULT
HISTORY PLOT
COLLECT DATA 0 150 1 <cr>
NODE/VARIABLES
ADD VARIABLE
Time
Displacement y
FIT
Static Solution
Period
What about damping? Physically, we know it is present. Lets see how
to model with damping.
FILES
OPEN d14
SAVE AS d15, OK
MATERIAL PROPERTIES
ISOTROPIC
DAMPING
STIFFNESS MATRIX MULT. = 1E-4 <cr>
OK
OK
SAVE
MAIN
JOBS
RUN
SUBMIT1
MONITOR
RESULTS
OPEN DEFAULT
HISTORY PLOT
NODE/VARIABLES
ADD VARIABLE
Time
Displacement y
FIT
Static Solution
Period
Dynamics D3 Beam Transient Analysis w/Contact
Here is a over hanging
cantilever beam. The beam is
rest and the load is placed on
the end at time t=0.
FILES
OPEN d15
SAVE AS d16, OK
MAIN
MESH GENERATION
CURVE TYPE, CIRCLE:CENTER,RADIUS
CURVES: ADD
5 0 0 <cr>
.2 <cr>
RETURN
MOVE
TRANS.
0 -.23 0 <cr>
CURVE
ALL: EXISTING
MAIN
10" X 1" X 1"
500 #
0.03"
MATERIAL PROPERTIES
ISOTROPIC
DAMPING
STIFFNESS MATRIX MULT. = 2E-5 <cr>
OK
OK
RETURN
CONTACT
CONTACT BODIES
DEFORMABLE, OK
ELEMENTS: ADD, ALL: EXISTING
NEW
RIGID, OK
CURVES ADD, ALL: EXISTING
RETURN (twice)
JOBS
MECHANICAL
CONTACT CONTROL
DISTANCE TOLERANCE = .01 <cr>
BIAS = 0.9 <cr>
SEPARATION FORCE = 1.0 <cr>
OK (thrice)
SAVE
RUN
SUBMIT(1), MONITOR, OK (twice)
RESULTS
OPEN DEFAULT
HISTORY PLOT
SET NODE
(Pick the one with point load& at overhang)
NODE/VARIABLES
ADD VARIABLE
Time
Displacement Y
FIT
Static Solution ? Period ?
0.03"
CHAPTER 11 Notes and Course Critique
The purpose of this appendix is to provide
pages for notes and the course critique.
Notes and Course Critique Class Notes
Class Notes
Class Notes Notes and Course Critique
Class Notes
Class Notes
Class Notes Notes and Course Critique
Class Notes
Class Notes
Class Critique Notes and Course Critique
Class Critique
Please use this form to provide feedback on your training program. Your comments will be reviewed, and
when possible included in the remainder of your course.
Lecture
excellent average poor
Is the level of technical detail appropriate?
Are the format and presentation correctly paced?
Are the discussions clear and easy to follow?
What changes do you suggest?
What additional information would you like?
Workshop
excellent average poor
Are the available problems relevant?
Was the technical assistance prompt and clear?
Was the equipment satisfactory?
What changes do you suggest?
What additional information would you like?
General
How would you change the balance of time spent on theory and workshop
no change more theory more workshop
Your Name:______________________________________Date:_______________________________
Instructor(s):___________________________________________________
Notes and Course Critique Class Critique

Mar 102

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Mar 102

Hochgeladen von

Copyright:

Verfügbare Formate

MA2003Z*ZZSM-MA102-NT1 1

MSC.Marc Advanced Course

MSC.Marc Advanced Course 43

Usual material data

Usual material data

MSC.Marc Advanced Course 69

Heat Transfer and Thermal Stresses Boundary conditions

MSC.Marc Advanced Course 133

Heat Transfer and Thermal Stresses Nonlinear Heat Transfer Analysis

MSC.Marc Advanced Course 139

0.0 1.0 2.0 3.0

MSC.Marc Advanced Course 183

Automatic Mesh Refinement Zienkiewicz-Zhu (Stress)

MSC.Marc Advanced Course 305

26926t 26955t pWt 6739 = = =

MSC.Marc Advanced Course 315

Statics S2 Cantilever Beam

MSC.Marc Advanced Course 349

Statics S6 Creep of a Tube

Statics S7 Interference Fit

Heat Transfer H3 Coupled Transient Cooling Fin

Heat Transfer H4 Dynamics with Friction Heating

MSC.Marc Advanced Course 429

Das könnte Ihnen auch gefallen