Ansys Ls-Dyna Users Guide

ANSYS LS-DYNA User's
Guide
ANSYS Release 10.0
002184
August 2005
ANSYS, Inc. and

ANSYS Europe,
Ltd. are UL
registered ISO
9001:2000
Companies.
ANSYS LS-DYNA User's Guide

ANSYS Release 10.0
ANSYS, Inc.
Southpointe
275 Technology Drive
Canonsburg, PA 15317
[email protected]
http://www.ansys.com
(T) 724-746-3304
(F) 724-514-9494
Copyright and Trademark Information

2005 SAS IP, Inc. All rights reserved. Unauthorized use, distribution or duplication is prohibited.
ANSYS, ANSYS Workbench, CFX, AUTODYN, and any and all ANSYS, Inc. product and service names are registered trademarks or trademarks of ANSYS, Inc.
or its subsidiaries located in the United States or other countries. ICEM CFD is a trademark licensed by ANSYS, Inc. All other trademarks or registered
trademarks are the property of their respective owners.
Disclaimer Notice
THIS ANSYS SOFTWARE PRODUCT AND PROGRAM DOCUMENTATION INCLUDE TRADE SECRETS AND ARE CONFIDENTIAL AND PROPRIETARY PRODUCTS
OF ANSYS, INC., ITS SUBSIDIARIES, OR LICENSORS. The software products and documentation are furnished by ANSYS, Inc., its subsidiaries, or affiliates
under a software license agreement that contains provisions concerning non-disclosure, copying, length and nature of use, compliance with exporting
laws, warranties, disclaimers, limitations of liability, and remedies, and other provisions. The software products and documentation may be used, disclosed,
transferred, or copied only in accordance with the terms and conditions of that software license agreement.
ANSYS, Inc. and ANSYS Europe, Ltd. are UL registered ISO 9001:2000 Companies.
U.S. GOVERNMENT RIGHTS

For U.S. Government users, except as specifically granted by the ANSYS, Inc. software license agreement, the use, duplication, or disclosure by the United
States Government is subject to restrictions stated in the ANSYS, Inc. software license agreement and FAR 12.212 (for non-DOD licenses).
Third-Party Software
See the online documentation in the product help files for the complete Legal Notice for ANSYS proprietary software and third-party software. The ANSYS
third-party software information is also available via download from the Customer Portal on the ANSYS web page. If you are unable to access the thirdparty legal notices, please contact ANSYS, Inc.
Published in the U.S.A.
Table of Contents
1. Introduction ........................................................................................................................................ 11
1.1. Overview of Steps in an Explicit Dynamic Analysis ......................................................................... 11
1.2. Commands Used in an Explicit Dynamic Analysis ........................................................................... 11
1.3. A Guide to Using this Document ................................................................................................... 13
1.4. Where to Find Explicit Dynamics Example Problems ...................................................................... 14
1.5. Additional Information ................................................................................................................. 14
2. Elements .............................................................................................................................................. 21
2.1. Solid and Shell Elements ............................................................................................................... 21
2.1.1. SOLID164 ............................................................................................................................. 21
2.1.2. SHELL163 ............................................................................................................................. 23
2.1.2.1. General Shell Formulations .......................................................................................... 23
2.1.2.2. Membrane Element Formulation ................................................................................. 24
2.1.2.3. Triangular Shell Formulations ...................................................................................... 24
2.1.3. PLANE162 ............................................................................................................................ 26
2.1.4. SOLID168 ............................................................................................................................. 28
2.2. Beam and Link Elements ............................................................................................................... 29
2.2.1. BEAM161 ............................................................................................................................. 29
2.2.2. LINK160 ............................................................................................................................. 210
2.2.3. LINK167 ............................................................................................................................. 210
2.3. Discrete Elements ....................................................................................................................... 210
2.3.1. COMBI165 Spring-Damper ................................................................................................. 210
2.3.2. MASS166 ........................................................................................................................... 211
2.4. General Element Capabilities ....................................................................................................... 211
3. Analysis Procedure .............................................................................................................................. 31
3.1. Build the Model ............................................................................................................................ 31
3.1.1. Define Element Types and Real Constants ............................................................................. 31
3.1.2. Specify Material Properties ................................................................................................... 32
3.1.3. Define the Model Geometry ................................................................................................. 32
3.1.4. Mesh the Model ................................................................................................................... 32
3.1.5. Define Contact Surfaces ....................................................................................................... 33
3.1.6. General Modeling Guidelines ............................................................................................... 34
3.2. Apply Loads and Obtain the Solution ............................................................................................ 34
3.2.1. Loads ................................................................................................................................... 34
3.2.2. Initial Velocities .................................................................................................................... 35
3.2.3. Constraints .......................................................................................................................... 36
3.2.4. DOF Coupling ...................................................................................................................... 36
3.2.5. Data Smoothing ................................................................................................................... 36
3.2.6. Specify Explicit Dynamics Controls ....................................................................................... 36
3.2.7. Save Database and Solve ...................................................................................................... 37
3.3. Review the Results ........................................................................................................................ 37
3.4. The Definition of Part .................................................................................................................... 37
3.4.1. Part Assemblies .................................................................................................................. 310
3.5. Adaptive Meshing ....................................................................................................................... 310
4. Loading ................................................................................................................................................ 41
4.1. General Loading Options .............................................................................................................. 41
4.1.1. Components ........................................................................................................................ 42
4.1.2. Array Parameters ................................................................................................................. 42
4.1.3. Applying Loads .................................................................................................................... 44
4.1.4. Data Curves ......................................................................................................................... 45
4.1.4.1. Using Data Curves with Material Models ...................................................................... 45
ANSYS LS-DYNA User's Guide . ANSYS Release 10.0 . 002184 . SAS IP, Inc.

4.1.4.2. Using Data Curves for Loading .................................................................................... 47
4.1.5. Defining Loads in a Local Coordinate System ........................................................................ 47
4.1.6. Specifying Birth and Death Times ......................................................................................... 47
4.2. Constraints and Initial Conditions .................................................................................................. 47
4.2.1. Constraints .......................................................................................................................... 48
4.2.2. Welds .................................................................................................................................. 49
4.2.3. Initial Velocity .................................................................................................................... 410
4.3. Coupling and Constraint Equations ............................................................................................. 411
4.4. Nonreflecting Boundaries ........................................................................................................... 411
4.5. Temperature Loading ................................................................................................................. 412
4.6. Dynamic Relaxation .................................................................................................................... 413
5. Solution Features ................................................................................................................................ 51
5.1. Solution Process ........................................................................................................................... 51
5.2. LS-DYNA Termination Controls ..................................................................................................... 51
5.3. Shared Memory Parallel Processing ............................................................................................... 52
5.4. Double Precision LS-DYNA ............................................................................................................ 52
5.5. Solution Control and Monitoring ................................................................................................... 53
5.6. Plotting Small Elements ................................................................................................................ 54
5.7. Editing the LS-DYNA Input File ...................................................................................................... 54
5.7.1. Method A ............................................................................................................................. 55
5.7.2. Method B ............................................................................................................................. 55
5.7.3. Using a Preexisting File.K ...................................................................................................... 56
6. Contact Surfaces .................................................................................................................................. 61
6.1. Contact Definitions ....................................................................................................................... 61
6.1.1. Listing, Plotting and Deleting Contact Entities ...................................................................... 64
6.2. Contact Options ........................................................................................................................... 65
6.2.1. Definition of Contact Types .................................................................................................. 65
6.2.2. Definition of Contact Options ............................................................................................... 66
6.3. Contact Search Methods ............................................................................................................... 69
6.3.1. Mesh Connectivity Tracking ................................................................................................. 69
6.3.2. Bucket Sort Method ........................................................................................................... 610
6.3.3. Limiting the Contact Search Domain .................................................................................. 610
6.4. Special Considerations for Shells ................................................................................................. 610
6.5. Controlling Contact Depth .......................................................................................................... 610
6.6. Contact Stiffness ......................................................................................................................... 611
6.6.1. Choice of Penalty Factor ..................................................................................................... 611
6.6.2. Symmetry Stiffness ............................................................................................................. 612
6.7. 2-D Contact Option ..................................................................................................................... 612
7. Material Models ................................................................................................................................... 71
7.1. Defining Explicit Dynamics Material Models .................................................................................. 72
7.2. Explicit Dynamics Material Model Descriptions .............................................................................. 73
7.2.1. Linear Elastic Models ............................................................................................................ 73
7.2.1.1. Isotropic Elastic Model ................................................................................................. 73
7.2.1.2. Orthotropic Elastic Model ............................................................................................ 74
7.2.1.3. Anisotropic Elastic Model ............................................................................................. 74
7.2.1.4. Elastic Fluid Model ....................................................................................................... 75
7.2.2. Nonlinear Elastic Models ...................................................................................................... 75
7.2.2.1. Blatz-Ko Rubber Elastic Model ...................................................................................... 75
7.2.2.2. Mooney-Rivlin Rubber Elastic Model ............................................................................ 76
7.2.2.3. Viscoelastic Model ....................................................................................................... 77
7.2.3. Nonlinear Inelastic Models ................................................................................................... 77
7.2.3.1. Bilinear Isotropic Model ............................................................................................... 77
vi

7.2.3.2. Temperature Dependent Bilinear Isotropic Model ........................................................ 78
7.2.3.3. Transversely Anisotropic Hardening Model .................................................................. 78
7.2.3.4. Transversely Anisotropic FLD Hardening Model ........................................................... 79
7.2.3.5. Bilinear Kinematic Model ........................................................................................... 710
7.2.3.6. Plastic Kinematic Model ............................................................................................. 711
7.2.3.7. 3-Parameter Barlat Model .......................................................................................... 711
7.2.3.8. Barlat Anisotropic Plasticity Model ............................................................................. 713
7.2.3.9. Rate Sensitive Power Law Plasticity Model .................................................................. 714
7.2.3.10. Strain Rate Dependent Plasticity Model .................................................................... 714
7.2.3.11. Piecewise Linear Plasticity Model ............................................................................. 715
7.2.3.12. Modified Piecewise Linear Plasticity Model ............................................................... 716
7.2.3.13. Composite Damage Model ...................................................................................... 718
7.2.3.14. Concrete Damage Model ......................................................................................... 718
7.2.3.15. Power Law Plasticity Model ...................................................................................... 719
7.2.3.16. Elastic Viscoplastic Thermal Model ........................................................................... 719
7.2.4. Pressure Dependent Plasticity Models ................................................................................ 721
7.2.4.1. Elastic-Plastic Hydrodynamic Model ........................................................................... 721
7.2.4.2. Geological Cap Model ................................................................................................ 722
7.2.5. Foam Models ..................................................................................................................... 724
7.2.5.1. Closed Cell Foam Model ............................................................................................ 724
7.2.5.2. Viscous Foam Model .................................................................................................. 725
7.2.5.3. Low Density Foam Model .......................................................................................... 725
7.2.5.4. Crushable Foam Model .............................................................................................. 726
7.2.5.5. Honeycomb Foam Model .......................................................................................... 726
7.2.6. Equation of State Models .................................................................................................... 727
7.2.6.1. Linear Polynomial Equation of State ........................................................................... 727
7.2.6.2. Gruneisen Equation of State ...................................................................................... 728
7.2.6.3. Tabulated Equation of State ....................................................................................... 729
7.2.6.4. Bamman Plasticity Model ........................................................................................... 729
7.2.6.5. Johnson-Cook Plasticity Model .................................................................................. 729
7.2.6.6. Null Material Model ................................................................................................... 731
7.2.6.7. Zerilli-Armstrong Model ............................................................................................. 731
7.2.6.8. Steinberg Model ........................................................................................................ 733
7.2.7. Discrete Element Models .................................................................................................... 735
7.2.7.1. Linear Elastic Spring Model ........................................................................................ 735
7.2.7.2. General Nonlinear Spring Model ................................................................................ 735
7.2.7.3. Nonlinear Elastic Spring Model .................................................................................. 736
7.2.7.4. Elastoplastic Spring Model ......................................................................................... 736
7.2.7.5. Inelastic Tension- or Compression-Only Spring Model ................................................ 736
7.2.7.6. Maxwell Viscosity Spring Model ................................................................................. 736
7.2.7.7. Linear Viscosity Damper Model .................................................................................. 737
7.2.7.8. Nonlinear Viscosity Damper Model ............................................................................ 737
7.2.7.9. Cable Model .............................................................................................................. 737
7.2.8. Other Models ..................................................................................................................... 738
7.2.8.1. Rigid Model ............................................................................................................... 738
8. Rigid Bodies ........................................................................................................................................ 81
8.1. Defining Rigid Bodies .................................................................................................................... 81
8.2. Specifying Inertia Properties ......................................................................................................... 81
8.3. Loading ........................................................................................................................................ 82
8.4. Switching Parts from Deformable to Rigid ..................................................................................... 82
8.5. Nodal Rigid Bodies ........................................................................................................................ 83
9. Hourglassing ....................................................................................................................................... 91
vii

10. Mass Scaling .................................................................................................................................... 101
11. Subcycling ....................................................................................................................................... 111
12. Postprocessing ................................................................................................................................ 121
12.1. Output Controls ........................................................................................................................ 121
12.1.1. Results (Jobname.RST) vs. History (Jobname.HIS) Files ....................................................... 121
12.1.2. Creating Components for POST26 .................................................................................... 121
12.1.3. Writing the Output Files for POST26 .................................................................................. 121
12.2. Using POST1 with ANSYS LS-DYNA ............................................................................................ 122
12.2.1. Animating Results ............................................................................................................ 122
12.2.2. Element Output Data ....................................................................................................... 123
12.2.3. Postprocessing after Adaptive Meshing ............................................................................ 124
12.3. Using POST26 with ANSYS LS-DYNA .......................................................................................... 125
12.3.1. Nodal and Element Solutions ............................................................................................ 126
12.3.2. Reading ASCII Files for Miscellaneous Output Data ............................................................ 126
12.3.3. Data Smoothing ............................................................................................................... 126
12.4. Finding Additional Information ................................................................................................. 127
13. Restarting ........................................................................................................................................ 131
13.1. The Restart Dump File ............................................................................................................... 131
13.2. The EDSTART Command ........................................................................................................... 131
13.2.1. A New Analysis ................................................................................................................. 131
13.2.2. A Simple Restart ............................................................................................................... 132
13.2.3. A Small Restart ................................................................................................................. 132
13.2.4. A Full Restart .................................................................................................................... 133
13.3. Effect on Output Files ................................................................................................................ 135
14. Explicit-to-Implicit Sequential Solution .......................................................................................... 141
14.1. Performing an Explicit-to-Implicit Sequential Solution ............................................................... 141
14.2. Troubleshooting a Springback Analysis ..................................................................................... 144
14.2.1. Springback Stabilization ................................................................................................... 145
15. Implicit-to-Explicit Sequential Solution .......................................................................................... 151
15.1. Structural Implicit-to-Explicit Solution for Preload ...................................................................... 151
15.1.1. Special Considerations for Thermal Loading ...................................................................... 154
15.2. Thermal Implicit-to-Explicit Solution .......................................................................................... 155
16. Arbitrary Lagrangian-Eulerian Formulation ................................................................................... 161
16.1. Performing an ALE Analysis ....................................................................................................... 163
17. Drop Test Module ............................................................................................................................ 171
17.1. Starting ANSYS With the Drop Test Module ............................................................................... 171
17.2. Typical Drop Test Procedure ...................................................................................................... 172
17.2.1. Basic Drop Test Analysis Procedure ................................................................................... 172
17.2.1.1. STEP 1: Create or import the model .......................................................................... 172
17.2.1.2. STEP 2: Set up the DTM ............................................................................................ 172
17.2.1.3. STEP 3: Define the magnitude of (g) ......................................................................... 173
17.2.1.4. STEP 4: Specify the drop height ................................................................................ 173
17.2.1.5. STEP 5: Orient the object .......................................................................................... 173
17.2.1.6. STEP 6: Specify solution controls .............................................................................. 173
17.2.1.7. STEP 7: Solve ........................................................................................................... 174
17.2.1.8. STEP 8: Animate results ............................................................................................ 174
17.2.1.9. STEP 9: Obtain Time-History Results ......................................................................... 174
17.2.2. Screen Coordinates Definition .......................................................................................... 174
17.2.3. Additional Notes on the Use of the DTM ........................................................................... 175
17.3. Advanced DTM Features ........................................................................................................... 176
17.3.1. Object Initial Velocity ....................................................................................................... 176
17.3.2. Modifying the Target ........................................................................................................ 176
viii

17.3.2.1. Target Position ........................................................................................................ 177
17.3.2.2. Target Size ............................................................................................................... 178
17.3.2.3. Target Orientation ................................................................................................... 178
17.3.2.4. Target Material Properties ........................................................................................ 178
17.3.2.5. Specifying Friction Coefficients ................................................................................ 179
17.4. Drop Test Set-up Dialog Box ...................................................................................................... 179
17.4.1. Using the Drop Test Set-up Dialog Box .............................................................................. 179
17.4.2. Basic Tab of the Drop Test Set-up Dialog Box .................................................................. 1710
17.4.3. Velocity Tab of the Drop Test Set-up Dialog Box .............................................................. 1712
17.4.4. Target Tab of the Drop Test Set-up Dialog Box ................................................................ 1713
17.4.5. Status Tab of the Drop Test Set-up Dialog Box ................................................................. 1715
17.5. Picking Nodes ......................................................................................................................... 1716
17.6. Postprocessing - Animation ..................................................................................................... 1716
17.7. Postprocessing - Graph and List Time-History Variables ............................................................ 1717
A. Comparison of Implicit and Explicit Methods ......................................................................................... A1
A.1. Time Integration .......................................................................................................................... A1
A.1.1. Implicit Time Integration ..................................................................................................... A1
A.1.2. Explicit Time Integration ...................................................................................................... A1
A.2. Stability Limit ............................................................................................................................... A2
A.2.1. Implicit Method ................................................................................................................... A2
A.2.2. Explicit Method ................................................................................................................... A2
A.3. Critical Time Step Size of a Rod ..................................................................................................... A3
A.4. ANSYS/LS-DYNA Time Step Size ................................................................................................... A3
B. Material Model Examples ...................................................................................................................... B1
B.1. ANSYS LS-DYNA Material Models .................................................................................................. B1
B.2. Material Model Examples .............................................................................................................. B3
B.2.1. Isotropic Elastic Example: High Carbon Steel ......................................................................... B3
B.2.2. Orthotropic Elastic Example: Aluminum Oxide ...................................................................... B3
B.2.3. Anisotropic Elastic Example: Cadmium ................................................................................. B3
B.2.4. Blatz-Ko Example: Rubber .................................................................................................... B3
B.2.5. Mooney-Rivlin Example: Rubber ........................................................................................... B4
B.2.6. Viscoelastic Example: Glass .................................................................................................. B4
B.2.7. Bilinear Isotropic Plasticity Example: Nickel Alloy .................................................................. B4
B.2.8. Transversely Anisotropic Elastic Plastic Example: 1010 Steel .................................................. B4
B.2.9. Transversely Anisotropic FLD Example: Stainless Steel .......................................................... B5
B.2.10. Bilinear Kinematic Plasticity Example: Titanium Alloy .......................................................... B5
B.2.11. Plastic Kinematic Example: 1018 Steel ................................................................................. B5
B.2.12. 3 Parameter Barlat Example: Aluminum 5182 ...................................................................... B5
B.2.13. Barlat Anisotropic Plasticity Example: 2008-T4 Aluminum ................................................... B6
B.2.14. Rate Sensitive Powerlaw Plasticity Example: A356 Aluminum .............................................. B6
B.2.15. Strain Rate Dependent Plasticity Example: 4140 Steel ......................................................... B6
B.2.16. Piecewise Linear Plasticity Example: High Carbon Steel ....................................................... B7
B.2.17. Modified Piecewise Linear Plasticity Example: PVC .............................................................. B7
B.2.18. Powerlaw Plasticity Example: Aluminum 1100 .................................................................... B8
B.2.19. Elastic Viscoplastic Thermal Example .................................................................................. B8
B.2.20. Geological Cap Example: SRI Dynamic Concrete ................................................................. B9
B.2.21. Johnson-Cook Linear Polynomial EOS Example: 1006 Steel ............................................... B10
B.2.22. Johnson-Cook Gruneisen EOS Example: OFHC Copper ...................................................... B10
B.2.23. Null Material Linear Polynomial EOS Example: Brass .......................................................... B11
B.2.24. Null Material Gruneisen EOS Example: Aluminum ............................................................. B11
B.2.25. Steinberg Gruneisen EOS Example: Stainless Steel ............................................................ B11
B.2.26. Cable Material Example: Steel ........................................................................................... B12
ix

B.2.27. Rigid Material Example: Steel ............................................................................................ B12
C. ANSYS LS-DYNA to LS-DYNA Command Mapping .................................................................................. C1
D. Thermal/Structural Preload Example ..................................................................................................... D1
Bibliography ................................................................................................................................................. 7
Index ................................................................................................................................................. Index1
List of Figures
2.1. Integration Points ............................................................................................................................... 26
4.1. Constrained Shell to Solid ................................................................................................................... 49
6.1. LS-DYNA Drawbead Representation .................................................................................................... 69
7.1. Surface of the Two-invariant Cap Model ............................................................................................ 722
9.1. Hourglass Deformations ..................................................................................................................... 91
11.1. Time Step Sizes Before and After Subcycling .................................................................................... 111
16.1. High Speed Impact of a Metal Bar .................................................................................................... 161
16.2. Lagrangian Impact Solution ............................................................................................................ 162
16.3. Eulerian Channel Flow Solution ....................................................................................................... 162
16.4. ALE Impact Solution ........................................................................................................................ 163
17.1. Two Views of the Target .................................................................................................................. 177
17.2. Drop Test Set-up Dialog Box - Basic Tab ......................................................................................... 1710
17.3. Drop Test Set-up Dialog Box - Velocity Tab .................................................................................... 1712
17.4. Drop Test Set-up Dialog Box - Target Tab ....................................................................................... 1713
17.5. Drop Test Set-up Dialog Box - Status Tab ....................................................................................... 1715
17.6. Graph and Time-History Variables Dialog Box ................................................................................ 1718
List of Tables
3.1. Loads Applicable in an Explicit Dynamics Analysis ............................................................................... 35
3.2. LS-DYNA Solution and Output Control Options ................................................................................... 36
Chapter 1: Introduction
ANSYS LS-DYNA combines the LS-DYNA explicit finite element program with the powerful pre- and postprocessing
capabilities of the ANSYS program. The explicit method of solution used by LS-DYNA provides fast solutions for
short-time, large deformation dynamics, quasi-static problems with large deformations and multiple nonlinearites,
and complex contact/impact problems. Using this integrated product, you can model your structure in ANSYS,
obtain the explicit dynamic solution via LS-DYNA, and review results using the standard ANSYS postprocessing
tools.
You can also transfer geometry and results information between ANSYS and ANSYS LS-DYNA to perform sequential
implicit-explicit / explicit-implicit analyses, such as those required for droptest, springback and other applications.
The following ANSYS LS-DYNA general information topics are available:
1.1. Overview of Steps in an Explicit Dynamic Analysis
1.2. Commands Used in an Explicit Dynamic Analysis
1.3. A Guide to Using this Document
1.4. Where to Find Explicit Dynamics Example Problems
1.5. Additional Information
1.1. Overview of Steps in an Explicit Dynamic Analysis

The procedure for an explicit dynamic analysis is similar to any other analysis that is available in the ANSYS program.
The three main steps are:
1.
Build the model (with the PREP7 preprocessor)
2.
Apply loads and obtain the solution (with the SOLUTION processor)
3.
Review the results (with the POST1 and POST26 postprocessors)
This document describes procedures and concepts that are unique to an explicit dynamic analysis performed
using the ANSYS LS-DYNA product. It does not describe all details of the three steps listed above. If you are familiar with the ANSYS program, you already know how to perform these steps, and this document will provide additional information you need to perform an explicit dynamic analysis. If you have not used ANSYS before, you
should review the following two manuals (in addition to this document) to learn the basic analysis procedures:
ANSYS Basic Analysis Guide
ANSYS Modeling and Meshing Guide
When using ANSYS LS-DYNA, it is recommended that you use the default settings provided by the program.
Many times, these settings are appropriate for solving the problems.
1.2. Commands Used in an Explicit Dynamic Analysis

You use the same set of commands to build a model and perform an explicit dynamic analysis that you use to
do any other type of ANSYS analysis. Likewise, you choose similar options from the ANSYS programs graphical
user interface (GUI) to build and solve models.
However, some commands are unique to an explicit dynamic analysis. These commands are:
EDADAPT
Activates adaptive meshing
EDALE
Assigns mesh smoothing to explicit dynamic elements that use the

ALE formulation
EDASMP
Creates a part assembly
EDBOUND
Defines a boundary plane for sliding or cyclic symmetry
EDBVIS
Specifies bulk viscosity coefficients
EDBX
Creates a box shaped volume to be used in a contact definition
EDCADAPT
Specifies adaptive meshing controls
EDCGEN
Specifies contact parameters
EDCLIST
Lists contact entity specifications
EDCMORE
Specifies additional contact parameters for a given contact definition
EDCNSTR
Defines various types of constraints
EDCONTACT
Specifies contact surface controls
EDCPU
Specifies CPU time limit
EDCRB
Merges two rigid bodies
EDCSC
Specifies whether subcycling will be used
EDCTS
Specifies mass scaling
EDCURVE
Specifies data curves
EDDAMP
Defines system damping
EDDBL
Selects numerical precision type
EDDC
Deletes or deactivates/reactivates contact entity specifications
EDDRELAX
Activates initialization to a prescribed geometry or dynamic relaxation

for the explicit analysis
EDDUMP
Specifies output frequency for the restart file (d3dump)
EDENERGY
Specifies energy dissipation controls
EDFPLOT
Specifies plotting of load symbols
EDGCALE
Defines global ALE controls
EDHGLS
Specifies the hourglass coefficient
EDHIST
Specifies time-history output
EDHTIME
Specifies the time-history output interval
EDINT
Specifies number of integration points for output
EDIS
Specifies stress initialization in a full restart analysis
EDIPART
Defines inertia for rigid parts
EDLCS
Defines a local coordinate system
EDLOAD
Specifies loads
EDMP
Defines material properties
EDNB
Defines a nonreflecting boundary
EDNDTSD
Smooths noisy data and provides a graphical representation of the

data
EDNROT
Applies a rotated coordinate nodal constraint
EDOPT
Specifies the type of output, ANSYS or LS-DYNA
EDOUT
Specifies LS-DYNA ASCII output files
EDPART
Creates, updates, or lists parts
EDPC
Selects and plots contact entities
EDPL
Plots time dependent load curves
EDPVEL
Applies initial velocities to parts or part assemblies
12
Section 1.3: A Guide to Using this Document

EDRC
Specifies rigid/deformable switch controls
EDRD
Switches a part from deformable to rigid or from rigid to deformable
EDREAD
Reads output from LS-DYNA ASCII files into POST26 variables
EDRI
Defines inertia properties for a new rigid body that is created when a
deformable part is switched to rigid
EDRST
Specifies time increment for output to the .RST file
EDSHELL
Specifies shell computation controls
EDSOLV
Specifies "explicit dynamic solution" as the subsequent status topic
EDSP
Specifies small penetration checking for contact entities
EDSTART
Specifies status (new or restart) of the analysis
EDTERM
Specifies termination criteria
EDTP
Plots elements based on their time step size
EDVEL
Applies initial velocities to nodes or node components
EDWELD
Defines a massless spotweld or generalized weld
EDWRITE
Writes explicit dynamics input to an LS-DYNA input file
PARTSEL
Selects a subset of parts
RIMPORT
Imports initial stresses from an explicit run into ANSYS
REXPORT
Exports displacements from an implicit run to ANSYS LS-DYNA
UPGEOM
Adds displacements from a previous analysis and updates the geometry to the deformed configuration
For detailed alphabetized descriptions of the ANSYS commands (including specific menu path information for
each command), see the ANSYS Commands Reference.
1.3. A Guide to Using this Document

This document contains both procedural and reference information. You may choose to read through it from
front to back. However, it may be more useful to read the chapters in a sequence that corresponds with the
process of planning and performing an explicit dynamic analysis.
Before you build a model, you must decide which element types and material models will best represent your
physical system. The following chapters provide the background information you need to make these decisions:
Chapter 2, Elements
Chapter 7, Material Models
After choosing appropriate element types and material models, you are ready to build the model. The typical
aspects of model building are presented in:
Chapter 3, Analysis Procedure
Chapter 6, Contact Surfaces
Chapter 8, Rigid Bodies
Chapter 4, Loading
Features related to solving and postprocessing are discussed in:
Chapter 5, Solution Features
Chapter 12, Postprocessing
13
Other more advanced capabilities that are not required to complete an analysis, but that may be useful in some
situations, are presented in:
Chapter 9, Hourglassing
Chapter 10, Mass Scaling
Chapter 11, Subcycling
Chapter 13, Restarting
Chapter 14, Explicit-to-Implicit Sequential Solution
Chapter 15, Implicit-to-Explicit Sequential Solution
Chapter 16, Arbitrary Lagrangian-Eulerian Formulation
Chapter 17, Drop Test Module
Finally, the appendices contain reference information on the following topics:
Appendix A:, Comparison of Implicit and Explicit Methods
Appendix B:, Material Model Examples
Appendix C:, ANSYS LS-DYNA to LS-DYNA Command Mapping
1.4. Where to Find Explicit Dynamics Example Problems

The Explicit Dynamics Tutorial describes a sample explicit dynamic analysis problem.
1.5. Additional Information

For a more detailed overview of the explicit dynamic analysis procedure, see the ANSYS Structural Analysis Guide.
For detailed information on explicit elements, see the ANSYS Elements Reference. For detailed theoretical information, see the Livermore Software Technology Corporations LS-DYNA Theoretical Manual.
14
Chapter 2: Elements
The following elements can be used in an explicit dynamic analysis:
LINK160
BEAM161
PLANE162
SHELL163
SOLID164
COMBI165
MASS166
LINK167
SOLID168
All of the explicit dynamic elements listed above are 3-D except for PLANE162, and all have reduced integration
by default when applicable (for example, reduced integration is not the default for mass or link elements). Reduced
integration means that the number of points for numerical integration in the element formation process is less
than that necessary for exact integration. Therefore, one integration point is the default formulation for both
solid and shell elements. Full integration options are also available for the solid and shell elements. For additional
details, see Chapter 9, Hourglassing and also the LS-DYNA Theoretical Manual.
These elements assume a linear displacement function; higher order elements with a quadratic displacement
function are not available. Therefore, the explicit dynamic elements are not available with extra shape functions,
midside nodes, or p-elements. Explicit elements with linear displacement functions and one point integration
are best suited for nonlinear applications with large deformations and material failure.
It should be noted that the explicit elements are not linked directly to material properties. For example, the
SOLID164 element supports more than 20 different material models, including elastic, plastic, rubber, and foam
models. Unless stated otherwise (see Chapter 6, Contact Surfaces), the minimum material property requirements
for all elements are density, Poisson's ratio, and elastic modulus. See Chapter 7, Material Models of this document
for detailed information on the material properties that can be used in an explicit dynamic analysis. See also the
ANSYS Elements Reference for a complete description of each element, including details of input and output
quantities.
The following explicit element topics are available:
2.1. Solid and Shell Elements
2.2. Beam and Link Elements
2.3. Discrete Elements
2.4. General Element Capabilities
2.1. Solid and Shell Elements

2.1.1. SOLID164
The SOLID164 element is an 8-node brick element. By default, it uses reduced (one point) integration plus viscous
hourglass control for faster element formulation. One point integration is advantageous due to savings on
computer time and robustness in cases of large deformations. A fully integrated solid formulation (KEYOPT(1) =
2) is also available; see the element description for SOLID164 in the ANSYS Elements Referenceand Section 3.3 of
Chapter 2: Elements
the LS-DYNA Theoretical Manual. If hourglass phenomenon is a concern, such as with foams, the fully integrated
formulation may perform better because hourglass control is not required; but it is about four times more costly
in terms of CPU time.
KEYOPT(5) defines the element continuum treatment. Two different formulations are available: Lagrangian (default)
and Arbitrary Lagrangian-Eulerian (ALE). See Arbitrary Lagrangian-Eulerian Formulation in the ANSYS LS-DYNA
User's Guide for more information.
Wedge, pyramid, and tetrahedra shaped SOLID164 elements are simply degenerate bricks (i.e., some of the nodes
are repeated). These shapes are often too stiff in bending and cause problems in some situations. Therefore,
these degenerate shapes should be avoided. If a tetrahedron mesh is desired, it is recommended that SOLID168
elements be used instead of degenerated SOLID164 elements.
You can use any of the following material models for solid elements:
Isotropic Elastic
Orthotropic Elastic
Anisotropic Elastic
Bilinear Kinematic
Plastic Kinematic
Viscoelastic
Blatz-Ko Rubber
Bilinear Isotropic
Power Law Plasticity
Strain Rate Dependent Plasticity
Composite Damage
Concrete Damage
Geological Cap
Piecewise Linear Plasticity
Honeycomb
Mooney-Rivlin Rubber
Barlat Anisotropic Plasticity
Elastic-Plastic Hydrodynamic
Rate Sensitive Powerlaw Plasticity
Elastic Viscoplastic Thermal
Closed Cell Foam
Low Density Foam
Viscous Foam
Crushable Foam
Johnson-Cook Plasticity
Null
Zerilli-Armstrong
22
Section 2.1: Solid and Shell Elements
Bamman
Steinberg
Elastic Fluid
2.1.2. SHELL163
Twelve different formulations are available for the SHELL163 element. Use KEYOPT(1) to specify the desired formulation. As with solid elements, the number of integration points per element directly impacts CPU time.
Therefore, for general analyses, the reduced integration shell formulations are recommended. The following is
an outline of the different formulations available for SHELL163. Coordinate systems for output of quantities associated with SHELL163 may vary for different formulations. See the ANSYS Elements Reference for more information.
2.1.2.1. General Shell Formulations
Belytschko-Tsay (KEYOPT(1) = 0 or 2) - default
Uses reduced integration (one point)
Very fast and is recommended for most applications

Should not be used when elements experience excessive warping
Belytschko-Wong-Chiang (KEYOPT(1) = 10)
Uses reduced integration (one point)
25% slower than Belytschko-Tsay

Generally gives correct results for warping
Belytschko-Leviathan (KEYOPT(1) = 8)
Reduced integration formulation (one point)
40% slower than Belytschko-Tsay

Includes physical hourglass control automatically
Fully integrated Belytschko-Tsay (KEYOPT(1) = 12)
Has four integration points in plane and does not need hourglass control.
Shear locking is remedied by assumed strain for the transverse shear.
2.5 times slower than reduced integration Belytschko-Tsay shell.
Recommended if hourglass modes are a problem in the analysis.
Hughes-Liu (KEYOPT(1) = 1, 6, 7, 11) = > four different formulations which can offset the mid-plane of
element away from the nodes.
KEYOPT(1) = 1 General Hughes-Liu. Has one-point integration and is 2.5 times slower that BelytschkoTsay.
KEYOPT(1) = 11 Fast (corotational) Hughes-Liu. Has one-point integration and is 1.5 times slower than
Belytschko-Tsay.
23
Chapter 2: Elements
KEYOPT(1) = 6 S/R Hughes-Liu. Has four integration points with no hourglassing but is 20 times slower
that Belytschko-Tsay.
KEYOPT(1) = 7 S/R corotational Hughes-Liu. Has four integration points with no hourglassing. 8.8 times
slower than Belytschko-Tsay. Recommended if hourglassing is a problem in an analysis.
2.1.2.2. Membrane Element Formulation
Belytschko-Tsay Membrane (KEYOPT(1) = 5)
Reduced (one-point) integration
Fast and recommended for most membrane applications

Good for fabrics where wrinkling is a concern (i.e., where large in-plane compressive stresses try to
collapse the thin fabric elements)
Fully integrated Belytschko-Tsay Membrane (KEYOPT(1) = 9)
Significantly slower than general membrane (KEYOPT(1) = 5)
Four integration points in plane
No hourglassing
2.1.2.3. Triangular Shell Formulations
C0 triangular shell (KEYOPT(1) = 4)
Formulation is rather stiff, so not recommended for entire mesh
Based on Mindlin-Reissner Plate Theory

Reduced integration formulation
BCIZ Triangular Shell (KEYOPT(1) = 3)
Based on Kirchhoff Plate Theory
Slower than C0 triangular shells
Reduced integration formulation
The description of SHELL163 in the ANSYS LS-DYNA User's Guide also has a complete list of available shell formulations.
Degenerate quadrilateral shell elements are prone to lock under transverse shear; therefore, C0 triangular shell
elements (based on work by Belytschko and coworkers) have been implemented. Triangular shells can be mixed
with quadrilateral shells within the same material property set, provided that the element sorting flag (the ITRST
field on the EDSHELL command) is set to 1 (which is the default).
You can use any of the following material models for shell elements:
Isotropic Elastic
Orthotropic Elastic
24
Bilinear Kinematic
Plastic Kinematic
Blatz-Ko Rubber
Bilinear Isotropic
Composite Damage
Modified Piecewise Linear Plasticity
Mooney-Rivlin Rubber (see note below)
3 Parameter Barlat Plasticity
Transversely Anisotropic Elastic Plastic
Transversely Anisotropic FLD
Bamman
Note When the Mooney-Rivlin Rubber material model is used with SHELL163 elements, the LS-DYNA
code will automatically use a total Lagrangian modification of the Belytschko-Tsay formulation instead
of using the formulation you specify via KEYOPT(1). This program-chosen formulation is required to address
the special needs of the hyperelastic material.
All shell element formulations can have an arbitrary number of integration points through the thickness. Typically,
2 integration points (default) are required through the thickness for elastic behavior, while 3 or more integration
points are required for plastic behavior. The number of integration points through the thickness is controlled
using the second real constant:
R,NSET,,R2 where R2 = number of integration points (NIP)
3-D plane stress constitutive subroutines are implemented for the shell elements; these update the stress tensor
such that the stress component normal to the shell mid surface is zero. The integration points are stacked vertically
at the centroid of the element, as shown in Figure 2.1: Integration Points.
25
Chapter 2: Elements
Figure 2.1 Integration Points
#
" !

Through-thickness directions at each node are initially normal to the element surface but rotate with the nodes.
The through-the-thickness integration is needed to calculate bending moments and in-plane forces. The distribution of strain is always linear; whereas the stress is more complicated in nature and depends on the material
law.
Two integration points are sufficient for linear elastic materials, while more points are required for nonlinear
materials. Stress output is given at the outermost integration points, not at the surfaces (despite the nomenclature
of postprocessors, which refer to top and bottom surfaces), so care is needed in interpretation of results. For
elastic materials, stresses can be extrapolated to the surfaces. For nonlinear materials the usual procedure is to
choose four or five integration points through the thickness and to ignore the inaccuracy (i.e., the difference in
stress between the surface and the outermost integration point). The locations of the outermost integration
points of Gauss quadrature are given in the following table:
Mid Plane
0
2 points
0.7746
0.8611
5 points
Outer Surface
3 points
4 points
Outermost Point
0.5774
0.9062
1.0000
Note Keep in mind that this integration could be made exactly beforehand when linear elastic material is used, but it is not done in ANSYS LS-DYNA because nonlinear behavior is generally being modeled.
Also, for elements using full integration, the output stress is the averaged stress values from the 2x2 integration points of the same layer.
2.1.3. PLANE162
The PLANE162 element is a 2-D, 4-node solid element that can be used either as a planar (X-Y plane) or as an
axisymmetric (Y-axis of symmetry) element. KEYOPT(3) allows you to specify a plane stress, axisymmetric, or
plane strain option for the element. For the axisymmetric element formulation, you can specify either an area
or volume weighted option using KEYOPT(2).
26

KEYOPT(5) defines the element continuum treatment. Two different formulations are available: Lagrangian (default)
and Arbitrary Lagrangian-Eulerian (ALE). See Arbitrary Lagrangian-Eulerian Formulation in the ANSYS LS-DYNA
User's Guide for more information.
The PLANE162 element is typically defined by four-nodes. A three-node triangle option is also available, but not
recommended because it is often too stiff. There are no real constants for this element. It is important to note
that models containing PLANE162 elements must consist only of this element type. ANSYS LS-DYNA does not
allow a mixed 2-D and 3-D finite element model. Furthermore, all PLANE162 elements in the model must be the
same type (plane stress, plane strain, or axisymmetric).
The material models available to use with this element will depend on the KEYOPT(3) setting. For KEYOPT(3) =
0, 1, or 2 (plane stress, plane strain, or axisymmetric), you can choose the following materials:
Isotropic Elastic
Orthotropic Elastic
Elastic Fluid
Viscoelastic
Bilinear Isotropic
Temperature Dependent Bilinear Isotropic
Bilinear Kinematic
Plastic Kinematic
Composite Damage
Bamman
For the plane stress option (KEYOPT(3) = 0), you can also choose the following materials:
3-Parameter Barlat Plasticity
For the axisymmetric and plane strain options (KEYOPT(3) = 1 or 2), you can also choose the following materials:
Blatz-Ko Rubber
Closed Cell Foam
Low Density Foam
Crushable Foam
27
Chapter 2: Elements
Honeycomb
Null
Zerilli-Armstrong
Steinberg
2.1.4. SOLID168
The SOLID168 element is a 10-node tetrahedron element. It uses five point integration; see the element description
for SOLID168 in the ANSYS Elements Reference.
You can use any of the following material models for solid elements:
Isotropic Elastic
Isotropic Elastic
Orthotropic Elastic
Anisotropic Elastic
Bilinear Kinematic
Plastic Kinematic
Viscoelastic
Blatz-Ko Rubber
Bilinear Isotropic
Composite Damage
Concrete Damage
Geological Cap
Honeycomb
Rate Sensitive Power Law Plasticity
Closed Cell Foam
Low Density Foam
Viscous Foam
Crushable Foam
Null
Zerilli-Armstrong
28
Section 2.2: Beam and Link Elements
Bamman
Steinberg
Elastic Fluid
SOLID168 elements are well suited to modeling irregular meshes such as those produced from various CAD/CAM
systems and the ANSYS Workbench. It is important to note, however, that models made up entirely of SOLID168
elements may not be as accurate as hexahedral (SOLID164) models. For this reason, we recommend using a
combination of SOLID168 and SOLID164 elements within a model as a good modeling practice. The SOLID168
elements could be used for the irregular portions of the model while the SOLID164 elements could be used to
mesh the more uniform features.
Although SOLID164 and SOLID168 elements can be used in the same model, you must be careful to develop an
adequate mesh transition between these two element types. This is due to the fact that SOLID168 elements
conatin mid-side nodes, which are not allowed in SOLID164 elements.
2.2. Beam and Link Elements

2.2.1. BEAM161
The BEAM161 element has two basic formulations: Hughes-Liu and Belytschko-Schwer. BEAM161 is best suited
for rigid body rotations because it does not generate any strains. Three nodes must be used to define the element;
a node at each endpoint and an orientation node are required. For both formulations, several beam cross sections
can be defined using KEYOPTs (4) and (5). In general, BEAM161 is efficient and robust for 2x2 gauss quadratic
integration (default). Several different integration formulations can be defined with KEYOPT(2).
The Hughes-Liu beam (default) is a conventionally integrated element that can model rectangular or circular
cross-sections using an array of integration points at the mid-span of the element. Alternatively, you can specify
a cross-section integration rule to model arbitrary cross-sections. The beam effectively generates a constant
moment along its length, so, as with the brick and shell elements, meshes need to be reasonably fine to achieve
adequate accuracy. Because of the location of the integration points, yielding is detected only at the element
center. Therefore, a cantilever beam model will yield at slightly too high a force because the fully plastic moment
must be developed at the center of the clamped element rather than at the outer surface.
The Belytschko-Schwer beam (KEYOPT(1) = 2, 4, or 5) is explicitly formulated and generates a moment which
varies linearly along the length of the beam. The elements have the "correct" elastic stresses and detect yielding
at their ends. For example, a cantilever loaded statically at its tip can be represented accurately in both elastic
and plastic regimes using a single element As with the Hughes-Liu beam, the mass is lumped onto the nodes,
so finer meshes may be required for dynamic problems where a correct mass distribution is important.
You can use any of the following material models for beam elements (with some restrictions for certain element
formulations):
Isotropic Elastic
Bilinear Kinematic
Plastic Kinematic
Viscoelastic
29
Chapter 2: Elements
2.2.2. LINK160
The LINK160 spar (truss) element is similar to the Belytschko-Schwer beam, but carries axial loads only. This element
assumes a straight bar, axially loaded at its ends, and of uniform properties from end to end. The materials you
can use for this element are isotropic elastic, plastic kinematic (rate dependent), and bilinear kinematic.
2.2.3. LINK167
The LINK167 element is a tension only spar that can be used to model cables. It is similar to a spring element in
that the relationships between force and deformation are input directly by the user. This element requires that
the cable option be specified with the EDMP command (see the EDMP command description).
2.3. Discrete Elements

2.3.1. COMBI165 Spring-Damper
Spring elements generate a force which depends on displacement; that is, change of length of the element. The
force is applied along the element axis. For example, a positive force (tension) acts along the positive axis direction
on node 1 and along the negative axis direction on node 2. By default, the element axis lies along the line from
node 1 to node 2. When the element rotates, the line of action of the force also rotates.
Damper elements are treated as a subset of spring elements: both linear viscous and nonlinear viscous dampers
can be modeled.
Rotational (torsional) springs and dampers are also available. These are selected by setting KEYOPT(1) = 1. The
rest of the input is the same as for translational springs; the given force-displacement relationship will be treated
as moment-rotation (rotation in radians), and the moments are applied about the element axis (positive clockwise).
Rotational springs affect only the rotational degrees of freedom of their nodes - they do not pin the nodes together.
The COMBI165 element can be used in combination with any other explicit element. However, because it does
not have mass, COMBI165 cannot be the only element type in an analysis. To represent a spring/mass system,
you must add mass by also defining MASS166 elements.
You cannot define both spring and damper properties for the same COMBI165 element. However, you can define
separate spring and damper elements that use the same nodes (that is, you can overlay two COMBI165 elements).
You can use any of the following material models for COMBI165:
Linear Elastic Spring
Linear Viscous Damper
Elastoplastic Spring
Nonlinear Elastic Spring
Nonlinear Viscous Damper
General Nonlinear Spring
Maxwell Viscoelastic Spring
Inelastic Tension or Compression-Only Spring
When using COMBI165, be sure to specify a unique set of real constants, element type, and material properties
(R, ET, and TB commands, respectively) for each part to ensure that parts are uniquely defined.
210
Section 2.4: General Element Capabilities
2.3.2. MASS166
The mass element is defined by a single node and a mass value (Force*Time2/Length). Mass elements can often
be used in a model to realistically characterize the mass of a structure without incorporating a large number of
solid or shell elements. In an automobile impact analysis, for example, mass elements can model engine components whose deformation behavior is not of primary interest. Such a use of mass elements will reduce the
number of elements required in an analysis, hence reducing the total computation time required to obtain a
solution.
You can also use the MASS166 element to define a lumped rotary inertia at a node. To use this option, set KEYOPT(1) = 1 in the MASS166 element definition and input six moment of inertia values (IXX, IXY, IXZ, IYY, IYZ, IZZ)
via the element real constants. You can not input a mass value for this option; therefore, you must define a second
mass element at the same node (with KEYOPT(1) = 0) to account for mass.
2.4. General Element Capabilities

The following element types can be declared as rigid bodies: LINK160, BEAM161, PLANE162, SHELL163, SOLID164,
LINK167, and SOLID168. Rigid bodies are covered in detail in Chapter 8, Rigid Bodies.
The mass of each solid, shell, and beam element is distributed equally on the nodes of the element. In the case
of shell and beam elements, a rotational inertia is also added to each node; a single value is used, so that the effect
is to distribute the mass spherically around the node.
211
212
Chapter 3: Analysis Procedure

The procedure for an explicit dynamics analysis consists of three main steps:
1.
Build the model
2.
Apply loads and obtain the solution
3.
Review the results
The following analysis procedure topics are available:

3.1. Build the Model
3.2. Apply Loads and Obtain the Solution
3.3. Review the Results
3.4. The Definition of Part
3.5. Adaptive Meshing
3.1. Build the Model

The first step in an explicit dynamic analysis is to create the model that will represent the physical system to be
analyzed. Use the PREP7 Preprocessor to build the model.
If you perform your analysis through the ANSYS graphical user interface (GUI), it is important to first set the
Preferences options (Main Menu> Preferences) to LS-DYNA Explicit so that menus are properly filtered to
show explicit dynamics input options. (Note that setting the preference to LS-DYNA explicit does not activate
an LS-DYNA solution. To do so, you must specify an explicit element type such as SHELL163.)
Once you have set the analysis preference, you can build the model as you normally would for any analysis type.
First, enter the preprocessor by issuing the /PREP7 command or by picking Main Menu> Preprocessor in the
GUI. In the preprocessor, perform the following tasks:
Define the element types and real constants
Specify material models
Define the model geometry
Mesh the model
Define contact surfaces
If you have never used an ANSYS product before, you should review the ANSYS Basic Analysis Guide and the ANSYS
Modeling and Meshing Guide to learn the general procedures used to build a model in ANSYS.
3.1.1. Define Element Types and Real Constants

Element types available for an explicit dynamic analysis are described briefly in Chapter 2, Elements. More detailed descriptions of each explicit element can be found in the ANSYS Elements Reference. It is recommended
that you read these descriptions carefully before deciding which element types to use in your model.
Once you have chosen appropriate elements to represent your physical system, you specify element types using
the ET command (or in the GUI pick Main Menu> Preprocessor> Element Type).
Each element description in the ANSYS Elements Reference lists all real constants that are associated with the
element. You must determine which real constants (if any) are required for each element type you include in the

model. Then specify the real constants using the R command (or in the GUI pick Main Menu> Preprocessor>
Real Constants).
3.1.2. Specify Material Properties

There are numerous material models available for use in an explicit dynamic analysis. You should refer to the
element descriptions in the ANSYS Elements Reference to find out which material models are valid for a particular
element. Also, review Chapter 7, Material Models of this document, which contains detailed descriptions for
all of the available material models.
Once you have decided which material model (or models) to use in your analysis, you must define all of the
properties associated with the model (as described in Chapter 7, Material Models). To do so in batch or command
mode, you typically use the MP, TB, and TBDATA commands (and in some cases, the EDMP command). Temperature dependent models may also require the MPTEMP, MPDATA, and TBTEMP commands. In the GUI,
material models are defined via the following menu path: Main Menu> Preprocessor> Material Props> Material Models
After you select the appropriate material model, you will be prompted to define the necessary properties for
that model. See Section 7.1: Defining Explicit Dynamics Material Models for more information on defining material models through the GUI.
In order to specify an orthotropic model that is not aligned with the global Cartesian coordinate system, you
must first define the local coordinate system with the EDLCS command (menu path: Main Menu> Preprocessor>
Material Props> Local CS> Create Local CS.)
For some material models, you may also need to use the EDCURVE command to define data curves associated
with the material (e.g., a stress-strain curve). (To access EDCURVE in the GUI, pick: Main Menu> Preprocessor>
Material Props> Curve Options.)
3.1.3. Define the Model Geometry

The easiest way to create the model geometry is with the solid modeling capabilities of the ANSYS program.
Refer to Solid Modeling in the ANSYS Modeling and Meshing Guide for detailed information on all solid modeling
functions.
For simple models (e.g., line elements only), you may choose to use the direct generation modeling method. By
this method, you define the nodes and elements of the model directly. See Direct Generation in the ANSYS
Modeling and Meshing Guide for details.
3.1.4. Mesh the Model

After building the solid model, you are ready to mesh the model with nodes and elements. The procedure for
generating the mesh is outlined in Generating the Mesh in the ANSYS Modeling and Meshing Guide. If you are not
familiar with ANSYS meshing, you should review that chapter before meshing an explicit dynamics model. Since
all of the details are contained in that chapter, we will discuss only a few key points in this section.
Meshing involves three main steps:
Set the element attributes
Set mesh controls
Generate the mesh
32
Section 3.1: Build the Model

To set the element attributes, you specify which previously defined element type, real constant set, and material
property set to use for subsequent meshing. Use the TYPE, MAT, and REAL commands, or menu paths:
Main Menu> Preprocessor> Meshing> Mesh Attributes> Default Attribs
Main Menu> Preprocessor> Modeling> Create> Elements> Elem Attributes
Mesh controls allow you to indicate the general size and shape of elements to use during meshing. There are
numerous mesh controls available in the ANSYS program (see the ANSYS Modeling and Meshing Guide for details).
In the GUI, these controls are accessed by picking Main Menu> Preprocessor> Meshing> Mesh Tool.
Keep the following points in mind when choosing mesh controls:
Avoid degenerate shell and solid element shapes (such as a triangular shell or a tetrahedral solid). They
are generally too stiff and less accurate than quadrilateral and hexahedral shapes.
Try to achieve a mesh with uniform element sizes (i.e., avoid areas with relatively small elements). A large
difference in element sizes can cause a small minimum time step size and, therefore, a long run time. If a
relatively few number of small elements are required to mesh a particular geometry, mass scaling (see
Chapter 10, Mass Scaling) can be used to increase the minimum time step.
Do not use the SmartSizing method of element size control (SMRTSIZE command), which can create a
large variation in element sizes within the mesh. Instead, use the ESIZE and related commands to control
element sizes.
Avoid poorly shaped elements which can cause hourglassing.
Do not make the mesh too coarse when reduced element formulations are used, or elements may experience hourglassing.
If hourglassing is a problem, try to use fully integrated elements in a part of the model or in the entire
model.
After specifying the desired mesh controls, mesh the model using typical meshing commands (AMESH, VMESH,
etc.). In the GUI, meshing functions are accessed by picking Main Menu> Preprocessor> Meshing> Mesh or
via the Mesh Tool (Main Menu> Preprocessor> Meshing> Mesh Tool).
3.1.5. Define Contact Surfaces

An explicit dynamic analysis often includes contact between surfaces. Chapter 6, Contact Surfaces of this document describes the types of contact available and the procedure to define contact. A brief overview of that
procedure is given here.
Defining contact involves four basic steps:
Determine the type of contact which best defines the physical model (EDCGEN command)
Identify contact surfaces (CM, or EDPART, or EDASMP commands, with the EDCGEN command)
Specify friction coefficient parameters (EDCGEN command)
Specify additional input required for the chosen contact type (EDCGEN and EDCONTACT commands)
If automatic contact is not being used, you must define contact surfaces by grouping the nodes on each surface
into a component with the CM command. Once these components are created, use the EDCGEN command to
specify contact between the desired surfaces (i.e., node components). You must also identify the type of contact
with the EDCGEN command. For the single surface contact algorithms, the outer surfaces of the contacting
bodies are determined by ANSYS LS-DYNA.
33

Note Contact surfaces can alternatively be identified by a PART number (using EDPART), or by an assembly of parts (using EDASMP) rather than by a node component. Part and assembly definitions are
discussed in Section 3.4: The Definition of Part.
The EDCGEN command is also used to specify friction coefficient parameters and other input specific to different
contact types. Another command, EDCONTACT is used to specify miscellaneous contact controls such as contact
penetration checking and sliding interface penalties. (Contact options are found in the GUI under Main Menu>
Preprocessor> LS-DYNA Options> Contact.)
3.1.6. General Modeling Guidelines

The following are guidelines that you should consider when planning your explicit dynamics model:
Use rigid bodies whenever possible to represent relatively stiff, unyielding objects in the model. Using rigid
bodies simplifies the solution and results in shorter run times. (See Chapter 8, Rigid Bodies for information
on defining rigid bodies.)
Use realistic values for material properties. For example, do not use an unrealistically high elastic modulus
to represent a rigid body, and do not use unrealistic thickness values for shell elements.
Consider using damping (EDDAMP command) to prevent unrealistic oscillations in your models structural
response. Refer to the EDDAMP command description in the ANSYS Commands Reference for detailed
information.
If you have performed 2-D dynamic analyses with the regular ANSYS program, consider extruding these
models to 3-D and analyzing them with ANSYS LS-DYNA. You may achieve more accurate results in a
shorter run time.
Note that the submodeling and substructuring features of the ANSYS program cannot be used in ANSYS
LS-DYNA.
3.2. Apply Loads and Obtain the Solution

To apply loads and obtain a solution, you need to enter the ANSYS solution processor. Issue the /SOLU command
or pick Main Menu> Solution in the GUI.
From the solution processor, you can apply loads, initial velocities, constraints, and DOF coupling to the model.
The following sections describe this process.
3.2.1. Loads
In an explicit dynamic analysis, all loads must be specified over time using component logic or part IDs, array
parameters, and the EDLOAD command. (In the GUI, load options are found under Main Menu> Solution>
Loading Options> Specify Loads.) Basic input for this command is a component name or part number and two
array parameter names or a load curve ID number (LCID). The component specified must contain the nodes or
elements on which the load is being applied. The array parameters specified must contain time varying load
data (one array for time values and one array for the corresponding load values; the two arrays must be the same
length). As an alternative to inputting the array parameters on the EDLOAD command, you can define the load
curve using the EDCURVE command (Main Menu> Solution> Loading Options> Curve Options) and input
the load curve ID on EDLOAD. Valid loads are shown in the table below.
34
Section 3.2: Apply Loads and Obtain the Solution
Table 3.1 Loads Applicable in an Explicit Dynamics Analysis

Load Type
Label
Displacements
UX, UY, UZ
Rotations
ROTX, ROTY, ROTZ
Forces
FX, FY, FZ
Moments
MX, MY, MZ
Velocities
VX, VY, VZ
Accelerations (on nodes)
AX, AY, AZ
Base Accelerations
ACLX, ACLY, ACLZ
Angular Velocities
OMGX, OMGY, OMGZ
Temperature
TEMP
Pressures (applied to elements)
PRESS
Displacements on Rigid Bodies
RBUX, RBUY, RBUZ
Rotations on Rigid Bodies
RBRX, RBRY, RBRZ
Velocities on Rigid Bodies
RBVX, RBVY, RBVZ
Forces on Rigid Bodies
RBFX, RBFY, RBFZ
Moments on Rigid Bodies
RBMX, RBMY, RBMZ
Angular Velocities on Rigid Bodies RBOX, RBOY, RBOZ
Note We recommend that you specify velocity time histories instead of displacement time histories.
A piecewise linear displacement time history may lead to discontinuous velocities and infinite accelerations.
The load symbol will appear automatically on the active graphics window. The load symbol is erased automatically
when you replot. To turn the display of this symbol on or off, issue EDFPLOT (or pick Main Menu> Solution>
Loading Options> Show Forces in the GUI).
To visualize the applied load curve, use the EDPL command (or pick Main Menu> Solution> Loading Options>
Plot Load Curve in the GUI).
For more detailed information on applying loads with the EDLOAD command, see Section 4.1: General Loading
Options.
3.2.2. Initial Velocities

In the ANSYS LS-DYNA program, initial velocities are defined using the EDVEL and EDPVEL commands. You can
use these commands to apply both translational and rotational velocities to various entities. Use EDVEL to apply
velocities to a nodal component or to a single node; use EDPVEL to apply velocities to a part or part assembly.
The EDVEL and EDPVEL commands provide two methods for specifying rotational velocities, Option = VGEN
and Option = VELO. The VGEN method applies a rigid body rotation to the specified entity around a specified
axis. The VELO method applies the rotational velocities directly to each node's rotation degree of freedom. Since
only shell and beam elements have rotation degrees of freedom, the rotations input with the VELO method are
only applicable to SHELL163 and BEAM161 elements.
You can access all the initial velocity options in the GUI by picking Main Menu> Solution> Initial Velocity. For
more information on how to apply initial velocities, see Section 4.2.3: Initial Velocity.
35
3.2.3. Constraints
In addition to loads and initial velocities, constraints can also be applied to the model. Using the D command,
you can apply constraints only to the displacement (UX, UY, UZ) and rotation (ROTX, ROTY, ROTZ) degrees of
freedom; the constraint value must be zero. (In the GUI, pick Main Menu> Solution> Constraints> Apply.)
Other specialized types of constraints are also available in an explicit dynamic analysis (such as symmetry
boundary planes and welded constraints). See Section 4.2.1: Constraints and Section 4.2.2: Welds for more information.
3.2.4. DOF Coupling

DOF coupling (CP command) and constraint equations (CE command) are also allowed in an explicit dynamic
analysis. (In the GUI, pick Main Menu> Preprocessor> Coupling/Ceqn.) Coupling is allowed only for the UX,
UY, and UZ degrees of freedom. Constraint equations are allowed only for the UX, UY, UZ, and ROTX, ROTY, ROTZ
degrees of freedom. For more information, see Section 4.3: Coupling and Constraint Equations.
3.2.5. Data Smoothing

If you're working with noisy data (such as an earthquake excitation), you may want to "smooth" that data to a
set of data that provides an accurate approximation of the data points.
To smooth data, you must first create four arrays. The first two arrays should contain the noisy data from the independent and the dependent variables. The second two arrays will be used to store the smoothed data. After
these arrays are created, you then smooth the data using the EDNDTSD command (Main Menu> Solution>
Loading Options> Smooth Data). See Section 12.3.3: Data Smoothing later in this document for more information on data smoothing.
3.2.6. Specify Explicit Dynamics Controls

Table 3.2: LS-DYNA Solution and Output Control Options shows some basic LS-DYNA solution and output
controls that you should specify for an explicit dynamic analysis.
Table 3.2 LS-DYNA Solution and Output Control Options

Option
Command
GUI Path
Termination Time
TIME
Main Menu> Solution> Time Controls> Solution Time
SHELL/BEAM Output
EDINT
Main Menu> Solution> Output Controls> Integ Pt Storage
Substep Output Controls
EDRST
Main Menu> Solution> Output Controls> File Output Freq
Output Interval (History File)
EDHTIME
Main Menu> Solution> Output Controls> File Output Freq
Energy Options
EDENERGY Main Menu> Solution> Analysis Options> Energy Options
The above table does not show all controls available in an explicit dynamic analysis. However, most of the default
settings for the LS-DYNA control options (output controls, file controls, damping options, etc.) are sufficient for
most explicit dynamic analyses and need not be modified.
A brief description of those options that are recommended follows:
Termination Time [TIME]

This option specifies time at the end of the analysis.
36
Section 3.4: The Definition of Part
Substep Output Controls [EDRST]

This option specifies the number of results written to the Jobname.RST file. Because explicit dynamics
analyses are only solved over very small time increments (i.e., 1e-7 seconds), only a relatively small number
of solutions should be written to the Jobname.RST file.
Output Interval [EDHTIME]

This option specifies the number of output steps for the history file (Jobname.HIS). The history file results
are typically saved for a small subset of nodes or elements [EDHIST], but at a much higher frequency than
the results file (Jobname.RST) results.
3.2.7. Save Database and Solve

After you have specified all loading and solution and output controls, save a backup copy of the database to a
named file (SAVE command or Utility Menu> File> Save as in the GUI). Next, start the solution calculations
(SOLVE command or Main Menu> Solution> Solve in the GUI).
For more details on explicit dynamic solution controls and other solution features, see Chapter 5, Solution
Features later in this document.
3.3. Review the Results

You can review explicit dynamic analysis results using POST1, the general postprocessor, and POST26, the timehistory processor.
POST1 is used to review results over the entire model at specific time-points.
POST26 is used to track specific nodal and element result items over a more detailed load history.
The following are basic guidelines for postprocessing explicit dynamic results.
The database must contain the same model for which the solution was calculated.
The results file (Jobname.RST) must be available for POST1.
The history file (Jobname.HIS) must be available for POST26.
All stresses and strains output from LS-DYNA are in the global Cartesian coordinate system. However, if
you are using composite materials, stresses can be in a local (element) coordinate system.
For more information on postprocessing explicit dynamic results, see Chapter 12, Postprocessing later in this
document. For a complete description of all postprocessing functions, see Chapter 4, An Overview of Postprocessing in the ANSYS Basic Analysis Guide.
3.4. The Definition of Part

A group of elements having a unique combination of MAT, TYPE, and REAL set IDs is designated as a part and is
assigned a PART ID. There are a number of ANSYS LS-DYNA commands which refer to the PART ID directly (for
example, EDCGEN, EDDC, EDLOAD, EDDAMP, EDCRB, and EDREAD). The PART IDs are automatically created
and written to the LS-DYNA input file Jobname.K at the start of the model solution by the ANSYS LS-DYNA
program. They can also be created and listed with the EDPART command.
When the program generates PART IDs, the IDs are created from a sequential list of selected elements. If any of
the MAT, TYPE, or REAL IDs change in the element list, the next PART ID number will be assigned to that group
of elements. For example, if the first ten elements have MAT = 3, TYPE = 1, and REAL = 2 then these ten elements
37

will be assigned part number 1; if the next one hundred elements have MAT = 2, TYPE = 2, and REAL = 1, these
elements will be assigned part number 2; and so on.
The part list can be created and listed using the CREATE/UPDATE/ADD/LIST options on the EDPART command.
It is also created and remains permanent if the SOLVE command is issued or if the EDWRITE command is used
to write the input file, Jobname.k. If the part list already exists, these latter commands will only update the list,
not overwrite it. For more information on the EDPART command, refer to the description of this command in
the ANSYS Commands Reference.
You can assign a specific PART ID to a group of elements by using the ADD option on EDPART. To do so, you
must first select a group of elements which have the same combination of MAT, TYPE, and REAL IDs (ESEL
command). You can name those elements as a component, or assign the PART ID to the currently selected set.
You can then use the EDPART,ADD option to specify an arbitrary number to be used as the PART ID for those
elements. It is important to note that user-specified PART IDs will not be overwritten by subsequent EDPART,CREATE or EDPART,UPDATE commands. However issuing EDPART,ADD with a different PART ID will overwrite existing PART IDs for the specified elements. To delete a user-specified PART ID completely, issue the EDPART,DELE,PARTID command. The deleted PART ID is immediately overwritten by a program-generated PART
ID (similar to EDPART,UPDATE).
EDPART,CREATE creates new PART IDs, but does not overwrite any user-specified PART IDs. These PART IDs can
be listed using the EDPART,LIST command. The list shows the status of parts at the time of creation or update
(EDPART,UPDATE). If the EDPART,CREATE command is issued repeatedly, the part list is overwritten. In order
to get the actual part list after modifications or additions to the model, issue the EDPART,UPDATE command.
This extends the existing part list without changing its order and allows elements to be added to an existing part
comprised of the same MAT, TYPE, and REAL IDs. Any combination of MAT, TYPE, and REAL which is not referred
to by any selected element will render the corresponding part unused. This will be evident from the zero value
printed on the fifth (USED) column of the part list. A warning is issued at the SOLVE or EDWRITE command if a
previously defined part-related command refers to an unused part.
Any combination of user-specified and automatically generated PART IDs can be used within a model. For example,
you may choose to input PART IDs (using the ADD option) for one or more groups of elements and then use
EDPART,CREATE or EDPART,UPDATE to define PART IDs for the rest of the model.
The following example shows a 14-element model with 2 MATs, 3 TYPEs and 3 REALs. In this model, element 10
originally had MAT = TYPE = REAL = 2, but was deleted after the part list was automatically created by issuing
EDPART,CREATE. An ELIST command produces the following element list:
ELEM
MAT TYP
REL
ESY
SEC
NODES
11
12
13
14
15
16
17
18
21
22
23
24
25
26
27
28
31
32
33
34
35
36
37
38
41
42
43
44
45
46
47
48
51
52
53
54
55
56
57
58
61
62
63
64
65
66
67
68
71
72
73
74
75
76
77
78
81
82
83
84
85
86
87
88
11
101
102
103
104
105
106
107
108
12
111
112
113
114
115
116
117
118
13
121
122
123
124
125
126
127
128
38
Section 3.4: The Definition of Part

ELEM
MAT TYP
REL
ESY
SEC
NODES
14
131
132
133
134
135
136
137
138
15
141
142
143
144
145
146
147
148
To define a user-specified PART ID, first select elements 3 and 5 and create an element component E35. EDPART,ADD,10,E35 specifies PART ID 10 for elements 3 and 5. PART ID 10 (for the component E35) will not change
if you issue EDPART,CREATE or EDPART,UPDATE after reselecting all of the elements in the model. Consider the
definition of a second user-specified PART ID 60 for element 14. Selecting element 14 and issuing EDPART,ADD,60
changes the PART ID for element 14 from 6 to 60. After selecting all of the elements, EDPART,UPDATE shows
the following part list.
PART
MAT
TYP
REAL
USED
10
2 - user defined
60
1 - user defined
In the above list, elements 1,2 and 4 have MAT = TYPE = REAL = 1 and form PART 1. Elements 3 and 5 also have
MAT = TYPE = REAL = 1 but have the user-specified PART ID 10. Element 14 with MAT = 1, TYPE = 3 and REAL =
2 has the user-specified PART ID 60. The PART IDs for the rest of elements were generated by the ANSYS LS-DYNA
program. Element 10 had MAT = TYPE = REAL = 2, but was deleted after the part list was created initially.
Therefore, USED for PART ID 4 is zero. The USED field in the part list indicates the number of elements having
the PART ID. USED is calculated from all currently selected elements. If the element selection changes, USED will
change. To get the correct number of USED, select the correct number of elements and issue EDPART,UPDATE.
Note For this example, issuing EDPART,UPDATE will not change the PART IDs in the part list. However,
issuing EDPART,CREATE will create a new part list. PART 5 will become the new PART 4 and PART 6 will
become the new PART 5. The user-specified PART IDs 10 and 60 will not change. Because the automatically
generated PART IDs can change as a result of EDPART,CREATE, this can invalidate previously defined
part-based loading, contact specifications, etc.
The following procedure is recommended:
a.
Build the model as usual until you have to specify a PART ID number on a part-related command.
b.
Create the part list (EDPART,CREATE) or specify PART IDs for some groups of elements (EDPART,ADD);
then list the parts (EDPART,LIST).
c.
Use the appropriate PART ID number(s) from the list for the command(s).
d.
Continue modeling.
e.
Update the part list (EDPART,UPDATE) if the elements or their attributes have changed, which includes
the addition of new elements.
f.
List the updated PART ID numbers and make use of them on further part-related command(s).
Issuing ELIST with the part listing key on (PTKEY = 1) will show the PART ID for each elements as follows.
39

ELEM
PART MAT TYP
REL
ESY
SEC
NODES
11
12
13
14
15
16
17
18
10
21
22
23
24
25
26
27
28
31
32
33
34
35
36
37
38
10
41
42
43
44
45
46
47
48
51
52
53
54
55
56
57
58
61
62
63
64
65
66
67
68
71
72
73
74
75
76
77
78
81
82
83
84
85
86
87
88
11
101
102
103
104
105
106
107
108
12
111
112
113
114
115
116
117
118
13
121
122
123
124
125
126
127
128
14
60
131
132
133
134
135
136
137
138
15
141
142
143
144
145
146
147
148
In addition to listing PART IDs in the element list, PART IDs can also be plotted with the EPLOT command by
turning on the part display key (/PNUM,PART,1). Element selection using the ESEL command can also be performed
based on PART IDs. These two capabilities replace the obsolete PARTSEL command.
The CDWRITE command writes the part information to the Jobname.CDB file. This information can then be
automatically read into ANSYS with the CDREAD command. However, if more than one Jobname.CDB file is
read, the part list from the last Jobname.CDB file overwrites the existing part list, if any. In general, you must issue
the EDPART,CREATE command to recreate the part list of the total model. This will affect all part-related commands
contained in the Jobname.CDB file. Therefore, you can join models, but not part-related inputs, which must be
modified using the newly-created PART ID numbers.
In limited cases, an update of the part list (EDPART,UPDATE) is possible. This requires that no used combination
of MAT/TYPE/REAL appears more than once in the list. However, partial changes to the part-related commands
may be necessary.
3.4.1. Part Assemblies

Part assemblies, which are created using the EDASMP command, are entities that are made up of several different
parts. Part assemblies can be used as input on several ANSYS LS-DYNA commands. They can be particularly
useful when defining contact between two entities consisting of several parts (see Chapter 6, Contact Surfaces
in the ANSYS LS-DYNA User's Guide). Using the EDASMP command, you can define as many as 16 parts in a part
assembly after you provide an ID number. The part assembly number can be any arbitrary number, but cannot
be a currently assigned PART ID number (use EDPART,LIST to determine the currently used numbers). You can
also list and delete part assemblies using the EDASMP,LIST and EDASMP,DELE options, respectively. As an example, to specify part assembly ID number 10 consisting of parts 1,2 and 4, the following command would be
issued:
EDASMP,ADD,10,1,2,4
3.5. Adaptive Meshing

In metal forming and high-speed impact analyses, a body may experience very large amounts of plastic deformation. Single point integration explicit elements, which are usually robust for large deformations, may give inaccurate results in these situations due to inadequate element aspect ratios. To counteract this problem, the ANSYS
310
Section 3.5: Adaptive Meshing

LS-DYNA program has the ability to automatically remesh a surface during an analysis to improve its integrity.
This capability, known as adaptive meshing, is controlled with the EDADAPT and EDCADAPT commands.
The EDADAPT command activates adaptive meshing within a specific PART ID. (Use the EDPART command to
create and list valid PART IDs.) For example, to turn on adaptive meshing for PART ID #1, you would issue the
following command:
EDADAPT,1,ON
Note Adaptive meshing is only valid for parts consisting of SHELL163 elements.
When adaptivity is turned on for a part, the mesh for that part will be automatically regenerated during an analysis to ensure adequate element aspect ratios are maintained throughout the deformation process. Adaptive
meshing is most commonly used in the analysis of large deformation processes such as metal forming (adaptive
meshing is typically used for the blank). To activate adaptive meshing for more than one part within a model,
you must reissue the EDADAPT command for each different PART ID. By default, adaptive meshing is off for all
parts within a model.
After specifying which parts will be adaptively meshed, you must also define adaptive meshing parameters using
the EDCADAPT command. The EDCADAPT command globally sets the control options for all PART IDs that are
to be adaptively meshed as defined by the EDADAPT command. The parameters controlled by the EDCADAPT
command are listed below.
Frequency (FREQ) - The time interval between which adaptive mesh refinements will take place. For example, if FREQ is set to 0.01, elements will be adaptively remeshed every .01 seconds (assuming the time
unit used is seconds) if they are above the specified angle tolerance. Since the default value for the FREQ
field is 0.0, this field must be specified when using adaptive meshing in an analysis.
Angle Tolerance (TOL) - The adaptive angle tolerance (in degrees) for which adaptive meshing will occur
(default = 1e31). The TOL field controls the aspect ratio between elements and is very important for ensuring accurate results. If the relative angle between elements exceeds the specified TOL value, the elements
will be automatically refined.
Adaptivity Option (OPT) - There are two different angle options available for mesh adaptivity. For OPT =
1, the angle change that is compared to the specified TOL value is computed based on the original mesh
configuration. For OPT = 2, the angle change that is compared to the specified TOL value is computed
based on the previously refined mesh.
Mesh Refinement Levels (MAXLVL) - The MAXLVL field controls the number of times an element can be
remeshed during the entire analysis. For a single original element, a MAXLVL of 1 would allow the creation
of one additional element, a MAXLVL of 2 would allow the creation of up to four additional elements, and
a MAXLVL of 3 would allow the creation of up to 16 additional elements. High values of MAXLVL will yield
more accurate results, but can dramatically increase the model size.
Remeshing Birth and Death Times (BTIME and DTIME) - The remeshing birth and death times control
when adaptive meshing is turned on and off in an analysis. For example, if you set BTIME = .01 and DTIME
= .1, adaptive meshing will only take place in the analysis between .01 and .1 seconds (assuming the time
unit used is seconds).
Interval of Remeshing Curve (LCID) - This data curve (specified on the EDCURVE command) defines the
interval of remeshing as a function of time. The abscissa of the data curve is time, and the ordinate is the
varied adaptive time interval. If this option is nonzero, the adaptive frequency (FREQ) will be replaced by
this data curve. Note, however, that a nonzero FREQ value is still required to initiate the first adaptive loop.
Minimum Element Size (ADPSIZE) - Minimum element size to be adapted based on element edge length.
If this parameter is undefined, the edge length limit is ignored.
311
One or Two Pass Option (ADPASS) - If ADPASS = 0, two pass adaptivity is used, and the calculation is repeated
after adaptive remeshing (this is the default). If ADPASS = 1, one pass adaptivity is used, and the calculation
is not repeated after adaptive remeshing. See Figures 30.9(a) and 30.9(b) in the ANSYS/LS-DYNA Theoretical
Manual for graphical representations of these two options.
Uniform Refinement Level Flag (IREFLG) - Values of 1, 2, 3, etc. allow 4, 16, 64, etc. elements, respectively,
to be created uniformly for each original element.
Penetration Flag (ADPENE) - Depending on whether the value of ADPENE is positive (approach) or negative
(penetrate), the program will adapt the mesh when the contact surfaces approach or penetrate the tooling
surface by this value. The tooling adaptive refinement is based on the curvature of the tooling. If ADPENE
is positive, the refinement generally occurs before contact takes place; consequently, it is possible to use
one pass adaptivity (ADPASS = 1).
Shell Thickness Level (ADPTH) - Absolute shell thickness level below which adaptivity should begin. This
option works only if the adaptive angle tolerance (TOL) is nonzero. If thickness based adaptive remeshing
is desired without angle change, set TOL to a large angle. (If ADPTH = 0.0, this option is not used.)
Maximum Element Limit (MAXEL) - Maximum number of elements at which adaptivity will be terminated.
Adaptivity is stopped if this number of elements is exceeded.
For most problems, you should do an initial analysis without adaptive meshing. If the results of the initial analysis
show a highly distorted mesh and if the solution appears to be incorrect, then you can rerun the analysis with
adaptive meshing turned on. Adaptive meshing may also be used when an analysis terminates in LS-DYNA with
a negative volume element error.
When adaptive meshing is turned on, the number of elements within the model will change during the solution.
After each adaptive loop, the mesh will be updated and a new results file will be written with the extension RSnn,
where nn is the adaptive mesh level. (An adaptive loop will occur at every time increment specified by FREQ or
at the time intervals specified by LCID.) For example, a model that is adaptively remeshed two times will produce
two results files, Jobname.RS01 and Jobname.RS02. For details on postprocessing adaptively meshed results,
see Section 12.2.3: Postprocessing after Adaptive Meshing.
Note A new results file will be created for each loop, even if the mesh does not change.
A new time history file having the extension HInn is also written each loop; however, these files may not be as
useful for postprocessing as the RSnn files. In addition, LS-DYNA creates a set of adaptive meshing files with the
root name "adapt". Therefore, you should not use "adapt" as your jobname when adaptive meshing is activated.
312
Chapter 4: Loading
After building the model, the next step is to apply loads to the structure in preparation for solution. In order to
properly model your structures behavior, it is necessary to apply loads with respect to a specified time interval.
This chapter discusses the following topics related to loads:
General loading options
How to apply, delete, and list general loads [EDLOAD]
How to plot the explicit dynamics load curve [EDPL]
How to define a data curve [EDCURVE]
The use of components or part IDs and array parameters
How to show and hide load symbols
Constraints and initial conditions
How to define sliding and cyclic symmetry planes [EDBOUND]
How to define other miscellaneous types of constraints [EDCNSTR]
How to define welds [EDWELD]
How constraints [D, EDNROT] are used in ANSYS LS-DYNA
How to apply initial velocities [EDVEL, EDPVEL] to your model
Coupling and constraint equations

Coupling degrees of freedom [CP]
Constraint equations between degrees of freedom [CE]
Nonreflecting boundaries [EDNB]
Temperature loading
Dynamic relaxation
The following loading topics are available:

4.1. General Loading Options
4.2. Constraints and Initial Conditions
4.3. Coupling and Constraint Equations
4.4. Nonreflecting Boundaries
4.5. Temperature Loading
4.6. Dynamic Relaxation
4.1. General Loading Options

Unlike most implicit analyses, all loads in an explicit analysis must be time-dependent in nature. Hence, when
using ANSYS LS-DYNA, many of the standard ANSYS commands are not valid. In particular, the F, SF, and BF
family of commands are not applicable in ANSYS LS-DYNA because they can only be used to specify time-independent loads. Additionally, the D command can only be used to define constrained nodes. For this reason, all
Chapter 4: Loading
loads in ANSYS LS-DYNA are applied using a pair of array parameters, one corresponding to the time and the
other corresponding to the loading condition.
Note Although nodal accelerations (AX, AY, AZ) and nodal velocities (VX, VY, VX) appear as degrees
of freedom, they are not physical DOFs and cannot be constrained with the D command. To apply these
nodal loads, use the EDLOAD command.
In ANSYS LS-DYNA, all loads are applied in one load step. This is much different from an implicit analysis where
loads are often applied in multiple steps. In ANSYS LS-DYNA, for certain kinds of loads, you can also specify when
a load is imposed on a body (birth time), and when the load is removed (death time) using the EDLOAD command.
Please refer to Birth Time, Death Time, and CID Support in the EDLOAD command to check the applicability of
the birth/death time.
To apply a load to your model, you will need to follow these steps:
Designate portions of the model that will receive the load as components (or parts, for rigid bodies)
Define array parameters containing time intervals and load data values
Specify load curves
Define the load direction using the EDLCS command if the load is not acting in the global coordinate
system
Apply loads to the model
4.1.1. Components
With the exception of loads on rigid bodies, all loads are applied to components in an explicit analysis. Therefore,
the first step in applying loads is to gather portions of your model into nodal or element groups called components.
Each component should consist of those portions of the model that will receive the same loads and that are related
to each other through such things as material properties, position in the model, expected behavior, etc.
For example, if you want to analyze a baseball being hit into a wall, you would probably define the nodes of the
ball as one component, the nodes on the bat as another component, and the nodes on the wall as a third component.
You may define as many components as you wish, and then apply loads to each of them. The components must
consist of either nodes or elements. (Components consisting of elements are only used when applying pressure
loads.) To define a component, you must first select only those entities that you want to be included in the
component. Then use the CM command or the following menu path in the GUI to define the component: Utility
Menu> Select> Comp/Assembly> Create Component.
See Selecting and Components in the ANSYS Basic Analysis Guide for more information on components.
For the case of rigid bodies, loads are applied to a specific part number rather than to components. This is due
to the fact that rigid bodies already consist of a specific set of nodes and elements as defined by the command
EDMP,RIGID,MAT.
4.1.2. Array Parameters

It is important to remember that in an explicit dynamic analysis the loading is applied to the structure over a
specified time interval. This is done so that you may observe the model's transient behavior for applied loads
over the specific time period. Therefore, you will not only define what type of load you would like to apply to
your model (FX, FY, FZ, ROTX, ROTY, ROTZ, etc.), but also the time history of the load for the time interval of interest.
42
Section 4.1: General Loading Options

Time intervals and their corresponding load values are grouped together and defined as array parameters. These
array parameters should be defined in sets of two, with each entry of the first array specifying a time value, and
each entry of the second array specifying a load value. The load entries in the second array correspond to the
time values specified in the first array.
In the GUI, use this menu path to define an array parameter: Utility Menu> Parameters> Array Parameters>
Define/Edit.
See Array Parameters in the ANSYS APDL Programmer's Guide for more information on how to define array parameters.
Note Load values for intermediate time points are obtained by linear interpolation. However, load
values outside of the specified time range are not extrapolated by the program. Therefore, you should
ensure that the load time range is at least equal to the solution time. Otherwise, the results near the
solution end time may be invalid due to premature load removal.
Once you define a set of array parameters that represent a time dependent load, you can apply the load by inputting the array parameters directly on the EDLOAD command. Or, you may choose to input the array parameters
on the EDCURVE command to define a load curve. The corresponding load curve ID (LCID) can then be input
on the EDLOAD command.
To illustrate the use of array parameters, consider the baseball example mentioned previously. Suppose that
you wish to examine the distortion of the ball from the moment of impact with the bat to one second later. Assume
that the displacement of the bat handle is known as a function of time and that the ball is initially traveling at
1600 in/sec (91 MPH) just prior to impact.
You will first need to define some nodal components that will be used to define the loading and contact surfaces.
Create a component named ball that contains all of the nodes of the ball, for which an initial velocity (EDVEL)
of 1600 in/sec will be applied (discussed later in this chapter). Then create a second component named ballsurf
which contains only the nodes on the surface of the ball. This component will be used later in the contact specification. You also need to define a third component named batsurf, which contains the nodes on the surface
of the bat. Contact algorithms are discussed in Chapter 6, Contact Surfaces.
nsel,s,node,...
cm,ball,node
nsel,s,node,...
cm,ballsurf,node
nsel,s,node,...
cm,batsurf,node
nsel,all
!
!
!
!
!
!
select
define
select
define
select
define
all
the
the
the
the
the
nodes comprising the ball

component ball
nodes on the ball surface
component ballsurf
nodes on the bat surface
component batsurf
Now select the nodes at the base of the bat handle (bathand) for which a displacement versus time loading curve
will be applied.
Define an array named time to store the time values. Remember to use time values that are consistent with all
loads, dimensions, and material properties in your model. Next, define an array named xdisp to contain the corresponding X displacements of the component bathand. Likewise, define arrays ydisp and zdisp to contain the
corresponding Y and Z displacements of bathand.
nsel,s,node,...
! select the nodes at the bat handle base
cm,bathand,node
! define the component bathand
nsel,all
*dim,time,,4
! dimension the array parameter time
*dim,xdisp,,4
! dimension the array parameter xdisp
*dim,ydisp,,4
! dimension the array parameter ydisp
*dim,zdisp,,4
! dimension the array parameter zdisp
time(1)=0,0.25,0.5,0.75,1 ! times at specified displacements
xdisp(1)=0,-1,-2,-1,3
! X displacement of the bat handle
43
Chapter 4: Loading
ydisp(1)=0,1,2,3,4
zdisp(1)=0,3,6,8,9
! Y displacement of the bat handle

! Z displacement of the bat handle
The example given is actually a simple version of a more complex phenomenon. In a more accurate simulation,
additional displacement locations (and corresponding load curves) should be defined to better simulate the
true motion of the bat handle. Further, the initial velocities of the nodes comprising the bat are all different. Finally,
the ball is a composite of several different materials and material models.
4.1.3. Applying Loads

Once you have defined the components and array parameters, you are ready to apply loads to your model [EDLOAD command]. In the GUI choose the following menu path: Main Menu> Solution> Loading Options>
Specify Loads.
You may now choose to add loads (via the ADD label on EDLOAD) such as forces, moments, nodal accelerations,
surface pressures, etc. Loads applied with the EDLOAD command are in the global Cartesian direction by default.
The dialog box in the GUI will give you a listing of all valid loads, as well as the components and parameters that
have been previously defined. You simply choose the desired load label along with the combination of components
(part numbers for rigid bodies) and array parameters (or previously defined load curve ID) you wish to use for
that load. It is important to note, as listed below, that all load labels are not valid for all components or part
numbers.
The following load labels are valid only for nodal components:
Forces: FX, FY, FZ
Moments: MX, MY, MZ
Displacements: UX, UY, UZ
Rotations: ROTX, ROTY, ROTZ
Velocities: VX, VY, VZ
Nodal Accelerations: AX, AY, AZ
Body Accelerations: ACLX, ACLY, ACLZ
Angular Velocities: OMGX, OMGY, OMGZ
Temperature: TEMP
Note Although V (X, Y, Z) and A (X, Y, Z) appear as DOFs, they are not actually physical DOFs. However,
these quantities are computed as DOF solutions and stored for postprocessing.
The following labels are valid only for rigid bodies (part numbers):
Forces: RBFX, RBFY, RBFZ
Moments: RBMX, RBMY, RBMZ
Displacements: RBUX, RBUY, RBUZ
Rotations: RBRX, RBRY, RBRZ
Velocities: RBVX, RBVY, RBVZ
Angular Velocities: RBOX, RBOY, RBOZ
The following label is valid only for element components:
Pressure: PRESS
Returning to our baseball example, the necessary displacement versus time load curves are automatically created
with the EDLOAD command from the time and x/y/zdisp array parameters.
44
Section 4.1: General Loading Options

edload,add,ux,,bathand,time,xdisp
edload,add,uy,,bathand,time,ydisp
edload,add,uz,,bathand,time,zdisp
! X displacement of bathand
! Y displacement of bathand
! Z displacement of bathand
Alternatively, the very stiff bat could have been modeled as a rigid body to simplify the required input and to
reduce the CPU time. For this case, the appropriate rigid body loads would be applied to valid part numbers
(and not to nodal components).
edload,add,rbux,,2,time,xdisp
edload,add,rbuy,,2,time,ydisp
edload,add,rbuz,,2,time,zdisp
! X displ. (if bat = part #2)

! Y displ. (if bat = part #2)
! Z displ. (if bat = part #2)
If the pressure that the bat imparts on the baseball was known as a function of time, the loading could be accomplished without defining any contact surfaces. Instead, the element component containing the elements on the
surface of the ball would be loaded with the PRESS label of the EDLOAD command.
edload,add,press,1,cover,battime,batload
In this last scenario, the elements contained in the element component cover are loaded on face number one
(face number is input in the KEY field) with the load curve produced by the battime and batload array parameters.
It is important to note that in the examples listed above, the pressure load uses the KEY field of the EDLOAD
command to specify a load key. Load keys (1,2,3, etc.) are associated with surface loads and are listed under
"surface loads" in the input data tables for each element type in the ANSYS Elements Reference. For most load
types other than pressure loads, you can use the KEY field to specify a coordinate system identification number,
CID, from the EDLCS command. The load will act in a direction you define on EDLCS, or in the global coordinate
system direction if CID is not specified. See Section 4.1.5: Defining Loads in a Local Coordinate System for more
information.
Note To avoid timing problems on some platforms, it is a good practice to always add a very small
time value (such as 1.0 10-6) to the value in the final item in the time array. For example, instead of the
value 3.0, such an array might contain the following value for the last item:
timeint(1)=0,1,2,3.00001
The addition of this very small padding factor does not affect the accuracy of the results.
In addition to adding loads, you can also list and delete loads with the commands EDLOAD,LIST and EDLOAD,DELE. You can display load symbols on the element plot with the EDFPLOT command, and you can plot
load curves with the EDPL command.
4.1.4. Data Curves

Data curves, which are defined with the EDCURVE command, have various uses in ANSYS LS-DYNA. They can
be used to define material data curves (for example, stress-strain) and load data curves (force-deflection) associated
with explicit dynamics material models. Data curves can also be used to define load curves that represent time
dependent loads (force, displacement, velocity, etc.). These load curves can then be input on the EDLOAD
command.
4.1.4.1. Using Data Curves with Material Models

Certain material models (for example, TB,PLAW or TB,HONEY) require specification of material property data
which may be a function of effective strain rate, plastic strain, or volumetric strain. For such data, property curves
need to be defined using the EDCURVE command prior to specifying the material behavior with the data table
[TBDATA] command. Data curves are also used for defining force deflection behavior in rigid-body and drawbead
contact.
45
Chapter 4: Loading
Similar to placing loads on components, data curves are grouped into array parameters and then associated
with a particular curve reference number (LCID) which can be used by a specified material model (PLAW, HONEY,
etc.) or contact type (RNTR, ROTR) and drawbead. Use the following steps to specify data curves:
1.
Define an array parameter that contains the abscissa values for the material or friction force behavior
(e.g., effective plastic strain, effective strain rate, displacement, etc.).
2.
Define a second array parameter that contains the ordinate values for the material behavior or friction
force (e.g., initial yield stress, elastic modulus, force, etc.).
3.
Specify the data curve that will be used to identify this data [EDCURVE]. During this step you will choose
a data curve ID number that will be used to associate this data with a particular material behavior when
setting up the data table [TBDATA].
After the parameters have been specified, use one of the following menu paths in the GUI to define a data curve:
Main Menu> Preprocessor> Material Props> Curve Options
Main Menu> Solution> Loading Options> Curve Options
Defined data curves can be listed using EDCURVE,LIST. They can be plotted using EDCURVE,PLOT and can be
deleted using EDCURVE,DELETE.
The following example shows how data curves can be used to define a piecewise linear plasticity curve
(TB,PLAW,,,,8) for steel:
! "3" was arbitrarily chosen as the material reference (MAT) number.
mp, ex,3,30.0e6
! elastic (Youngs) modulus (psi)
mp,dens,3,7.33e-4
! mass density (lbf-sec^2/in^4)
mp,nuxy,3,0.30
! Poissons ratio (unitless)
!Note: First convert engineering stress versus engineering strain data
!
into true stress versus true (hencky) strain data. Then subtract
!
off the elastic true strain from the total true strain to find
!
the plastic true strain, which is used with the total true stress
!
in LS-DYNA *MAT_PIECEWISE_LINEAR_PLASTICITY material model #24.
!-----------------------------------------------------------------------! Stress-Strain Data used with Piecewise Linear Plasticity (Power Law 8):
!-----------------------------------------------------------------------!
Total
Total
Total
Total
Elastic
Plastic
! Stress/
Eng.
Eng.
True
True
True
True
! Strain
Stress
Strain
Stress
Strain
Strain
Strain
! Point
(psi)
(in/in)
(psi)
(in/in)
(in/in)
(in/in)
!-----------------------------------------------------------------------!
1
0
0.0000
0
0.0000
0.0000
0.0000
!
2
60,000
0.0020
60,120
0.0020
0.0020
0.0000
!
3
77,500
0.0325
80,020
0.0320
0.0027
0.0293
!
4
83,300
0.0835
90,260
0.0802
0.0030
0.0772
!
5
98,000
0.1735
115,000
0.1600
0.0038
0.1562
!
6
98,300
0.2710
124,940
0.2398
0.0042
0.2356
!
7
76,400
1.2255
170,030
0.8000
0.0057
0.7943
!-----------------------------------------------------------------------!Note: The first point on the stress/strain curve is NOT entered.
!
Start with the second point (where ordinate = yield stress).
!
Also, please follow the limits imposed by the *SET command.
*dim,strn,,6 ! define array for effective plastic true strain data
*dim,strs,,6 ! define array for effective total true stress data
strn(1)= 0.0, 0.0293, 0.0772, 0.1562, 0.2356, 0.7943 ! strain (in/in)
strs(1)= 60120., 80020., 90260., 115000., 124940., 170030. ! stress (psi)
edcurve,add,1,strn,strs ! curve #1: abscissa=strain & ordinate=stress
tb,plaw,3,,,8
! specify power law #8 for material (MAT) #3
tbdata,6,1
! use load curve #1 for stress/strain data
!Note: If desired, a plastic failure strain can be defined. Further,
!
strain rate effects can be included by specifying the necessary
!
strain rate parameters or the load curve defining the strain rate
!
scaling effect on the yield stress. Please refer to Chapter 7 of
!
this guide for a complete description of this material model.
46
Section 4.2: Constraints and Initial Conditions
4.1.4.2. Using Data Curves for Loading

In addition to their use with certain material models, data curves (also known as load curves) can be used in the
definition of time dependent loads. The procedure to define a load curve is the same as described above for
material data curves, except that the first array parameter must contain the time values and the second array
parameter must contain the corresponding load values. After you define a load curve with the EDCURVE command,
the corresponding load curve reference number (LCID) can be input on the EDLOAD command.
The following 3-step example shows how load curves are used with the EDLOAD command.
! Step 1: Define Array Parameters
*dim,time,,5
time(1)=0,.025,.05,.075,.1
*dim,yforce,,5yforce(1)=0,100,200,300,400
!
! Step 2: Define Load Curve and Corresponding LCID (#11)
edcurve,add,11,time,yforce
!
! Step 3: Refer to LCID on the EDLOAD Command
edload,add,fy,,comp,,,,11,1.0
As shown above, the LCID is referenced in the ninth field of the EDLOAD command. It is important to note that
if you use an LCID with the EDLOAD command, array parameters should not also be specified to define the
load. Using an LCID instead of array parameters on the EDLOAD command is particularly useful when a specific
load curve is to be applied to multiple components or load labels.
4.1.5. Defining Loads in a Local Coordinate System

ANSYS LS-DYNA allows you to apply a load to a component or part in any coordinate direction for which a local
coordinate system has been defined. This is accomplished using the KEY field of the EDLOAD command in
conjunction with the local coordinate system (EDLCS command). Once a local coordinate system has been
defined with the EDLCS command (EDLCS,ADD,CID,X1,Y1,Z1,X2,Y2,Z2, X3,Y3,Z3), the local coordinate system
ID (CID) is used in the KEY field of the EDLOAD command to define the direction in which the load acts. For force
and moment loads (Lab = FX, MX, etc. on EDLOAD), the load will be applied in the direction of the local coordinate
system defined by EDLCS. For prescribed motion degrees of freedom (Lab = UX, ROTX, VX, AX, etc. on EDLOAD),
the motion will act in the direction of a vector from point (X1, Y1, Z1) to point (Z2, Y2, Z2) as input on EDLCS. If
CID is not specified, the load will act in the global Cartesian coordinate direction. Some load types do not support
the use of local coordinate systems; see Birth Time, Death Time, and CID Support Table in the Notes section of
the EDLOAD command for more information.
4.1.6. Specifying Birth and Death Times

For each load definition, you can specify a birth time and a death time by using the BTIME and DTIME fields of
the EDLOAD command (EDLOAD,ADD,Lab,KEY,Cname,Par1,Par2,PHASE,LCID,SCALE,BTIME,DTIME). These
options allow you to activate a load at any time during an analysis and then deactivate it at a later time. Such an
option can be very useful in applications such as a multistage forming process where more than one load is applied
in succession. Some load types do not support birth and death time; see Birth Time, Death Time, and CID Support
Table in the Notes section of the EDLOAD command for more information.
4.2. Constraints and Initial Conditions

Before initiating the solution, you will need to apply constraints to your model. Additionally, you may wish to
set initial velocities for moving objects.
47
Chapter 4: Loading
4.2.1. Constraints
Unlike the ANSYS implicit program, ANSYS LS-DYNA does distinguish between zero and nonzero constraints.
Nonzero constraints are handled as loads (together with a load curve; see earlier discussion in this chapter). Only
zero constraints can be applied with the D command, i.e., the value specified must always be 0 (zero). No other
values are valid, as the D constraints are only used to fix certain portions of the model. You may use symmetry/antisymmetry boundary conditions to apply these zero constraints.
You can use the EDNROT command to apply zero constraints in terms of a rotated nodal coordinate system.
You must first define the local coordinate system with the EDLCS command.
When modeling small symmetric sections of a geometry, you may need to define sliding or cyclic symmetry. You
can use the EDBOUND command to define symmetry boundary planes for sliding or cyclic symmetry. You use
nodal components to identify the boundaries and a direction vector to define a normal (sliding symmetry) or
axis of rotation (cyclic symmetry).
Various other types of constraints can be modeled in ANSYS LS-DYNA using the EDCNSTR command. The
available constraint types are extra node set (ENS), nodal rigid body (NRB), shell edge to solid (STS), and rivet
(RIVET). In the GUI, you access these constraints by picking:
Main Menu> Solution> Constraints> Apply> Additional Nodal
The extra node set constraint type (EDCNSTR,ADD,ENS) allows the addition of nodes (via a nodal component)
to an existing rigid body that was defined with the EDMP command. The nodal component that is added must
not be attached to any other rigid body. The extra nodes that are added to a rigid body may be located anywhere
in the model, and may have coordinates outside those of the original rigid body. The ENS option has many potential applications, including placing nodes where joints will be attached between rigid bodies, defining nodes
where point loads will be applied, and defining a lumped mass at a specific location.
Unlike typical rigid bodies that are defined with the EDMP command, nodal rigid bodies defined with the EDCNSTR,ADD,NRB command are not associated with a part number, but are based on a node component. The NRB
option can be advantageous when modeling rigid (welded) joints in a model. For a rigid joint, portions of different
flexible components (having different MAT IDs) act together as a rigid body. It is difficult to define this type of
rigid body with a unique MAT ID (and corresponding part number). However, the rigid joint can be easily defined
using a nodal rigid body. Because nodal rigid bodies are not associated with a part number, other options that
use rigid bodies (such as loads applied with the EDLOAD command ) cannot be used with a nodal rigid body.
The shell to solid edge option (EDCNSTR,ADD,STS) ties regions of solid elements to regions of shell elements.
As shown in Figure 4.1: Constrained Shell to Solid, a single shell node may be tied to up to nine solid nodes
that define a 'fiber' vector. The solid element nodes that define the fiber remain linear (straight) throughout the
analysis but can move relative to each other in the fiber direction. The shell node must be coincident with one
of the solid element nodes along the fiber.
48
Section 4.2: Constraints and Initial Conditions
Figure 4.1 Constrained Shell to Solid

& $21$0 !()%&
# '
&%$$ "!

#
4
3
5

7
6

8
`a4DB 77P YWVTC4RQ7P4IGEFDB 1@

PC XU XAU SH 6P B H 9E C A
9
9$
Similar to spotwelds defined with the EDWELD command, the RIVET option (EDCNSTR,ADD,RIVET) defines a
massless rigid constraint between two nodes that are noncoincident. Unlike a spotweld, however, failure cannot
be specified for a rivet. When a rivet is defined, the distance between the nodes is kept constant throughout any
motion that occurs during a simulation. Nodes connected by a rivet cannot be part of any other constraints
specified in the model.
4.2.2. Welds
In an explicit dynamic analysis, it is common to model components that are physically welded together. This is
especially true in automotive applications where parts are often assembled using welds. In such cases, welded
constraints can be modeled in ANSYS LS-DYNA using the EDWELD command. Two different types of welds can
be modeled: massless spotwelds and generalized welds. Nodes connected by the EDWELD command cannot
be constrained in any other way.
For a massless spotweld, you must specify two noncoincident nodes. You can define failure within the spotweld
by inputting failure parameters on the EDWELD command. This failure is based on the following relationship:
fn
S
n
exp n
fs
+
S
s
exp s
In the GUI, use the following menu path to define a massless spotweld: Main Menu> Preprocessor> LS-DYNA
Options> Spotweld> Massless Spotwld
A generalized weld is typically used to model a long-welded section of two parts. For a generalized weld, you
must specify a valid nodal component. Coincident nodes are permitted. However, if you use coincident nodes,
you must specify a local coordinate system (EDLCS) which will be used for output data. Failure within the generalized weld is also defined with the failure parameters on the EDWELD command using the same relationship
as shown in the equation above.
49
Chapter 4: Loading
In the GUI, use the following menu path to define a generalized weld: Main Menu> Preprocessor> LS-DYNA
Options> Spotweld> Genrlizd Spotwld
4.2.3. Initial Velocity

In transient dynamic simulations, it is often necessary to specify initial conditions. In the ANSYS LS-DYNA program,
initial velocities are defined using the EDVEL and EDPVEL commands. You can use these commands to apply
both translational and rotational velocities to various entities. Use EDVEL to apply velocities to a nodal component
or to a single node; use EDPVEL to apply velocities to a part or part assembly.
The EDVEL and EDPVEL commands provide two methods for specifying rotational velocities, Option = VGEN
and Option = VELO. The VGEN method applies a rigid body rotation to an entity (node component, part, etc.)
around a specified axis. The VELO method applies the rotational velocities directly to each node's rotation degree
of freedom. Since only shell and beam elements have rotation degrees of freedom, the rotations input with the
VELO method are only applicable to SHELL163 and BEAM161 elements. For both the VGEN and VELO methods,
the translational velocities are always defined relative to the global Cartesian coordinate system.
Note To model a rotating body, with or without translations, you should use Option = VGEN since
this method applies a rigid body rotation.
Due to the LS-DYNA architecture, the two methods for defining initial velocities, Option = VELO and Option =
VGEN, cannot be used in the same analysis.
To define an initial velocity, use the following procedure:
1.
Define the entity to which you want to apply the initial velocity. This may be a single node, a nodal
component [CM], a part [EDPART], or a part assembly [EDASMP].
2.
Decide whether the VELO or VGEN option is appropriate for your application.
3.
Specify the initial velocity [EDVEL or EDPVEL] by using one of the following menu paths in the GUI:
Main Menu> Solution> Initial Velocity> On Nodes> w/Nodal Rotate (VELO option)
Main Menu> Solution> Initial Velocity> On Nodes> w/Axial Rotate (VGEN option)
Main Menu> Solution> Initial Velocity> On Parts> w/Nodal Rotate (VELO option)
Main Menu> Solution> Initial Velocity> On Parts> w/Axial Rotate (VGEN option)
For either the VELO or the VGEN option, input the translational velocities relative to the global Cartesian coordinate
system (VX, VY, and VZ fields on the EDVEL or EDPVEL command). For the VELO option, input the nodal rotations
in terms of the global Cartesian coordinate system (OMEGAX, OMEGAY, and OMEGAZ fields). For the VGEN option,
input the magnitude of the angular velocity (OMEGAX), the coordinate axis for rotation (XC, YC, and ZC fields) and
directional angles (ANGX, ANGY, and ANGZ fields) relative to the global X, Y, and Z axes.
If you do not specify initial velocities using EDVEL or EDPVEL, all initial velocities in the model will be zero.
Likewise, if you only specify the Cname field on the EDVEL command (e.g., EDVEL,VGEN,Cname) or the PID field
on the EDPVEL command, (e.g., EDPVEL,VGEN,PID), zero initial velocities will be applied since the remaining
fields on these two commands default to zero.
If you wish to change an initial velocity previously specified using EDVEL, just respecify a new velocity with the
same component name or node number. This new value will overwrite the old data for the component or node.
To list or delete initial velocities previously applied to node components or nodes, use the EDVEL,LIST and
EDVEL,DELE commands (in the GUI, use Main Menu> Solution> Initial Velocity> On Nodes> List and Main
Menu> Solution> Initial Velocity> On Nodes> Delete).
410
Section 4.4: Nonreflecting Boundaries

Similarly, if you wish to change an initial velocity previously specified using EDPVEL, just respecify a new velocity
with the same part or part assembly ID number. This new value will overwrite the old data for the part or part
assembly. To list or delete initial velocities previously applied to parts or part assemblies, use the EDPVEL,LIST
and EDPVEL,DELE commands (in the GUI, use Main Menu> Solution> Initial Velocity> On Parts> List and
Main Menu> Solution> Initial Velocity> On Parts> Delete).
For our baseball example, the initial velocity of the baseball would be specified as follows:
edvel,velo,ball,,,-1600.0 ! Component ball initially moving
! in -Z direction at 1600 in/sec
Or, assuming the component ball was part number 1 in the model, you could specify the velocity of the baseball
as:
edpvel,velo,1,,,-1600.0 ! Part number 1 initially moving
! in -Z direction at 1600 in/sec
4.3. Coupling and Constraint Equations

The CP-family of commands (CP, CPDELE, CPINTF, CPLGEN, CPLIST, CPNGEN, CPSGEN) can be used to define,
modify, delete, list, and generate coupling between different degrees of freedom (DOFs) of a structure. Similarly,
the CE-family of commands (CE, CEDELE, CEINTF, CELIST, CERIG) can be used to define, modify, delete, and list
constraint equations between different DOFs of a structure. When used in the ANSYS LS-DYNA explicit dynamics
program, the CP and CE commands can be used only with UX, UY, and UZ DOFs (rotational DOFs are not allowed).
Since the rotational DOFs (ROTX, ROTY, ROTZ) are not allowed, the CP family of commands should not be used
in an explicit analysis to model rigid body behavior that involves rotations. If CP is used in this manner, it could
lead to nonphysical responses.
Also, be aware that a coupled set containing nodes that are not coincident or nodes that are not along the line
of the coupled degree of freedom does not produce a moment constraint. This means that if the structure rotates,
the set of coupled nodes will rotate. Only the applied forces and the reaction forces will satisfy the moment
equilibrium in the model. For each node in the coupled set, the result of satisfying the moment will depend on
the distance from the node to the center of the coupled set, and the direction of displacement will depend on
the resulting moment. This may lead to a nonphysical response in some cases.
See Coupling and Constraint Equations in the ANSYS Modeling and Meshing Guide for details on defining coupling
and constraint equations.
4.4. Nonreflecting Boundaries

When modeling geomechanical systems, an infinite domain is often required to represent the ground or other
large solid body. For this type of analysis, you can use nonreflecting boundaries at the exterior of the model to
limit the overall size of the model. You apply these boundaries to the surface of SOLID164 and SOLID168 elements
that are being used to model an infinite domain. Nonreflecting boundaries will prevent artificial stress wave reflections generated at the boundary from reentering the model and contaminating the results.
When nonreflecting boundaries are included in the model, LS-DYNA computes an impedance matching function
for all boundary segments based on an assumption of linear material behavior. Therefore, the finite element
mesh should be constructed so that all significant nonlinear behavior is contained within the discrete model.
To define a nonreflecting boundary, select the nodes (NSEL command) that make up the desired boundary along
the external surface of a SOLID164 or SOLID168 body, then define a component from these nodes (CM command).
Use the EDNB command to apply the nonreflecting boundary to that nodal component, and to turn on the
411
Chapter 4: Loading
dilatational and shear activation flags. For example, the following command defines a nonreflecting boundary
on the component ground:
EDBOUND,ADD,GROUND,1,1
To list and delete defined nonreflecting boundaries, use the commands EDNB,LIST and EDNB,DELE, respectively.
4.5. Temperature Loading

You may need to define temperature loading in an explicit dynamic analysis in order to accommodate temperature dependent materials, or to include the effects of thermally induced stresses. Temperature loading is applicable to the PLANE162, SHELL163, SOLID164, and SOLID168 element types. The ANSYS LS-DYNA program offers
several types of temperature loading:
Time-varying temperature applied to a nodal component (EDLOAD)
Constant temperature applied to all nodes in the model (TUNIF / BFUNIF)
Temperature results from an ANSYS thermal analysis applied as non-uniform temperature loads (that do
not vary with time) in a subsequent explicit dynamic analysis (LDREAD; also requires a sequential solution)
The first method utilizes the EDLOAD,,TEMP command and the general loading procedure to apply a timevarying temperature to a specific nodal component. You must define two array parameters to represent the
load; the first parameter contains the time values and the second parameter contains the temperature values.
You may choose to define a load curve with these parameters or to input them directly on the EDLOAD command.
You can also scale the temperature values using the SCALE parameter on EDLOAD. See Section 4.1: General
Loading Options for a complete description of how to apply this type of load.
The second method allows you to apply a uniform constant temperature to all nodes in the model. This method
may be used to model a structure subjected to steady-state thermal loading. You can use the TUNIF command
or the BFUNIF,TEMP command to apply this type of temperature load.
The third method allows you to apply the temperatures calculated in an ANSYS thermal analysis as loads in an
explicit dynamic analysis. This method is useful for modeling temperature-dependent phenomena such as forging.
To use this method, you must perform an implicit-to-explicit sequential solution. In the explicit phase, you use
the LDREAD command to read the temperature data from the thermal (implicit) analysis results file (Jobname.RTH)
and apply them as loads to the nodes in the model. You can only transfer the temperatures from one specified
time point in the thermal analysis. Refer to Chapter 15, Implicit-to-Explicit Sequential Solution of this guide for
a detailed description of this loading procedure.
For all three methods of temperature loading, a reference temperature can be input via the TREF command.
The thermal loading is defined as the difference between the applied temperature and the reference temperature.
If the reference temperature is not specifically defined, it defaults to zero.
In order for any of these thermal loads to take effect, you must use the temperature dependent bilinear isotropic
material model or the elastic viscoplastic thermal material model. The temperature dependent bilinear isotropic
model can be used to represent a thermoelastic material (if needed) by omitting the yield strength and tangent
modulus. Please refer to Chapter 7, Material Models of this guide for complete descriptions of these material
models.
Note It is important to note that the EDLOAD method of temperature loading cannot be used together
with the LDREAD or TUNIF / BFUNIF methods of temperature loading. In addition, EDLOAD cannot be
used to list or delete temperatures applied with LDREAD, TUNIF, or BFUNIF.
412
Section 4.6: Dynamic Relaxation

You can use LDREAD and TUNIF (or BFUNIF) together in the same explicit dynamic analysis. When LDREAD is
issued, it applies temperature loads to the selected nodes, overwriting any temperature loads defined by TUNIF
or BFUNIF. Any nodes that are unselected at the time LDREAD is issued will take on the temperature load defined
by the TUNIF or BFUNIF command. To delete the temperature loads defined by LDREAD (at selected nodes or
all nodes), use the BFDELE command. To list the temperature loads defined by LDREAD (at selected nodes or
all nodes), use the BFLIST command.
4.6. Dynamic Relaxation

In order to conduct implicit-to-explicit sequential solutions (see Chapter 15, Implicit-to-Explicit Sequential
Solution), dynamic relaxation capability has been added to the ANSYS LS-DYNA program (EDDRELAX command).
True dynamic relaxation (EDDRELAX,DYNA) allows an explicit solver to conduct a static analysis by increasing
the damping until the kinetic energy drops to zero. When an implicit solver is used to provide the preload (EDDRELAX,ANSYS), a slightly different approach is taken, in that the stress initialization is based on a prescribed
geometry (i.e., the nodal displacement results from the implicit solution). In this latter case, the explicit solver
only uses 101 time steps to apply the preload. In the former case, the solver will check the kinetic energy every
250 cycles (by default) until the kinetic energy from the applied preload is dissipated. ANSYS LS-DYNA supports
both methods, which occur in pseudo time before the transient portion of the analysis begins at time zero. The
EDLOAD command specifies the analysis type through the PHASE label.
EDLOAD , ADD, Lab, , Cname, Par1, Par2, PHASE
PHASE:
Load curve for transient analysis only (default) or for implicit-to-explicit

sequential solutions
Load curve for dynamic relaxation only
Load curve for both transient and dynamic relaxation analyses
There are essentially five different types of analyses that can be conducted with ANSYS LS-DYNA concerning
dynamic relaxation. They are discussed below:
1.
Transient Dynamic Analysis Only (default - EDDRELAX,OFF): In this case, the PHASE parameter of the
EDLOAD command is set to zero and dynamic relaxation is not used. This is the default setting.
2.
Stress Initialization Only with No Transient Analysis (EDDRELAX,DYNA): This case basically uses the ANSYS
LS-DYNA explicit solver to approximate the solution of a static analysis that is really best handled by an
implicit solver such as ANSYS. The PHASE parameter on the EDLOAD command is set to one. The termination time (TIME command) must be set to zero to prevent unloading of the static load. True dynamic
relaxation is used.
3.
Stress Initialization with a No-Load Transient Analysis (EDDRELAX,DYNA): This case is the same as the
previous case, except that the termination time is set to the desired value. At time zero, the structure is
instantly unloaded and free to vibrate.
4.
Stress Initialization with a Loaded Transient Analysis (EDDRELAX,DYNA): This is similar to the previous
case, except that the load curve specified by the EDLOAD command (via PHASE = 2) will be used in both
pseudo time for the dynamic relaxation to obtain the preload and in real time for the transient analysis.
If a ramped load curve is used instead of a constant load curve, then the structure will unload at time
zero and reload again. Therefore, if a ramped load curve is needed for the explicit solver to accurately
converge on the static solution, it would be better to use two EDLOAD commands, the first (with PHASE
= 1) to apply the preload using a ramped load curve and the second (with PHASE = 0) to continue the
load with a load curve that is not ramped.
413
Chapter 4: Loading
Note If you issue EDLOAD multiple times for the same component (or Part ID) and the same
load label, the values associated with the last EDLOAD command are used. Consequently, you
cannot use multiple values of the PHASE parameter for a given component (or Part ID) and load
label. To work around this limitation, you must use duplicate nodal components for the stress
initialization and transient curves. (This workaround is applicable only if you use components to
define the load.)
5.
414
Implicit-to-Explicit Sequential Solutions (EDDRELAX,ANSYS): This case uses the displacement results
from the ANSYS implicit solver to apply the preload for the ANSYS LS-DYNA explicit solver via stress initialization to a prescribed geometry. The PHASE parameter on the EDLOAD command must be set to
zero. This scenario is explained in detail in Chapter 15, Implicit-to-Explicit Sequential Solution of this
guide. Fields 2-6 of the EDDRELAX command are ignored for this case.
Chapter 5: Solution Features

The solution procedure for an explicit dynamic analysis is similar to an implicit analysis. However, the ANSYS LSDYNA program offers several solution features that are unique to an explicit analysis.
The following ANSYS LS-DYNA solution topics are available:
5.1. Solution Process
5.2. LS-DYNA Termination Controls
5.3. Shared Memory Parallel Processing
5.4. Double Precision LS-DYNA
5.5. Solution Control and Monitoring
5.6. Plotting Small Elements
5.7. Editing the LS-DYNA Input File
5.1. Solution Process

After the model has been built (i.e., after completing element definitions, real constant and material property
specifications, modeling, meshing, boundary/initial conditions, loading, and termination controls), you can start
the solution process by issuing the SOLVE command (in the GUI, pick Main Menu> Solution> Solve).
At this point, several steps are performed by the ANSYS LS-DYNA program:
1.
The header records, including geometric quantities (such as nodes and elements), are written on the
two results files Jobname.RST and Jobname.HIS. (At this point, the ANSYS LS-DYNA database should
contain all relevant information. Before issuing the SOLVE command, you should issue a SAVE command
to write all model information to the file Jobname.DB.)
2.
The input file for the LS-DYNA program, Jobname.K, is written using all information that has been entered
so far.
3.
Control is transferred from the ANSYS LS-DYNA program to the LS-DYNA program. The solutions produced
by the LS-DYNA solver are written to the results files Jobname.RST and Jobname.HIS. If the EDOPT,ADD,,BOTH option was specified before issuing the SOLVE command, results files for the LS-POST
postprocessor (files d3plot and d3thdt) will also be written.
When the solution is complete without errors or warnings, the ANSYS LS-DYNA GUI notifies the user that the
solution is done, and control is transferred back to the ANSYS LS-DYNA program. The results can be viewed using
the POST1 and POST26 processors of the ANSYS LS-DYNA program. If errors or warnings are produced, messages
pop-up and are also displayed in the ANSYS Output Window. The messages state the number of errors and
warnings that were produced, and make reference to the LS-DYNA message file, where details of the errors and
warnings are written. These same details are also written to the LS-DYNA d3hsp file.
5.2. LS-DYNA Termination Controls

The point at which the LS-DYNA solution terminates will depend on the termination controls you specify when
setting up the model. Several types of termination controls are available:
Termination time - Use the TIME command to specify an end-time for the analysis. The calculation will
stop when the accumulation of time steps reaches that end-time.
CPU time limit - You can use the EDCPU command to specify a CPU time limit (in seconds). The calculation
will stop when that time limit is reached.
Termination criteria - You can use the EDTERM command to specify that the solution stop when a specific
node or rigid body reaches a certain position, or when a specific node comes into contact with another
surface. You may set up multiple criteria using this command. (See the EDTERM command for details.)
You should always specify an analysis end-time using the TIME command. The other termination controls are
optional. The solution will terminate when any one of the specified termination criteria is met.
When you are performing a small restart (EDSTART,2) or a full restart (EDSTART,3) analysis, if the previous analysis terminated due to a satisfied termination criteria that was set by EDTERM, you must change or delete that
criteria so that it will not cause the restart to terminate immediately.
5.3. Shared Memory Parallel Processing

For large models, you can use the shared memory parallel processing (SMP) capabilities of ANSYS LS-DYNA to
shorten the elapsed time necessary to run an analysis. To use this feature, you must have a machine with at least
two processors, and you must purchase the appropriate number of ANSYS LS-DYNA SMP licenses. Please contact
your ANSYS sales representative for more information on purchasing the appropriate licenses.
To use the SMP capabilities you need to:
1.
Specify the number of processors to be used. You can set this value via the config100.ans file, via the
/CONFIG command, or via the SETNPROC macro. For more information on all of these methods, please
refer to Section 15.1: Activating Parallel Processing in the ANSYS Advanced Analysis Techniques Guide. If
you do not set the number of processors to be greater than one, you will not be able to see the appropriate menu selections in the GUI to specify SMP, nor will the EDRUN command settings have any effect.
2.
Issue the EDRUN command to specify an SMP run.
When you are using shared memory parallel processing, the calculations may be executed in different order,
depending on CPU availability and the workload on each CPU. Because of this, you may see slight differences in
the results when you run the same job multiple times. To avoid these differences, you can specify that consistency
be maintained (EDRUN,SMP,1). Maintaining consistency can result in an increase of up to 15% in CPU time.
If you are using all CPUs on a platform (ANSYS recommends always running one fewer CPU than is available),
you should close all other applications on those CPUs. Any calculations or machine resources that compete with
the ANSYS LS-DYNA application running in SMP mode will reduce the SMP performance significantly.
The SMP capability is expected to scale linearly when used with up to 8 processors. When used with more than
8 processors, any additional scalability is minimal; the larger the problem, the higher the potential for scalability.)
5.4. Double Precision LS-DYNA

For more accurate results, you can use the double precision capabilities of ANSYS LS-DYNA. This feature is especially
useful in sequential explicit-to-implicit spring back types of analyses.
Note The double precision version of LS-DYNA may be up to 20% slower than the single precision
version. Results may also vary based on problem specifications.
To use this feature, you must have the double precision version of LS-DYNA available on your machine. The
double precision version of LS-DYNA is not available on the HP-UX 10.2, SGI 32-bit, IBM32-bit, and Linux 32-bit
platforms.
Use the EDDBL (EDDBL,1) command to specify the double precision version .
52
Section 5.5: Solution Control and Monitoring

You can check the numeric version status by issuing the EDDBL,STATUS command or the GUI dialog box for the
command (accessed through Main Menu> Solution> Analysis Options> Double Precision).
5.5. Solution Control and Monitoring

The ANSYS LS-DYNA program allows certain sense switch controls that enable you to interrupt the solution
process and to check the solution status. This section describes how to use these sense switch controls.
While the LS-DYNA execution is in progress, type CTRL-C in the console window of the ANSYS LS-DYNA program.
This interrupts the LS-DYNA program execution and lets you enter program control commands (i.e., switches)
in the console window of ANSYS LS-DYNA. (Unlike in ANSYS, you are not thrown out of the ANSYS LS-DYNA GUI
by the CTRL-C interrupt, and the LS-DYNA solver will continue to run in the background.) A typical segment of
output from the console window is shown below:
______________________________________________________________
dt of cycle
1 is controlled by solid
element
1
time...........................
0.00000E+00
time step......................
3.30496E-05
kinetic energy.................
2.82787E+03
internal energy................
1.00000E-20
spring and damper energy.......
1.00000E-20
hourglass energy ..............
0.00000E+00
system damping energy..........
0.00000E+00
sliding interface energy.......
0.00000E+00
external work..................
0.00000E+00
eroded kinetic energy..........
0.00000E+00
eroded internal energy.........
0.00000E+00
total energy...................
2.82787E+03
total energy / initial energy..
1.00000E+00
energy ratio w/o eroded energy.
1.00000E+00
global x velocity..............
0.00000E+00
global y velocity.............. -6.23477E+01
global z velocity..............
0.00000E+00
cpu time per zone cycle............
6666660 nanoseconds
average cpu time per zone cycle....
6666660 nanoseconds
average clock time per zone cycle.. 15167664 nanoseconds
estimated total cpu time
=
6 sec (
0 hrs
estimated cpu time to complete
=
6 sec (
0 hrs
estimated total clock time
=
13 sec (
0 hrs
estimated clock time to complete =
13 sec (
0 hrs
enter sense switch:
______________________________________________________________
0
0
0
0
mins)
mins)
mins)
mins)
At this point, you can enter one of the following four switches:
sw1
ANSYS LS-DYNA terminates. A restart file is written.
sw2
ANSYS LS-DYNA responds with time and cycle numbers and continues to run. This allows
you to see how far the solution has actually progressed.
sw3
ANSYS LS-DYNA writes a restart file and continues to run.
sw4
ANSYS LS-DYNA writes a results data set and continues to run.
The first estimate of CPU time reported in the console window (before the issuance of CTRL-C) is usually inaccurate
(see estimated total cpu time in the output example shown above). You can use CTRL-C to interrupt the execution
of the LS-DYNA solver, then type the sense switch sw2 to get a better estimate of the execution time and cycle
numbers. The ANSYS LS-DYNA program writes all important messages (errors, warnings, failed elements, contact
problems, etc.) to the ANSYS LS-DYNA console window and to the LS-DYNA ASCII output file d3hsp.
53

On a UNIX system, the progress of the ANSYS LS-DYNA solution can also be checked by reading the last lines of
certain ASCII output files. To do so, use another window to go to the directory where the ANSYS LS-DYNA program
was started. Type the following command at the system prompt:
tail -m filename
where filename = d3hsp or glstat or matsum. For m (the number of lines to be viewed), a value of 30 should be
sufficient.
The d3hsp file will always be created by ANSYS LS-DYNA. The files glstat and matsum will be created only if the
following commands were issued in the PREP7 or SOLUTION processors of ANSYS LS-DYNA:
EDOUT,GLSTAT
EDOUT,MATSUM
(See Chapter 12, Postprocessing of this manual for more information on how output control commands are
used.)
The files d3hsp, glstat and matsum are ASCII files. As the solution progresses, these files are written at requested
time points. By looking at the current "time" value of the solution from the last 30 lines of any one of these files
and comparing this "time" value against the specified "end time" value on the TIME command, you can determine
the progress of the solution.
5.6. Plotting Small Elements

As outlined in Comparison of Implicit and Explicit Methods, ANSYS LS-DYNA automatically calculates the critical
time step size of each element in a model based on its material properties and size. The overall time step for an
entire model is then based on the smallest critical time step value of all the elements within the model.
You can plot explicit dynamic elements based on their time step sizes using the EDTP command. This allows
you to monitor areas of the model that have small elements. There are three options available (OPTION field of
EDTP) for plotting elements based on time step size.
Option 1 - Plot the elements with the smallest time step sizes. Each element is shaded red or yellow based
on its time step size. The number of elements plotted is based on the VALUE1 field (default = 100) of the
EDTP command.
Option 2 - Plot and list the elements with the smallest time step sizes. This option produces the same plot
as Option 1, and also provides a listing of the smallest elements.
Option 3 - This option produces a plot similar to Option 1, except that all selected elements are plotted.
Elements beyond the number specified by VALUE1 are blue and translucent, with the degree of translucency
based on the VALUE2 field of the EDTP command.
In the plot produced by EDTP the elements are shaded red (smallest), yellow (intermediate), or blue (largest)
based on their time step sizes.
Note Care should be taken when using the EDTP command for large models because the time step
size calculations may take a significant amount of CPU time.
5.7. Editing the LS-DYNA Input File

Most of the major LS-DYNA capabilities are supported in the ANSYS LS-DYNA program and can be conveniently
accessed through the graphical user interface of this program. It should be noted, however, that there are a
54
Section 5.7: Editing the LS-DYNA Input File

number of other capabilities in the LS-DYNA program that cannot be directly accessed through the GUI of the
ANSYS LS-DYNA program. Some examples are:
several material properties, such as fabric, soil, etc.
certain element types, such as seat belt
application of constraints on rigid bodies in the local coordinate system
Although these capabilities cannot be directly accessed, users who are familiar with the LS-DYNA program can
still use these features indirectly in the ANSYS LS-DYNA program. The procedure for including additional features
is outlined below.
Note Only the LS-DYNA features that are described in the ANSYS LS-DYNA documentation and available
through the ANSYS LS-DYNA interface are supported features. Although you can access other features
as described below, features that you implement by editing the LS-DYNA input file are not supported.
After modeling has been completed, you can issue the EDWRITE command instead of issuing the SOLVE command. (In the GUI, pick Main Menu> Solution> Write Jobname.k.) EDWRITE will create the LS-DYNA input file
Jobname.K. (Note that if the analysis is a small restart, the input file will be named Jobname.R, and if it is a full
restart, the input file will be named Jobname_nn.K.)
The difference between the EDWRITE command and the SOLVE command is that step number 3 discussed
under Solution Process at the beginning of this chapter will not be executed. That is, Jobname.K will be written
(along with headers to the ANSYS results files), but the solution process of the LS-DYNA program will not be
started. After issuing EDWRITE, use one of the following two methods to include additional functionality in the
explicit dynamic analysis.
5.7.1. Method A
1.
Exit the ANSYS LS-DYNA program.
2.
Edit the LS-DYNA input file Jobname.K to incorporate the items that are not supported directly through
the ANSYS LS-DYNA program.
3.
Execute the LS-DYNA executable separately in the same directory where the two results files, Jobname.RST and Jobname.HIS, reside. The solution produced by the LS-DYNA executable will be appended
to these results files.
4.
After completion of the LS-DYNA execution, enter the ANSYS LS-DYNA program and view the results
using the postprocessors (POST1 and POST26).
5.7.2. Method B
1.
Using another window, go to the directory where the ANSYS LS-DYNA program was started. Do not exit
the ANSYS LS-DYNA program in the original window.
2.
Edit the LS-DYNA input file Jobname.K to incorporate the items that are not supported directly through
the ANSYS LS-DYNA program.
3.
Execute the LS-DYNA executable separately in the same directory (where the two results files, Jobname.RST and Jobname.HIS, reside). The solution produced by the LS-DYNA executable will be appended
to these results files. (Remember, you have not exited the ANSYS LS-DYNA program in the original window.)
4.
After completion of the LS-DYNA execution, go back to the original window and view the results using
the ANSYS LS-DYNA postprocessors.
55

For both Method A and Method B, you execute the LS-DYNA program by running the lsdynaxx script (where xx
is the ANSYS release number) with the correct arguments:
lsdynaxx i = jobname.k pr = product name
For example, if you have ANSYS Mechanical with LS-DYNA at ANSYS Release 10.0 and your input file is crashtest.k,
you would type:
lsdyna100 i=crashtest.k pr=ansysds
Note The lsdyna100 script is only available on UNIX platforms. On the PC, use the ls970.exe (or
ls970_DP.exe) executable instead to run LS-DYNA.
You may need additional command line arguments, depending on what type of analysis you are running. For
an implicit-to-explicit sequential solution (REXPORT and EDDRELAX commands; see Chapter 15, Implicit-toExplicit Sequential Solution), the argument m=drelax is also required. For a restart analysis, (EDSTART command;
see Chapter 13, Restarting), the argument r=d3dumpnn is also required, where nn is the number of the dump
file (01, 02, ... 99) to be used as a starting point. For any type of analysis, you can specify the memory to be used
by including the command line argument memory=n, where n is the desired amount of memory in words.
When using these methods to modify the file Jobname.K, nodes and elements must not be changed. Also, note
that the ANSYS LS-DYNA database is not updated with the changes you make in file Jobname.K. Thus, if an attempt
is made to view the modified items in the PREP7, POST1 or POST26 processors after the solution is complete,
the original version, and not the modified version, will be shown. (For these reasons, editing the LS-DYNA keyword
input file, Jobname.K, is NOT a supported feature of ANSYS LS-DYNA.) In all cases, the complete set of results
can be postprocessed in the LS-POST postprocessor (assuming you requested LS-DYNA results files via the EDWRITE command).
5.7.3. Using a Preexisting File.K

In a typical ANSYS LS-DYNA analysis, the results file headers are written when you issue the SOLVE command
(or when you issue EDWRITE). However, the LS-DYNA solver also has the ability to create the .RST and .HIS results
file headers when only a file.K input file exists. This allows you to take a preexisting file.K, solve it using LS-DYNA,
and then postprocess the results with the ANSYS LS-DYNA program.
To use this method, first make sure that the model defined in the .K file does not contain any features that ANSYS
LS-DYNA does not support. Then edit the .K file so that the *DATABASE_FORMAT command is set to either 1 for
ANSYS only results files, or 2 for ANSYS and LS-DYNA results files. Next, run the lsdynaxx script as described
above.
You can then postprocess the results using POST1 and POST26. If the corresponding ANSYS database does not
exist, element attributes (material properties, real constants, etc.) will not be available in the results files. However,
you should still be able to postprocess the results using the ANSYS postprocessors. If you requested LS-DYNA
results files, you will also be able to postprocess the results using LS-POST.
56
Chapter 6: Contact Surfaces

Contact surfaces in ANSYS LS-DYNA allow you to represent a wide range of types of interaction between components in a model. This chapter provides guidelines for defining physically realistic contact in an explicit dynamic
analysis.
It should be noted that contact is represented differently in an explicit dynamic analysis than it is in other types
of ANSYS analyses. In other analyses, the contact is represented by actual contact elements. For explicit dynamics,
there are no contact elements. You simply indicate the contact surfaces, the type of contact between them, and
other parameters related to the contact type.
The following explicit dynamics contact topics are available:
6.1. Contact Definitions
6.2. Contact Options
6.3. Contact Search Methods
6.4. Special Considerations for Shells
6.5. Controlling Contact Depth
6.6. Contact Stiffness
6.7. 2-D Contact Option
6.1. Contact Definitions

Due to complicated large deformation dynamics which typically occur during an explicit dynamic analysis, determining contact between components in a model can be extremely difficult. For this reason, special features
have been included in the ANSYS LS-DYNA program to make defining contact between surfaces as efficient as
possible. All contact surfaces are defined in ANSYS LS-DYNA through the use of the EDCGEN command.
Use the following steps when using the EDCGEN command:
STEP 1: Determine the type of contact surface which best defines your physical model.
STEP 2: Identify contact entities.
STEP 3: Specify friction coefficient parameters.
STEP 4: Specify any additional input which is required for a given contact type.
STEP 5: Specify birth and death times for the contact definition.
STEP 1: Determining Contact Type

In order to adequately describe interaction between complex geometries during large deformation contact and
dynamic impact, a large number of contact surface options have been incorporated into the ANSYS LS-DYNA
product. These contact types, which include node-to-surface, surface-to-surface, single surface, single edge,
eroding, tied, tiebreak, drawbead, and rigid contact options, are discussed in detail later in this chapter (see
Section 6.2: Contact Options). For most typical analyses, the following contact options are recommended.
Automatic Single Surface (ASSC): Contact is established when a surface of one body contacts itself or the
surface of another body. This type is easy to use because no contact or target surface definitions are required. It is efficient for self-contacting problems or large deformation problems where general areas of
contact are not known beforehand.
Automatic General (AG): This contact type is similar to Automatic Single Surface in that contact is established
when a surface of one body contacts itself or the surface of another body. Automatic General contact is
also easy to define since you do not need to specify contact and target surfaces. The main advantage of

AG contact is that it is based on newer contact algorithms; consequently, its performance may be better
than ASSC contact in some applications.
Node-to-Surface (NTS): Contact is established when a contacting node penetrates a target surface. This
type is commonly used for general contact between two surfaces and is most efficient when a smaller
surface comes into contact with a larger surface, such as a thin rod impacting a flat plate.
Surface-to-Surface: (STS) Contact is established when a surface of one body penetrates the surface of another body. This type is commonly used for arbitrary bodies that have large contact areas and is very efficient for bodies that experience large amounts of relative sliding with friction, such as a block sliding on
a plane.
STEP 2: Defining Contact Entities

With the exception of the single surface (ASSC, SS, and ESS), automatic general (AG), and single edge contact
(SE), all of the contact options require you to identify contact and target surfaces for which contact can occur.
You can do this by using nodal components, PART ID definitions, and/or part assembly ID definitions. When using
contact or target components, use select logic and the CM command to group the nodes (only nodal components
are valid). Then, as shown in the sample input listing below, use the EDCGEN command to specify contact
between the desired components such as the ball and bat surfaces of the example problem discussed in Chapter 4,
Loading:
NSEL,S,NODE,.... ! Select a set of nodes on ball surface
CM,BALLSURF,NODE ! Place selected nodes into component
! BALLSURF
NSEL,S,NODE,.... ! Select a set of nodes on bat surface
CM,BATSURF,NODE ! Place selected nodes into component
! BATSURF
EDCGEN,NTS,BALLSURF,BATSURF,.25,.23 ! Generate node-to-surface
! contact between components
! BALLSURF and BATSURF
Alternatively, contact surfaces can be specified using part numbers or part assembly numbers that are currently
defined within the finite element model. Part assembly numbers are defined with the EDASMP command.
The input line below demonstrates how the EDCGEN command is used to define contact between two different
parts / part assemblies in a model:
EDCGEN,STS,1,2,.25,.23 ! Generate surface-to-surface contact
! between PARTS 1 and 2
Additionally, contact can be defined between contact and target surfaces using a combination of PART / part
assembly and component definitions, as illustrated below:
EDCGEN,NTS, N1,2,.3,.28 ! Generate node-to-surface contact
! between component N1 and PART 2
EDCGEN,ESTS,1,N2,.15,.15 ! Generate eroding surface-to-surface
! contact between PART 1 and component N2
! between PART 1 and itself
You can also use the EDCGEN command for defining contact between part assemblies as shown in the following
sample input:
! between part assemblies 5 and 6
For the specialized case of single surface contact (ASSC, AG, ESS, and SS), no contact or target surface definitions
are required. As explained later in this section, single surface contact is the most general type of contact definition
as all external surfaces within a model can be in contact at any point during an analysis. The program will ignore
62
Section 6.1: Contact Definitions

any contact and target surfaces defined for single surface contact and will issue a warning message upon execution
of the EDCGEN command. A typical command for single surface contact is presented below:
EDCGEN,ASSC,,,.34,.34 ! Generate automatic single surface
! contact for the entire model
Note--When defining contact entities in an explicit analysis, no initial penetrations are allowed. Therefore, use
great care when defining contact components.
STEP 3: Specifying Friction Coefficient Parameters

The frictional coefficient used for contact is determined from the static friction coefficient (FS), the dynamic
friction coefficient (FD), and the exponential decay coefficient (DC). (FS, FD, and DC can be input on the EDCGEN
command.) The frictional coefficient is assumed to be dependent on the relative velocity, Vrel, of the surfaces in
contact:
c = FD + (FS FD) eDC i Vrel

The coefficient for viscous friction, VC (also input on EDCGEN), can be used to limit the friction force to a maximum.
A limiting force is computed:
Flim = VC i A cont
where Acont is the area of the segment contacted by the node in contact. The suggested value for VC is to use
the yield stress in shear:
VC =
o
3
where o is the yield stress of the contacted material.

In order to avoid undesirable oscillation in contact, e.g., for sheet forming simulation, a contact damping perpendicular to the contacting s
urfaces is applied. The contact damping coefficient is calculated as follows:
VDC
i crit
100
VDC is the viscous damping coefficient (input on EDCGEN command as percent of critical damping; e.g., VDC =
20 indicates 20%). crit is determined in the following fashion by ANSYS LS-DYNA:
crit = 2m
where
m = min(mslave , mmaster )
= ki
mslave + mmaster
; k is interface stiffness
mslave i mmaster
63
STEP 4: Specifying Additional Input

For eroding, rigid, tiebreak, and drawbead contact, additional information (fields V1-V4 on the EDCGEN command)
may also be required. This information varies with contact type, as outlined below.
Eroding surface contacts (ENTS, ESS, and ESTS) are used when surfaces of solid elements fail and contact needs
to be redefined with the remaining internal elements. For eroding contact, V1-V3 are defined as follows: the
boundary condition symmetry option key (V1) determines whether a symmetry condition will be retained along
a surface where elements fail; the interior erosion option key (V2) specifies whether erosion can subsequently
occur along internal surfaces when the exterior surface fails; the adjacent material key (V3) determines whether
solid element faces are included for erosion along free boundaries.
Rigid body contact (RNTR and ROTR) is normally used for multi-body dynamics. For rigid body contact, a user
defined force deflection curve is used to resist penetration rather than a linear stiffness. Therefore, for rigid body
contact, you must specify a data curve ID (V1), a force calculation method type (V2) applicable to the specified
rigid contact option, and an unloading stiffness value (V3).
Tiebreak contact (TSTS and TNTS) is used to define contact failure when surfaces are glued together. For both
types of tiebreak contact, tensile (V1) and shear (V2) failure stresses are used to calculate the failure criterion and
are required for input. For tiebreak nodes-to-surface contact, exponents for the normal force (V3) and the shear
force (V4) can also be specified to define the failure criterion.
Drawbead contact (DRAWBEAD) is typically used in specialized cases to represent drawbeads, which help restrain
a blank during drawing operations. In drawbead contact, you must input a load curve ID (V1), which gives the
bending component of the restraining force as a function of drawbead displacement, and a drawbead depth
(V3). You may optionally include a curve ID (V2), which gives the normal restraining force as a function of drawbead
displacement and the number of equally spaced integration points (V4) along the drawbead.
STEP 5: Specifying Birth and Death Times

For each contact definition, you can specify a birth time and a death time by using fields BTIME and DTIME on
the EDCGEN command. This allows you to activate contact at any time during the transient analysis, then deactivate it at a later time.
6.1.1. Listing, Plotting and Deleting Contact Entities

Once you define contact using the EDCGEN command, you can list, plot, or delete contact definitions. Use the
EDCLIST command to list all currently defined contact entities. A sample listing is shown below. Each contact
definition is given a contact entity reference number that can be used when plotting contact definitions.
CURRENT EXPLICIT DYNAMIC CONTACT DEFINITIONS
1 General Surface to Surface Contact Defined Between:
Nodal contact Entity N1
And Nodal Target Entity N2
FS = 0.10000 FD = 0.08000 DC = 0.00000 VC = 0.00000 VDC = 0.0000
2 Automatic Single Surface Contact Defined For:
All External Surfaces of Model
FS = 0.20000 FD = 0.15000 DC = 0.00000 VC = 0.00000 VDC = 0.0000
You can use the EDPC command to select and plot contact entities. The plot will consist of nodes or elements,
depending on the method that was used to define the contact surfaces (that is, components or parts). You can
specify a minimum contact entity number, maximum contact entity number, and contact entity number increment
by using the MIN, MAX, and INC fields of the EDPC command. Hence, by issuing the command EDPC,1,2,1 for
the above listing, the defined entities for the STS and ASSC contact would be selected and plotted. It is important
to note that for single surface contact definitions, all external surfaces in the model will be selected and plotted.
64
Section 6.2: Contact Options

Note EDPC selects the nodes and/or elements of the specified contact entities. Therefore, after plotting
the contact entities, you must reselect all nodes and elements required for subsequent operations (such
as SOLVE). Use the NSEL,ALL and ESEL,ALL commands (or other appropriate forms of these commands).
If contact has been incorrectly specified, you can delete it using the EDDC command. To delete a specific contact
definition, issue the command EDDC,DELE,Ctype,Cont,Targ, where Ctype is the contact type and Cont and
Targ are the contact and target parts or components for which contact has been defined. To delete all current
contact definitions, issue the command EDDC,DELE,ALL.
You cannot delete contact specifications in a small restart analysis (EDSTART,2). However, you can use the EDDC
command to deactivate (EDDC,DACT,Ctype,Cont,Targ) or reactivate (EDDC,RACT,Ctype,Cont,Targ) contact
in a small restart. This capability is useful when you know at what stages in an analysis certain types of contact
will occur. To use this feature you must define all contact specifications (EDCGEN command) in the new analysis,
and you must perform at least one small restart. Once a contact specification is defined, you can deactivate it in
a restart when it is not needed, and reactivate it in a later restart when it is needed. By considering contact only
when it is needed, you can achieve significant CPU time savings. (Remember that if you delete contact in a new
analysis using EDDC,DELE, its definition is removed from the database and you cannot reactivate it in a later restart.)
The EDDC command is not supported in an explicit dynamic full restart analysis (EDSTART,3). Thus, you cannot
delete, deactivate, or reactivate contact specifications in a full restart that were defined in a previous analysis.
6.2. Contact Options

In order to adequately characterize the complex interaction between surfaces in an explicit dynamic analysis,
twenty-four different contact types have been incorporated in the ANSYS LS-DYNA program (see the table below).
With such a large number of contact options, it is extremely important for a user to have an understanding of
each contact type so that the proper contact option can be used to realistically model a physical phenomenon.
Therefore, in this section, each of the contact options available in the ANSYS LS-DYNA program will be discussed.
Table 6-1 Contact Types
Single surface
Nodes to surface
Surface to surface
General (Normal)
SS
NTS
STS, OSTS
Automatic
ASSC, AG, ASS2D
ANTS
ASTS
Rigid
RNTR
ROTR
Tied
TDNS
TDSS, TSES
Tied with failure
TNTS
TSTS
ENTS
ESTS
Eroding
ESS
Edge
SE
Drawbead
DRAWBEAD
Forming
FNTS
FSTS, FOSS
6.2.1. Definition of Contact Types

As shown in the columns of Table 6-1, you can choose from three basic contact types in the ANSYS LS-DYNA
program: single surface contact, nodes-to-surface contact, and surface-to-surface contact.
1.
Single Surface Contact (SS, ASSC, AG, ASS2D, ESS, SE)
65

Single surface contact is established when a surface of one body contacts itself or the surface of another
body. In single surface contact, the ANSYS LS-DYNA program automatically determines which surfaces
within a model may come into contact. Therefore, single surface contact is the simplest type to define
because no contact or target surface definitions are required. When it is defined, single surface contact
allows all external surfaces within a model to come into contact. This option can be very powerful for
self-contact or large deformation problems when general areas of contact are not known beforehand.
Unlike implicit modeling, where over-defining contact will significantly increase computation time, using
single surface contact in an explicit analysis will cause only minor increases in CPU time. Most impact
and crash-dynamic applications will require single surface contact to be defined. Since automatic general
(AG) contact is very robust and includes shell edge (SE) contact as well as improved beam contact, it is
recommended as the first choice for self contact and large deformation problems when the contact
conditions are not easy to predict.
2.
Node-to-Surface Contact (NTS, ANTS, RNTR, TDNS, TNTS, ENTS, DRAWBEAD, FNTS)
Node-to-surface contact is a contact type which is established when a contacting node penetrates a
target surface. This type of contact is commonly used for general contact between two surfaces. Use the
same rules as in ANSYS implicit to determine which surfaces are target or contact:
The flat or concave surface is the target. The convex surface is the contact surface.
The coarser mesh is the target surface. The finer mesh is the contact surface.
In the case of Drawbead contact, the bead is always the nodal contact surface and the blank is always
the target surface.
3.
Surface-to-Surface Contact (STS, OSTS, ASTS, ROTR, TDSS, TSTS, ESTS, FSTS, FOSS, TSES)
Surface-to-surface contact is established when a surface of one body penetrates the surface of another
body. Surface-to-surface contact is the most general type of contact as it is commonly used for bodies
that have arbitrary shapes with relatively large contact areas. This type of contact is most efficient for
bodies that experience large amounts of relative sliding, such as a block sliding on a plane or a sphere
sliding within a groove.
6.2.2. Definition of Contact Options

For each of the three contact types, there are often several contact options available. In the ANSYS LS-DYNA
program, the following options may be used:
1.
General (Normal) Contact (SS, NTS, STS, OSTS)

Although the general contact options use the simplest contact algorithms, they still are used for a wide
range of applications. In fact, the NTS and STS options are two of the three recommended options for
ANSYS LS-DYNA. The primary advantage of using the general contact algorithms is that they are extremely
fast and robust. The only concern when working with the general contact options deals with contact
surface orientation. The contact surface orientation defines which side of a surface is solid and which
side is air'. When using solid elements, ANSYS LS-DYNA automatically sets the contact orientation correctly
for the general contact options. For shell elements, however, you must set the contact surface orientation
when using general contact. To do this, set the ORIE field of the EDCONTACT command to 2 to activate
automatic reorientation of the contact surfaces. It is important to note, the reorientation will occur only
if there is no initial penetration of the shell surface.
2.
Automatic Contact (ASSC, AG, ASS2D, ANTS, ASTS)

Along with the general contact family, the automatic contact options are the most commonly used
contact algorithms. The main difference between the automatic and general options is that the contact
66
Section 6.2: Contact Options

surface orientation for shell elements is automatically determined by the automatic contact algorithms.
In automatic contact, checks are made for contact on both sides of shell elements. Therefore, the contact
search depth is always limited. If this is considered to be the reason for breaking through a contact surface,
try using general contact with an infinite or a large search depth. See Section 6.5: Controlling Contact
Depth later in this chapter.
3.
Eroding Contact (ESS, ENTS, ESTS)

The eroding contact options are needed when the elements forming one or both exterior surfaces experience material failure during contact. Contact is allowed to continue with the remaining interior elements. Eroding contact should be used with solid elements in penetration problems and other applications
which experience surface failure. When using eroding contact, a symmetry plane option (V1), an interior/exterior node erosion option (V2), and an adjacent material treatment option (V3) must be specified
on the EDCGEN command.
4.
Rigid Contact (RNTR, ROTR)

The rigid body contact options are similar to the general contact options NTS and OSTS except that,
rather than using a linear stiffness to resist penetration, a user-defined force-deflection curve is used.
These contacts are typically used for multi-body dynamics where all of the bodies are rigid. Rigid body
contacts are beneficial because they can allow inclusion of energy absorption without the need for
modeling deformable elements. However, rigid body contact (RNTR, ROTR) cannot be used with deformable
bodies. The contact of a rigid body to a deformable body must be defined with general, automatic, or
eroding options. For the rigid options, the data curve ID (V1), force calculation method (V2), and unloading
option (V3) must be specified on the EDCGEN command.
5.
Tied Contact (TDNS, TDSS, TSES)

The tied contact options actually 'glue' the contact nodes (surfaces) to the target surfaces. The contact
and target surfaces must be initially coplanar so that during initialization, an isoparametric position of
the contact node (surface) within the target segment is calculated. Thereafter, upon application of loads
or initial velocities, the contact nodes (surfaces) are forced to maintain their isoparametric position
within the target surface. The effect of tied contact is that the target surfaces can deform and the slave
nodes are forced to follow that deformation. When defining tied contact, the body with the coarser mesh
should always be defined as the target surface. Only translational degrees of freedom (UX, UY, and UZ)
are affected by tied contact.
6.
Tiebreak (Tied with Failure) Contact (TNTS, TSTS)

Tiebreak contact is identical to tied contact except that the contact nodes (surfaces) are tied to the target
surfaces only until a failure criterion is reached. This is done by 'pinning' the contact nodes (surfaces) to
the target using a penalty stiffness; after the failure criterion is exceeded, the contact nodes (surfaces)
are allowed to slide relative to, or separate from, the target surface. The tiebreak contact options are
typically used to represent spot-welded or bolted connections. The main difference between TNTS and
TSTS is that the TSTS failure is based on a failure stress and the TNTS is based on a failure force. When
using TSTS, the normal (V1) and shear (V2) failure stresses must be specified using the EDCGEN command.
For the TNTS option, the normal (V1) and shear (V2) failure forces and normal (V3) and shear (V4) force
exponents must be supplied using the EDCGEN command. Failure of the connection will occur when:
m1
fn
fn,fail
m2
f
+ s
fs,fail
67
7.
Edge Contact (SE)

The single edge contact option should be used when contact occurs orthogonal to the shell surface
normal direction. This contact option does not require contact or target surface definitions. Single edge
contact is often used in sheet metal applications which have their surface normals orthogonal to the
impact direction.
8.
Drawbead Contact (DRAWBEAD)

Drawbead Contact is typically used in metal forming operations in which special care must be taken to
restrain the blank. During drawing and stamping simulations, it is common for the blank to lose contact
with the forming surfaces. Drawbead contact, which allows implementation of bending and frictional
restraining forces, helps to ensure that the blank will remain in contact the entire length of the drawbead
depth.
68
Section 6.3: Contact Search Methods
Figure 6.1 LS-DYNA Drawbead Representation
9.
Forming Contact (FNTS, FSTS, FOSS)

The forming contact options are primarily used in metal forming applications. For these contact types,
the tools and dies are typically defined as the target (master), while the workpiece is defined as the
contact (slave) surface. Mesh connectivity is not required for the forming contact option, but the orientation of the tooling meshes must be in the same direction. The forming contact options are based on
the Automatic contact types and are therefore very robust in metal forming applications.
6.3. Contact Search Methods

There are two different contact search algorithms used by ANSYS LS-DYNA for determining which target surface
is being contacted by which contact surface. These methods are outlined below:
6.3.1. Mesh Connectivity Tracking

In mesh connectivity tracking, the contact search algorithm uses shared nodes of neighboring element segments
to identify possible sources of contact. Therefore, when a target segment is no longer in contact with a contact
surface node, the neighboring segments are checked. The mesh connectivity method is beneficial because it is
very fast, but has the disadvantage of requiring that the mesh must be continuous for the contact algorithm to
work correctly. Therefore, you should specify different contact sets for distinct regions. The mesh connectivity
method is used by the NTS, OSTS, TSTS, TNTS, and TDNS contact options. However, by setting the SHTK field of
the EDCONTACT command to be positive, contact options NTS, OSTS, and TDNS will use the bucket sort method.
69
6.3.2. Bucket Sort Method

The bucket sort algorithm, which is used by all the contact options other than those specified above, divides the
target surface region into cubes (buckets). Contacting nodes can contact any segment of the target surface in
the same bucket or adjacent buckets. The bucket sort method is extremely robust, but can be somewhat slower
than the mesh connectivity tracking method if the target surface contains a large number of elements.
6.3.3. Limiting the Contact Search Domain

You can limit the overall region for which a contact search will be executed by defining a contact box domain.
When you specify a contact box, a contact search will only be performed within the volume specified by the
coordinates of the box. An advantage of this feature is that it minimizes CPU time when you know the potential
contact area of two bodies beforehand. The contact box option is only valid when defining contact with parts
or part assemblies. The contact box is defined using the EDBX command.
EDBX,Option,BOXID,XMIN,XMAX,YMIN,YMAX,ZMIN,ZMAX
Option
ADD, DELETE, LIST
BOXID
User defined List ID number
XMIN
Minimum x coordinate
XMAX
Maximum x coordinate
YMIN
Minimum y coordinate
YMAX
Maximum y coordinate
ZMIN
Minimum z coordinate
ZMAX
Maximum z coordinate
Once defined, a BOXID can be used in the BOXID1 and BOXID2 fields of the EDCGEN command. BOXID1 corresponds to the contact box and BOXID2 corresponds to the target contact box.
EDCGEN,NTS,Cont,Targ, ... DTIME,BOXID1,BOXID2
6.4. Special Considerations for Shells

Modeling rigid bodies with shell elements requires special consideration. For the automatic nodes-to-surface
(ANTS), automatic single surface (ASSC), automatic surface-to-surface (ASTS), and single surface (SS) contact
definitions, shell thickness is taken into account for the contact formulation for both the determination of the
contact surface as well as the search depth. Therefore, it is necessary to ensure that realistic thicknesses are
specified for the rigid body shells. A thickness that is too small may result in the loss of contact, and a value that
is too large may result in degradation of speed in the contact sorting algorithm. The shell thickness contact option
field SHTK of the EDCONTACT command is ignored for the above contact types.
Setting the SHTK field of the EDCONTACT command to 1 or 2 has several affects on the NTS, STS, and OSTS
contact options. The first is that the contact depth is directly computed (see the equations in Section 6.5: Controlling Contact Depth) based on element thickness for shells and edge length for solids and cannot be controlled
by the user. Second, as explained above, the contact search algorithm becomes Bucket Sort, so mesh connectivity
is not necessary for the contact to work correctly.
6.5. Controlling Contact Depth

For the contact options STS, NTS, OSTS, TNTS, and TSTS, you must be sure that spurious contact is not defined
between components in a model. For these contact types, a contact depth of 1 x 1010 (nearly infinite) is assumed
610
Section 6.6: Contact Stiffness

by the ANSYS LS-DYNA program. Therefore, any time a contact node passes behind a target surface (or viceversa), contact is defined and a force proportional to the contact depth is generated. In an explicit dynamic
analysis, it is common for unwanted spurious contact to be defined because of the geometry of parts which may
be in relative motion. In cases where the contact is not genuine, and the contact depth is relatively large, the
contact forces can become large and cause the model to become unstable. For this reason, the ANSYS LS-DYNA
program allows you to specify a maximum contact depth, beyond which the contact penetrations are considered
spurious and will be ignored. To control the contact penetration distance, set the PENCHK field of the EDCONTACT
command to a value of 1 or 2.
The PENCHK field discussed above controls penetration checking for all contact definitions in the model of types
STS, NTS, OSTS, TNTS, and TSTS. You can change the penetration checking of each individual contact definition
of these same types by using the EDSP command.
For all of the contact options other than STS, NTS, OSTS, TNTS, and TSTS, the contact penetration depth is automatically limited by the element thickness and cannot be adjusted by the user. The expressions for contact depth
in shell and solid elements are listed below:
shell elements:
contact depth = min[shell thickness, 0.4 min side length, 0.5
area ]
solid elements:
volume
contact depth = min
, 0 .5
area
area
6.6. Contact Stiffness

6.6.1. Choice of Penalty Factor
A stiffness relationship between two bodies must be established for contact to occur. Without a contact stiffness,
bodies will pass through one another. The relationship is generated through an 'elastic spring' that is put between
the two bodies, where the contact force is equal to the product of the contact stiffness (k) and the penetration
(). The amount of penetration (), or incompatibility, between the two bodies is therefore dependent on the
stiffness k. Ideally, there should be no penetration, but this implies that k = , which will lead to numerical instabilities. The value of k that is used depends on the relative stiffness of the bodies in contact. In the ANSYS LSDYNA program, the contact stiffness is determined by the following relationships:
k=
k=
fs Area2 K
Volume
fs Area K
Minimum Diagonal
for segments on solid elements
for segments on shell elements
Area = area of contact segment

K = bulk modulus of contacted element
fs = penalty factor (0.1 by default)
In almost all cases, the contact stiffness parameter automatically determined by the ANSYS LS-DYNA program
will provide excellent results. The contact stiffness, however, can be changed for all contact surfaces by entering
611

a new value for the penalty factor, fs, with the SFSI field of the EDCONTACT command. In practice, increasing
the value of SFSI above 0.1 will typically cause instabilities to occur.
6.6.2. Symmetry Stiffness

If the contact stresses of contact and target surfaces are badly mismatched due to large differences in material
properties or element size, then instabilities or unrealistic behavior may arise. This can be avoided by adjusting
the penalty stiffness option, PENO of the EDCONTACT command. For example, by setting PENO to 1, both the
contact and target stiffnesses are taken into account and the least stresses of the surfaces are used.
If there are very large differences in the contact stiffnesses, the program scales them to closer values, overwriting
user input for the scale factor.
Together with the related mass, a closed contact element is a spring-mass system for which the current stability
criterion yields a critical time step in the explicit analysis. The program reports the least one of the time steps
with a message. If the actual time step size used is larger than the one listed in this message, then you should
scale the penalty value of the offending surface with the EDCONTACT command or reduce the actual time step
size with the EDCTS command. If the difference between the two time steps is small, this approach may not be
necessary.
In addition to using the SFI and PENO options of the EDCONTACT command to control penalty stiffness between
contact surfaces, individual contact (slave) and target (master) stiffness values can be adjusted using the EDCMORE
command. Unlike the EDCONTACT command, which is applied to all contact definitions within a model, the
EDCMORE command allows the specification of additional contact parameters for a single contact surface specification. The Val1 and Val2 fields of the EDCMORE command can be used to adjust the slave and master
penalty stiffnesses in cases where the default contact stiffness determined by the ANSYS LS-DYNA program is
found to be inadequate.
6.7. 2-D Contact Option

In order to define contact between PLANE162 elements, a specialized 2-D contact option exists in the ANSYS LSDYNA program. The option, ASS2D, is a single surface contact option (similar to ASSC). Similar to the contact
used for the 3-D elements, the ASS2D option is specified using the EDCGEN command:
EDCGEN,ASS2D,Cont,,FS,FD,DC, ... BTIME,DTIME
For 2-D contact, only part assemblies can be used to define the contact component. The only other fields used
with the ASS2D option are FS, FD, DC, BTIME, and DTIME. All of the other fields of the EDCGEN command will
be ignored for 2-D contact.
612
Chapter 7: Material Models

ANSYS LS-DYNA includes over 40 material models that can be used to represent a wide range of material behavior. The available materials are listed below. The models and the procedure for using them are outlined in detail
in the remainder of this chapter. For additional information on each material model, refer to Appendix B:, Material Model Examples or to Chapter 16 of the LS-DYNA Theoretical Manual. (The LS-DYNA material number corresponding to each material model is shown in parentheses below, where applicable.)
Linear Elastic Models
Isotropic (#1)
Orthotropic (#2)
Anisotropic (#2)
Elastic Fluid (# 1)
Nonlinear Elastic Models
Blatz-Ko Rubber (#7)
Mooney-Rivlin Rubber (#27)
Viscoelastic (#6)
Nonlinear Inelastic Models
Bilinear Isotropic (#3)
Temperature Dependent Bilinear Isotropic (#4)
Transversely Anisotropic Elastic Plastic (#37)
Transversely Anisotropic FLD (# 39)
Bilinear Kinematic (#3)
Plastic Kinematic (#3)
3-Parameter Barlat (#36)
Barlat Anisotropic Plasticity (#33)
Rate Sensitive Power Law Plasticity (#64)
Strain Rate Dependent Plasticity (#19)
Piecewise Linear Plasticity (rate dependent) (#24)
Modified Piecewise Linear Plasticity (rate dependent) (#123)
Composite Damage (#22)
Concrete Damage (#72)
Power Law Plasticity (#18)
Elastic Viscoplastic Thermal Model (#106)
Pressure Dependent Plasticity Models
Elastic-Plastic Hydrodynamic (# 10)

Geological Cap (# 25)
Foam Models
Closed Cell Foam (# 53)
Viscous Foam (# 62)
Low Density Foam (# 57)
Crushable Foam (# 63)
Honeycomb (# 26)
Models Requiring Equations of State
Bamman Plasticity (# 51)
Johnson-Cook Plasticity (# 15)
Null (# 9)
Zerilli-Armstrong (# 65)
Steinberg (# 11)
Discrete Element Models
Linear Elastic Spring
General Nonlinear Spring
Nonlinear Elastic Spring
Elastoplastic Spring
Inelastic Tension or Compression-Only Spring
Maxwell Viscosity Spring
Linear Viscosity Damper
Nonlinear Viscosity Damper
Cable (# 71)
Rigid Model
Rigid (# 20)
7.1. Defining Explicit Dynamics Material Models

You define material models using the ANSYS commands MP, MPTEMP, MPDATA, TB, TBTEMP, and TBDATA,
and the ANSYS LS-DYNA command EDMP. The next section, Explicit Dynamics Material Model Descriptions,
describes how you define the individual material properties of each material model via commands.
Defining material models via the GUI is much more direct than doing so via commands:
1.
72
Choose menu path Main Menu> Preprocessor> Material Props> Material Models. The Define Material Model Behavior dialog box appears.
Section 7.2: Explicit Dynamics Material Model Descriptions

Note If you do not first define an ANSYS LS-DYNA element type, you will not have access to
the ANSYS LS-DYNA material models.
2.
In the Material Models Available window on the right, double-click on LS-DYNA, then on one of the
material model categories: Linear, Nonlinear, Equation of State, Discrete Element Properties, or Rigid
Material.
3.
Double-click on a material subcategory. For example, under the Nonlinear category, the subcategories
are: Elastic, Inelastic, and Foam Material Models.
4.
Continue to double-click on any further material subcategories until a data input dialog box appears.
The options that appear in this dialog box will be all of the individual material property options that are
valid for the material model you have selected.
5.
Enter the desired values, and click on OK. The model that you defined is listed by model type and number
in the Material Models Defined window on the left.
You can later edit the values by double-clicking on the material model listed on the left side in the Material
Models Defined window until the associated data input dialog box appears. There, you can edit the data and
click on OK again.
You can copy the contents of an existing material model by choosing Edit> Copy and specifying the new model
number. The copied material model is listed on the left side in the Material Models Defined window with the
new number. Its contents are identical to the model whose contents you copied.
You can delete a material model by clicking on the model number to highlight it, then choosing Edit> Delete.
For further information on defining materials using the GUI, see Section 1.1.4.4: Material Model Interface in the
ANSYS Basic Analysis Guide. Also, see Section 4.1.1.10: Using Tree Structures in the ANSYS Operations Guide for
information specifically on the tree structure layout of the material model interface.
If you define, edit, copy, or delete material models via the GUI, ANSYS will automatically issue the correct commands and write them to the log file.
7.2. Explicit Dynamics Material Model Descriptions

This section describes each material model in detail. Wherever load curve ID is mentioned as required input,
you actually input a material data curve ID. Material data curves are defined using the EDCURVE command as
outlined in Chapter 4, Loading. All applicable properties for a material model appear in that models dialog
box when you are working interactively. The corresponding commands are provided here for when you are
working in batch or command mode. Be sure to define all properties that are listed for the material model; do
not define extraneous properties.
7.2.1. Linear Elastic Models

7.2.1.1. Isotropic Elastic Model
Isotropic material model. Use the MP command to input the required values:
MP,DENS - density
MP,EX - elastic modulus
MP,NUXY - Poissons ratio
For an example input listing, see Section B.2.1: Isotropic Elastic Example: High Carbon Steel.
73
7.2.1.2. Orthotropic Elastic Model

Orthotropic material model. Use the MP command to input the required values.
MP,DENS - density
MP,EX - elastic moduli (also EY, EZ); one value required
MP,NUXY - Minor Poissons ratio (also NUYZ, NUXZ); one value required or MP,PRXY - Major Poissons ratio
(also PRYZ, PRXZ); one value required
MP,GXY - shear moduli (also GYZ and GXZ); one value required
If only one value is specified for a given material property family (i.e., EX), the other values will be automatically
defined (e.g., EY = EZ = EX). To specify a material coordinate system, use the EDLCS and EDMP,ORTHO commands.
If no material coordinate system is specified, material properties are locally orthotropic with material axes defined
by nodes I, J, and L of the element (see the figure below). For a multi-layer composite laminate, use the TB,COMP
command instead and specify the lamina (ply) properties as SHELL163 real constants. See Section 7.2.3.13:
Composite Damage Model for more information.
For an example input listing, see Section B.2.2: Orthotropic Elastic Example: Aluminum Oxide.
7.2.1.3. Anisotropic Elastic Model

This material description requires the full elasticity matrix. Because of symmetry, only 21 constants are required.
This material is only valid for SOLID164 and SOLID168 elements.

%#

$#

#
"

!

)10(&
)'
74

Input density with MP. The constants are input in upper triangular form by means of the TB,ANEL command. To
specify a material orientation axis, use the EDLCS and EDMP,ORTHO commands. If no material coordinate system
is specified, material properties are locally orthotropic with material axes defined by nodes I, J, and L of the element
(see figure of element coordinate system above).
MP,DENS - density
TB,ANEL
TBDATA, 1, C11, C12, C22, C13, C23, C33
TBDATA, 7, C14, C24, C34, C44, C15, C25
TBDATA, 13, C35, C45, C55, C16, C26, C36
TBDATA, 19, C46, C56, C66
When you list out the data table information [TBLIST] for this material type, the constants will be listed in lower
triangular form [D] instead of upper triangular form [C]. This discrepancy is not a computational error; the material data is correctly transferred to the LS-DYNA program.
For an example input listing, see Section B.2.3: Anisotropic Elastic Example: Cadmium.
7.2.1.4. Elastic Fluid Model

Use this option to model containers filled with fluid that undergo dynamic impact loading. You input density
using the MP command (DENS). Elastic fluid is then defined as a material type using the EDMP command:
MP,DENS
EDMP,FLUID,MAT,VAL1
The fluid model requires that the bulk modulus be specified. You can input the bulk modulus in the VAL1 field
of the above command. As an alternative to inputting the bulk modulus directly using EDMP, you can input the
elastic modulus (EX) and Poissons ratio (NUXY) using the MP command, and the program will calculate the bulk
modulus as shown below.
MP,EX
MP,NUXY
K=
E
3(1 2 )
If VAL1 (on EDMP), EX, and NUXY are all specified, VAL1 will be used as the bulk modulus.
7.2.2. Nonlinear Elastic Models

7.2.2.1. Blatz-Ko Rubber Elastic Model
The hyperelastic continuum rubber model defined by Blatz and Ko. This model uses the second Piola-Kirchhoff
stress:
1

1
Sij = G Cij V 1 2 ij
V
75

G is the shear modulus, V is the relative volume, is the Poissons ratio, Cij is the right Cauchy-Green strain tensor,
and ij is the Kronecker delta. Use the MP command to input density (DENS) and shear modulus (GXY).
For an example input listing, see Section B.2.4: Blatz-Ko Example: Rubber.
7.2.2.2. Mooney-Rivlin Rubber Elastic Model

Incompressible rubber material model; nearly identical to the 2-parameter existing ANSYS Mooney-Rivlin model.
The strain energy density function is defined in terms of input parameters C10, C01 and :
1
W = C10 (I1 3) + C01(I 2 3) + C 1 + D(I 3 1)2
I2
c
C
C = 10 + C01
2
C (5 2) + C01(11 5)
D = 10
2(1 2 )
I1, I2, and I3 are invariants of the right Cauchy-Green Tensor.
Input Poissons ratio () and density with the MP command. (For Poisson's ratio, a value greater than .49 is recommended; smaller values may not work.) Input the Mooney-Rivlin constants with the TB and TBDATA commands.
Data at only one temperature is permitted and must be specified in locations 1 and 2 for the data table:
TB,MOONEY,,,,0
TBDATA, 1, C10
TBDATA, 2, C01
As an alternative to inputting C10 and C01 directly, you can set these constants to zero and supply tabulated
uniaxial data via a load curve. The program will calculate the constants based on the experimental data input in
locations 3-6 of the TBDATA command. To use this input method, you must set the TBOPT field of the TB command
to 2:
TB,MOONEY,,,,2
TBDATA, 1, C10 (set to zero to use experimental data)
TBDATA, 2, C01 (set to zero to use experimental data)
TBDATA, 3, C3 (specimen gauge length L0)
TBDATA, 4, C4 (specimen gauge width)
TBDATA, 5, C5 (specimen thickness)
TBDATA, 6, C6 (load curve ID)
The load curve definition that provides the uniaxial data should give the change in gauge length L versus the
corresponding force. In compression, both the force and the change in gauge length must be specified as negative values. In tension, the force and change in gauge length should be input as positive values. The principal
stretch ratio in the uniaxial direction, 1, is then given by:
76
1 =
L o + L
Lo
where L0 is the initial length and L is the actual length.

Alternatively, you can input the stress versus strain curve by setting the gauge length, thickness, and width equal
to 1.0 and defining the engineering strain in place of the change in gauge length and the nominal (engineering)
stress in place of the force.
The least squares fit to the experimental data is performed during the initialization phase of the solution in ANSYS
LS-DYNA.
For an example input listing, see Section B.2.5: Mooney-Rivlin Example: Rubber.
7.2.2.3. Viscoelastic Model

Linear viscoelastic material model introduced by Herrmann and Peterson. The model assumes the deviatoric
behavior:
ij ( )
t
ij = 2o ( t )
d

where the shear relaxation modulus is given by:
( t ) = G + (G 0 - G ) e - t
In the model, elastic bulk behavior is assumed when calculating the incrementally integrated pressure from the
volume, V: p = K ln V. The parameters G , G0 , K (Bulk modulus) and are required to define the linear viscoelastic
material model. Input these values with TB,EVISC and locations 46, 47, 48, and 61 of the TBDATA command:
TB,EVISC
TBDATA, 46, G0
TBDATA, 47, G
TBDATA, 48, K
TBDATA, 61, 1/
Note For this material option, you must specify density (DENS) with the MP command.
For an example input listing, see Section B.2.6: Viscoelastic Example: Glass.
7.2.3. Nonlinear Inelastic Models

7.2.3.1. Bilinear Isotropic Model
Classical bilinear isotropic hardening model (strain rate independent) that uses two slopes (elastic and plastic)
to represent the stress-strain behavior of a material. Specify the stress-strain behavior at only one temperature.
(A temperature dependent model is also available; see Temperature Dependent Bilinear Isotropic Model.) Input
elastic modulus (EX), Poissons ratio (NUXY), and density (DENS) with the MP command. The program calculates
77

the bulk modulus (K) using the EX and NUXY values that you input. Input the yield strength and tangent slope
with TB,BISO and locations 1 and 2 of the TBDATA command:
TB,BISO
TBDATA, 1, Y (yield stress)
TBDATA, 2, Etan (tangent modulus)
For an example input listing, see Section B.2.7: Bilinear Isotropic Plasticity Example: Nickel Alloy.
7.2.3.2. Temperature Dependent Bilinear Isotropic Model

Classical bilinear isotropic hardening model that is strain rate independent and uses two slopes (elastic and
plastic) to represent the stress-strain behavior of a material. The stress-strain behavior can be specified at up to
six different temperatures. If the stress-strain behavior is specified at only one temperature, the Bilinear Isotropic
(which is both strain rate and temperature independent) material model is assumed. You can use this model to
represent a thermo-elastic material by inputting a large value for yield strength.
Input density (DENS) with the MP command. (The density is temperature independent.) Input elastic modulus
(EX), Poissons ratio (NUXY), and thermal expansion coefficient (ALPX) with the MPTEMP and MPDATA commands.
(These properties are temperature dependent.) Input the yield strength and tangent slope with the TB,BISO,,NTEMP
command, TBTEMP commands, and locations 1 and 2 of the TBDATA command. The yield strength and tangent
modulus must be defined relative to the same temperature data that are used on the MPTEMP command.
MP,DENS
MPTEMP,1,TEMP1,TEMP2, . . . ,TEMPntemp
MPDATA,EX,,1,EX1, EX2, . . . ,EXntemp
MPDATA,NUXY,,1,NUXY1,NUXY2, . . . ,NUXYntemp
MPDATA,ALPX,,1,ALPX1,ALPX2, . . . ,ALPXntemp
TB, BISO,,NTEMP (NTEMP can be 2 to 6)
TBTEMP,TEMP1 (first temperature point)
TBTEMP,TEMP2 (second temperature point)
. . . (continue the pattern NTEMP times)
TBTEMP,TEMPntemp (last temperature point)
Note For this material model, you must supply data over a temperature range wide enough to cover
the actual temperatures that will occur in the analysis. Otherwise, the analysis will terminate.
7.2.3.3. Transversely Anisotropic Hardening Model

Fully iterative anisotropic plasticity model available for shell and 2D elements only. In this model the yield
function given by Hill[3.] is reduced to the following for the case of plane stress:
2
F() = y = 11 + 2
22
78
2R
2R + 1 2
1122 + 2
12
R +1
R +1

P
where R is the anisotropic hardening parameter which is the ratio of the in-plane plastic strain rate, 22 , to the
P
out of plane plastic strain rate, 33 :

R=
P
22
P
33
The stress-strain behavior can be specified at only one temperature. Input the elastic modulus (EX), density
(DENS), and Poissons ratio (NUXY) with the MP command. Input the yield stress, tangent modulus, anisotropic
hardening parameter, and a load curve ID for effective yield stress versus effective plastic strain with TB,PLAW,,,,7
and locations 1 - 4 of the TBDATA command:
TB,PLAW,,,,7
TBDATA, 3, R (anisotropic hardening parameter)
TBDATA, 4, LCID (load curve ID for yield stress vs. plastic strain)
For an example input listing, see Section B.2.8: Transversely Anisotropic Elastic Plastic Example: 1010 Steel.
7.2.3.4. Transversely Anisotropic FLD Hardening Model

This material model is used for simulating the sheet metal forming of anisotropic materials. Only transversely
anisotropic materials can be considered. For this model, the dependence of the flow stress with the effective
plastic strain can be modeled using a defined load curve (using EDCURVE). In addition, you can define a forming
limit diagram (also using EDCURVE). This diagram will be used by the ANSYS LS-DYNA program to compute the
maximum strain ratio that the material can experience.
79
This plasticity model is only available for shell and 2-D elements. The model directly follows the plasticity theory
introduced in the Transversely Anisotropic Elastic Plastic model described earlier in this section. You can refer
to that model for the theoretical basis.
To utilize the Transversely Anisotropic FLD model, you must input the density (DENS), elastic modulus (EX), and
Poisson's ratio (NUXY) with the MP command. As shown below, the additional input parameters are specified
using TB,PLAW,,,,10 and locations 1-5 of the TBDATA command.
TB,PLAW,,,,10
TBDATA,1,y (yield stress)
TBDATA,2,Etan (tangent modulus)
TBDATA,3,R (anisotropic hardening parameter)
TBDATA,4,LCID1 (load curve for defining effective stress vs. plastic strain)
TBDATA,5,LCID2 (load curve for defining FLD)
For an example input listing, see Section B.2.9: Transversely Anisotropic FLD Example: Stainless Steel.
7.2.3.5. Bilinear Kinematic Model

(strain rate independent)
Classical bilinear kinematic hardening model that uses two slopes (elastic and plastic) to represent the stressstrain behavior of a material. The stress-strain behavior can be specified at only one temperature. Input elastic
modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the yield strength and
tangent slope with TB,BKIN and locations 1 and 2 of the TBDATA command:
710

TB,BKIN
For an example input listing, see Section B.2.10: Bilinear Kinematic Plasticity Example: Titanium Alloy.
7.2.3.6. Plastic Kinematic Model

Isotropic, kinematic, or a combination of isotropic and kinematic hardening models with strain rate dependency
and failure. Isotropic and kinematic contributions may be varied by adjusting the hardening parameter between
0 (kinematic hardening only) and 1 (isotropic hardening only). Strain rate is accounted for using the CowperSymonds model which scales the yield stress by the strain rate dependent factor as shown below:
1
1 + P ( + E eff )
Y =
p p
C 0

where 0 is the initial yield stress, is the strain rate, C and P are the Cowper-Symonds strain rate parameters,
peff is the effective plastic strain, and Ep is the plastic hardening modulus which is given by
Ep =
Etan E
E Etan
The stress-strain behavior can only be specified at one temperature. Input the elastic modulus (EX), density
(DENS), and Poissons ratio (NUXY) with the MP command. Input the yield stress, tangent slope, hardening
parameter, strain rate parameters C and P, and the failure strain with TB,PLAW,,,,1 and locations 1 - 6 of the TBDATA
command:
TB,PLAW,,,,1
TBDATA, 3, (hardening parameter)
TBDATA, 4, C (strain rate parameter)
TBDATA, 5, P (strain rate parameter)
TBDATA, 6, f (failure strain)
For an example input listing, see Section B.2.11: Plastic Kinematic Example: 1018 Steel.
7.2.3.7. 3-Parameter Barlat Model

Anisotropic plasticity model developed by Barlat and Lian[1.] used for modeling aluminum sheets under plane
stress conditions. Both exponential and linear hardening rules are available. The anisotropic yield criterion for
plane stress is defined as:
2( Y )m = a K1 + K 2 m+ a K1 K 2 m + c 2K 2 m
where Y is the yield stress, a and c are anisotropic material constants, m is Barlat exponent, and K1 and K2 are
defined by:
711
K1 =
xx + h yy
2
2
xx h yy
2 2
K2 =
+ p xy
2
where h and p are additional anisotropic material constants. For the exponential hardening option, the material
yield strength is given by:
y = k (0 + p )n
where k is the strength coefficient, 0 is the initial strain at yield, p is the plastic strain, and n is the hardening
coefficient. All of the anisotropic material constants, excluding p which is determined implicitly, are determined
from Barlat and Lian width to thickness strain ratio (R) values as shown:
a=2 2
R00
R90
1 + R00 1 + R90
c=2a
R00 1 + R90
1 + R00 R90
h=
The width to thickness strain ratio for any angle can be calculated from:
R =
2mm
Y
xx yy
Above, is the uniaxial tension in the direction. The stress-strain behavior can be specified at only one temperature. Input the elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input
the hardening rule type, HR (equal to 1 for linear or 2 for exponential), tangent modulus (for HR = 1) or strength
coefficient (for HR = 2), yield stress (for HR = 1) or hardening coefficient (for HR = 2), the Barlat exponent, m, the
width to thickness strain ratio values, R00, R45, and R90, and the orthotropic material axes with TB,PLAW,,,,3 and
locations 1 - 8 of the TBDATA command:
TB,PLAW,,,,3
TBDATA, 1, HR (hardening rule type)
TBDATA, 2, Etan or k (tangent modulus or strength coefficient)
TBDATA, 3, Y or n (yield stress or hardening coefficient)
TBDATA, 4, m (Barlat exponent)
TBDATA, 5, R00
TBDATA, 6, R45
TBDATA, 7, R90
TBDATA,8,CSID (defines orthotropic material axes)
712

The last input item, CSID, has two valid values: 0 (default) and 2. If CSID = 0, the local coordinate system will be
defined by nodes I, J, and L of the element:
If CSID = 2, material axes are determined by a local coordinate system specified with the EDLCS command (see
the command description for the details of how the axes are oriented). Before specifying the material property,
you must define the local coordinate system using EDLCS, and then issue EDMP,ORTHO,,VAL1, where VAL1 is
the coordinate system ID from EDLCS.
For an example input listing, see Section B.2.12: 3 Parameter Barlat Example: Aluminum 5182.
7.2.3.8. Barlat Anisotropic Plasticity Model

Anisotropic plasticity model developed by Barlat, Lege, and Brem[2.] used for modeling material behavior in
forming processes. The anisotropic yield function is defined as:
= S1 S2 m + S2 S3 m + S3 S1 m
where m is the flow potential exponent and Si are the principal values of the symmetric matrix Sij.
S xx = 1/ 3 [c ( xx yy ) b (zz xx )]
S yy = 1/ 3 [a ( yy zz ) c ( xx yy )]
Szz = 1/ 3 [b (zz xx ) a ( yy zz )]
S yz = f yz
Szx = g zx
S xy = f xy
where a, b, c, f, g, and h represent the anisotropic material constants. When a=b=c=f=g=h=1, isotropic material
behavior is modeled and the yield surface reduces to the Tresca surface for m = 1 and the von Mises surface for
m = 2 or 4. For this material option, the yield strength is given by:
713
y = k (P + 0 ) n
where k is the strength coefficient, P is the plastic strain, 0 is the initial strain at yield, and n is the hardening
coefficient. The stress-strain behavior can be specified at only one temperature.
Input the elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the
strength coefficient, the initial strain at yield, the hardening coefficient, the flow potential exponent, and the
Barlat anisotropic constants a-h with TB,PLAW,,,,6 and locations 1 - 10 of the TBDATA command.
TB,PLAW,,,,6
TBDATA, 1, k (strength coefficient)
TBDATA, 2, 0 (initial strain)
TBDATA, 3, n (hardening coefficient)
TBDATA, 4, m (flow potential (Barlat) exponent)
TBDATA, 5, a
TBDATA, 6, b
TBDATA, 7, c
TBDATA, 8, f
TBDATA, 9, g
TBDATA, 10, h
For an example input listing, see Section B.2.13: Barlat Anisotropic Plasticity Example: 2008-T4 Aluminum.
7.2.3.9. Rate Sensitive Power Law Plasticity Model

Strain rate dependent plasticity model typically used for superplastic forming analyses. The material model follows
a Ramburgh-Osgood constitutive relationship of the form:
yy = k mn
where is the strain, is the strain rate, k is the material constant, m is the hardening coefficient, and n is the
strain rate sensitivity coefficient. The stress-strain behavior can be specified at only one temperature. Input the
elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the material constant,
hardening coefficient, strain rate sensitivity coefficient, and initial strain rate with TB,PLAW,,,,4 and locations 1 4 of the TBDATA command:
TB,PLAW,,,,4
TBDATA, 1, k (material constant)
TBDATA, 2, m (hardening coefficient)
TBDATA, 3, n (strain rate sensitivity coefficient)
TBDATA, 4, (initial strain rate)
0
For an example input listing, see Section B.2.14: Rate Sensitive Powerlaw Plasticity Example: A356 Aluminum.
7.2.3.10. Strain Rate Dependent Plasticity Model

Strain rate dependent isotropic plasticity model used mainly for metal and plastic forming analyses.
In this model, a load curve is used to describe the initial yield strength, 0, as a function of effective strain rate.
The yield stress for this material model is defined as:
714

eff
Y = 0 + Ehp
where 0 is the initial yield strength, is the effective strain rate, (p)eff is the effective plastic strain, and Eh is
given by:
Eh =
EEtan
E Etan
The stress-strain behavior can be specified at only one temperature. Input the elastic modulus (EX), density
(DENS), and Poissons ratio (NUXY) using the MP command. Input the load curve ID for defining the initial yield
stress vs. effective strain rate, the tangent modulus, the load curve ID for defining the elastic modulus vs. effective
strain rate (optional), the load curve ID for defining the tangent modulus vs. effective strain rate (optional), and
the load curve ID for defining the von Mises stress at failure vs. effective strain rate using TB,PLAW,,,,5 and locations
1 - 5 of the TBDATA command. For shell elements, you have the option of specifying Mn Time in location 6, instead
of LCID4 in location 5, to define failure of the material. Mn Time is the minimum time step size for automatic
element deletion.
TB,PLAW,,,,5
TBDATA, 1, LCID1 (load curve ID for defining the initial yield stress vs. effective strain rate)
TBDATA, 2, Etan (tangent (plastic hardening) modulus)
TBDATA, 3, LCID2 (the load curve ID for defining the elastic modulus vs. effective strain rate)
TBDATA, 4, LCID3 (the load curve ID for defining the tangent modulus vs. effective strain rate)
TBDATA, 5, LCID4 (the load curve ID for defining von Mises stress at failure vs. effective strain rate)
TBDATA, 6, Mn Time (minimum time step size for automatic element deletion, for shell elements only)
For an example input listing, see Section B.2.15: Strain Rate Dependent Plasticity Example: 4140 Steel.
7.2.3.11. Piecewise Linear Plasticity Model

This model provides a multilinear elastic-plastic material option that allows stress vs. strain curve input and strain
rate dependency. This is a very commonly used plasticity law, especially for steel. Failure based on plastic strain
can also be modeled with this material.
The piecewise linear plasticity model provides three different methods to account for the strain rate. The first
method utilizes the Cowper-Symonds model, which scales the yield stress as shown:
1
P
P
P
P
y (eff , eff ) = y (eff ) 1 + ( eff )P
(P )
where is the effective plastic strain rate, C and P are strain rate parameters, and y eff is the yield stress
without considering strain rate. This yield stress quantity can be input as either y and Etan (TBDATA entries 1
and 2) or as a total true stress vs. effective plastic strain curve, LCID1 (TBDATA entry 6).
The second method uses the LCID2 curve (TBDATA entry 7) to scale the yield stress with respect to strain rate.
If LCID2 is input, it is used instead of the Cowper-Symonds scaling term defined by the strain rate parameters C
and P.
715

The third method used to account for the strain rate dependency requires that you define separate stress vs.
effective plastic strain curves at specific strain rates. Using TBDATA entries 8 - 27, you may input stress vs. effective
plastic strain behavior for up to ten different strain rates (Rate1 and LCID3, Rate2 and LCID4, etc.). It is important
that you input all strain rate definitions in ascending order. Note that if Rate1 and LCID3 are specified, TBDATA
entries 4-7 (C, P, LCID1, and LCID2) are automatically ignored.
Input the elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the yield
stress, the tangent modulus, the effective plastic strain at failure, and the remaining parameters (listed below)
with TB,PLAW,,,,8 and locations 1 - 27 of the TBDATA command:
TB, PLAW,,,,8
TBDATA,1, Y (yield stress)
TBDATA,2, Etan (tangent modulus)
TBDATA,3, F (effective plastic strain at failure)
TBDATA,4, C (strain rate parameter)
TBDATA,5, P (strain rate parameter)
TBDATA,6, LCID1 (load curve defining total true stress vs. effective plastic strain)
TBDATA,7, LCID2 (load curve that scales LCID1 with respect to strain rate)
TBDATA,8, Rate1 (strain rate value 1)
TBDATA,9, LCID3 (stress vs. effective plastic strain curve for Rate1)
If load curve LCID1 (total true stress vs. effective plastic strain) is input, the yield stress (Y) and tangent modulus
(Etan) values will be overwritten. If LCID2 (the scaling curve) is input, strain rate parameters C and P will be overwritten. If C and P, LCID2, and Rate1 (first strain rate value) are all set to zero, strain rate dependency is ignored.
For an example input listing, see Section B.2.16: Piecewise Linear Plasticity Example: High Carbon Steel. See also
Section 4.1.4.1: Using Data Curves with Material Models for an example of how to input a true stress vs. effective
plastic strain curve for this material model.
7.2.3.12. Modified Piecewise Linear Plasticity Model

This model provides a multilinear elastic-plastic material option that allows stress vs. strain curve input and strain
rate dependency. This model is similar to piecewise linear plasticity (TB,PLAW,,,,8), but has an enhanced failure
716

criteria. Failure is based on effective plastic strain, plastic thinning, or the major principal in-plane strain component
(see TBDATA entries 8 - 10: EPSTHIN, EPSMAJ, and NUMINT).
The modified piecewise linear plasticity model uses the same three methods to account for the strain rate as the
piecewise linear plasticity model (see Section 7.2.3.11: Piecewise Linear Plasticity Model for details). To account
for strain rate, do one of the following:
Input the strain rate parameters C and P (TBDATA entries 4 and 5) to use the Cowper-Symonds model.
Input the LCID2 curve (TBDATA entry 7) to scale the yield stress with respect to strain.
Define separate stress vs. strain curves for different strain rates (TBDATA entries 11 - 30).
Input the elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the remaining parameters with TB,PLAW,,,,11 and locations 1 - 30 of the TBDATA command:
TB, PLAW,,,,11
TBDATA,1, Y (yield stress)
TBDATA,2, Etan (tangent modulus)
TBDATA,3, F (effective plastic strain at failure)
TBDATA,4, C (strain rate parameter)
TBDATA,5, P (strain rate parameter)
TBDATA,6, LCID1 (load curve defining total true stress vs. effective plastic strain)
TBDATA,7, LCID2 (load curve that scales LCID1 with respect to strain rate)
TBDATA,8, EPSTHIN (thinning plastic strain at failure)
TBDATA,9, EPSMAJ (major in-plane strain at failure)
TBDATA,10, NUMINT (number of through-thickness integration points that must fail for element failure)
If load curve LCID1 (total true stress vs. effective plastic strain) is input, the yield stress (Y) and tangent modulus
(Etan) values will be overwritten. If LCID2 (the scaling curve) is input, strain rate parameters C and P will be overwritten. If C and P, LCID2, and Rate1 (first strain rate value) are all set to zero, strain rate dependency is ignored.
For an example input listing, see Section B.2.17: Modified Piecewise Linear Plasticity Example: PVC.
717
7.2.3.13. Composite Damage Model

Material model developed by Chang & Chang[4.] for the failure of composite materials. The following five parameters are used in the model:
S1 = longitudinal tensile strength
S2 = transverse tensile strength
S12 = shear strength
C2 = transverse compressive strength
= nonlinear shear stress parameter
All parameters are determined experimentally. Input the elastic moduli (EX, EY, EZ), shear moduli (GXY, GYZ,
GXZ), density (DENS), and Poissons ratios (NUXY, NUYZ, NUXZ) with the MP command. Input the bulk modulus
at compressive failure, the shear strength, the longitudinal tensile strength, transverse tensile strength, transverse
compressive strength, and nonlinear shear stress parameter with TB,COMP and locations 1 - 6 of the TBDATA
command:
TB, COMP
TBDATA, 1, KFAIL (bulk modulus of material in compressive failure)
TBDATA, 2, S12 (shear strength)
TBDATA, 3, S1 (longitudinal tensile strength)
TBDATA, 4, S2 (transverse tensile strength)
TBDATA, 5, C2 (transverse compressive strength)
TBDATA, 6, (nonlinear shear stress parameter)
Note Please refer to the LS-DYNA Theoretical Manual for more details about LS-DYNA Material model
#22 (Composite Damage). This material model is required for multi-layer composite laminates, even if
the failure capabilities are not used. The lamina (ply) properties are specified as SHELL163 real constants.
7.2.3.14. Concrete Damage Model

The Concrete Damage model is used to analyze buried steel reinforced concrete subject to impact loadings. This
model requires constants for the concrete and reinforced materials, and also requires tabulated equation of state
constants. (Refer to Section 7.2.6: Equation of State Models for a detailed description of the tabulated equation
of state.) Input the density (DENS) and Poissons ratio (NUXY) with the MP command. Input all values shown
below with the TB,CONCR,,,,2 command and locations 1 78 of the TBDATA command:
TB,CONCR,,,,2
TBDATA,1,f (maximum principal stress for failure)
TBDATA,2,A0 (cohesion constant)
TBDATA,3,A1 (pressure hardening coefficient)
TBDATA,4,A2 (pressure hardening coefficient)
TBDATA,5,A0Y (cohesion for yield)
TBDATA,6,A1Y (pressure hardening coefficient for yield limit)
TBDATA,7,A2Y (pressure hardening coefficient for yield limit)
TBDATA,8,A1F (pressure hardening coefficient for failed material)
TBDATA,9,A2F (pressure hardening coefficient for failed material)
TBDATA,10,B1 (damage scale factor)
TBDATA,11,B2 (damage scale factor for uniaxial tension)
TBDATA,12,B3 (damage scale factor for triaxial tension)
TBDATA,13,PER (percent reinforcement)
718

TBDATA,14,Er (elastic moduli of reinforcement)
TBDATA,15,NUXYr (Poissons ratio of reinforcement)
TBDATA,16,y (initial yield stress)
TBDATA,17,Etan (tangent modulus)
TBDATA,18,LCP (load curve ID giving rate sensitivity for principal material)
TBDATA,19,LCR (load curve ID giving rate sensitivity for reinforcement)
TBDATA,20-32,1-13 (tabulated damage functions 1-13)
TBDATA,33-45,1-13 (tabulated scale factors 1-13)
TBDATA,46,GAMA (temperature constant)
TBDATA,47,E0 (initial internal energy)
TBDATA,48,V0 (initial relative volume)
TBDATA,49-58,v1-v10 (volumetric strain data values 1-10; natural log of the relative volume)
TBDATA,59-68,C1-C10 (volumetric pressure values at vi)
TBDATA,69-78,T1-T10 (temperature values at vi)
7.2.3.15. Power Law Plasticity Model

Strain rate dependent plasticity model typically used for metal and plastic forming analyses. Elastoplastic behavior with isotropic hardening is provided with this model. The material model has a Power Law constitutive relationship which includes the Cowper-Symonds multiplier to account for strain rate:
1
P
Y = 1 + k[e + P ] n
eff
C

where is the strain rate, C and P are the Cowper-Symonds strain rate parameters, e is the elastic strain to yield,
eff
( p )
is the effective plastic strain, k is the strength coefficient, and n is the hardening coefficient. The stressstrain behavior can be specified at only one temperature. Input the elastic modulus (EX), density (DENS), and
Poissons ratio (NUXY) with the MP command. Input the strength coefficient, hardening coefficient, and strain
rate parameters C and P with TB,PLAW,,,,2 and locations 1 - 4 of the TBDATA command:
TB, PLAW,,,,2
TBDATA, 1, k (strength coefficient)
TBDATA, 2, n (hardening coefficient)
TBDATA, 3, C (strain rate parameter)
TBDATA, 4, P (strain rate parameter)
For an example input listing, see Section B.2.18: Powerlaw Plasticity Example: Aluminum 1100.
7.2.3.16. Elastic Viscoplastic Thermal Model

This material allows an elastic viscoplastic material definition with thermal effects. The uniaxial stress-stress-strain
curve has the form:
719
(p , p ) = 0 + Qr1 (1 exp( Cr1 p )) + Qr 2 (1 exp( Cr 2 p f ))

eff eff
eff
ef
+Q x1 (1 exp( Cx1 p )) + Q x 2 (1 exp( Cx 2 p ))
eff
eff
+ Vk p
eff
Vm
where 0 is the initial yield stress, peff is the effective plastic strain, and the other parameters are as described in
the TBDATA input shown below.
Strain rate is accounted for using the Cowper and Symonds model (where C and P are the Cowper-Symonds
strain rate parameters), which scales the yield stress with the factor:
p
1 + eff
C
To convert from these constants, set the viscoelastic constants, Vk and Vm, to the following values:
1
1
Vk = p
C
Vm =
1
p
You must input density (DENS) using the MP command. You must also input the elastic modulus (EX), Poisson's
ratio (NUXY), and the coefficient of thermal expansion (ALPX) by either using the MP command or by inputting
load curves in locations 13, 14, and 18 of the TBDATA command. You can use the load curve input to represent
each quantity as a function of temperature. Input the remaining parameters (described below) with the
TB,PLAW,,,,12 command and locations 1 - 18 of the TBDATA command.
TB,PLAW,,,,12
TBDATA,1, 0 (initial yield stress)
TBDATA,2, QR1 (isotropic hardening parameter)
TBDATA,3, CR1 (isotropic hardening parameter)
TBDATA,4, QR2 (isotropic hardening parameter)
TBDATA,5, CR2 (isotropic hardening parameter)
TBDATA,6, QX1 (kinematic hardening parameter)
TBDATA,7, CX1 (kinematic hardening parameter)
TBDATA,8, QX2 (kinematic hardening parameter)
TBDATA,9, CX2 (kinematic hardening parameter)
TBDATA,10, Vk (viscous parameter)
TBDATA,11, Vm (viscous parameter)
TBDATA,12, LCID1 (load curve ID for total true stress vs. effective plastic strain (if specified, TBDATA 2-9 are
ignored))
TBDATA,13, LCID2 (load curve ID for elastic modulus vs. temperature)
TBDATA,14, LCID3 (load curve ID for Poisson's ratio vs. temperature)
TBDATA,15, LCID4 (load curve ID for initial yield stress vs. temperature; if input, 0 is ignored)
720

TBDATA,16, LCID5 (load curve ID for scaling factor on isotropic hardening parameters (QR1 and QR2) vs.
temperature)
TBDATA,17, LCID6 (load curve ID for scaling factor on kinematic hardening parameters (QX1 and QX2) vs.
temperature)
TBDATA,18, LCID7 (load curve ID for thermal coefficient of expansion vs. temperature)
Note If the total true stress vs. effective plastic strain is input (LCID1 in location 12 of TBDATA), then
TBDATA locations 2-9 are ignored. Also, if LCID2, LCID3, or LCID7 is specified, then the corresponding
MP command is ignored. (For example, LCID2 overrides the command MP,EX.)
For an example input listing, see Section B.2.19: Elastic Viscoplastic Thermal Example.
7.2.4. Pressure Dependent Plasticity Models

7.2.4.1. Elastic-Plastic Hydrodynamic Model
This model is used for modeling materials undergoing large amounts of strain, where the plastic behavior can
either be defined from a set of data points or from the yield stress and tangent modulus. If the effective true
plastic strain and effective true stress data values are not specified, the yield strength will be calculated as shown
below (based on isotropic hardening):
y = 0 + Eh p
You can calculate the plastic hardening modulus, Eh, in terms of the Youngs modulus and tangent modulus:
Eh =
E tE
E Et
If the effective true plastic strain and stress values are specified, the stress-strain behavior is defined by the data
points along the effective true stress vs. true plastic strain curve. Up to 16 data points may be specified. Linear
extrapolation will be used if the strain values exceed the maximum input values; therefore, you should input
values to cover the full range of strains expected in the analysis. Input the density (DENS), elastic modulus (EX),
and shear modulus (GXY) with the MP command. Input the parameters described below with TB,PLAW,,,,9 and
locations 1-45 of the TBDATA command:
TB,PLAW,,,,9
TBDATA,1, 0 (initial yield stress)
TBDATA,2,Eh (hardening modulus)
TBDATA,3,PC (pressure cutoff value)
TBDATA,4, f (failure strain)
TBDATA,5-20,1-16 (effective strain data curve values)
TBDATA,21-36,1-16 (effective stress data curve values)
TBDATA,37,Co (linear polynomial equation of state constant)
TBDATA,38,C1 (linear polynomial equation of state constant)
721

TBDATA,44,Eo (initial internal energy)
TBDATA,45,Vo (initial relative volume)
Note that constants 37 through 45 specified on the TBDATA command are the same as the linear polynomial
equation of state model. Please refer to Section 7.2.6: Equation of State Models later in this chapter for more
details.
7.2.4.2. Geological Cap Model

This model is an inviscid, two invariant geologic cap model which can be used for geomechanical problems or
for materials such as concrete. In this model, the two invariant cap theory is extended to include nonlinear kinematic hardening. A discussion of the extended cap model and its parameters is given below.
The cap model is formulated in terms of the invariants of the stress tensor. The square root of the second invariant
J2D
of the deviatoric stress tensor,

J2D
, is found from the deviatoric stresses (s) as
1
Sij Sij
2
and is the objective scalar measure of the distortional or shearing stress. The first invariant of the stress, J1, is the
trace of the stress tensor.
Figure 7.1 Surface of the Two-invariant Cap Model

$ "
%#!
"
2D
1 space, as shown in Figure 7.1: Surface of the
The cap model consists of three surfaces in pressure
Two-invariant Cap Model. Surface f1 is the failure envelope, f2 is the cap surface, and f3 is the tension cutoff. The
functional form of f1 is
f1 =
J2D min(Fe (J1), Tmises )
where Fe is given by
Fe (J1) exp( J1) + J1

722

and
Tmises X ( n ) L ( n )
J
2D
1 space, and therefore does not harden unless kinematic
This failure envelope surface is fixed in
hardening is present. Next, there is a cap surface f2 in the figure, with f2 given by
f2 = J2D Fc (J1 , )
where Fc is defined by
Fc (J1 , )
1
R
[X ( ) L ( )]2 J1 L ( ) 2
X() is the intersection of the cap surface with the J1 axis
X( ) = + RFe ( )
and L() is defined by
if > 0
L ( )
0 if 0
The hardening parameter is related to the plastic volume change through the hardening law
p
p = W {1 exp[ D ( X( ) X 0 )]}
Geometrically, is seen in the figure as the J1 coordinate of the intersection of the cap surface and the failure
surface. Finally, there is the tension cutoff surface, denoted f3 in the figure. The function f3 is given by
f3 T J1
where T is the input material parameter which specifies the maximum hydrostatic tension sustainable by the
2D
1 space is then bounded by the failure envelope surface above, the
material. The elastic domain in
tension cutoff surface on the left, and the cap surface on the right.
Input the density (DENS) and shear modulus (GXY) with the MP command. Input the parameters described below
with the TB,GCAP command and locations 113 of the TBDATA command.
TB,GCAP
TBDATA,1,K (bulk modulus)
TBDATA,2, (failure envelope parameter)
TBDATA,3, (failure envelope linear coefficient)
TBDATA,4, (failure envelope exponential coefficient)
TBDATA,5, (failure envelope exponent)
TBDATA,6,R (cap surface axis ratio)
723

TBDATA,7,D (hardening law exponent)
TBDATA,8,W (hardening law coefficient)
TBDATA,9,X0 (hardening law exponent)
TBDATA,10,C (kinematic hardening coefficient)
TBDATA,11,N (kinematic hardening parameter)
TBDATA,12,Ftype (formulation flag: use 1 for soil or concrete, 2 for rock )
TBDATA,13,Toff (tension cutoff value; Toff < 0, positive in compression)
Please refer to Material Model 25 in the LS-DYNA Theoretical Manual for more information about this material
model.
For an example input listing, see Section B.2.20: Geological Cap Example: SRI Dynamic Concrete.
7.2.5. Foam Models

7.2.5.1. Closed Cell Foam Model
Rigid, closed cell, low density polyurethane foam material model often used for modeling impact limiters in
automobile designs. This material model is very similar to the honeycomb model in that the components of the
stress tensor are uncoupled until volumetric compaction is achieved. Unlike honeycomb, however, the closed
cell foam model is isotropic in nature and includes the effects of confined air pressure. The material model defines
stress to be:
sk
ij = ij + ij [po /(1 + )]
where (ij)sk is the skeletal stress, po is the initial foam pressure, is the ratio of foam to polymer density, ij is the
Kronecker delta, and is the volumetric strain which is defined by:
= V 1 + 0
where V is the relative volume and o is the initial volumetric strain. The yield condition is applied to the principal
trial stresses and is defined by:
Y = a + b (1 + c )
where a, b, and c are user defined input constants. The stress-strain behavior can be specified at only one temperature. Input the elastic modulus (EX) and density (DENS) with the MP command. Poissons ratio for this
model is assumed to be zero. Input the yield stress constants a, b, and c, the initial foam pressure, the ratio of
foam to polymer density, and the initial volumetric strain with TB,FOAM,,,,1 and locations 1 - 6 of the TBDATA
command:
TB, FOAM,,,,1
TBDATA, 1, a
TBDATA, 2, b
TBDATA, 3, c
TBDATA, 4, po (initial foam pressure)
TBDATA, 5, (ratio of foam to polymer density)
TBDATA, 6, o (initial volumetric strain)
724
7.2.5.2. Viscous Foam Model

Energy absorbing foam material used in crash-simulation models. The model consists of a nonlinear elastic
stiffness in parallel with a viscous damper. The elastic stiffness is intended to limit total crush while the viscosity
absorbs energy. The elastic stiffness, E , and initial viscous coefficient, V , are both nonlinear functions of the
relative volume, V:
E1 = E1V n1
V2 = V2 1 V n 2
where E1 is the initial elastic stiffness, V2 is the initial viscous coefficient, and n1 and n2 are the Power Laws for
the elastic stiffness and viscous coefficient, respectively. The stress-strain behavior can be specified at only one
temperature. Input the initial elastic stiffness (EX), Poissons ratio (NUXY), and density (DENS) with the MP command. Input the Power Law for the elastic stiffness, the initial viscous coefficient, the elastic stiffness for viscosity
(required to prevent time step problems), and the pearly for viscosity with TB,FOAM,,,,3 and locations 1 - 4 of
the TBDATA command:
TB, FOAM,,,,3
TBDATA, 1, n1 (Power Law for the elastic stiffness)
TBDATA, 2, V2 (initial viscous coefficient)
TBDATA, 3, E1 (elastic stiffness for viscosity)
TBDATA, 4, n2 (Power Law for the viscous coefficient)
7.2.5.3. Low Density Foam Model

Highly compressible (urethane) foam material model often used for padded materials such as seat cushions. In
compression, the model assumes hysteresis unloading behavior with possible energy dissipation. In tension the
material model behaves linearly until tearing occurs. For uniaxial loading, the model assumes that there is no
coupling in transverse directions. By using input shape factor controls (a hysteresis unloading factor (HU), a decay
constant (), and a shape factor for unloading) the observed unloading behavior of foams can be closely approximated. The stress-strain behavior can be specified at only one temperature. Input the elastic modulus (EX) and
density (DENS) with the MP command. Input the load curve ID for nominal stress vs. strain, the tension cutoff
(tearing) stress, the hysteresis unloading factor, the decay constant, the viscous coefficient, the shape factor for
unloading, the failure option when the cutoff stress is reached, and the bulk viscosity action flag with TB,FOAM,,,,2
and locations 1 - 8 of the TBDATA command:
TB, FOAM,,,,2
TBDATA, 1, LCID (load curve ID for stress-strain behavior)
TBDATA, 2, TC (tension cutoff stress, default=1E20)
TBDATA, 3, HU (hysteretic unloading factor: between 1.0 - no energy dissipation (default) and 0.0 - full energy
dissipation)
TBDATA, 4, (decay constant)
TBDATA, 5, DAMP (viscous coefficient, values between 0.05 and 0.5 are recommended)
TBDATA, 6, SHAPE (shape unloading factor, default = 1)
TBDATA, 7, FAIL (failure option when cutoff stress is reached: 0.0 - tensile stress remains at cutoff value, 1.0
- tensile stress reset to zero)
TBDATA, 8, BVFLAG (bulk viscosity action flag: 0.0 - no bulk viscosity (recommended) and 1.0 - bulk viscosity
active)
725
7.2.5.4. Crushable Foam Model

Material model used for crushable foams in side impact and other applications where cyclic behavior is unimportant. The foam model is strain rate dependent and crushes one-dimensionally with a Poissons ratio that is
essentially zero. In the formulation, the elastic modulus is considered constant and the stress is updated assuming
elastic behavior:
n +1/ 2
n +1 = n + E ij
ij
ij
tn +1/ 2
where ij is the strain rate, E is the elastic modulus, and t is time. The model includes a tensile stress cutoff value
which defines failure under tensile loads. For stresses below the tensile cutoff value, the model predicts similar
response between tensile and compressive loading. It is important to have a nonzero value for the cutoff stress
to prevent deterioration of the material under small tensile loads. Input the elastic modulus (EX), density (DENS),
and Poissons ratio (NUXY) with the MP command. Input the load curve ID for defining stress vs. volumetric strain,
the tension cutoff value, and the viscous damping coefficient with TB,FOAM,,,,4 and locations 1 - 3 of the TBDATA
command:
TB, FOAM,,,,4
TBDATA, 1, LCID (load curve ID for defining yield stress vs. volumetric strain)
TBDATA, 2, TC (tension cutoff value)
TBDATA, 3, DAMP (viscous damping coefficient, values between 0.05 and 0.5 are recommended)
7.2.5.5. Honeycomb Foam Model

Orthotropic material model used for honeycomb materials. The behavior of the model before compaction is
orthotropic where the components of the stress tensor are uncoupled. The elastic moduli are considered to vary
linearly with relative volume as shown below:
Eaa = Eaau + (E Eaau )

Ebb = Ebbu + (E Ebbu )
Ecc = Eccu + (E Eccu )
Gab = Gabu + (G Gabu )
Gbc = Gbcu + (G Gbcu )
Gca = Gcau + (G Gcau )
where
G=
E
= Elastic shear modulus for fully compacted honeycomb material
o
2(1 + )
and
1 V
= max min
,1, 0
1 Vf
V = the relative volume (defined as ratio of current to original volume)

Vf = relative volume of the fully compacted honeycomb
726

To specify a material coordinate system, use the EDLCS and EDMP,ORTHO commands. If no material coordinate
system is specified, material properties are locally orthotropic with material axes defined by nodes I, J, and L of
the element (see Orthotropic Elastic Model for a graphic depiction).
Load curves can be used to input the magnitude of the average stress as the relative volume changes. Each curve
must have the same abscissa values. Curves can be defined either as a function of relative volume (V) or volumetric
strain (1 - V). Input the elastic modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command.
Input all values shown below with TB,HONEY and locations 1 - 17 of the TBDATA command:
TB, HONEY
TBDATA, 1, Y (yield stress for fully compacted honeycomb)
TBDATA, 2, Vf (relative volume of fully compacted honeycomb)
TBDATA, 3, (material viscosity coefficient, default = 0.05)
TBDATA, 4, Eaau (Elastic modulus in aa direction for uncompressed configuration)
TBDATA, 5, Ebbu (Elastic modulus in bb direction for uncompressed configuration)
TBDATA, 6, Eccu (Elastic modulus in cc direction for uncompressed configuration)
TBDATA, 7, Gabu (Shear modulus in ab direction for uncompressed configuration)
TBDATA, 8, Gbcu (Shear modulus in bc direction for uncompressed configuration)
TBDATA, 9, Gcau (Shear modulus in ca direction for uncompressed configuration)
TBDATA, 10, LCA (Load curve ID for aa direction stress vs. relative volume or volumetric strain)
TBDATA, 11, LCB (Load curve ID for bb direction stress vs. relative volume or volumetric strain)
TBDATA, 12, LCC (Load curve ID for cc direction stress vs. relative volume or volumetric strain)
TBDATA, 13, LCS (Load curve ID for shear yield vs. relative volume or volumetric strain)
TBDATA, 14, LCAB (Load curve ID for ab direction stress vs. relative volume or volumetric strain)
TBDATA, 15, LCBC (Load curve ID for bc direction stress vs. relative volume or volumetric strain)
TBDATA, 16, LCCA (Load curve ID for ca direction stress vs. relative volume or volumetric strain)
TBDATA, 17, LCRS (Load curve ID for strain rate effects. This input is optional. The curves defined above are
scaled using this curve)
7.2.6. Equation of State Models

There are 3 different equation of state models available in ANSYS LS-DYNA:
Linear polynomial
Gruneisen
Tabulated
These equation of state models are used in conjunction with certain material models, such as the Johnson-Cook
Plasticity and Zerilli-Armstrong models. You identify the equation of state to be used by inputting the appropriate
number in the EOSOPT field of the TB command.
7.2.6.1. Linear Polynomial Equation of State

(EOSOPT = 1)
This equation of state is linear in internal energy. The pressure is given by:
P = C0 + C1 + C22 + C33 + (C4 + C5 + C62 )E
727
where terms C22 and C62 are set to zero if

density.
< 0, =
1
0
and 0 is the ratio of current density to initial
Input the required constants using the TBDATA command. The starting location (LOC) for the constants depends
on which material model you are using. (See the description for the specific material model to find out what
starting location value should be used.)
TBDATA,LOC,C0
TBDATA,LOC+1,C1
TBDATA,LOC+2,C2
TBDATA,LOC+3,C3
TBDATA,LOC+4,C4
TBDATA,LOC+5,C5
TBDATA,LOC+6,C6
TBDATA,LOC+7,E0 (initial internal energy)
TBDATA,LOC+8,V0 (initial relative volume)
7.2.6.2. Gruneisen Equation of State

(EOSOPT = 2)
This equation of state defines the pressure volume relationship in one of two ways, depending on whether the
material is compressed or expanded.
The Gruneisen equation of state with cubic shock velocity-particle velocity defines pressure for compressed
materials as:
p=

a
0C2 1 + 1 0 2
2
2

2
3
S3
1 (S1 1) S2
+1
( + 1)2
+ ( 0 + a )E
and for expanded materials as:
p = 0C2 + ( 0 + a )E
where C is the intercept of the vs-vp curve; S1, S2, and S3 are the coefficients of the slope of the vs-vp curve, 0
=
is the Gruneisen gamma, a is the first order volume correction to 0, and
1
0
TBDATA,LOC,C
TBDATA,LOC+1,S1
728

TBDATA,LOC+2,S2
TBDATA,LOC+3,S3
TBDATA,LOC+4,0
TBDATA,LOC+5,A
7.2.6.3. Tabulated Equation of State

(EOSOPT = 3)
The tabulated equation of state model is linear in internal energy. Pressure is defined by:
P = Ci ( vi ) + Ti ( vi )E
TBDATA,LOC,GAMA (temperature constant)
TBDATA,LOC+3 to LOC+12,v1-v10 (volumetric strain data values 1-10, natural log of the relative volume)
TBDATA,LOC+13 to LOC+22,C1-C10 (volumetric pressure values at vi)
TBDATA,LOC+23 to LOC+32,T1-T10 (temperature values at vi)
7.2.6.4. Bamman Plasticity Model

This is a fairly complex material model used in metal forming processes with strain rate and temperature dependent
plasticity. The Bamman model does not require an additional equation of state model (EOSOPT is not used) because
internal equation of state variables are specified in fields 21 through 26 of the TBDATA command (as shown
below).
Input the density (DENS), elastic modulus (EX) and Poisson's ratio (NUXY) with the MP command. Input the
parameters described below with TB,EOS and locations 1-26 of the TBDATA command:
TB,EOS,,,,4
TBDATA,1,TI (initial temperature)
TBDATA,2,HC (heat generation coefficient)
TBDATA,3-20,C1-C18 (constants for flow stress)
TBDATA,21-26,A1-A6 (internal equation of state variables)
For a complete discussion of the kinematics associated with this model, as well as a more detailed description
of the required input constants, refer to the description of Material Model 51 in Chapter 16 of the LS-DYNA Theoretical Manual.
7.2.6.5. Johnson-Cook Plasticity Model

This model, also called the viscoplastic model, is a strain-rate and adiabatic (heat conduction is neglected) temperature-dependent plasticity model. This model is suitable for problems where strain rates vary over a large
729

range, and temperature changes due to plastic dissipation cause material softening. The model may be used for
shell and solid elements. For solid elements, you need an equation of state (discussed later in this section).
Johnson and Cook express the flow stress as
n
y = ( A + B p )(1 + c ln *)(1 T *m )
where:
A, B, C, n, and m = material constant
p = effective plastic strain
p
*=
0
T * = homologous temperature =
T Troom
Tmelt Troom
The strain at fracture is given by:
f = D1 + D2 exp D3 * 1 + D4 ln * 1 + D5 T *
Where * is the ratio of pressure divided by effective stress:
*=
p
eff
Fracture occurs when the damage parameter
D=
p
f
reaches the value of 1.

Input the Youngs modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the
parameters described in the above equations with TB,EOS and locations 1 - 15 of the TBDATA command:
TB,EOS,,,,1,EOSOPT
TBDATA,1,A
TBDATA,2,B
TBDATA,3,n
TBDATA,4,C
TBDATA,5,m
TBDATA,6,Tmelt
TBDATA,7,Troom
TBDATA,8, * (effective plastic strain rate)
TBDATA,9,CP (specific heat)
TBDATA,10,pressure cutoff
TBDATA,11,D1
730

TBDATA,12,D2
TBDATA,13,D3
TBDATA,14,D4
TBDATA,15,D5
When using the Johnson-Cook Plasticity Model, you can define one of three types of equations of state: Linear
Polynomial (EOSOPT = 1), Gruneisen (EOSOPT = 2), and Tabulated (EOSOPT = 3). Each has its own set of required
constants that are input starting with position (LOC) 16 of the TBDATA command. See the descriptions of the
three equation of state models in Section 7.2.6: Equation of State Models.
For an example input listing using the Linear Polynomial EOS, see Section B.2.21: Johnson-Cook Linear Polynomial EOS Example: 1006 Steel.
For an example input listing using the Gruneisen EOS, see Section B.2.22: Johnson-Cook Gruneisen EOS Example:
OFHC Copper.
7.2.6.6. Null Material Model

This material allows equations of state to be considered without computing deviatoric stresses. Optionally, you
can define a viscosity. Erosion in tension and compression is possible. The Youngs modulus and Poissons ratio
are used only for setting the contact interface stiffness, so you should use reasonable values.
Input the Youngs modulus (EX), density (DENS), and Poissons ratio (NUXY) with the MP command. Input the
parameters listed below with TB,EOS and locations 1 - 4 of the TBDATA command:
TB, EOS,,,,2, EOSOPT
TBDATA, 1, Pressure Cutoff ( 0.0)
TBDATA, 2, Viscosity Coefficient (optional)
TBDATA, 3, Relative volume for erosion in tension (V/Vo)
TBDATA, 4, Relative volume for erosion in compression (V/Vo)
When using the Null Model, you can define one of three types of equations of state: Linear Polynomial (EOSOPT
= 1), Gruneisen (EOSOPT = 2), and Tabulated (EOSOPT = 3). Each has its own set of required constants that are
input starting with position (LOC) 16 of the TBDATA command. See the descriptions of the three equation of
state models in Section 7.2.6: Equation of State Models.
For an example input listing using the Linear Polynomial EOS, see Section B.2.23: Null Material Linear Polynomial
EOS Example: Brass.
For an example input listing using the Gruneisen EOS, see Section B.2.24: Null Material Gruneisen EOS Example:
Aluminum.
7.2.6.7. Zerilli-Armstrong Model

This model is used in metal forming processes and high speed impact applications where the stress depends on
strain, strain rate, and temperature. The Zerilli-Armstrong model expresses the yield stress as follows.
For FCC Metals:
(T )
y = C1 + C2 (p )0.5 e( C3 + C4 In (*)) T + C5
(293)
731

and for BCC Metals:
(T )
y = C1 + C2 e( C3 + C4 In (*)) T + C5 (p )n + C6
(293)
where:
(T )
2
= B1 + B2 T + B3 T
(293)
and the relationship between heat capacity and temperature is given by:
Cp = G1 + G2T + G3T2 + G4T3
An equation of state must also be defined with the Zerilli-Armstrong model.
Input the density (DENS), elastic modulus (EX) and Poissons ratio (NUXY) with the MP command. Input the
parameters described below with TB,EOS and locations 1-19 of the TBDATA command:
TB,EOS,,,,3,EOSOPT
TBDATA,1,o (initial strain)
TBDATA,2,N (strain exponent for BCC metals)
TBDATA,3,Troom (room temperature)
TBDATA,4,PC (pressure cutoff value)
TBDATA,5,SPALL (Spall type. Set SPALL = 1.0 to use minimum pressure limit; SPALL = 2.0 to use maximum
principal stress; SPALL = 3.0 to use minimum pressure cutoff.)
TBDATA,6,C1 (flow stress coefficient)
TBDATA,12,FAIL (failure strain)
TBDATA,13,B1 (temperature coefficient)
TBDATA,16,G1 (heat capacity coefficient)
When using the Zerilli-Armstrong Model, you can define one of three types of equations of state: Linear Polynomial (EOSOPT = 1), Gruneisen (EOSOPT = 2), and Tabulated (EOSOPT = 3). Each has its own set of required constants
that are input starting with position (LOC) 20 of the TBDATA command. See the descriptions of the three
equation of state models in Section 7.2.6: Equation of State Models.
732
7.2.6.8. Steinberg Model

This material model is used for modeling high strain rate effects in solid elements with failure. The yield strength
is a function of both temperature and pressure. The model must be defined with an equation of state.
The Steinberg material model was originally developed for treating plasticity at very high strain rates (105 s-1)
where the influence of strain rate variation on the material's yield stress begins to saturate out. In this model,
the shear modulus (G) and the yield strength (y) both increase with pressure but decrease with temperature.
Both G and y approach zero as the melting temperature of the material is reached.
In this model, the shear modulus before the material melts is defined by:
E - Ec
fE /E E
G = G o 1 + bpV 1/3 - h i
- 300 e i m i
3R
where Go, b, h, and f are experimentally determined material constants, p is the pressure, V is the relative volume,
and Ec is the cold compression energy:
x
E c (x) =
pdx
900 R exp (ax )

(1 x )
2 ( o a 1 / 2)
x =1V
and Em is the melting energy:
E m (x) = E c ( x ) + 3RTm ( x )
which is in terms of the melting temperature Tm(x):
Tm (x) =
Tmo exp (2ax )

V
2 ( o a 1/ 3)
and the melting temperature at = o, Tmo.

In the above equations, R is defined by
R =
R
A
where R is the gas constant and A is the atomic weight. If R is not defined, LS-DYNA computes it with R in the
cm-gram-microsecond system of units.
The yield strength is given by:
Ei - E c
fE / E E
i
m
i
- 300 e
3R
y = 1 + b pV 1/3 - h
o
733

if Em exceeds Ei. Here,
is given by:
y = 1 + ( i + p )
o
where i is the initial plastic strain. Whenever o exceeds m, o is set equal to m. After the material melts,
y and G are set to one half their initial values.
If the coefficients EC0-EC9 are not defined in the input, LS-DYNA will fit the cold compression energy to a ten
term polynomial expansion either as a function of or , depending on the input variable FLAG:
E c (i ) =
E c ( i ) =
EC i
i
i=0
9
EC i
i
i=0
where ECi is the ith coefficient, and:
1
o
A linear least squares method is used to perform the fit.

A choice of three spall models is offered to represent material splitting, cracking, and failure under tensile loads.
The pressure limit model, SPALL = 1, limits the hydrostatic tension to the specified value, pcut. If pressures more
tensile than this limit are calculated, the pressure is reset to pcut. This option is not strictly a spall model, since
the deviatoric stresses are unaffected by the pressure reaching the tensile cutoff, and the pressure cutoff value,
pcut, remains unchanged throughout the analysis. The maximum principal stress spall model, SPALL = 2, detects
spall if the maximum principal stress, max, exceeds the limiting value -pcut. Note that the negative sign is required
because pcut is measured positive in compression, while max is positive in tension. Once spall is detected with
this model, the deviatoric stresses are reset to zero, and no hydrostatic tension (p < 0) is permitted. If tensile
pressures are calculated, they are reset to 0 in the spalled material. Thus, the spalled material behaves as a rubble
or incohesive material. The hydrostatic tension spall model, SPALL = 3, detects spall if the pressure becomes
more tensile than the specified limit, pcut. Once spall is detected the deviatoric stresses are reset to zero, and
nonzero values of pressure are required to be compressive (positive). If hydrostatic tension (p < 0) is subsequently
calculated, the pressure is reset to 0 for that element.
Input the shear modulus (GXY) and density (DENS) with the MP command. Input the parameters described in
the above equations with TB,EOS and locations 128 of the TBDATA command.
TB,EOS,,,5,EOSOPT
734

TBDATA,1,o (initial yield stress)
TBDATA,2, (hardening coefficient)
TBDATA,3,n (hardening parameter)
TBDATA,4,i (initial plastic strain)
TBDATA,5,m (maximum yield stress)
TBDATA,6,b (shear modulus parameter)
TBDATA,7,b' (Steinberg yield strength parameter)
TBDATA,8,h (Steinberg yield strength parameter)
TBDATA,9,f (Steinberg exponential coefficient)
TBDATA,10,A (atomic weight)
TBDATA,11,Tmo (absolute melting temperature)
TBDATA,12,o (melting temperature parameter)
TBDATA,13,a (melting temperature parameter)
TBDATA,14,PC (pressure cutoff, pcut)
TBDATA,15,SPALL (Spall type. Set SPALL = 1 to use the pressure limit spall model; set SPALL = 2 to use the
maximum principal stress spall model; set SPALL = 3 to use the hydrostatic tension spall model.)
TBDATA,16,FLAG (cold compression energy flag)
TBDATA,17,MMN (minimum bulk modulus)
TBDATA,18,MMX (maximum bulk modulus)
TBDATA,19-28,EC0-EC9 (cold compression energy constants)
When using the Steinberg Model, you can define one of three types of equations of state: Linear Polynomial
(EOSOPT = 1), Gruneisen (EOSOPT = 2), and Tabulated (EOSOPT = 3). Each has its own set of required constants
that are input starting with position (LOC) 29 of the TBDATA command. See the descriptions of the three
equation of state models in Section 7.2.6: Equation of State Models.
For an example input listing using the Gruneisen EOS, see Section B.2.25: Steinberg Gruneisen EOS Example:
Stainless Steel.
7.2.7. Discrete Element Models

7.2.7.1. Linear Elastic Spring Model
This model provides a translational or rotational elastic spring located between two nodes. Input the spring
elastic stiffness with the TB,DISCRETE,,,,0 and location 1 of the TBDATA command:
TB,DISCRETE,,,,0
TBDATA,1,KE (Elastic stiffness (force/displacement) or (moment/rotation))
7.2.7.2. General Nonlinear Spring Model

This model provides general nonlinear translational or rotational spring with arbitrary loading and unloading
definitions located between two nodes. Hardening and softening can be defined. Input the model parameters
with the TB,DISCRETE,,,,5 and locations 1 to 5 of the TBDATA command:
TB,DISCRETE,,,,5
TBDATA,1,LCDL (Load curve ID describing force vs. displacement or moment vs. rotation relationship for
loading)
TBDATA,2,LCDU (Load curve ID describing force vs. displacement or moment vs. rotation relationship for
unloading)
TBDATA,3,BETA (Hardening parameter)
735

TBDATA,4,TYI (Initial yield force in tension (= 0: Tensile and compressive yield with strain softening; 0:
Kinematic hardening without strain softening; = 1: Isotropic hardening without strain softening.))
TBDATA,5,CYI (Initial yield force in compression (< 0))
7.2.7.3. Nonlinear Elastic Spring Model

This model provides nonlinear elastic translational or rotational spring with arbitrary force/displacement or
moment/rotation dependency. Input the model parameters with the TB,DISCRETE,,,,3 and locations 1 and 2 of
the TBDATA command:
TB,DISCRETE,,,,3
TBDATA,1,LCID (Load curve ID (force/displacement or moment/rotation))
TBDATA,2,LCR (Optional load curve ID describing scale factor on force or moment as a function of relative
velocity or rotational velocity, respectively.)
7.2.7.4. Elastoplastic Spring Model

This model provides elastoplastic translational or rotational spring with isotropic hardening located between
two nodes. Input the spring elastic stiffness, tangent stiffness and yield force with the TB,DISCRETE,,,,2 and locations
1 to 3 of the TBDATA command:
TB,DISCRETE,,,,2
TBDATA,1,KP (Elastic stiffness (force/displacement) or (moment/rotation))
TBDATA,2,KT (Tangent Stiffness (force/displacement or moment/rotation))
TBDATA,3,FY (Yield (force) or (moment))
7.2.7.5. Inelastic Tension- or Compression-Only Spring Model

This model provides an inelastic tension or compression only, translational or rotational spring located between
two nodes. Optionally, the user can define unloading stiffness instead of the maximum loading stiffness. Input
the model parameters with the TB,DISCRETE,,,,7 and locations 1 to 3 of the TBDATA command:
TB,DISCRETE,,,,7
TBDATA,1,LCFD (Load curve ID describing arbitrary force/torque vs. displacement/twist relationship. Must
be defined in the positive force-displacement quadrant, regardless of whether the spring acts in tension or
compression.)
TBDATA,2,KU (Unloading stiffness. The maximum of KU and the maximum loading stiffness in the
force/displacement or the moment/twist curve is used for unloading.)
TBDATA,3,CTF (Compression or tension indicator: 1.0 is tension only, 0.0 is compression only (default), 1.0
is compression only.)
7.2.7.6. Maxwell Viscosity Spring Model

This model provides a three parameter Maxwell viscoelastic translational or rotational spring located between
two nodes. Optionally, a cutoff time with a remaining force/moment can be defined. Input the model parameters
with the TB,DISCRETE,,,,6 and locations 1 to 6 of the TBDATA command:
TB,DISCRETE,,,,6
TBDATA,1,Ko (Short time stiffness)
TBDATA,2,KI (Long time stiffness)
TBDATA,3,BETA (Decay parameter)
TBDATA,4,TC (Cutoff time, after which a constant force/moment is transmitted)
TBDATA,5,FC (Force/moment after cutoff time)
736

TBDATA,6,COPT (Time implementation option)
7.2.7.7. Linear Viscosity Damper Model

This model provides linear translational or rotational damper located between two nodes. Input the damping
constant with the TB,DISCRETE,,,,1 and location 1 of the TBDATA command:
TB,DISCRETE,,,,1
TBDATA,1,DC (Damping constant (force/displacement rate) or (moment/rotation rate))
7.2.7.8. Nonlinear Viscosity Damper Model

This model provides a nonlinear damping spring with arbitrary force/velocity or moment/rotational velocity
dependency located between two nodes. The load curve must define the response in the negative and positive
quadrants and pass through point (0, 0). Input load curve ID with the TB,DISCRETE,,,,4 and location 1 of the TBDATA
command:
TB,DISCRETE,,,,4
TBDATA,1,LCID (Load curve ID describing force vs. rate-of-displacement relationship or moment vs. rate-ofrotation relationship. The load curve must define the response in the negative and positive quadrants and
pass through point (0, 0). )
7.2.7.9. Cable Model

Use this model to realistically model elastic cables. No force will develop in compression. The force, F, generated
by the cable is nonzero if and only if the cable is in tension. The force is given by:
F = K x max(L,0.)
where L is the change in length
L = current length - (initial length - offset)
and the stiffness is defined as:
K=
E area
(initial length offset )
The area and offset are defined by means of the real constants for LINK167. For a slack cable, the offset should
be a negative length. For an initial tensile force, the offset should be positive. If a load curve is specified, Youngs
modulus is ignored and the load curve is used instead. The points on the load curve are defined as engineering
stress versus engineering strain, i.e., the change in length over the initial length.
Use the MP and EDMP commands to input the required values:
MP,DENS
MP,EX
EDMP,CABLE,MAT,Load Curve ID
For an example input listing, see Section B.2.26: Cable Material Example: Steel.
737
7.2.8. Other Models

7.2.8.1. Rigid Model
Rigid bodies are defined for a material type using the command EDMP. For example, to define a rigid body for
material 2, issue: EDMP,RIGID,2. All elements defined with the specified material number are then considered
to be part of that rigid body. This material number, along with the element type and real constant type numbers
of the elements, will define the PART ID of the rigid body. This PART ID can then be used to define loads and
constraints to the rigid body (as described in Chapter 4, Loading). Elements within a rigid body do not have to
be linked by mesh connectivity. Therefore, to represent more than one independent rigid body in a model,
multiple rigid material types must be specified. Two independent rigid bodies, however, cannot share a common
node.
Along with the EDMP command, you must use the MP command to specify Young's modulus (Ex), Poissons
ratio (NUXY), and density (DENS) for the rigid bodys material type. Realistic material property values are required
so that the stiffness of the contact surfaces can be calculated by the program. For this reason, never use unrealistically high values of Youngs modulus or density for a rigid body in an explicit dynamic analysis. A rigid body
cannot be stiffened because it is perfectly rigid.
Because the motion of the mass center of a rigid body is transferred to its nodes, constraints should not be applied
to rigid bodies using the D command. The constraints and initial velocities of one node of a rigid body will be
transferred to the center of mass of the body. However, if more than one node is constrained, it is hard to determine
which constraints will be used. To properly apply constraints to a rigid body, the translational (VAL1) and rotational (VAL2) constraint parameter fields of the EDMP command are used as shown below:
VAL1 - Translational constraint parameter (relative to global Cartesian coordinates)
No constraints (default)
Constrained X displacement
Constrained Y displacement
Constrained Z displacement
Constrained X and Y displacements
Constrained Y and Z displacements
Constrained Z and X displacements
Constrained X, Y, and Z displacements
VAL2 - Rotational constraint parameter (relative to global Cartesian coordinates)
No constraints (default)
Constrained X rotation
Constrained Y rotation
Constrained Z rotation
Constrained X and Y rotations
Constrained Y and Z rotations
Constrained Z and X rotations
Constrained X, Y, and Z rotations
For example, the command EDMP,RIGID,2,7,7 would constrain the rigid body elements of material 2 in all degrees
of freedom.
738

After defining a rigid body, you can assign inertia properties, mass, and an initial velocity vector to that body
using the EDIPART command. If you do not specifically define the inertia properties for a rigid body, the program
will calculate them based on the finite element model.
For an example input listing, see Section B.2.27: Rigid Material Example: Steel.
739
740
Chapter 8: Rigid Bodies

Using rigid bodies to define stiff parts in a finite element model can greatly reduce the computation time required
to perform an explicit analysis. When rigid bodies are defined, all of the degrees of freedom of the nodes in the
body are coupled to the body's center of mass. Hence, a rigid body has only six degrees of freedom regardless
of the number of nodes that define it. By default, the mass, center of mass, and inertia properties of each rigid
body are calculated from the bodys volume and density of elements. The forces and moments acting on the
body are summed from the nodal forces and moments at each time step. The motion of the body is then calculated,
and displacements are transferred to the nodes.
The following rigid body topics are available:
8.1. Defining Rigid Bodies
8.2. Specifying Inertia Properties
8.3. Loading
8.4. Switching Parts from Deformable to Rigid
8.5. Nodal Rigid Bodies
8.1. Defining Rigid Bodies

In ANSYS LS-DYNA, a rigid body that remains rigid for the duration of the analysis is designated as a material
model. Use the EDMP command to specify this type of rigid body. For information on defining rigid bodies as
a material model, see Chapter 7, Material Models.
In addition, two rigid bodies can be merged together so that they act as a single rigid body using the EDCRB
command. Unlike the definition of rigid bodies which are based on material numbers, rigid body constraints are
based on PART ID and a constraint equation number. Therefore, to place a rigid body constraint between two
bodies, the command EDCRB,ADD,NEQN,PARTM,PARTS should be issued, where NEQN is the constraint equation
reference number, PARTM is the master rigid body PART ID number and PARTS is the slave rigid body PART ID
number. Care should be taken not to issue more than one EDCRB command with the same NEQN value, as only
the last NEQN referenced will be used. When using the EDCRB command, the second rigid body noted becomes
absorbed into the first, so any subsequent references to the second body are meaningless.
8.2. Specifying Inertia Properties

By default, the program calculates the inertia properties of each rigid body. However, you may find it useful to
assign a specific center of gravity, mass, initial velocity (in a local or global system of coordinates) and a specific
inertia tensor to a rigid body instead of relying on the values calculated from the finite element model during
the solution. You can use the command EDIPART to assign any of these characteristics to the rigid body. The
command format is
EDIPART,PART,Option,Cvect,TM,IRCS,Ivect,Vvect,CID
where:
PART is the part ID for which the inertia is defined;
Option is the option to be performed (ADD, DELETE, or LIST);
Cvect is a vector containing coordinates of the center of the mass of the part;
TM is the translational mass;
IRCS is a flag for the inertia tensor reference coordinate system;
Ivect is a vector containing the components of the inertia tensor;
Vvect is a vector containing the initial velocity of the rigid body;
CID is a local coordinate system ID.
Chapter 8: Rigid Bodies

If you are working in the ANSYS LS-DYNA GUI, you can input all of the above parameters, including the required
array parameters, by picking Main Menu> Preprocessor> LS-DYNA Options> Inertia Options> Define Inertia.
For batch input, you must dimension (*DIM) and fill the array parameters prior to issuing the EDIPART command.
The example input below demonstrates how to define inertia properties for part number 2, which is a rigid body.
! Define parameter input
tm = 0.6300E-03
! Translation mass
ircs = 0
! Coordinate system flag
cid = 0
! Coordinate system ID
!
/prep7
!
! Create a model that includes multiple parts
! with part number 2 being a rigid body
.
.
.
edmp,rigid, . . .
! Define rigid body (by MAT number)
edpart,create
! Create part list
!
!Dimension required arrays
*dim,coord2,,3
*dim,velo2,,6
*dim,inert2,,6
!
! Fill arrays
coord2(1) = 0.2450E+01, 0.5000E+00, 0.5000E+00
velo2(1) = 0, 10.0E-7, 0
inert2(1) = 0.3150E-03, 0, 0, 0.2851E-03, 0, 0.2851E-03
!
EDIPART,2,ADD,coord2,tm,ircs,inert2,velo2,cid
!
8.3. Loading
As described in Chapter 4, Loading, displacements and velocities are applied to rigid bodies using the EDLOAD
command. It is important to note that all rigid body displacements and velocities are applied to a PART ID rather
than to a component of nodes. For example, a typical EDLOAD command may look like this:
EDLOAD,ADD,RBUX, ,2,PAR1,PAR2
This command defines a UX displacement load on the rigid body identified as PART number 2. The PART number
is input in the Cname field of the EDLOAD command. (Note that for other load types, a component name is input
in this field. However, for the rigid body load labels, a PART number must be input instead of a component name.)
The PART number must correspond to a rigid body that has been defined with the EDMP command.
8.4. Switching Parts from Deformable to Rigid

In some dynamic applications, long duration, large rigid body motions arise that are prohibitively expensive to
simulate if the majority of elements in the model are deformable. One such example would be an automotive
rollover where the time duration of the rollover would dominate the CPU cost relative to the impact that occurs
much later. To improve efficiency for this class of applications, ANSYS LS-DYNA offers the capability to switch a
subset of materials from a deformable state to a rigid state, and then back to deformable. By switching the deformable parts to rigid during the rigid body motion stage, you can achieve significant savings in computation
time.
Deformable/rigid switching is inherently related to performing a restart. You need to stop the analysis, define
the part switch, then restart the analysis again. Although you would not typically use part switching in the original analysis, you must set a flag in the new analysis input to let LS-DYNA know that all materials in the model
82
Section 8.5: Nodal Rigid Bodies

have the potential to become rigid bodies sometime during the calculation. Issuing the EDRD or EDRI command
in the original analysis will set this flag.
Note Parts that are defined as rigid using the command EDMP,RIGID are permanently rigid and cannot
be changed to deformable.
To switch a deformable part to a rigid part, issue
EDRD,D2R,PART,MRB
where PART is the part number, and MRB indicates a master rigid part to which the switched part should be
merged. (In the GUI, go to Main Menu> Solution> Rigid-Deformable> Switch.) If you do not wish to merge
the part to another rigid body, simply leave MRB blank. Once a deformable part is switched to rigid using the
above command, you can switch it back to deformable in a subsequent restart by issuing
EDRD,R2D,PART
In a new analysis, if no deformable materials are going to be switched to rigid but switching may occur in a
subsequent restart analysis, then issue EDRD,D2R with no additional arguments. (In the GUI, go to Main Menu>
Solution> Rigid-Deformable> Switch and select Initialize as the action.)
When you switch a deformable part to rigid, you can define the inertia properties for the rigid part by issuing
the EDRI command (Main Menu> Solution> Rigid-Deformable> Inertia Property). If you do not define the
inertia properties, they will be calculated by the program.
After a deformable-to-rigid switch, some constraints defined for the deformable part will become invalid. To
avoid instabilities in the calculations, you should use the EDRC command (Main Menu> Solution> Rigid-Deformable> Controls) to change the state of these constraint definitions. Please refer to the EDRC command
description for more details.
For rigid body switching to work properly, the choice of element formulation is critical. In the current LS-DYNA
implementation, Hughes-Liu shell and beam elements cannot be used with part switching because of the strain
and stress update algorithms used in these elements. The element formulations that cannot be used include
KEYOPT(1) = 1, 6, 7, and 11 for SHELL163 and KEYOPT(1) = 0 and 1 for BEAM161.
A special consideration for SOLID164 elements is that, upon switching from a rigid state to deformable, the element
stresses are zeroed out to eliminate spurious behavior.
8.5. Nodal Rigid Bodies

Unlike typical rigid bodies that are defined with the EDMP command, nodal rigid bodies are not associated with
a part number. A nodal rigid body is defined by the command EDCNSTR,ADD,NRB,COMP1, where COMP1 is a
node component. Nodal rigid bodies are typically used for modeling rigid (welded) joints in which different
flexible components (having different MAT IDs) act together as a rigid body. Because nodal rigid bodies are not
associated with a part number, other options that use rigid bodies discussed earlier in this chapter (such as loads
applied with the EDLOAD command ) cannot be used with a nodal rigid body. See Section 4.2.1: Constraints
and the EDCNSTR command for more information on using nodal rigid bodies.
83
84
Chapter 9: Hourglassing
Despite being robust for large deformations and saving extensive amounts of computer time, the one-point
(reduced) integration solid and shell elements used in ANSYS LS-DYNA are prone to zero-energy modes. These
modes, commonly referred to as hourglassing modes, are oscillatory in nature and tend to have periods that are
much shorter than those of the overall structural response (i.e., they result in mathematical states that are not
physically possible). They typically have no stiffness and give a zigzag appearance to a mesh (see Figure 9.1: Hourglass Deformations) known as hourglass deformations. The occurrence of hourglass deformations in an analysis
can invalidate the results and should always be minimized.
Figure 9.1 Hourglass Deformations
Undeformed mesh and deformed mesh with hourglassing effect

Hourglassing can affect brick and quadrilateral shell and 2-D elements, but not triangular shell, triangular 2-D,
or beam elements.
Good modeling practices normally prevent hourglassing from becoming significant. The general principles are
to use a uniform mesh and to avoid concentrated loads on a single point. Since one excited element transfers
the hourglassing mode to its neighbors, all point loads should be spread over an area of several neighboring
nodes. In general, refining the overall mesh will almost always significantly reduce the effects of hourglassing.
ANSYS LS-DYNA offers a number of internal hourglass controls. The idea behind these methods is (1) to add
stiffness which resists hourglass modes but not rigid body motions and linear deformation fields, or (2) to damp
velocities in the direction of hourglass modes.
One method for controlling hourglassing modes is to adjust the model's bulk viscosity. Hourglass deformations
are resisted by a structures bulk viscosity, which is automatically calculated by the program. It is possible to increase the bulk viscosity of the model by adjusting the linear (LVCO) and quadratic (QVCO) coefficients of the
EDBVIS command. It is generally not recommended, however, to dramatically change the default values of the
EDBVIS command because of adverse effects they will have on the global modes of the structure.
Another more generally applicable solution to hourglassing problems is to use the fully integrated formulations
of SHELL163 and SOLID164. Fully integrated elements will never experience hourglassing modes. However, these
options are more costly (in CPU time) than other element formulations, and they may lead to unrealistically stiff
results (locking) for problems involving incompressible behavior, metal plasticity, and bending. The locking is
remedied in SHELL163 by using an assumed strain field.
Chapter 9: Hourglassing
Hourglassing deformations can also be resisted by adding elastic stiffness to a model. Hourglassing can be a
problem with small displacement situations, particularly when dynamic relaxation is used. In these cases, it is
often beneficial to add elastic stiffness to the model instead of using bulk viscosity methods. This can be done
by increasing the hourglassing coefficient (HGCO) of the EDHGLS command. However, use care when increasing
these coefficients because they may over-stiffen the model's response in large deformation problems and cause
instabilities when HGCO exceeds 0.15.
The hourglass control methods discussed so far have been for the entire model. The final method of hourglass
control is to locally increase a portion of the model's stiffness by using the command EDMP,HGLS. For this
command, the material number, the hourglass control type (viscous or stiffness), the hourglass coefficient, and
bulk viscosity coefficients must be specified. (The defaults for hourglass coefficient and bulk viscosities should
be sufficient.) Using this method, hourglass control is specified for a given material and not for the entire model.
This allows resisting hourglass deformations in high risk areas of the model without dramatically changing the
stiffness characteristics of the entire model.
When performing an explicit dynamics analysis with reduced integration elements, it is always important to
determine whether hourglassing effects have significantly degraded the results. As a general guideline, the
hourglassing energy should not exceed 10% of the internal energy. The hourglass energy can be compared to
the internal energy by reviewing the ASCII files GLSTAT and MATSUM and can be plotted in POST26 (see
Chapter 12, Postprocessing). To make sure that hourglass energy results are reported in these files, the HGEN
field of the EDENERGY command must be set to 1.
92
Chapter 10: Mass Scaling

As discussed in Appendix A and shown in the figure below, the minimum time step size for explicit time integration
depends on the minimum element length, lmin, and the sonic speed, c. (The example depicted represents a 2-D
continuum.) Note that for a given set of material properties, the minimum time step size tmin, is controlled by
the smallest element dimension in the model (I2 in this case). Also, for a given mesh, the minimum time step size
is dependent upon sonic speed, which is a function of material properties (density, elastic modulus, and Poisson's
ratio). You can use the EDTP command to check the minimum time step size in your model.

l
l
tmin = min = 2
c
c
c=
E
(1 2 ) i
= Poissons ratio
= specific mass density
E = Youngs Modulus
In the ANSYS LS-DYNA program, you can control the minimum time step size by including mass scaling in the
analysis. It may be necessary to use mass scaling if the program calculated time step is too small. When mass
scaling is requested, element density is adjusted to achieve a user specified time step size:
2
t specified
(1 2 ) i i
li
E
i =
( t specified )2 i E
l2 i (1 2 )
i
in element i
The EDCTS command is used to specify mass scaling in ANSYS LS-DYNA. Use this command to apply mass scaling
to the program in one of two ways, depending on the value of the specified time step size DTMS:
If DTMS is positive, the same time step size is used for all elements, and mass scaling is applied to all elements.
If DTMS is negative, mass scaling is applied only to elements whose calculated time step size is smaller
that DTMS.
Of the two methods listed above, the second is typically more efficient and therefore recommended. Although
proper use of mass scaling will add a small amount of mass to the model and slightly change a structure's center
of mass, the CPU reduction it achieves far outweighs the minor errors introduced. For example, when using mass
Chapter 10: Mass Scaling

scaling, it is common to achieve a 50% decrease in CPU time while adding only .001% mass to the model. Care
must be taken not to add too much mass to a model in which inertial effects are significant.
The time step size calculated for the elements is multiplied by a scale factor (usually 0.9). The users input for the
mass scaling affects the time step size before scaling. To control the time step size after mass scaling, use the
TSSFAC parameter of the EDCTS command.
102
Chapter 11: Subcycling

Subcycling, also known as mixed time integration, can be used to speed up an analysis when element sizes
within a model vary significantly. Relatively small elements will result in using a small time step size for all of the
elements within a model, even the larger ones. If subcycling is turned on, elements are sorted based on their
time step size into groups. Each group is assigned a step size that is some even multiple of the smallest element
step size (see Figure 11.1: Time Step Sizes Before and After Subcycling). Thus, the minimum time step size is
increased for the smallest element and the groups with larger elements are processed every 2nd, 3rd, 4th, etc.,
step depending on their size. Subcycling is turned on in the program using the EDCSC command.
Figure 11.1 Time Step Sizes Before and After Subcycling

tmin determines time step size for all elements tfmin n*tmin
Elements are sorted three times by the program:

1.
By element number in ascending order.
2.
By part number for large vector blocks.
3.
By connectivity to ensure disjointedness for right hand side vectorization.
There are two advantages to using subcycling:
Greatly speeds-up problems with differences in element size.
Allows local mesh refinement without penalty.
Subcycling supports the following element classes and contact options:
Solid elements
Beam elements
Shell elements
Penalty based contact algorithms
Discrete elements are intentionally excluded since these elements generally contribute insignificantly to the
calculational costs.
Chapter 11: Subcycling

The interface stiffnesses used in the contact algorithms are based on the minimum value of the slave node or
master segment stiffness and, consequently, the time step size determination for elements on either side of the
interface is assumed to be decoupled; thus, scaling penalty values to larger values when subcycling is active can
be a dangerous exercise.
Nodes that are included in constraint equations, rigid bodies, or in contact with rigid walls are always assigned
the smallest time step sizes.
112
Chapter 12: Postprocessing

You can review ANSYS LS-DYNA results using the two ANSYS postprocessors, POST1 and POST26. Use POST1 to
view results for the entire model at specific time points, or to view animated results. Use POST26 to view the
results of a specific component of the model at a larger number of time points over a period of time. With an
explicit dynamic analysis, you will typically want to view animated results (POST1) and time-history results
(POST26).
Note Experienced LS-DYNA users may also use the LSTC postprocessor LS-POST. However, this processor
is not supported by ANSYS, Inc.
The following postprocessing topics are available:
12.1. Output Controls
12.2. Using POST1 with ANSYS LS-DYNA
12.4. Finding Additional Information
12.1. Output Controls

12.1.1. Results (Jobname.RST) vs. History (Jobname.HIS) Files
Results available for postprocessing depend on the information written to the Jobname.RST and Jobname.HIS
files via the EDRST and EDHTIME commands (Main Menu> Solution> Output Controls> File Output Freq).
Note the distinction between the Jobname.RST and the Jobname.HIS files: The Jobname.RST file, used
primarily in POST1 processing, contains solutions for the entire model, but captured at a relatively small number
of time points. Generally, the Jobname.RST file contains just enough solutions to animate the results. The Jobname.HIS file, used in POST26, contains solutions captured at a relatively large number of time points, but for
only a portion of the model. (Capturing a large number of solutions for the entire model would quickly fill up
disk space.) As a comparison, the number of time steps in the Jobname.RST file is usually less than 100; in the
Jobname.HIS file, it is usually 1000 or more.
12.1.2. Creating Components for POST26

Before reviewing results in POST26, you may create element and/or nodal components within your model. For
example, during either the PREP7 or the SOLUTION phases of your analysis, choose a set of elements for which
you want to review results. Create a component consisting only of those elements. You can also create a component consisting of a specified set of nodes. These components can be created via the GUI or by issuing the
following commands:
ESEL,S,MAT,,1 !Select elements of material 1.
CM,elm1,elem !Create element component elm1.
NSLE !Select nodes from elements.
CM,nod1,node !Create nodal component nod1.
Element and nodal components should be limited to conserve disk space. For details on creating components,
see the ANSYS Basic Analysis Guide.
12.1.3. Writing the Output Files for POST26

Before using POST26 to review results, you must direct ANSYS LS-DYNA to write the appropriate information to
the Jobname.HIS and other output files. While in either the PREP7 or SOLUTION phases of your analysis, specify
the number of time steps, the element and nodal components to be analyzed, and the ASCII files to be written.

You can specify this information via the GUI (Main Menu> Solution> Output Controls) or by issuing the following
commands:
EDHTIME,NSTEPS
EDHIST,elm1
EDHIST,nod1
EDOUT,GLSTAT
EDOUT,MATSUM
EDOUT,SPCFORC
EDOUT,RCFORC
EDOUT,SLEOUT
EDOUT,NODOUT
EDOUT,RBDOUT
!Specify the number of time steps

! for time-history results.
!Specify the name of the element component.
!Specify the name of the nodal component.
!Write ASCII file GLSTAT (Global time step
! and energy statistics).
!Write ASCII file MATSUM (Energy information
! for each PART).
!Write ASCII file SPCFORC (Single point
! constraint (Reaction) forces).
!Write ASCII file RCFORC (Resultant interface
! force data).
!Write ASCII file SLEOUT (Sliding interface
! energy).
!Write ASCII file NODOUT (Nodal data).
!Write ASCII file RBDOUT (Rigid body data).
Note to LS-DYNA Users--If you are familiar with other ASCII files created by the LS-DYNA program, you may want
to issue EDOUT,ALL to write all possible ASCII files.

The first step in POST1 is to read data from the results file into the database. If the database does not already
contain model data, issue the RESUME command to read the database file (or in the GUI pick Utility Menu>
File> Resume Jobname.db). Then issue the SET command to read in the desired set of results (or in the GUI
pick Main Menu> Postproc> Read Results> By Load Step or By Time/Freq).
The typical POST1 operations available in ANSYS are also available in ANSYS LS-DYNA. You can display the deformed shape, show contour and vector displays, and obtain tabular listings just as you would in ANSYS. For
details on these and other POST1 postprocessing functions, see the ANSYS Basic Analysis Guide.
When using POST1, you can rotate stress results into a defined coordinate system using the RSYS command.
Stress data will be rotated during printout, display, or element table operations. RSYS supports stress output for
all explicit element types except BEAM161, COMBI165, and composite SHELL163 (KEYOPT(3) = 1). To use RSYS
in models that contain these element types, you must unselect the unsupported elements before issuing the
RSYS command.
Note Strain data results cannot be rotated for any explicit elements. If you request strain results when
RSYS is not set to the global Cartesian coordinate system, the printing or plotting command will be ignored.
When you display results in POST1, failed elements will automatically be removed from the display. Failed elements
are elements that exceed the specified failure criteria (for example, failure strain). In some cases, the mesh connectivity may appear to be lost due to the number of failed elements in the model at a given time step. However,
the elements are still in the selected set, and the model will behave as expected physically. (During the LS-DYNA
solution, failed elements are removed from the solution after they fail.)
12.2.1. Animating Results

You can also display animated shapes using POST1 capabilities; for example, you could animate the element
centroidal stresses superimposed on the deformed shape. After you have read in a set of results from the Jobname.RST file, pick Utility Menu> PlotCtrls> Animate> Over Results. Then use the push-button animation
controls in the dialog box to start, stop, resume, and otherwise control the animation.
122
Section 12.2: Using POST1 with ANSYS LS-DYNA

If you are using command input, use the appropriate plotting command (for example, PLESOL) followed by the
ANDATA command (as shown below).
PLESOL,Item,Comp
ANDATA
This command method reads in all results data and animates the plot. For faster plotting, use the INRES command
to select specific types of solution data before issuing the above commands:
INRES,Item
12.2.2. Element Output Data

Remember that the results available for POST1 processing differ from element to element. For a complete description of output data for each explicit dynamics element (LINK160, BEAM161, PLANE162, SHELL163, SOLID164,
COMBI165, MASS166, LINK167, SOLID168), see the ANSYS Elements Reference.
For the PLANE162, SHELL163, SOLID164, and SOLID168 elements, the available strain output depends on the
type of material models used. For all material models, total strains (total strain components, total principle strains,
total strain intensity, and total equivalent strain) are always available. For the following material models, elastic
strains are also available.
Isotropic Elastic
Bilinear Isotropic
Bilinear Kinematic
Viscoelastic
Powerlaw Plasticity
For the following plastic material models, equivalent plastic strain is available. However, plastic strain components
and principal plastic strains are not available.
Bilinear Isotropic
Bilinear Kinematic
Temperature Dependent Bilinear Isotropic
Steinberg
Powerlaw Plasticity
Geological Cap
Bamman
Zerilli-Armstrong
Modified Piecewise Linear Plasticity
123

Below are some additional notes about the explicit dynamics elements:
With SOLID164, stress and strain solutions are saved only at the element centroid, regardless of whether
you use 1-point or 8-point integration (KEYOPT(1)).
With BEAM161 and SHELL163 elements, stress and strain solutions are saved only at the element center
(for each layer for shell elements), regardless of whether you use full or reduced integration.
When using BEAM161 and SHELL163 elements, you must specify both the number of integration points
for which you want ANSYS LS-DYNA to calculate the results (using real constants) and the number of these
points for which you want results stored (using the EDINT command). By default, the number of integration
points (NIP) is set to 2 for shell elements; you can specify up to 5 points (layers) for the Gauss integration
rule. For more than 5 layers, you need to use either the trapezoidal integration rule or the user-defined
integration rule. (Note--The trapezoidal rule is not recommended for less than twenty integration points,
especially in bending.) Also by default, the number of layers you specify using EDINT,SHELLIP is set to 3
for shell elements. If NIP = 2, the results at the bottom correspond to integration point 1, the results at
the top correspond to integration point 2, and the results at the middle are an average of the top and
bottom results. For beam elements, results are saved for 4 integration points by default (BEAMIP = 4 on
the EDINT command). For the resultant beam formulation (KEYOPT(1) = 2), there is no stress output, regardless of the BEAMIP setting. If you set BEAMIP = 0, no stress output is written for any of the beam elements. In this case, the beams will not appear in any POST1 plots because the program assumes they are
failed elements.
With SHELL163 elements use the LAYER,NUM command to specify the layer for which you want results.
Layers are numbered starting from the bottom layer and moving up for stress data. However, for strain
data, layer 1 is the bottom layer and layer 2 is the top layer, no matter how many layers exist. Strain information is saved only at these two layers. Further, data is only available at the middle of the layer, not at the
top or bottom surface of a layer (or the element, for that matter). To get results close to the element surface,
specify a large number of integration points through the thickness of the shell. However, saving data for
all of these layers can become expensive.
SHELL163 is not affected by the SHELL,LOC command. By default, output data for the top layer is displayed
when plotting data (PLNSOL, PLESOL), but data for the top and bottom layers appears on printed results.
12.2.3. Postprocessing after Adaptive Meshing

If you use adaptive meshing in an explicit dynamic analysis, the analysis will typically produce multiple results
files. The procedure for postprocessing these files is slightly different than postprocessing a single results file.
Adaptive meshing can be used in ANSYS LS-DYNA to refine the mesh of shell elements during a large deformation
analysis (see Section 3.5: Adaptive Meshing for details on how to use adaptive meshing). When adaptive meshing
is included in an analysis, the number of elements in the model may increase during the solution as the mesh is
improved to conform to a user specified element aspect ratio. As the number of elements in the model changes,
the extension of the results file is updated to represent a new finite element mesh for the model.
The name of the results file extension is incremented as follows:
Jobname.RS01 - First mesh refinement
Jobname.RS02 - Second mesh refinement
Jobname.RS03 - Third mesh refinement
etc.
In order to postprocess multiple results files, you must specify the appropriate filename and extension (FILE
command) to examine a particular time step. You should not resume the original model database (.DB file), as
124
Section 12.3: Using POST26 with ANSYS LS-DYNA

the mesh in that database will not match what is in the results files. The sample input below demonstrates a
typical postprocessing session.
/POST1
FILE,Jobname,RS01 !
!
SET,LIST
!
!
SET,1,2
!
!
PRNSOL,. . .
!
Specify the results file for the first

mesh refinement
List out the time steps that were written
for the first mesh refinement
Set the results to a particular time
step value
Print or display the results
Caution: The element attributes are not saved on the RSnn file. Therefore, care must be taken when
using select logic during postprocessing with these files. Also, you should not attempt to perform a
solution using only the information contained in these files, nor should you attempt to save a new
database from these files.
You should not use the standard method for animating results (ANDATA command) after adaptive meshing
because it is only valid for animating a single results file (Jobname.RST). Instead, you must use the ANMRES
command to animate results when multiple results files are present (Jobname.RS01, Jobname.RS02, etc.).
Before starting the animation, you must use the /FILNAME command to specify the jobname that the results
will be animated across. Because results are animated across multiple files with the same jobname, simply specifying the FILE command will not allow multiple results animations.
The following sample input demonstrates a typical animation session using ANMRES. In this example, the jobname
is 'impact' and the animation runs from results file impact.rs01 to impact.rs10. The results file increment is two
and the time delay is 0.5 seconds.
/FILNAME,impact
/POST1
FILE,impact,rs01
SET,last
PLNSOL,s,eqv
ANMRES,.5,1,10,2
In the GUI, the ANMRES command is accessed from the utility menu: Utility Menu> PlotCtrls> Animate> Over
Results. Choose the options to animate over multiple results files and then use the push-button animation
controls in the dialog box to start, stop, resume, and otherwise control the animation.

The typical POST26 operations available in ANSYS are also available in ANSYS LS-DYNA. You can obtain two types
of results using POST26:
Nodal and element solutions
Miscellaneous output data, such as different energies, work, and reaction forces, stored in ASCII files
(GLSTAT, RCFORC, SLEOUT, MATSUM, SPCFORC, NODOUT, and RBDOUT)
Remember when reviewing results in POST26 that you must use the Jobname.HIS file to have adequate results
for a useful time-history record. The Jobname.HIS file must be loaded into POST26 using the FILE command
(Main Menu> TimeHist Postpro> Settings> File):
FILE,Jobname,HIS
If you do not load the Jobname.HIS file into POST26, the results will be read from the Jobname.RST file, which
may provide inadequate data for time-history postprocessing. (This is because the EDHTIME setting is usually
much larger than the EDRST setting.)
125
12.3.1. Nodal and Element Solutions

POST26 works with tables of result items versus time, known as variables. Each variable is assigned a reference
number, with variable number 1 reserved for time. To generate results, define and then graph or list the variables
via the GUI (Main Menu> TimeHist Postpro) or by using the following commands:
NSOL
Specifies nodal data be stored from the results file
ESOL
Specifies element data be stored from the results file
STORE
Stores data into variables
PLVAR
Displays results in graph format
PRVAR
Lists variables vs. time
Element and nodal solution results can be obtained from both the Jobname.HIS and Jobname.RST. Care must
be taken when reading from the .HIS file, however, such that only nodes and elements included in the component
specified by the EDHIST,NCOMP are specified.
For storing and displaying solutions at different integration points for the BEAM161 and SHELL163 elements,
use the LAYERP26 (Main Menu> TimeHist Postpro> Define Variables) and the STORE command (Main Menu>
TimeHist Postpro> Store Data) for each layer, as shown below for SHELL163:
LAYERP26,num1
ESOL
NSOL
STORE,MERGE
LAYERP26,num2
ESOL
NSOL
STORE,MERGE
!Specify the first layer number

!Store element data for first layer
!Store nodal data for first layer
!Store results for first layer
!Specify the next layer number
!Store element data for next layer
!Store nodal data for next layer
!Store results for next layer (in addition to
! results for first layer)
.
.
.
LAYERP26,numn
ESOL
NSOL
STORE,MERGE
!Specify the nth layer number

!Store element data for nth layer
!Store nodal data for nth layer
!Store results for nth layer (in addition to
! results for previous layers)
PRVAR (or PLVAR, etc.) !Plot results for all layers
You must specify STORE,MERGE for each layer, or your final plot will contain data from the last layer only.
12.3.2. Reading ASCII Files for Miscellaneous Output Data

To review information written to the GLSTAT, MATSUM, SPCFORC, RCFORC, SLEOUT, NODOUT, and RBDOUT
files, first issue FILE,Jobname,HIS (Main Menu> TimeHist Postpro> Settings> File) to obtain the time information, then issue the EDREAD command (Main Menu> TimeHist Postpro> Read LSDYNA Data). After EDREAD,
you must issue the STORE command (Main Menu> TimeHist Postpro> Store Data) to store the data in time
history variables. Once stored, the data can be analyzed using standard POST26 procedures.
12.3.3. Data Smoothing

If youre working with noisy data (such as an earthquake excitation), you may want to smooth that data to a
smaller set of data that provides an accurate approximation of the data points.
Four arrays are required for smoothing data. The first two contain the noisy data from the independent and the
dependent variables, respectively; the second two will contain the smoothed data (after smoothing takes place)
from the independent and dependent variables, respectively. You must always create the first two vectors (*DIM)
126
Section 12.4: Finding Additional Information

and fill these vectors with the noisy data before smoothing the data. If you are working in interactive mode,
ANSYS automatically creates the third and fourth vector, but if you are working in batch mode, you must also
create these vectors (*DIM) before smoothing the data (ANSYS will fill these with the smoothed data).
Once these arrays have been created, you can smooth the data using the EDNDTSD command (Main Menu>
TimeHist Postpro> Smooth Data). You can choose to smooth all or some of the data points using the DATAP
field, and you can choose how high the fitting order for the smoothed curve is to be using the FITPT field. DATAP
defaults to all points, and FITPT defaults to one-half of the data points. To plot the results, you can choose to
plot unsmoothed, smoothed, or both sets of data.
12.4. Finding Additional Information

For more information on postprocessing, including POST1 and POST26, see the ANSYS Basic Analysis Guide. For
more information on any of the commands discussed in this chapter, see the appropriate discussion in the ANSYS
Commands Reference.
127
128
Chapter 13: Restarting

Restarting means performing an analysis that continues from a previous analysis. The restart can begin from the
end of the previous analysis or from partway through it. Reasons for doing a restart include:
To complete an analysis that was terminated or that requires more than the user-defined CPU limit.
To perform an analysis in stages and monitor results at the end of each stage.
To diagnose an analysis that went wrong.
To change the model and continue calculations.
The ability to perform an analysis in stages provides great flexibility for explicit dynamic applications. After the
completion of each stage, a restart "dump" file is written. This file contains all information required to continue
the analysis. Results can be checked at each stage by postprocessing the output in the normal way. You may
then modify the model and continue the analysis. For example, you may decide to delete excessively distorted
elements, materials that are no longer important, or contact surfaces that are no longer needed. You may also
choose to change loading or consider other materials that were not present in the previous analysis. You can
also change output frequencies of the various results files. By continuously monitoring and fine-tuning the
analysis, you can increase efficiency and reduce the chance of wasting computer time on an incorrect analysis.
Restarting can also be used to diagnose an analysis that encountered problems. You can restart at a point before
the problem (numerical difficulties or error message) occurred, and request more frequent outputs to the results
files. By viewing the development of the error, you may be able to identify where the first symptoms appeared
and what caused them.
The following restart topics are available:
13.1. The Restart Dump File
13.2. The EDSTART Command
13.3. Effect on Output Files
13.1. The Restart Dump File

By default, LS-DYNA writes a restart "dump" file (d3dump) at the end of each analysis. The dump file is a binary
file that contains a complete LS-DYNA database for use in a restart. You can use the EDDUMP command to request
that additional restart dump files be written at a specified interval during the analysis. This will give you more
options for choosing a point from which to initiate a restart. The restart files are written sequentially as d3dump01,
d3dump02, etc. Care should be taken not to request too many dump files because they can be quite large.
13.2. The EDSTART Command

The EDSTART command specifies the status (new or restart) of an explicit dynamic analysis (in the GUI, pick
Main Menu> Solution> Analysis Options> Restart Option). There are four analysis types available: a new
analysis (the default), a simple restart, a small restart, or a full restart.
13.2.1. A New Analysis

For a new analysis, you can use the EDSTART command to change the memory to be used (for example, you
can increase the value if more memory is required by the LS-DYNA solver) or to change the scale factor for binary
file sizes.
13.2.2. A Simple Restart

A simple restart is one for which the database (Jobname.DB) has not been changed. You would typically run a
simple restart when the ANSYS LS-DYNA solution was prematurely interrupted by a user defined CPU limit or by
issuing a sense switch control SW1 (after CTRL-C). For a problem which was prematurely interrupted, enter the
solution processor and issue EDSTART,1,,,d3dumpnn followed by the SOLVE command. The analysis will then
be continued and all results will be appended to the results files Jobname.RST and Jobname.HIS.
13.2.3. A Small Restart

Use a small restart when minor changes to the database are necessary. For this type of restart, you must issue
the command EDSTART,2,,,d3dumpnn followed by any commands appropriate for changing the database, then
issue the SOLVE command. The types of changes you can make to the database during a small restart are listed
below.
Reset the termination time (TIME)
Reset the output file interval (EDRST, EDHTIME)
Specify additional ASCII files for output (EDOUT)
Set more displacement constraints (D)
Change initial velocities (EDVEL, EDPVEL)
Change loading curves (EDCURVE)
Change LS-DYNA numerical controls such as:
global mass damping (EDDAMP)
dynamic relaxation control (EDDRELAX)
contact small penetration control (EDSP)
time step control (EDCTS)
Change the termination criteria (EDTERM)
Delete, deactivate, or reactivate contact entities (EDDC)
Delete elements (EDELE)
Clear meshes (LCLEAR, ACLEAR, VCLEAR)
Change selected set of parts (PARTSEL)
Switch parts from rigid-to-deformable or deformable-to-rigid (EDRD, EDRC)
Change the restart dump file frequency (EDDUMP)
Only the commands mentioned above can be used in a small restart analysis (refer to the ANSYS Commands
Reference for details on their usage). Because some of these commands are applicable to a restart as well as a
new analysis, it is important that you issue EDSTART,2 first so that subsequent commands are processed correctly
for the restart.
In a small restart, you should generally extend the time of the calculation (TIME command). If the previous analysis (new or restart) finished at the specified end-time and no new time is input for the subsequent restart, the
restart analysis will stop immediately with only one step. You may also need to modify termination criteria that
were set in the previous analysis using the EDTERM command. If the previous analysis terminated due to one
132
Section 13.2: The EDSTART Command

of these criteria, that specific criterion must be modified so that it will not cause the restart to terminate immediately.
In some cases, the usage of a command in the restart depends on initial settings in the original analysis. For example, in order to perform a rigid-deformable switch in the restart analysis, you must issue the EDRD command
in the original analysis, even if no switch is being made. Furthermore, inertia properties for parts that are switched
in the restart must be specified in the original analysis (EDRI command). Another example is mass scaling; in
order to use mass scaling (EDCTS command) in the restart, mass scaling must be active in the original analysis.
Limitations related to restarts are discussed in each restart related command description.
When you issue the SOLVE command to initiate the restart analysis, ANSYS LS-DYNA creates a text file named
Jobname.R which is used as input to LS-DYNA. This file contains only the changes that were made to the model
for the restart. If you want to run LS-DYNA directly, you can use the EDWRITE command to output this file, then
specify it as the input file on the LS-DYNA execution command.
The results for the small restart analysis will be appended to all the results files. The restart solutions are numbered
as load step 2, 3, etc. in Jobname.RST. In other results files, results are appended according to the time value.
Restart dump files (d3dumpnn) are consecutively numbered beginning from the last number. (All modifications
to the database during the restart will be reflected in subsequent restart dump files.)
Note When postprocessing the restart results, take care not to select parts that were unselected in
the restart analysis. If you do select such parts, there will be a mismatch in the database since the associated
element definitions still exist, but no postprocessing data is written for the unselected parts.
After the first small restart analysis, you may choose to perform an additional restart, or even a series of restart
analyses. Be sure to issue EDSTART,2 with a different d3dump file at the beginning of each restart. The general
procedure for multiple restarts is outlined below.
1.
Create the initial model and run a new analysis.
2.
Postprocess the results.
3.
Issue EDSTART,2 with an appropriate d3dump file.
4.
Issue the commands needed to change the model.
5.
Issue the SOLVE command
6.
Postprocess the restart analysis results.
7.
Repeat step 3-6, as needed.
13.2.4. A Full Restart

A full restart is appropriate when many changes to the database are needed. For example, you may need to
consider more materials, to remove portions of the model, or to apply different loading conditions.
To initiate a full restart, you must issue the command EDSTART,3 to indicate that subsequent commands apply
to the full restart. For example, assume that the previous analysis was run using the input file Jobname.K, and
it produced a restart dump file named d3dump01. You would issue the command EDSTART,3,,,d3dump01,
then make any necessary changes to the model using commands available in the ANSYS LS-DYNA product. (A
few of the ANSYS LS-DYNA commands are not supported in a new restart; these are discussed below.)
When the EDSTART,3 command is issued, the jobname is automatically changed to Jobname_01 to avoid
overwriting the previous results and database. In a full restart, LS-DYNA creates entirely new results files instead
of appending to the existing results (as is done in other types of restart analyses). The advantage of the full restart
is that the changed database and result files match each other.
133

An important step in performing a full restart is specifying stress initialization via the EDIS command. You must
carry forward some results (deformed nodal positions and stresses/strains) from the previous analysis. Typically,
you would specify stress initialization for some or all of the parts that will be kept in the full restart. To do this,
you must issue the command EDIS,ADD,PIDN,PIDO for each part that will be initialized. If part IDs have changed
in the full restart due to changes in the model, you need to specify the new part ID in the PIDN field, and the old
part ID from the previous analysis in the PIDO field. If you issue EDIS with no arguments, stress initialization is
performed for all parts that were defined in the previous analysis (that is, parts having the same part ID); this
option is only appropriate if part IDs have not changed and you want all parts to be initialized.
When you issue SOLVE to initiate the full restart solution (or when you issue EDWRITE), the complete database
is written as an LS-DYNA input file, Jobname_01.K. When LS-DYNA starts execution, it uses the information
contained in files Jobname_01.K and d3dump01 to initialize any portions of the model which were specified
by EDIS commands. The deformed positions and velocities of the nodes on the elements of each part, and the
stresses and strains in the elements (and the rigid body properties, if the material of the part is rigid) are assigned
at this time.
Note Portions of the model that are not initialized will have no initial strains and stresses. However, if
initialized and non-initialized parts share common nodes, those nodes will be considered by the initialized
part. This will cause a sudden strain in the non-initialized part.
During the initialization, it is assumed that each part being initialized has the same characteristics (that is, the
same number of elements, in the same order, with the same topology) in the full restart analysis (Jobname_01.DB)
as it did in the previous analysis (Jobname.DB). If this is not the case, the part cannot be initialized. (Note that
the parts may have different identifying numbers, as stated above.) To avoid a mismatch in parts, follow these
recommendations when creating or changing the model:
If you anticipate the need to delete certain elements from the model in a future full restart analysis, use
a different element type number, material number, or real constant number for those elements in the
original analysis, even if they share the same attributes with other elements in the model. This will cause
a unique part number to be assigned to those elements so that they may be deleted later without altering
the makeup of any other parts in the model.
If you need to add more elements to the model during the full restart, use a different element type number,
material number, or real constant number for those elements, even if they share the same attributes with
other elements in the full restart analysis. Again, this will cause a unique part number to be assigned to
the new elements, and previously existing parts will not be altered.
If you do not follow the above guidelines, you may inadvertently create parts in the restart analysis that do not
match parts from the previous analysis. In this case, stress initialization may fail for those parts.
For discrete elements (COMBI165), the initialization is all or nothing. If you initialize any discrete elements in
the full restart, all discrete elements will be initialized.
Although you can change almost any aspect of the analysis in a full restart, a few features are not supported or
are supported in a limited way, as described below.
Contact specifications: You cannot add or delete any contact specifications (EDCGEN and EDDC) in the
full restart. However, you can list contact specifications (EDCLIST) that were defined in the previous analysis.
Initial velocities: You cannot change initial velocities (EDVEL and EDPVEL) in the full restart. For any portions
of the model that are carried over from the previous analysis, the velocities at the beginning of the full
restart will be the same as the velocities at the end of the previous analysis. You cannot define initial velocities for new nodes or parts added in the full restart; the initial velocity of new model entities are assumed
134
Section 13.3: Effect on Output Files

to be zero. You can list the initial velocities defined in the previous analysis by using the commands
EDVEL,LIST and EDPVEL,LIST.
Adaptive meshing: Adaptive meshing (EDADAPT and EDCADAPT) is not supported in a full restart. In
addition, a full restart is not possible if adaptive meshing was used in the previous analysis.
Mass scaling: As in a small restart, mass scaling (EDCTS) is only supported in a full restart if it was active
in the previous analysis.
You can perform multiple full restart analyses and you can mix full restart analyses with other restart analysis
types (simple or small). Just issue EDSTART,3 with a different d3dump file at the beginning of each full restart.
Each time EDSTART,3 is issued, the jobname will be changed from the current jobname to Jobname_nn (nn =
01, 02,) automatically.
13.3. Effect on Output Files

For both the simple restart and small restart, results are appended to the output files of the previous analysis.
For a simple restart, all outputs are shown as substeps of load step 1 in the Jobname.RST file (similar to a new
analysis). For a small restart, different stages of restarts are shown as different load steps in Jobname.RST. For
both the simple restart and small restart, the history file (Jobname.HIS) and ASCII files (glstat, matsum, etc.) are
appended with consecutive time markings.
In a full restart, new results files are created using the numbered jobname of the restart (Jobname_nn.RST and
Jobname_nn.HIS). ASCII output files, however, are not renamed and will be overwritten. If you need to keep
any ASCII output files from the previous analysis, you must save them to another name prior to issuing SOLVE
in the full restart. Time is continuous and is not reset to zero in any of the output files (Jobname_nn.RST, Jobname_nn.HIS, glstat, matsum, etc.). The results in Jobname_nn.RST are saved as substeps of load step 1.
135
136
Chapter 14: Explicit-to-Implicit Sequential

Solution
The simulation of some engineering processes requires the capabilities of both implicit (ANSYS) and explicit
ANSYS LS-DYNA analyses. To solve these problems, you need to use both solution methods, e.g., an explicit
solution followed by an implicit solution or vice-versa. ANSYS LS-DYNA is an explicit dynamics program intended
to solve short duration dynamic problems. If an engineering process contains phases that are essentially static
or quasi-static (such as a preload before a dynamic phase or a springback after a metal forming phase), then
these phases are best analyzed using the ANSYS implicit code. Procedures combining the ANSYS implicit solver
with the ANSYS LS-DYNA explicit solver provide an extremely powerful tool that can be used to simulate many
complex physical phenomena. In this chapter, we will describe explicit-to-implicit procedures, while implicit-toexplicit processes are outlined in Chapter 15, Implicit-to-Explicit Sequential Solution.
The following explicit-to-implicit analysis topics are available:
14.1. Performing an Explicit-to-Implicit Sequential Solution
14.2. Troubleshooting a Springback Analysis
14.1. Performing an Explicit-to-Implicit Sequential Solution

In sheet metal forming operations, springback deformation is an essential parameter that significantly complicates
the design of forming tools. Springback deformation is best defined as the dimensional change of the formed
part from that of the die which occurs from elastic deformations during unloading. In most dynamic metal
forming operations, the highly nonlinear deformation processes tend to generate a large amount of elastic strain
energy in the blank material. This elastic energy, which becomes stored in the blank while it is in dynamic contact
with the die components, is subsequently released after the forming pressure is removed. This release of energy,
which is the driving force for springback, generally causes the blank to deform towards its original geometry.
Hence, the final part shape in a sheet metal forming process not only depends upon the contours of the dies,
but also on the amount of elastic energy stored in the part while it is being plastically deformed. Since the amount
of elastic energy stored in a part is a function of many process parameters such as material properties and interfacial loads, predicting springback during forming is extremely complicated. This poses significant problems to
designers and analysts who must accurately assess the amount of springback that will occur during a forming
process so that a final desired part shape can be obtained. Herein lies the great advantage of performing an explicit-to-implicit sequential solution. By simulating the dynamic forming process explicitly, and then modeling
the springback deformations implicitly, stringent design tolerances can be attained for sheet metal forming
operations.
In an explicit-to-implicit sequential solution, you must first run the ANSYS LS-DYNA program to simulate the
metal forming process. In the explicit metal forming analysis, the sheet metal blank being deformed must consist
of SHELL163 and/or SOLID164 elements. Then the deformed shape, stresses, and thicknesses of these elements
only are transferred into the corresponding implicit ANSYS elements (SHELL181 and SOLID185). Once valid
boundary conditions are specified on the blank material, implicit simulation of the elastic springback within the
workpiece can be performed.
Note Only elastic properties are considered in the implicit phase. This implies that the analysis is purely
elastic unloading from the formed state (configuration). The displacements are interpreted as the changes
in these quantities from the formed configuration.
A detailed description of the explicit-to-implicit solution procedure follows.
Chapter 14: Explicit-to-Implicit Sequential Solution

1.
Run the explicit analysis as described earlier, using Jobname1. You must use SHELL163 and/or SOLID164
to model the working piece in order to analyze the springback effect in a subsequent ANSYS implicit
analysis. In addition, you must use one of the following element formulations for the SHELL163 elements:
KEYOPT(1) = 1, 2, 6, 7, 8, 9, 10, 11, or 12 (10 and 12 are recommended). Solve and finish the analysis.
You should always check your explicit analysis solution from ANSYS LS-DYNA carefully before proceeding
with the ANSYS implicit analysis. Specifically, check whether there is any undesirable dynamic effect left
in the structure at the end of the explicit run (using POST26).
2.
Save the explicit analysis database to file Jobname1.DB.

Command(s): SAVE
GUI: Utility Menu> File> Save as
Note If you do not save your Jobname1.DB file at this point, then the database for this explicit
run will not be saved. Only the database file for the subsequent implicit run will be saved.
3.
Change to Jobname2 to prevent the explicit results files from being overwritten.
Command(s): /FILNAME,Jobname2
GUI: Utility Menu> File> Change Jobname
4.
Reenter the preprocessor.

Command(s): /PREP7
GUI: Main Menu> Preprocessor
5.
Convert explicit element types to corresponding companion implicit element types. (Note that the 2-D
explicit element, PLANE162, and composite shell element SHELL163 with KEYOPT(3) = 1, cannot be used
in a sequential solution.) The companion explicit-implicit element type pairs are:
Explicit Element Type
Implicit Element Type
LINK160
LINK8
BEAM161
BEAM4
SHELL163
SHELL181
SOLID164
SOLID185
COMBI165
COMBIN14
MASS166
MASS21
LINK167
LINK10
Although all explicit element types are converted, only SHELL163 and SOLID164 data (stresses and
thicknesses for SHELL163, and stresses for SOLID164) are transferred to SHELL181 and SOLID185 (via the
RIMPORT command; see step 12).
Command(s): ETCHG,ETI
GUI: Main Menu> Preprocessor> Elem Type> Switch Elem Type
6.
142
Redefine the key options, real constants, material properties, boundary conditions, and loading values
on any implicit elements that are converted from explicit element types. (For SHELL163 elements that
were converted to SHELL181, you do NOT need to redefine the real constants, but you do need to redefine
the other values. For SOLID164 elements that were converted to SOLID185 elements, you need to specify
the uniform reduced integration option by setting KEYOPT(2) = 1.) The TYPE, REAL, and MAT numbers
from the explicit elements are retained, but the actual key option and real constant values are reset to
zero or the default settings.
Command(s): KEYOPT, R, MP, etc.
GUI: Main Menu> Preprocessor> Element Type/Real Constants/Material Properties/Loads
Section 14.1: Performing an Explicit-to-Implicit Sequential Solution

Note Only linear elastic material properties (as specified with the MP command) can remain
active in the ANSYS implicit phase. Delete any inelastic material properties (as specified with the
TB command) from the ANSYS LS-DYNA run.
7.
Turn off shape checking because elements may have undergone considerable deformation during the
explicit analysis.
Command(s): SHPP,OFF
GUI: Main Menu> Preprocessor> Checking Ctrls> Shape Checking
8.
Redefine the implicit elements to the deformed configuration.

Command(s): UPGEOM
GUI: Main Menu> Preprocessor> Modeling> Update Geometry
9.
Unselect or delete any unnecessary elements (mainly those making up any rigid bodies from the explicit
analysis), or convert them to null elements. Any explicit elements that are not either unselected, deleted,
converted to null elements, or converted to implicit will remain active in ANSYS, which will produce an
error and terminate the analysis. Also, if the rigid bodies in the explicit analysis were made up of SHELL163
or SOLID164 elements, these elements must be unselected, deleted, or converted to NULL elements
before importing stresses and thicknesses (from SHELL163 to SHELL181, or SOLID164 to SOLID185) by
the RIMPORT command (See Step 12); otherwise, the implicit analysis will be terminated.
Command(s): ESEL, EDELE
GUI: Utility Menu> Select> Entities or
Main Menu> Preprocessor> Delete> Elements
10. Reenter the solution processor.

Command(s): /SOLU
GUI: Main Menu> Solution
11. Set any necessary constraints on the model by modifying or adding to the boundary conditions defined
during the explicit analysis (for example, in a metal forming analysis, you need to constrain the blank).
Command(s): D, etc.
GUI: Main Menu> Solution> Loads> Apply> Structural-Displacement> On Nodes, etc.
12. Import stresses and changed thicknesses (from SHELL163 to SHELL181, and from SOLID164 to SOLID185).
For SOLID164, only the stresses are transferred. For SHELL163, the stresses and thicknesses are transferred.
The deformed integration point thicknesses are averaged before being transferred to the implicit analysis. These thicknesses override those defined using real constants.
Command(s): RIMPORT
GUI: Main Menu> Solution> Loads> Apply> Structural-Other> Import Stress
13. Turn large deformation effects on.
Command(s): NLGEOM,ON
GUI: Main Menu> Solution> Soln Controls
14. Solve and finish the analysis.
Command(s): SOLVE, FINISH
GUI: Main Menu> Solution> Solve> Current LS
Main Menu> Finish
Once you have solved the analysis, you can use any of the standard ANSYS postprocessing functions to review
your results.
The following is a sample input stream for performing an explicit-to-implicit sequential solution.
/batch,list
resume,stamp1,db
/filename,stamp2
! resume explicit database (explicit

! problem run previously)
! change jobname so explicit results
143
Chapter 14: Explicit-to-Implicit Sequential Solution

! are not overwritten
/prep7
etchg,eti
r,1,.01,.01,.01,.01
ddel,all
tbdelete,all,all
upgeom,,,,stamp1,rst
d,n1,ux,0.0,,,,uy,uz
d,n2,ux,0.0,,,,uy
d,n3,ux,0.0
esel,s,ename,,181
!
!
!
!
!
!
!
!
!
!
!
convert explicit to implicit elements

change real constants (deformed
thicknesses used by RIMPORT)
delete constraint loads from
explicit analysis
delete material models (retain
linear material properties)
redefine the implicit elements
at the deformed geometry
Set necessary constraints on the model
(at least 3 non-colinear nodes)
!
!
!
!
Select only corresponding SHELL181

elements (No former rigid bodies)
Select only SHELL181 nodes
Turn off shape checking
nsle,s,1
shpp,off
finish
/solution
nlgeom,on
! Turn on nonlinear geometry
rimport,dyna,,,,,stamp1,rst ! Import stresses and thicknesses from
! explicit analysis
save
solve
finish
14.2. Troubleshooting a Springback Analysis

Springback analyses are sometimes numerically difficult to solve. Some typical causes of convergence difficulties
and actions you can take to resolve them are listed below.
Dynamic effects due to a high punch speed used in the explicit (LS-DYNA) run.
Use a lower punch speed in the explicit run. You can do this by redefining the tool speed in the explicit
analysis (EDLOAD command).
Slenderness of the structure; the ratio of part size to its thickness is large.
A more refined mesh may reduce this slenderness effect.
The model contains local instabilities: the part, being very thin, tends to wrinkle along edges or sharp
curvatures.
Use a refined mesh to accurately model curvatures of the formed part.
The model is improperly constrained.

Verify the boundary conditions in the springback analysis. This requires some engineering judgment on
how and where to apply the boundary conditions in the springback phase. When the tools are removed,
the part should be free to move except for restraints against rigid body movement. It is important to
simulate this carefully in the finite element model. A commonly recommended approach is to use displacement constraints at a minimum of three non-colinear nodes. One node should be restricted from motion
in all directions, another node should be restricted from motion in two directions, and the last node should
be restricted from motion in only one direction. You must select these three nodes carefully so that the
expected springback behavior is not prevented. The magnitude and direction of the displacements in
the springback analysis should be viewed as relative quantities to the plane formed by these three nodes.
If you have checked all of the above conditions and the analysis is still not converging, try the following methods
(in the order listed):
144
Attempt incremental springback by using a higher number of substeps.

Section 14.2: Troubleshooting a Springback Analysis
Use a linear springback analysis (if such an approximation is acceptable) with the command NLGEOM,OFF.
Invoke springback stabilization (which is described below).
14.2.1. Springback Stabilization

If convergence difficulties persist, you should activate springback stabilization using the SPSCALE and MSCALE
arguments of the RIMPORT command. The stabilization method is analogous to attaching artificial springs to
all unconstrained degrees of freedom. The stiffnesses of these springs are directly proportional to the stiffnesses
of the elements.
In each substep, ANSYS will assign the artificial springs large default stiffnesses to simulate a total restraint of
the motion of the model at the beginning of each load step. The artificial stiffness is scaled down exponentially
with the progress of iterations. (The picture below illustrates rapid decline of artificial spring stiffness over an iterative cycle.) The substep is considered to be converged only when the equilibrium convergence tolerances
are met (as in a normal nonlinear analysis) and the artificial stiffness has reached an insignificantly small value.
This may slow down the conversion of the solution in some cases; however, the benefit is more robust behavior
of the model.
The SPSCALE parameter on the RIMPORT command is a scale factor that you input to scale (up or down) the
default initial stiffness of the springs. (You must input SPSCALE to activate springback stabilization.) You can
also specify an acceptable small stiffness value using the MSCALE parameter on the RIMPORT command.
The springback stabilization method is valid only when the geometric nonlinear option is active (NLGEOM,ON).
145
146
Chapter 15: Implicit-to-Explicit Sequential

Solution
In Chapter 14, Explicit-to-Implicit Sequential Solution, we explained how to perform explicit-to-implicit sequential
solutions for analyzing springback in metal forming processes. This chapter discusses another form of sequential
solution, implicit-to-explicit. There are two types of implicit-to-explicit sequential solutions. The first type, which
involves performing an ANSYS implicit structural analysis followed by an ANSYS LS-DYNA explicit dynamic analysis, is used primarily for applying a preload prior to the transient phase of the analysis. (If this preload contains
thermally induced strains, the implicit structural analysis needs to contain temperatures from an implicit heat
transfer analysis.) The second type is a thermal-to-explicit sequential solution in which the implicit phase is an
ANSYS thermal analysis. This type of analysis is useful for modeling temperature-dependent phenomena such
as forging.
The following implicit-to-explicit analysis topics are available:
15.1. Structural Implicit-to-Explicit Solution for Preload
15.2. Thermal Implicit-to-Explicit Solution
15.1. Structural Implicit-to-Explicit Solution for Preload

Unlike explicit-to-implicit solutions, which are used only in forming applications, structural implicit-to-explicit
sequential solutions can be used for a broad range of engineering problems where a structure's initial stress
state affects its dynamic response. The following is a short list of applications in which this type of implicit-toexplicit solution may be useful.
Droptest simulation of prestressed consumer goods
Bird-strike of a rotating engine blade (and blade containment)
Mounting and bearing loads on a turbine
Pressure vessels with initial internal pressure
Golf club striking a preloaded (wound) golf ball
Dynamic response of bolted joints
Dynamic response of a turbine engine under thermal straining
In an implicit-to-explicit sequential solution, you must first run an ANSYS implicit structural analysis to apply a
preload to the structure being analyzed. In this implicit analysis, completely constrain all of the nodes of any
elements that will only be used in the explicit analysis (e.g., the bird in a bird-strike problem). The nodal displacements and rotations from the ANSYS implicit solution are written to the ANSYS LS-DYNA dynamic relaxation file
drelax.
Note Temperatures from the ANSYS implicit structural solution are also written to the drelax file, but
are not used by LS-DYNA. See Section 15.1.1: Special Considerations for Thermal Loading for more information on how temperature loads are handled.
After defining additional loads, initial velocities, different material models (e.g., adding plasticity), etc., the explicit
dynamic analysis can be conducted. The first part of this analysis uses the displacement results stored in the
drelax file to do a stress initialization to a prescribed geometry. This preload is applied in pseudo time over 101
time steps to damp out any kinetic energy. The transient portion of the analysis then begins at time zero with a
stable preloaded structure.
Chapter 15: Implicit-to-Explicit Sequential Solution

A detailed description of the implicit-to-explicit solution procedure follows.
1.
Run the implicit analysis as described earlier, using Jobname1. Keep in mind that this analysis must be
small strain with linear material behavior. The only element types that can be used for an implicit-toexplicit sequential solution are:
LINK8
BEAM4
SHELL181
SOLID185
COMBIN14
MASS21
LINK10
2.
Define any additional nodes and elements that are necessary to complete the explicit solution (for example, the bird in a bird-strike simulation, or a rigid surface that a phone would impact in a droptest).
These additional nodes and elements may not be part of the implicit analysis, but they need to be defined
here nonetheless. These additional nodes must be constrained (using D,ALL,ALL,0).
Command(s): N, E
GUI: Main Menu> Preprocessor> Modeling> Create> Nodes or Elements
3.
Solve and finish the analysis.

GUI: Main Menu> Solution> Solve
Main Menu> Finish
4.
Save the implicit analysis database to file Jobname1.DB.

Command(s): SAVE
Note If you do not save your Jobname1.DB file at this point, then the database for this implicit
run will not be saved. Only the database file for the subsequent explicit run will be saved.
5.
Change to Jobname2 to prevent the implicit results files from being overwritten.
6.

Command(s): /PREP7
7.
Convert implicit element types to corresponding companion explicit element types. Note that the 2-D
explicit element, PLANE162, cannot be used in this type of sequential solution. (PLANE162 is allowed in
a thermal implicit-to-explicit sequential solution; see Section 15.2: Thermal Implicit-to-Explicit Solution
for details.) The corresponding companion implicit-explicit element type pairs are:
LINK8
LINK160
BEAM4
BEAM161
SHELL181
SHELL163
SOLID185
SOLID164
COMBIN14
COMBI165
MASS21
152
MASS166
Section 15.1: Structural Implicit-to-Explicit Solution for Preload

LINK10
LINK167
Command(s): ETCHG,ITE
GUI: Main Menu> Preprocessor> Element Type> Switch Elem Type
Implicit elements not listed above can also be used, as long as they are defined by the same number of
nodes, but they will not automatically be converted to explicit elements when ETCHG is issued. These
elements must be converted manually using EMODIF. Higher-order implicit elements can also be used,
but must also be converted manually using EMODIF with the corner nodes only. Do NOT delete or unselect
the midside nodes - these nodes must be written to the LS-DYNA input file. The drelax file contains
solutions for these nodes, but the ANSYS LS-DYNA explicit elements do not use these nodes in their
definition.
Command(s): EMODIF
GUI: Main Menu> Preprocessor> Modeling> Move/Modify> Nodes
Note Element types LINK8 and LINK10 lack a third node; however, their corresponding companion explicit element types, LINK160 and LINK167, require a third (orientation) node. If you
are using element types LINK8 or LINK10, you must first convert the element type using ETCHG,ITE,
and then manually define the third node of LINK160 or LINK167 elements using N and EMODIF.
Also, if you are converting BEAM4 to BEAM161, you may need to manually define the third node
of BEAM161 elements as well. However, BEAM4 allows you to define a third, optional node. If
you have defined this third node on BEAM4, then the conversion to BEAM161 will be completed
automatically when you issue ETCHG,ITE. If you did not define the third node on BEAM4, then
you must manually define it on BEAM161 using N and EMODIF.
8.
Redefine the key options, real constants, boundary conditions, and loading values on the explicit elements.
The TYPE, REAL, and MAT numbers from the implicit elements are retained, but the actual key option
and real constant values are reset to zero or the default settings.
GUI: Main Menu> Preprocessor> Element Type, Real Constants, Material Props, or LS-DYNA
Options
9.
Remove constraints from the additional nodes or elements defined in Step 2, above.
Command(s): DDELE
GUI: Main Menu> Preprocessor> LS-DYNA Options> Constraints> Delete

Command(s): /SOLU
11. Read nodal displacements, rotations, and temperatures from the implicit results file, and write this information to an ASCII LS-DYNA file, drelax.
Command(s): REXPORT
GUI: Main Menu> Solution> Constraints> Read Disp
12. Initialize the structure to the prescribed geometry according to the displacements and rotations contained
in the drelax file. In this step, LS-DYNA applies the load information (displacements and rotations) from
the drelax file to the original geometry and calculates the deformed geometry, which it then uses as a
starting point for the explicit analysis.
Command(s): EDDRELAX
GUI: Main Menu> Solution> Analysis Options> Dynamic Relax
13. Apply any necessary loading for the explicit run.
153

Command(s): EDVEL, EDLOAD, EDCURVE, etc.
GUI: Main Menu> Solution> Initial Velocity
Main Menu> Solution> Loading Options> Specify Loads
14. Solve and finish the explicit dynamics analysis. You can then return to the implicit solution, if necessary.
The following is a sample input stream for performing an implicit-to-explicit sequential solution.
/batch,list
resume,drop1,db
/filename,drop2
/prep7
etchg,ite
mp,dens,1,.0216
ddel,all
tb,plaw,1,,,8
edmp,rigid,2,7,7
edcgen,assc
nsel,s,loc,x,0
nsel,a,loc,y,0
d,all,ux,0
!
!
!
!
Resume implicit database (implicit

problem run previously)
Change jobname so implicit results are
not overwritten
!
!
!
!
!
!
!
!
!
!
Convert implicit to explicit elements

Change material properties
Delete constraint loads from implicit
analysis
Define explicit dynamics nonlinear
material models
Change an existing material to a rigid body
Specify contact algorithms (if any)
Select geometry for new constraints to be
specified
! Set necessary constraints on the rigid

! body, etc.
d,all,uy,0
finish
/solution
rexport,dyna,,,,,drop1,rst ! Create DRELAX file from implicit
! results
eddrelax,ansys
! Specify stress initialization by prescribed
! geometry
edpart,create
! Create parts for loading
edpart,list
! List parts
edvel,
! Apply initial velocities
edload,add,rbvx,,2,time,load,0 ! Apply phase = 0 loads
save
solve
finish
15.1.1. Special Considerations for Thermal Loading

Most types of temperature loads that are applied in the implicit structural analysis will be carried over to the
subsequent explicit analysis; these include temperature loads applied with the BF, BFK, BFL, BFA, BFV, LDREAD,
TUNIF, BFUNIF, and TREF commands. If you do not want to include the temperature loads in the explicit analysis,
you can delete them in the ANSYS structural analysis phase (before you convert the model to explicit elements
via ETCHG), or you can delete them in the explicit analysis.
In the explicit analysis phase, you can use the BFLIST command to list the temperature loads mentioned above.
You can then use the BFDELE command to delete any temperatures originally defined with the LDREAD command
or one of the "BF" type commands listed above. If you do not want to include a uniform temperature carried
over by TUNIF (or BFUNIF,TEMP) you should set TUNIF to the reference temperature (which is the temperature
specified on the TREF command). Any temperature loads that are not deleted will be written to the LS-DYNA
input file, Jobname.K, automatically when the SOLVE or EDWRITE command is issued.
In order for temperature loads to take effect in the explicit analysis, you must use the Temperature Dependent
Bilinear Isotropic material model or the Elastic Viscoplastic Thermal material model for portions of the model
that are subjected to temperature loading. Note that only the SHELL163 and SOLID164 elements can accept
temperature loading in a structural implicit-to-explicit sequential solution.
154
Section 15.2: Thermal Implicit-to-Explicit Solution

Note The temperature loads in the implicit analysis are also written to the drelax file when REXPORT
is issued (Step 12 above). However, the temperatures contained in the drelax file are not used in the LSDYNA calculation.
Appendix D:, Thermal/Structural Preload Example contains a complete example of a thermal/structural preload
followed by a structural transient dynamic analysis.
Currently, ANSYS LS-DYNA does not support applying a thermal transient in a model that includes a thermal
preload.
15.2. Thermal Implicit-to-Explicit Solution

The following is a detailed description of how to perform a thermal implicit-to-explicit sequential solution. Since
the temperature loads are step-applied at the beginning of the transient, the analysis should be run for a sufficient
amount of time to reduce the effects of the thermal shock loading. If only temperature-dependent material behavior is desired without the effects of thermally induced strain, set the coefficient of thermal expansion equal
to zero (ALPX = 0) so that the analysis will be stable from the beginning of the transient. This is a convenient way
to vary the modulus of elasticity in a model with only one material definition.
1.
Run the implicit thermal analysis using Jobname1. The only element types that can be used for a thermal
implicit-to-explicit sequential solution are:
PLANE55
SOLID70
SHELL57
Keep in mind that you cannot combine 2-D and 3-D elements in the subsequent explicit analysis.
Therefore, you should use only 2-D or only 3-D elements in the thermal analysis phase.
2.
You may define any additional nodes and elements that are necessary to complete the explicit solution.
These additional nodes and elements may not be part of the thermal analysis, but they can be defined
here if desired. These additional nodes must be constrained to a constant temperature (using
D,NODE,TEMP,VALUE).
Command(s): N, E
3.
Solve and finish the analysis.

GUI: Main Menu> Solution> Solve
Main Menu> Finish
4.
Save the thermal analysis database to file Jobname1.DB.

Command(s): SAVE
Note If you do not save your Jobname1.DB file at this point, then the database for this thermal
run will not be saved. Only the database file for the subsequent explicit run will be saved.
5.
Change to Jobname2 to prevent the thermal results files from being overwritten.
6.

Command(s): /PREP7
155

7.
Convert thermal element types to corresponding companion explicit element types. The corresponding
companion thermal-explicit element type pairs are:
Thermal (Implicit) Element Type
PLANE55
PLANE162
SHELL57
SHELL163
SOLID70
SOLID164
Command(s): ETCHG,TTE
GUI: Main Menu> Preprocessor> Element Type> Switch Elem Type
8.
Redefine the key options, real constants, and material property values on the explicit elements. The TYPE,
REAL, and MAT numbers from the implicit elements are retained, but the actual key option and real
constant values are reset to zero or the default settings. Keep in mind that for temperature loading to
take effect, you must use the Temperature Dependent Bilinear Isotropic material model in the explicit
analysis.
GUI: Main Menu> Preprocessor> Element Type> Real Constants> Material Properties
9.
Define additional nodes and elements, if needed. You may have already completed this task in the
thermal analysis portion of the sequential solution (see Step 2). However, if desired, you could wait until
the explicit analysis to complete this step.
Command(s): N, E

Command(s): /SOLU
11. Read temperature results from the thermal analysis results file (Jobname1.RTH) and apply them as loads
in the explicit analysis. You can only transfer the temperatures from one time point in the thermal analysis; specify this point by inputting the time value or the corresponding load step and substep numbers
on the LDREAD command. If you issue LDREAD more than once, the temperature results specified by
the last command will be used.
Command(s): LDREAD,TEMP
GUI: Main Menu> Solution> Loading Options> Temp From ANSYS
The temperatures read in are applied as body loads to the nodes in the explicit model. You can use the
BFLIST command to list these temperature loads and the BFDELE command to delete them.
Command(s): BFLIST, BFDELE
GUI: Main Menu> Solution> Loading Options> List Temps/Delete Temps
12. Apply a uniform temperature load, if needed. You can use the TUNIF command (or BFUNIF,TEMP) to
apply a uniform temperature load to all nodes in the model. The temperatures read in by LDREAD will
overwrite the TUNIF temperature for all nodes that were present in the thermal analysis. Therefore,
TUNIF would typically be used only when additional nodes and elements have been defined in the explicit portion of the sequential solution.
Note If you used TUNIF (or BFUNIF,TEMP) to define a uniform temperature in the thermal
analysis, the same uniform temperature will be applied automatically to any new nodes added
in the explicit analysis phase. If you do not want to apply a uniform temperature in the explicit
analysis, you should set TUNIF to zero (or to the TREF temperature, if it is nonzero).
Command(s): TUNIF
GUI: Main Menu> Solution> Loading Options> Uniform Temp
156

13. Define a reference temperature. The thermal loading is defined as the difference between the applied
temperature and the reference temperature. If the reference temperature is not specifically defined, it
defaults to zero.
Command(s): TREF
GUI: Main Menu> Solution> Loading Options> Reference Temp
14. Apply any additional loading needed for the explicit run.
Command(s): EDVEL, EDLOAD, EDCURVE, etc.
GUI: Main Menu> Solution> Initial Velocity
Main Menu> Solution> Loading Options> Specify Loads
15. Solve and finish the explicit dynamic analysis.
The following is a sample input stream for performing a thermal implicit-to-explicit sequential solution.
/clear
/title, Thermal Implicit-to-Explicit Sequential Solution
/view,,1,2,3
/plopts,info,1
/filenam,thermal
/prep7
et,1,SOLID70
r,1
mp,kxx,1,16.3
mp,c,1,502.0
block,0,1,0,1,0,11
esize,0.5
vmesh,all
finish
/solu
antype,static
time,1.0
! define unique name for thermal analysis
! thermal brick element

! thermal conductivity, W/m-C
! specific heat, J/kg-C
! create and mesh model
! steady-state heat transfer analysis
nsel,s,loc,z,0
d,all,temp,300.0
! 300 Celsius constraint at far end of beam
nsel,s,loc,z,11
cm,ntip,node
esln
cm,etip,elem
bf,all,hgen,20.0e3
! 20,000 W/m^3 heat generation at near end ...
esel,all
nsel,s,ext
sf,all,conv,6.0,0.0
nsel,all
! exterior beam nodes

! hf=6.0 W/m^2-C and Tbulk=0 degrees C
save
solve
fini
/post1
set,last
plnsol,temp
/wait,3
fini
/filnam,explicit
! define unique name for explicit analysis
/prep7
157

bfdele,all,all
sfdele,all,all
ddele,all,all
! delete thermal analysis heat generation loads

! delete thermal analysis convection loads
! delete thermal analysis temp. constraints
etchg,tte
! convert thermal element types to explicit
mp,dens,1,8030.0
edmp,hgls,1,5
! kg/m^3 - density cannot be temp. dependent

! add hourglass control
! Define temperature-dependent BISO model

! Specify material temperatures, degrees Celsius
mptemp,1, 0, 500, 700, 900, 1100, 1200
! Modulus of elasticity, N/m^2 (Pascal)
mpdata,ex,1,1, 1.93e11, 1.5e11, 1.0e11, 0.9e11, 0.8e11, 0.7e11
! Poisson's ratio, unitless
mpdata,nuxy,1,1, 0.3, 0.31, 0.32, 0.33, 0.34, 0.35
! Coefficient of thermal expansion, m/m-C
mpdata,alpx,1,1, 1.78e-5, 2.5e-5, 5e-5, 5.5e-5, 6.0e-5, 6.5e-5
tb,biso,1,6
tbtemp,0
tbdata,1,30000,3e6
tbtemp,500
tbdata,1,20000,2e6
tbtemp,700
tbdata,1,10000,1e6
tbtemp,900
tbdata,1,9000,0.9e6
tbtemp,1100
tbdata,1,8000,0.8e6
tbtemp,1200
tbdata,1,7000,0.7e6
edpart,create
eddamp,all,,0.01
fini
/solu
ldread,temp,last,,,,thermal,rth
bflist
nsel,s,loc,z,0
d,all,ux,0.0,,,,uy,uz
esel,all
nsel,all
time,0.15
! apply temperature body loads
! structural displacement constraints
! use sufficient time (see below)
! Note: Since the temperatures are step-applied at the beginning of

!
the transient, a thermal shock condition is created. You
!
should run the analysis for a sufficient amount of time
!
before adding any structural loads in order for the structure
!
to stabilize (i.e., stop vibrating). Animation plots of both
!
stress and strain are helpful in verifying this condition.
!
!
Alternately, if only temperature-dependent material properties
!
(e.g., modulus of elasticity) are desired, you can avoid the
!
thermal shock condition by setting the coefficient of thermal
!
expansion equal to zero (ALPX=0). Large plastic strains can
!
develop due to a thermal shock scenario, so care must be
!
used when running a thermal-to-explicit sequential solution.
edrst,100
edhtime,500
edhist,ntip
edhist,etip
edout,glstat
save
158

solve
fini
/post1
set,last
/dscale,,10
plnsol,epto,x,2
! 10 times displacement magnification

! X and Y strains are the same ...
159
1510
Chapter 16: Arbitrary Lagrangian-Eulerian

Formulation
The Arbitrary Lagrangian-Eulerian (ALE) element formulation is a standard numerical approach for solving large
strain deformation problems encountered in metal forming and high-speed impact applications. The general
concept of the ALE formulation is that an arbitrary referential domain is defined for the description of motion
that is different from the material (Lagrangian) and spatial (Eulerian) domains.
In a pure Lagrangian system, the mesh deforms with the material being modeled so that there is no material
flow between elements. The Lagrangian approach is well suited for moderately large strain problems where
mesh distortion and element entanglement are not a significant problem. The advantage of the Lagrangian
approach is that the free surface of the material is automatically captured by the mesh. The main disadvantage
of the Lagrangian approach is that problems develop in physical situations that involve highly deformed surfaces.
Consider the high speed impact problem of a metal bar shown in Figure 16.1: High Speed Impact of a Metal
Bar. As shown in Figure 16.2: Lagrangian Impact Solution, the Lagrangian approach begins to break down
along the impact surface between the metal and the rigid wall. The mesh in this region becomes highly distorted
and will not yield accurate results. The best solution for increasing the accuracy of the Lagrangian approach in
highly deformed regions is to use an adaptive meshing procedure. Adaptive procedures, however, are computationally expensive and are still not fully developed in three dimensional problems. Other disadvantages of the
Lagrangian approach are that only one material can be modeled in each element and that new (damaged) surfaces
cannot be created.
Figure 16.1 High Speed Impact of a Metal Bar
Chapter 16: Arbitrary Lagrangian-Eulerian Formulation
Figure 16.2 Lagrangian Impact Solution
In a Eulerian based formulation, the mesh is stationary and the material flows through the mesh. The Eulerian
approach originated in the fluid dynamics field and is best suited for very large deformation flow problems, such
as the channel flow depicted in Figure 16.3: Eulerian Channel Flow Solution. In the Eulerian method, new
(damaged) surfaces are automatically created. The greatest disadvantage of the Eulerian approach is that a fine
mesh is required to capture the material response, making the method very computationally expensive. This is
particularly true for problems that contain regions where the structural response is desired and the strains are
relatively small.
Figure 16.3 Eulerian Channel Flow Solution
The Arbitrary Lagrangian-Eulerian (ALE) approach is a very effective alternative for simulating large deformation
problems. In its most basic sense, the ALE method defines that the mesh motion is independent of the motion
of the material being analyzed. Although the mesh motion may be arbitrary, it typically deforms with the material in 'near Lagrangian' flow fields. The greatest advantage of the ALE method is that it allows smoothing of a
162
Section 16.1: Performing an ALE Analysis

distorted mesh without performing a complete remesh. This smoothing allows the free surface of the material
to be followed automatically without encountering the distortional errors of the Lagrangian approach (see Figure 16.4: ALE Impact Solution).
The main difficulty of the ALE method is the path dependent behavior of the plastic flow being modeled. Due
to the path dependence, the relative motion between the mesh and the material must be accounted for in the
material constitutive equations. In addition, the ALE method does not allow new (damaged) surfaces to be created
and is limited to geometries where the material flow is relatively predictable.
Figure 16.4 ALE Impact Solution
For more information about the ALE formulation, refer to Livermore Software Technology Corporations LS-DYNA
Theoretical Manual.
16.1. Performing an ALE Analysis

The ALE formulation is available in the ANSYS LS-DYNA program only for PLANE162 and SOLID164 elements.
For both of these element types, KEYOPT(5) defines the element continuum treatment. The default KEYOPT(5)
setting is zero, which defines the material continuum to be Lagrangian. Setting KEYOPT(5) = 1 defines the ALE
formulation. The current implementation of ALE in ANSYS LS-DYNA does not allow mixing of materials within
an element.
In general, the 2-D ALE formulation (PLANE162) is not as robust as the 3-D formulation. Therefore, it may be advantageous to model 2-D problems with SOLID164 elements and then apply nodal motion constraints in the
depth direction.
As described earlier, the greatest advantage of the ALE method is that it allows smoothing of a distorted mesh
without performing a complete remesh. The algorithms for moving the mesh relative to the material control the
range of the problems that can be solved by an ALE formulation. In the ANSYS LS-DYNA program, smoothing
may be applied via the EDALE command. Using this command, you must define at least one of four different
weighting factor options:
163
Chapter 16: Arbitrary Lagrangian-Eulerian Formulation

AFAC - Simple average smoothing weight factor. For this method, the coordinates of a node are the simple
average of the coordinates of its surrounding nodes.

BFAC - Volume-weighted smoothing weight factor. This method uses a volume-weighted average of the
coordinates of the centroids of the elements surrounding a node.

DFAC - Equipotential smoothing weight factor. Equipotential zoning is a method of making a structured mesh
for finite difference or finite element calculations by using the solutions of Laplace equations as the mesh
lines. The same method can be used to smooth selected points in an unstructured 3-D mesh provided that
it is at least locally structured.
EFAC - Equilibrium smoothing weight factor (available only for PLANE162). For this method, artificial springs
are attached to each ALE element node. The springs are used to adjust the position of each node from the
equilibrium solution. This approach can overcome possible calculation instabilities found in the other
smoothing methods.
The EDALE command also allows you to define start and end times for ALE smoothing.
You can define two additional ALE options with the EDGCALE command:
ADV - Number of cycles between advection
METH - Advection method (donor cell or Van Leer)
In general, it is not worthwhile to advect an element unless at least twenty percent of its volume will be transported, because the gain in the time step size will not offset the cost of the advection calculations. It is best to begin
an ALE analysis with a Van Leer advection technique (METH = 1).
If you are working in the GUI, all ALE related options are accessed by picking Main Menu> Solution> Analysis
Options> ALE Options.
Note In order to activate the ALE formulation for a given set of elements, you must specify a smoothing
weight factor (EDALE command), and you must specify the cycles between advections (NADV argument
on the EDGCALE command). If you do not supply both of these inputs, the default Lagrangian formulation
will be used for all elements, including those for which KEYOPT(5) = 1.
The remap step maps the solution from a distorted Lagrangian mesh onto the new mesh. The underlying assumptions of the remap step are 1) the topology of the mesh is fixed (that is, the element nodal connectivity
remains unchanged), and 2) the mesh motion during a step is less than the characteristic lengths of the surrounding elements. The algorithms for performing the remap step are taken from the computational fluid dynamics
community and are referred to as advection algorithms.
The donor cell algorithm is a first order Godunov method applied to the advection equation. Aside from its first
order accuracy, it is a good advection algorithm; it is stable, monotonic and simple. The Van Leer algorithm is a
higher order Godunov method that improves the estimates of the initial values of left and right states for the
Riemann problem at the nodes. The donor cell algorithm assumes that the distribution of the initial value function
is constant over an element. Van Leer replaces the piecewise constant distribution with a higher order interpolation function that is subject to an element level conservation constraint.
164
Chapter 17: Drop Test Module

The Drop Test Module (DTM) is an optional add-on feature to ANSYS LS-DYNA product. The DTM greatly simplifies
the procedure for simulating a drop test. A drop test involves orienting an object with respect to an assumed
gravitational field and allowing it to drop from some specified height under the influence of gravity onto a flat,
rigid surface (the target). In a typical drop test, the object is dropped from rest and the target lies in a plane
perpendicular to the direction of the acceleration due to gravity (g).
The DTM provides a streamlined GUI that guides you through the steps involved in a drop test analysis. It allows
you to quickly orient your model relative to the target, specify solution controls, and perform the explicit dynamic
analysis of the test. The analysis predicts deformations and stresses in the object over a user-specified time interval
to characterize the implications of the impact of the object with the target.
The Drop Test Module is intended for use in the ANSYS Graphical User Interface (GUI) environment. GUI features
pertaining to the Drop Test Module have been implemented using the ANSYS User Interface Design Language
(UIDL), along with the Tool Command Language and Toolkit (Tcl/Tk).
The Drop Test Module is specially designed so that a user having little experience with explicit dynamics can
easily perform a drop test analysis. Thus, you will find many advantages to using the Drop Test Module. For example, the module automatically creates the target. The DTM also offers options to specify initial translational
and angular object velocities, contact surface friction coefficients, and a target orientation that is not perpendicular to the direction of g. For most drop tests, the analysis can be started at a time just prior to impact. This reduces
computation time. When this option is used, the DTM automatically calculates the object's translational velocity
and applies it at the start of the analysis. For all drop test cases, drag forces acting on the object as it moves
through the air are neglected.
The following Drop Test Module topics are available:
17.1. Starting ANSYS With the Drop Test Module
17.2. Typical Drop Test Procedure
17.3. Advanced DTM Features
17.4. Drop Test Set-up Dialog Box
17.5. Picking Nodes
17.6. Postprocessing - Animation
17.7. Postprocessing - Graph and List Time-History Variables
17.1. Starting ANSYS With the Drop Test Module

To start ANSYS to include both ANSYS LS-DYNA and the DTM option, do one of the following:
If you are starting ANSYS from a command line (UNIX) or DOS prompt (Windows), include -dtm in the line
as well as a productname that represents an ANSYS LS-DYNA product. For example, to start the ANSYS
LS-DYNA product with the DTM in interactive, graphics mode, type the following in a command line:
ansys100 -p dyna -dtm -g
If you are starting ANSYS using the ANSYS Launcher, on the Launch tab choose an ANSYS LS-DYNA license
and select the Drop Test Module for ANSYS LS-DYNA add-on.
Refer to Chapter 3, Running the ANSYS Program in the ANSYS Operations Guide for more information on starting
ANSYS.
You can also have ANSYS automatically include the DTM option by setting the following environment variable:

ANSYS100_DTM=ON
See the ANSYS Installation and Configuration Guide for your platform for more details.
17.2. Typical Drop Test Procedure

The step-by-step procedure described in this section can be used as a guide for most types of drop test analyses.
A typical test would involve dropping an object from some height in a gravitational field onto a flat, rigid surface
(target), neglecting surface friction. The basic procedure outlined here assumes that the object has a zero initial
velocity, and the object is being dropped onto a target that lies in a plane which is normal to the direction of the
acceleration due to gravity. More advanced features, such as inclusion of surface friction effects, specification of
a nonzero initial velocity (translational and/or angular), and modification of the target size, properties, and orientation, are discussed in Section 17.3: Advanced DTM Features in this user's guide.
Before you begin, it is important for you to understand how the screen coordinates are used within the Drop
Test Module. In the DTM, the screen coordinates are always defined so that the screen Y axis (vertical) is in the
direction opposite to the acceleration due to gravity (g). Thus, screen coordinates are used as a convenient
method for keeping track of the relationship between the object to be dropped and the gravitational field. See
Section 17.2.2: Screen Coordinates Definition in this user's guide for more details.
17.2.1. Basic Drop Test Analysis Procedure

17.2.1.1. STEP 1: Create or import the model
Before entering the Drop Test Module, you must build or import the model of the object to be dropped. All elements should be defined and all material properties, including damping, should be specified. Only elements
that are compatible with LS-DYNA (that is, explicit dynamic elements) should be included in the model. Make
sure that there are not materials defined with either density or modulus of elasticity equal to zero. Otherwise,
you will not be able to set up the DTM. See Chapter 2, Elements in this user's guide for more information.
To avoid time increment degeneration problems, follow these guidelines when building the model:
Avoid triangular, tetrahedron, and prism elements
Avoid small elements
Avoid acute-angled elements
Use as uniform a mesh as possible
We recommend that a certain degree of damping be applied to the model to reduce the oscillatory response.
You can apply both alpha (mass weighted) and beta (stiffness weighted) damping using the EDDAMP command.
Note After completing the model and before entering the DTM, it is good practice to save the existing
database under a unique name (pick Utility Menu> File> Save As).
17.2.1.2. STEP 2: Set up the DTM

You set up your analysis in the DTM by choosing Main Menu> Drop Test> Set Up. The Drop Test Set-up dialog
box appears that includes several tabbed pages. The most common drop test analysis parameters are on the
Basic tab, while parameters you may occasionally need to modify are included on most of the remaining tabs.
A description of each tab is presented below along with the tasks you can perform within the tab:
Basic tab -- Specify magnitude of acceleration due to gravity (g), object drop height, object orientation, analysis
start time, run time after impact, and the number of times data is written to the results file and to the time history
172
Section 17.2: Typical Drop Test Procedure

output file. See Section 17.4.2: Basic Tab of the Drop Test Set-up Dialog Box in this user's guide for more information.
Velocity tab -- Modify initial translational and/or angular velocity of the object. See Section 17.4.3: Velocity Tab
of the Drop Test Set-up Dialog Box in this user's guide for more information.
Target tab -- Modify dimensions and material properties of the target, rotate the target, and define contact and
friction properties for the target. See Section 17.4.4: Target Tab of the Drop Test Set-up Dialog Box in this user's
guide for more information.
Status tab -- Display the values currently set for acceleration due to gravity, initial translational and angular velocity, and the estimated time at the end of solution. See Section 17.4.5: Status Tab of the Drop Test Set-up Dialog
Box in this user's guide for more information.
While the Drop Test Set-up dialog box is displayed, no other ANSYS menus are available until you choose OK
or Cancel.
The remainder of this section (STEP 3 through STEP 9) discusses the options you would use during a typical drop
test analysis. Advanced options that are not commonly used are covered in Section 17.3: Advanced DTM Features
in this user's guide.
17.2.1.3. STEP 3: Define the magnitude of (g)

You must specify the magnitude of (g) by choosing the Basic tab. Under the Gravity heading, you pick a (g)
value corresponding to the test specific system of units (386.4, 32.2, etc.), or you can type your own value.
17.2.1.4. STEP 4: Specify the drop height

You must specify the drop height on the Basic tab. Under the Drop Height heading, you type the drop height,
and specify a reference point. The drop height is measured along the screen Y axis from the center of the top
face of the target to some height reference point within the model. Reference point options include the lowest
object point, that is, the point on the object with the minimum screen Y coordinate value, or the object's center
of gravity as calculated by ANSYS. The drop height that you input must be in length units that are consistent
with the rest of the analysis (gravity for example).
17.2.1.5. STEP 5: Orient the object

In the DTM, the screen coordinates are automatically defined with the Y axis directed opposite to the direction
of (g). You can orient the object with respect to the gravitational field under the Set Orientation heading on
the Basic tab. There you have the option of orienting the object by using the DTM Rotate tool, or by picking
nodes on the object that define a vector parallel to the screen Y direction.
17.2.1.6. STEP 6: Specify solution controls

You can access solution control options on the Basic tab. Under the Solution time heading, you can specify
when the analysis starts (near impact time or at drop time), as well as the run duration time after impact. Under
the Number of Results Output heading, you can specify the number of times results are written to the results
file (file.RST) and to the time-history file (file.HIS).
For more information on the tasks involved in STEP 3 through STEP 6, see Section 17.4.2: Basic Tab of the Drop
Test Set-up Dialog Box in this user's guide.
173
17.2.1.7. STEP 7: Solve

Before solving, you should make sure that you are satisfied with the information in the Drop Test Set-up dialog
box. Perform any of the following verification steps:
1.
Review the information under all of the tabs.
2.
Click on the Status tab since this displays your most updated input data.
3.
View the object and the target in the ANSYS Graphics window.
If you choose the Status tab, ANSYS attempts to automatically create the target (unless you have already created
it by choosing the Target tab) and displays a statement in black print at the end of the Status tab page indicating
that all the set-up data is acceptable and you can proceed to solve the analysis, or a statement in yellow print
indicating that you should proceed with caution because the data produced error messages, or a statement
displayed in red indicating that the set-up data is unacceptable.
If you receive either the second or third statement regarding error messages, check them in the ANSYS Output
window. See Section 17.4.5: Status Tab of the Drop Test Set-up Dialog Box in this user's guide for more information
on using the Status tab. Make changes if necessary, then choose OK. At this time, all of the standard ANSYS
menus become usable again. You then initiate the LS-DYNA solution by choosing Main Menu> Drop Test>
Solve.
Note A database that you save from the current version of ANSYS LS-DYNA DTM is not compatible
with a previous version of ANSYS LS-DYNA DTM.
17.2.1.8. STEP 8: Animate results

Once the solution is completed, you can choose Main Menu> Drop Test> Animate Results to animate dynamic
results (such as von Mises stresses) for the dropped object over the analysis time duration. For animating results,
it is recommended that you choose to start the analysis at drop time. See Section 17.6: Postprocessing - Animation
in this user's guide for information on the Animate Over Results dialog box, and see Chapter 15, Animation
in the ANSYS Basic Analysis Guide for general information on producing animations in ANSYS.
17.2.1.9. STEP 9: Obtain Time-History Results

In addition to animating the results, you can also review the analysis results at specific points in the model as a
function of time by choosing Main Menu> Drop Test> Time History. See Chapter 12, Postprocessing and
Section 17.7: Postprocessing - Graph and List Time-History Variables in this user's guide for more information.
17.2.2. Screen Coordinates Definition

When you first enter the DTM, screen Cartesian coordinates, which are fixed, are defined automatically, and are
represented graphically in the lower left corner of the ANSYS Graphics window. Screen coordinates are used as
a convenient way of keeping track of the relationship between the object to be dropped and the gravitational
field. There are three important points you should understand about screen coordinates:
In the DTM, screen coordinates are always defined so that the screen Y axis is in the direction opposite to
the acceleration due to gravity, (g).
Screen coordinates are defined with respect to global Cartesian coordinates.
Object coordinates can be rotated with respect to the screen coordinates.
174

The screen coordinates are always defined so that the Y axis is in the screen up direction. The object coordinate
system is the coordinate system in which the finite element model was defined. You can reorient the object coordinate system using the DTM Rotate tool. You can define a vector to be parallel to the screen Y coordinate
system by picking two nodes on the object. See Section 17.4.2: Basic Tab of the Drop Test Set-up Dialog Box in
this user's guide for more information. By redefining the object Y direction, you are specifying the relationship
between the object and the gravitational field, because the screen Y direction is always assumed to be up or
opposite to the direction of the acceleration due to gravity, (g).
The screen coordinates also determine the default drop test viewing orientation, which is defined as follows:
The screen X direction is "right" (to the viewer's right).
The screen Y direction is "up" on the screen.
The screen Z axis is along the normal from the screen to the viewer, or "out".
17.2.3. Additional Notes on the Use of the DTM

This section contains some notes of interest and practices to follow that will help you to use the DTM more effectively. Some of these points were mentioned briefly in previous sections.
During initialization, a keyword is created that is related to the filtering of menu picks in the DTM. The
keyword is DTFILT. You should avoid using this word as a keyword or a parameter name, as it could interfere
with the filtering of menu options.
Within the DTM, a component containing all the nodes in the object is automatically created and given
the name _dtdrpob. This component name is used in issuing commands associated with LS-DYNA. You
should avoid using this name.
Parameters _dtcgnum and _dtlownum are used for time-history postprocessing plots and lists. You should
avoid using these names.
The DTM assumes that there are no constraints imposed on the object being dropped. However, coupling
of components within an object is allowed.
Some parameter names used internally by the Drop Test Module begin with the underscore character (_),
so you should avoid using parameter names that begin with an underscore. (This is true for any ANSYS
analysis.)
If you do not choose to orient the object, the default assumption is the current screen Y axis view, which
is directed opposite to the acceleration due to gravity (g).
You can modify any input set up parameter any number of times before entering the solution phase. All
parameters in the Drop Test Module are subject to redefinition after being initially defined. You will see
that whenever an input parameter that affects the initial target/object orientation is changed, the target
is recreated automatically to reflect the change when you click on the Target tab or the Status tab.
After you specify a drop height, you will notice that, by default, the target is located nearly in contact with
the object. This is due to the default option to begin the analysis near impact time. The target is created
with just a small separation between the object and target. You may want to adjust the Solution time
option to begin the analysis at drop time. After doing so, you will observe that the redefined target is
at a position consistent with the specified drop height. However, while this option provides full animation
of the drop scenario, it will require more computation time.
To use the option to start the analysis near impact time, the initial angular velocity must be zero.
The Drop Test Module does not alter the model of the object in any way. The DTM does create a model
of a target after you specify a new orientation for the object.
175
17.3. Advanced DTM Features

This section discusses advanced features of the Drop Test Module that are provided as a convenience, but may
not be required in a typical drop test analysis. These features include options to specify initial translational and
angular velocities, modify the target's position and properties, and specify contact surface friction coefficients.
Details on these features are included in the following sections.
17.3.1. Object Initial Velocity

By default, the translational and angular velocities of the object being dropped are assumed to be zero at the
time of the drop. You can impose an initial translational and/or angular velocity on the object being dropped;
however, in certain cases when you specify initial angular velocities, as discussed below, the near impact time
starting option is not valid and the GUI prevents you from making this choice. The near impact time starting
option is further discussed in Section 17.4.2: Basic Tab of the Drop Test Set-up Dialog Box , under Solution Time
in this user's guide.
You should always input initial velocities corresponding to the time of the "drop", even if the analysis is set to
begin near impact. The program will calculate the initial translational velocity at the start of the analysis based
on the initial velocity input and the acceleration due to gravity.
The initial velocities that you specify are applied via the EDVEL command. This command is discussed further
under Section 4.2.3: Initial Velocity in this user's guide.
You can specify initial translational or angular velocity by choosing Main Menu> Drop Test> Set Up then
choosing the Velocity tab. Under either the Translational Velocity or Angular Velocity headings, you can
choose to input the velocity in terms of screen coordinates or object coordinates. After choosing the desired
option, simply input the X, Y, and Z components of the velocity. If you enter the velocity relative to the screen
coordinates, the components are resolved into the object coordinates internally by the DTM before being input
on the EDVEL command. See Section 17.4.3: Velocity Tab of the Drop Test Set-up Dialog Box in this user's guide
for more information.
17.3.2. Modifying the Target

The target is one rigid SOLID164 brick element, which is created automatically in the Drop Test Module and is
always assumed to be stationary. Figure 17.1: Two Views of the Target(a) shows the target in the default drop
test viewing orientation, and Figure 17.1: Two Views of the Target(b) is an alternate view of the target that
more clearly shows its geometry. Figure 17.1: Two Views of the Target(b) shows that the target has two faces
that are much larger than the other four. The two large faces lie in planes that are parallel to each other. The
large face that has an outward normal vector directed along the screen Y axis (by default) is the top face of the
target.
176
Section 17.3: Advanced DTM Features
Figure 17.1 Two Views of the Target
It is expected that the default target position, size, orientation, material properties, and contact parameters will
be acceptable for most analyses. But if necessary, you can change the characteristics of the target by following
the procedures outlined in the following sections.
17.3.2.1. Target Position

You can change the target position by redefining the reference for the target center.
To change the target position, choose Main Menu> Drop Test> Set Up then choose the Target tab. Under the
Dimensions heading, there are two options from which you can choose for centering the target in the screen
X-Z plane. You can align the target's center of gravity either underneath the object's center of gravity, or underneath the object's lowest Y coordinate point. See Section 17.4.4: Target Tab of the Drop Test Set-up Dialog Box
in this user's guide for more information.
177
17.3.2.2. Target Size

The default target size is based on the characteristic target length TL. The program calculates TL based on the
overall dimensions of the object being dropped. By default, the sides of the two large square faces of the target
are of length 2TL, and the thickness of the target in the direction normal to these two faces is 0.1TL.
You can change the size of the target by choosing Main Menu> Drop Test> Set Up then choosing the Target
tab. Under the Dimensions heading, you can enter two target scaling factors, one for scaling the target length
(Ls) and one for scaling the target thickness (Ts). Thus, the sides of the two large square faces are of length 2LsTL,
and the target height is 0.1TsTL. As an example, If you input Ls=2.0 and Ts=0.5, then the target is recreated with
sides that are twice the default length and a thickness that is half its default value. Using this option to change
the target size does not alter the location of the target center point. If you change the target's size, the target is
re-drawn automatically. See Section 17.4.4: Target Tab of the Drop Test Set-up Dialog Box in this user's guide for
more information.
17.3.2.3. Target Orientation

By default, the target is oriented so that its two large faces are normal to the screen Y axis, and therefore normal
to (g). But you may need to change the target orientation for your particular model. For example, you may want
to simulate the object being dropped onto an inclined plane. To modify the target orientation, choose Main
Menu> Drop Test> Set Up, choose the Target tab. Under the Orientation angle heading, you type an angle
in degrees that is used to rotate the target about the screen Z axis (which is normal to the screen and toward
the viewer) using a right-hand rule. This is ideal for creating an inclined plane target. If the target is not in its
default orientation when this option is chosen, it is initially returned to its default orientation so that it is always
modified beginning from the default. If you change the target's orientation, the target is re-drawn automatically.
See Section 17.4.4: Target Tab of the Drop Test Set-up Dialog Box in this user's guide for more information.
17.3.2.4. Target Material Properties

Although the target is always defined to be rigid, a material density, modulus of elasticity, and Poisson's ratio
are still required for the LS-DYNA solution. Material properties are used for a contact stiffness calculation and a
time step calculation. Details on the contact stiffness and time step can be found under Chapter 6, Contact
Surfaces in this user's guide.
By default, the target is defined to be rigid using the EDMP command:
EDMP,RIGID,MAT,7,7
The material number (MAT) used is one greater than the highest material number used by the model of the object
being dropped. The default property values used for the target depend on the material properties used for the
object. The Drop Test Module searches through all materials belonging to the object, finds the one with the
lowest elastic modulus, and stores the corresponding material number. The modulus, density, and Poissons ratio
corresponding to this material number are also specified as the modulus, density, and Poissons ratio of the target
material.
If you know the actual target material properties, it is preferred that you input them directly. To specify the
properties, choose Main Menu> Drop Test> Set Up, choose the Target tab, then, under the Material Properties
heading, enter the Young's modulus, density, and Poisson's ratio. See Section 17.4.4: Target Tab of the Drop Test
Set-up Dialog Box in this user's guide for more information.
178
Section 17.4: Drop Test Set-up Dialog Box
17.3.2.5. Specifying Friction Coefficients

In the Drop Test Module, automatic single surface contact (AG) is defined for the entire model. The contact is
specified via the EDCGEN command.
You can change the static friction coefficient (FS), the dynamic friction coefficient (FD), and the exponential decay
coefficient (DC). These coefficients apply to all contact surfaces, including the target surface. You can input values
for these coefficients by choosing Main Menu> Drop Test> Set Up, then choosing the Target tab.
If you do not change parameters FS, FD, or DC, then they have default values of 0.0. See Section 17.4.4: Target
Tab of the Drop Test Set-up Dialog Box in this user's guide for more information.
17.4. Drop Test Set-up Dialog Box

17.4.1. Using the Drop Test Set-up Dialog Box
You use the Drop Test Set-up dialog box for inputting parameters for an ANSYS LS-DYNA analysis that uses the
Drop Test Module (DTM). You access it by choosing Main Menu> Drop Test> Set Up.
The Drop Test Set-up dialog box consists of four tabbed pages, each of which contains a group of related
DTM set-up information. The tabs, in order from left to right, are Basic, Velocity, Target, and Status. Each information group is logically arranged on a tab; the most basic and frequently addressed parameters appear on the
first tab, with each of the next two tabs providing more advanced parameters. Most DTM analyses involve using
the default settings for parameters on these two tabs. You do however have the option of modifying these
parameters as needed. The Status tab displays the values currently set for acceleration due to gravity, initial
translational and angular velocity, and the estimated time at the end of solution. This tab is located furthest to
the right because you are encouraged to examine its contents as a final set up task, before committing the
parameters to the ANSYS database (by clicking OK in the dialog box).
Once you are satisfied with the settings on the Basic tab, you do not need to progress through the remaining
tabs unless you want to change some of the advanced parameters. As soon as you click OK on any tab of the
dialog box, the settings are applied to the ANSYS database and the dialog box closes. The target is then created
and loads are applied.
Note When you make changes to any tabbed page, your changes are applied to the ANSYS database
only when you click OK to close the dialog box.
179
17.4.2. Basic Tab of the Drop Test Set-up Dialog Box

Figure 17.2 Drop Test Set-up Dialog Box - Basic Tab
Use the Basic tab to specify the options described below.

Gravity -- Specify the magnitude of gravity.
Magnitude of g drop down list and text field: You have the option to select one of the standard values for
gravity or type in a customized value. The values for gravity you can choose from the drop down list are:
zero (default)
386.4
32.2
9.81
9810
Any value that you choose must have units that are consistent with the units used in the model of the object
being dropped. You only need to input the magnitude of gravity. The direction is assumed to be in the negative
Y direction, referenced to the screen coordinates. The object's coordinate components of the gravity loading
are imposed on the object internally by the DTM, using three separate EDLOAD commands just before the
problem is solved.
Drop Height -- Specify values for the drop height and the height reference point.
Height field: (default = 0) You type the value of the drop height in units consistent with the units of length
used in creating the object being dropped. The height is always measured along the Y axis (referenced
to the screen coordinates) from the center of the top face of the target to some height reference point in
the object.
Reference drop down list: You choose the reference point from the following options:
1710
Lowest Object Point (default): Reference point is the point on the object with the minimum screen
Y coordinate value. If this option is used, then the node on the object with the minimum Y coordinate
value is the reference point. If there is more than one node at this Y coordinate, then one of them is
chosen arbitrarily to be the reference point.
Object CG: Reference point is the object's center of gravity as calculated by ANSYS. If the center of
gravity has been previously calculated for some other purpose, then the information is stored, and
the calculation is not repeated.
If the analysis is set to begin near impact time (default), then you will notice that the separation between the
center of the target and the reference point, measured along the screen Y axis, is not equal to the specified drop
height. This is because the target is created so that there is only a small separation, which is the case just before
impact. However, if the analysis is set to begin at drop time, the separation between the target and the reference
point equals the drop height.
By default, the target is oriented such that its top face is normal to the screen Y axis. If the orientation of the target
is modified from the default orientation (see Section 17.3.2.3: Target Orientation in this User's Guide), the location
of the center of the top face of the target at the time of the drop is not altered. Therefore, the definition of the
drop height outlined above is not changed.
Note You should always define the drop height so that all nodes on the object are above the plane of
the top face of the target prior to the drop.
Set Orientation -- Specify the orientation of the model with respect to the gravitational field.
Rotate button: You click this button to access the DTM Rotate dialog box, which allows you to dynamically
rotate the object on the screen. This dialog box contains a Help button that you can click for more information.
Pick Nodes button: You click this button to access a picking dialog box, which allows you to define a
vector either by picking two nodes, or by picking one node and having the object's center of gravity be
the second node. The up direction is defined as the direction going from the first node that you pick to
the second node that you pick. This dialog box contains a Help button that you can click for more information.
Solution Time -- Specify when the drop test analysis will start and how long it will run.
Start analysis near impact time radio button (default): You click this option to save computation time
that would be expended as the object drops through the air. To use this option, the angular velocity must
be zero. The velocity at the start of the analysis is calculated automatically assuming the object is a rigid
body and there is no resistance due to the air. If an initial translational velocity is applied, it is taken into
account in calculating the velocity at the start of the analysis.
Start analysis at drop time radio button: You click this option to include the entire time it takes for the
object to drop from the specified height in the analysis. Although extra computation time is expended
that is not directly involved with the drop test impact, this option accounts for rigid body rotation and is
useful if you want to produce animations that include frames of the entire drop test, starting from when
the object drops.
Run time after impact field: You type the time in seconds that you want the analysis to run after the
object hits the target.
Number of Results Output -- Specify the number of times that the output is written to the results file and to
the time-history file.
On Results file field: (default = 100) You type the number of times that the output is written to results
file Jobname.RST. See the EDRST command for details.
On time history file field: (default = 1000) You type the number of times that the output is written to
time-history file Jobname.HIS. See the EDHTIME command for details. Note here that a component containing the lowest point and the node closest to the object's center of gravity is automatically created for
time-history postprocessing. You may define additional nodes. Refer to the EDHIST command for details.
1711

Action Buttons
OK: Applies the settings on the Drop Test Set-up dialog box to the ANSYS database and closes the dialog
box.
Cancel: Cancels any changes that you have made on the Drop Test Set-up dialog box since the last time
that you clicked the OK button.
Help: Displays this help topic.
Related Information -- For further information about the options on the Basic tab, see the following sections
in this user's guide:

Section 17.3.2: Modifying the Target
Section 17.5: Picking Nodes
Section 17.3.1: Object Initial Velocity
Section 17.4.1: Using the Drop Test Set-up Dialog Box
17.4.3. Velocity Tab of the Drop Test Set-up Dialog Box

Figure 17.3 Drop Test Set-up Dialog Box - Velocity Tab
Use the Velocity tab to specify the object velocities, as described below.
Translational Velocity -- Specify the initial translational velocity of the object being dropped at the beginning
of the drop time, even if the analysis is set to begin near impact. If you apply a translational velocity other than
zero, and choose to start the analysis near impact time, the initial angular velocity must be zero. If these conditions
are not met, ANSYS will automatically start the analysis at drop time, and it will automatically recreate the target.
Velocity relative to drop down list: You choose either object coordinate system or screen coordinate
system (default). If you enter the velocity relative to the screen coordinate system, the components are
resolved into the object coordinate system internally by the DTM.
X Component field: (default = 0) The X component of the translational velocity.
Y Component field: (default = 0) The Y component of the translational velocity.
Z Component field: (default = 0) The Z component of the translational velocity.
Angular Velocity -- Specify the initial angular velocity of the object being dropped. The angular velocity is about
the object centroid in units of radians/(unit time). Applying an angular velocity other than zero is only valid when
1712

starting the analysis at drop time. ANSYS automatically sets this option when you choose Solve, and it automatically recreates the target to reflect this change.
Velocity relative to drop down list: You choose either object coordinate system (default) or screen coordinate system. If you enter the velocity relative to the screen coordinate system, the components are
resolved into the object coordinate system internally by the DTM before being input on the EDVEL
command.
X Component field (default = 0): The X component of the angular velocity.
Y Component field (default = 0): The Y component of the angular velocity.
Z Component field (default = 0): The Z component of the angular velocity.
Action Buttons
box.
Related Information -- For further information about the options on the Velocity tab, see the following sections
Section 17.3.1: Object Initial Velocity

17.4.4. Target Tab of the Drop Test Set-up Dialog Box

Figure 17.4 Drop Test Set-up Dialog Box - Target Tab
When you choose the Target tab, the target is automatically created.
Also, use the Target tab to specify the options described below.
Dimensions -- Specify the target's position and size.
Center target using drop down list. You can change the position of the target by changing the location
of the point on the center of the top face of the target in the screen X-Z plane. The point's Y coordinate
1713

is fixed when the drop height is specified. You can designate this point with one of the following options
in the drop down list:
Lowest Object Point: (default) The lowest object point is the point on the object with the minimum
screen Y coordinate value. If only a single node exists on the object at this point, then the screen X
coordinate value of this node is used as the point's X coordinate, and the screen Z value is used as the
point's Z value. If there is more than one node at the minimum Y point, then the X coordinate of the
point is found by first selecting only those nodes located at the minimum Y point. Then, the point's X
value is equal to the sum of the maximum and minimum screen X values in the selected set of nodes,
divided by 2. Similarly, the point's Z value is equal to the sum of the maximum and minimum screen
Z values in the selected set of nodes, divided by 2.
Object CG: The screen X value at the object center of gravity is used as the point's X value, and the
screen Z value of the cg is used as the point's Z value.
Length scale factor field: (default = 1.0) You type a scale factor that changes the length of the target from
the default length. The default length is determined by ANSYS, based on the overall dimensions of the
object being dropped.
Thickness scale factor field: (default = 1.0) You type a scale factor that changes the thickness of the target
from the default thickness. The default thickness is equal to 0.1 times the default length.
Orientation angle -- Specify the rotation of the target about the screen Z axis (normal to the screen and pointing
toward you).
Rotation about screen Z field: (default = 0) You type the rotation angle in degrees about the screen Z axis.
The drop height loses its meaning when the normal to the targets large faces is perpendicular to the screen Y
axis. Therefore, if you modify the orientation so that the angle is greater than 89o or less than -89o with the screen
Y axis, then the input is considered to be invalid. A warning is printed to the screen, and the target is returned
to its previous orientation.
Material Properties -- Specify the material properties of the target. Material properties are used for a contact
stiffness calculation and a time step calculation.
Young's Modulus field: (default = object's lowest Young's modulus) You type the target's Young's modulus in units consistent with the analysis.
Density field: (default = object's highest density) You type the target's density in units consistent with
the analysis.
Poisson's Ratio field: (default = object's Poisson's ratio) You type the target's Poisson's ratio.
Contact -- Specify the contact friction coefficients described below.
Static friction coeff field: (default = 0.0) You type the static friction coefficient, which is FS as described
in the Notes section of the EDCGEN command.
Dynamic friction coeff field: (default = 0.0) You type the dynamic friction coefficient, which is FD as described in the Notes section of the EDCGEN command.
Exponential decay coeff field: (default = 0.0) You type the exponential decay coefficient, which is DC as
described in the Notes section of the EDCGEN command.
Action Buttons
1714
box.
Related Information -- For further information about the options on the Target tab, see the following sections
in this user's guide, except as noted:
Section 17.3.2: Modifying the Target

Section 6.2.1: Definition of Contact Types
17.4.5. Status Tab of the Drop Test Set-up Dialog Box

Figure 17.5 Drop Test Set-up Dialog Box - Status Tab
Use the Status tab to check the values currently set for acceleration due to gravity, initial translational and rotational velocity, and the total time at the end of solution. One of the following statements is also displayed:
Successful set-up. Ready to drop. [black print] This statement indicates that all the current set-up data
is acceptable and you can click OK to commit the settings.
Proceed with caution. See error messages. [yellow print] This statement indicates that some of the
current set-up data could be problematic in enabling a solution, or is otherwise non-standard (for example,
a zero entry detected for the value of gravity). You should examine the error messages in the ANSYS
Output window to determine what is causing them to occur. Then, if necessary, change settings as appropriate and recheck the Status tab until the first message in black appears for problem situations. Click OK
to commit the settings.
Unsuccessful set-up. See error messages. [red print] This statement indicates that the current set-up
data has produced errors that are serious enough to prevent a solution. The target will not be created in
this situation. You should examine the error messages in the ANSYS Output window to determine the
causes of these problems. Then change settings as appropriate and recheck the Status tab until the first
message in black appears. Click OK to commit the settings.
It is recommended that you check this information in the Status tab along with a view of the object and the
target in the in the ANSYS Graphics window before you click OK in the Drop Test Set-up dialog box as this
commits all the set up data to the ANSYS database.
Action Buttons
1715
box.
Related Information -- For further information about the options on the Status tab, see the following section
17.5. Picking Nodes

The Pick Nodes dialog box appears after you have clicked on the Pick Nodes... button on the Basic tab of the
Drop Test Set-up dialog box. You use the Pick Nodes dialog box in specifying the orientation of the model with
respect to the gravitational field by picking two nodes to define a vector in the screen y direction, or by picking
one node and having the object's center of gravity (cg) be the second node. The up direction is defined as the
direction going from the first node that you pick to the second node that you pick.
Pick Two Nodes option: To specify the model's orientation by picking two nodes:
1.
Pick the first node.
2.
Choose OK in the Pick Nodes dialog box.
3.
Pick the second node.
4.
The object's orientation changes according to the vector defined by the two nodes that you picked. The
object is aligned such that the vector you created will be parallel to the Y direction of the screen coordinate
system.
Pick One Node and Use CG option: To specify the model's orientation by picking one node and using
the object's center of gravity (calculated by ANSYS):
1.
Pick the first node.
2.
3.
Choose OK again in the Pick Nodes dialog box because the second node is the object's center of
gravity by default.
The object's orientation changes according to the vector defined by the node that you picked and the
object's center of gravity.
Related Information -- For further information related to the Pick Nodes dialog box, see the following sections:
Section 17.4.2: Basic Tab of the Drop Test Set-up Dialog Box in this user's guide
Chapter 5, Graphical Picking in the ANSYS Operations Guide
17.6. Postprocessing - Animation

You access the Animate Over Results dialog box after solution by choosing Main Menu> Drop Test> Animate
Results.
1716
Section 17.7: Postprocessing - Graph and List Time-History Variables

The Animate Over Results dialog box is based on the ANSYS ANDATA command and allows you to specify
animation options as described below.
Model result data radio buttons -- Specify the type of results data to be used for the animation sequence. Options
are:
Current Load Step (default).
Load Step Range.
Result Set Range.
Range Minimum, Maximum text fields -- Specify the first and last data sets over which the animation should
span. If the Minimum field is left blank or equals 0, the default is the first data set. If the Maximum field is left
blank or equals 0, the default is the last data set.
Increment result set text field -- Specify the increment between result data set (default = 1).
Include last SBST (substep) for each LDST (load step) check box -- Specify whether or not to force the last
result data set in a load step to be animated. The default is unchecked, meaning that the last data set is not included in the animation.
Animation time delay (sec) text field -- Specify the time delay between frames during the animation (default
= 0.5 seconds).
Display Type two-column selection list -- Specify what parameter is to be animated in the left column (for example,
Stress, Stress-total, Stress-plastic), and the specific subset of the chosen parameter in the right column (for example,
1st Principal Stress, von Mises Stress).
Action Buttons
OK: Applies the settings on the Animate Over Results dialog box to the ANSYS database and closes the
dialog box.
Cancel: Cancels any changes that you have made on the Animate Over Results dialog box since the last
time that you clicked the OK button.
Help: Displays information on the ANDATA command from the ANSYS Commands Reference.
After you choose OK, the animation will be created and will be displayed. An Animation Controller dialog box
will also be displayed that includes controls for starting and stopping the animation, as well as changing its direction. This dialog box includes its own Help button where more detailed information is available.
Related Information -- For further information related to the Animate Over Results dialog box, see the following
sections:
Section 17.2.1.8: STEP 8: Animate results in this user's guide.

Chapter 15, Animation in the ANSYS Basic Analysis Guide.
17.7. Postprocessing - Graph and List Time-History Variables

You access the Graph Time-History Variables dialog box or the List Time-History Variables dialog box after
solution by choosing Main Menu> Drop Test> Time History.
1717
Figure 17.6 Graph and Time-History Variables Dialog Box
Note The contents of the Graph Time-History Variables dialog box and the List Time-History
Variables dialog box are the same so this help topic describes both of them. Only the Graph TimeHistory Variables dialog box is shown above.
Use the applicable dialog box to display the time-history results of a drop test analysis as described below:
Nodes to graph/list drop down list: You select one of the following choices as the node(s) to graph or list:
CG and Lowest Pt.: (default) Both the center of gravity and the lowest point of the object, as initially
specified will be graphed or listed.
Center of grav.: Only the center of gravity of the object will be graphed or listed.
Initial Low Pt.: Only the lowest point of the object, as initially specified will be graphed or listed.
Item and Comp to be graphed/listed two-column selection list: You first select Displacement (default), Velocity,
or Acceleration as the item to be graphed or listed, then select the x, y, or z component of the chosen item referenced to the object or screen coordinate system. The default item and component is displacement in the Y
direction referenced to the object coordinate system.
Action Buttons
OK: Applies the settings on the Graph/List Time-History Variables dialog box to the ANSYS database
and closes the dialog box.
Cancel: Cancels any changes that you have made on the Graph/List Time-History Variables dialog box
since the last time that you clicked the OK button.
Related Information -- For further information related to the Graph/List Time-History Variables dialog box,
see the following sections in this user's guide:
1718

Section 17.2.2: Screen Coordinates Definition
Appendix A. Comparison of Implicit and Explicit

Methods
This appendix addresses some fundamental differences in the implementation of the implicit solution method
versus the explicit solution method. These topics are available:
A.1. Time Integration

A.1.1. Implicit Time Integration
Inertia effects of mass and damping ([C] and [M]) are typically not included for implicit time integration. Average
acceleration displacements evaluated at time t+t are given by:
{ut + t } = [K ]1 {Fta t }
+
For linear problems, the solution is unconditionally stable when the stiffness matrix [K] is linear, and large time
steps can be taken.
For nonlinear problems:
The solution is obtained using a series of linear approximations (Newton-Raphson method).
The solution requires inversion of the nonlinear stiffness matrix [K].
Small iterative time steps are required to achieve convergence.
Convergence tools are provided, but convergence is not guaranteed for highly nonlinear problems.
A.1.2. Explicit Time Integration

For the explicit method, a central difference time integration method is used. Accelerations evaluated at time t
are given by:
{at } = [M]1 ({Ftext } {Ftint })

where
{Ftext }
is the applied external and body force vector
{Ftint }
is the internal force vector which is given by:
Fint = ( BT n d + Fhg ) + Fcontact )
Fhg is the hourglass resistance force, and Fcont is the contact force.
The velocities and displacements are then evaluated:
Appendix A. Comparison of Implicit and Explicit Methods
{v t + t/2 } = {v t t/2 } + {at } t t

{ut + t } = {ut } + {v t + t / 2 } t t + t / 2
where
t t + t / 2 = .5 ( t t + t t + t )
and
t t t / 2 = .5 ( t t t t + t )
The geometry is updated by adding the displacement increments to the initial geometry {X0}:
{ x t + t } = { xo } + {ut + t }
For nonlinear problems:
A lumped mass matrix is required for simple inversion.
The equations become uncoupled and can be solved for directly (explicitly).
No inversion of the stiffness matrix is required. All nonlinearities (including contact) are included in the
internal force vector.
The major computational expense is in calculating the internal forces.
No convergence checks are needed since the equations are uncoupled.
Very small time steps are required to maintain the stability limit.
A.2. Stability Limit

A.2.1. Implicit Method
For linear problems, the implicit solution is always stable; that is, the time step can be arbitrarily large.
For nonlinear problems, the time step size may become small due to convergence difficulties.
A.2.2. Explicit Method

The explicit solution is only stable if the time step size is smaller than the critical time step size:
t t crit =
2
max
where max = largest natural circular frequency.

Due to this very small time step size, the explicit method is useful only for very short transients.
A2
Section A.4: ANSYS/LS-DYNA Time Step Size
A.3. Critical Time Step Size of a Rod

The critical time step is based on the Courant-Friedrichs-Levy criterion. As an example, we will compute the
critical time step size of a rod.
The natural frequency is given by:
max = 2
c
l
where
c=
(c is the wave propagation velocity).

Substituting max into the equation for critical time step yields:
t =
l
c
t is the time needed for the wave to propagate through the rod of length l. Note that the critical time step size
for explicit time integration depends on element length and material properties (sonic speed).
A.4. ANSYS/LS-DYNA Time Step Size

ANSYS LS-DYNA checks all elements when calculating the required time step. For stability reasons, a scale factor
of 0.9 (default) is used to decrease the time step:
t = 0 . 9
l
c
The characteristic length, l, and the wave propagation velocity, c, are dependent on element type.
For beam elements:
l = length of the element, and
c=
For shell elements:
I=
A
max (I1,I 2,I 3,I 4 )
For triangular shell elements:
A3
Appendix A. Comparison of Implicit and Explicit Methods
I=
2A
max (I1,I 2,I 3 )
For all shell elements:
c=
A4
E
(1 2 )
Appendix B. Material Model Examples

This appendix provides example input for most of the material models available in ANSYS LS-DYNA.
Note The input values given in this appendix are provided as examples only. The validity of these example materials is not guaranteed by ANSYS, Inc. It is the users responsibility to ensure that material
data used in an ANSYS LS-DYNA analysis are accurate and appropriate for the structure being modeled.
The following material model topics are available:
B.1. ANSYS LS-DYNA Material Models

The following table lists ANSYS LS-DYNA material models and their corresponding LS-DYNA commands. The LSDYNA material model number is also listed. The Example column indicates whether example material input is
provided for each model. The example input listings are shown in the section that follows this table.
ANSYS Material
Model
LS-DYNA Command
LS-DYNA
MAT #
Example
Isotropic Elastic
*MAT_ELASTIC
Yes
Orthotropic Elastic
*MAT_ORTHOTROPIC_ELASTIC
Yes
Anisotropic Elastic
*MAT_ANISOTROPIC_ELASTIC
Yes
Elastic Fluid
*MAT_ELASTIC_FLUID
No
Blatz-Ko Rubber
*MAT_BLATZ-KO_RUBBER
Yes
Mooney-Rivlin Rub- *MAT_MOONEY-RIVLIN_RUBBER

ber
27
Yes
Viscoelastic
*MAT_VISCOELASTIC
Yes
Bilinear Isotropic
Plasticity
*MAT_PLASTIC_KINEMATIC
Yes
Temperature De*MAT_ELASTIC_PLASTIC_THERMAL
pendent Bilinear Isotropic
No
Transversely Aniso- *MAT_TRANSVERSELY_ANISOTROPIC _ELASTIC_PLASTIC

tropic Elastic Plastic
37
Yes
Transversely Aniso- *MAT_FLD_TRANSVERSELY _ANISOTROPIC

tropic FLD
39
Yes
Bilinear Kinematic
Yes
Plastic Kinematic
Yes
3 Parameter Barlat
Plasticity
*MAT_3-PARAMETER_BARLAT
36
Yes
Barlat Anisotropic
Plasticity
*MAT_BARLAT_ANISOTROPIC _PLASTICITY
33
Yes
Rate Sensitive
*MAT_RATE_SENSITIVE_POWERLAW _PLASTICITY
Powerlaw Plasticity
64
Yes
Strain Rate Depend- *MAT_STRAIN_RATE_DEPENDENT _PLASTICITY

ent Plasticity
19
Yes
Piecewise Linear
Plasticity
24
Yes
*MAT_PIECEWISE_LINEAR_PLASTICITY

ANSYS Material
Model
LS-DYNA Command
LS-DYNA
MAT #
Example
Modified Piecewise *MAT_MODIFIED_PIECEWISE_LINEAR _PLASTICITY

Linear Plasticity
123
Yes
Composite Damage *MAT_COMPOSITE_DAMAGE
22
No
Concrete Damage
72
No
Powerlaw Plasticity *MAT_POWER_LAW_PLASTICITY
18
Yes
Elastic Viscoplastic
Thermal
106
Yes
Elastic-Plastic Hydro- *MAT_ELASTIC_PLASTIC_HYDRO

dynamic
10
No
Geological Cap
*MAT_GEOLOGICAL_CAP_MODEL
25
Yes
Closed Cell Foam
*MAT_CLOSED_CELL_FOAM
53
No
Viscous Foam
*MAT_VISCOUS_FOAM
62
No
Low Density Foam
*MAT_LOW_DENSITY_FOAM
57
No
Crushable Foam
*MAT_CRUSHABLE_FOAM
63
No
Honeycomb Foam
*MAT_HONEYCOMB
26
No
Tabulated EOS
*EOS_TABULATED
Bamman
*MAT_BAMMAN
*MAT_CONCRETE_DAMAGE
*MAT_ELASTIC_VISCOPLASTIC _THERMAL
No
51
No
Johnson-Cook Linear *MAT_JOHNSON_COOK

Polynomial EOS
*EOS_LINEAR_POLYNOMIAL
15
Yes
Johnson-Cook
Gruneisen EOS
15
Yes
Null Linear Polynomi- *MAT_NULL

al EOS
*EOS_LINEAR_POLYNOMIAL
Yes
Null Gruneisen EOS *MAT_NULL
Yes
65
No
Steinberg Gruneisen *MAT_STEINBERG

EOS
*EOS_GRUNEISEN
11
Yes
Linear Elastic Spring *MAT_SPRING_ELASTIC
N/A
No
General Nonlinear
Spring
*MAT_SPRING_GENERAL_NONLINEAR
N/A
No
Nonlinear Elastic
Spring
*MAT_SPRING_NONLINEAR_ELASTIC
N/A
No
Elastoplastic Spring *MAT_SPRING_ELASTOPLASTIC
N/A
No
Inelastic Tension or *MAT_SPRING_INELASTIC

Compression-only
Spring
N/A
No
Maxwell Viscosity
Spring
*MAT_SPRING_MAXWELL
N/A
No
Linear Viscosity
Damper
*MAT_DAMPER_VISCOUS
N/A
No
*MAT_JOHNSON_COOK
*EOS_GRUNEISEN
*EOS_GRUNEISEN
Zerilli-Armstrong
B2
*MAT_ZERILLI_ARMSTRONG
Section B.2: Material Model Examples

ANSYS Material
Model
LS-DYNA Command
LS-DYNA
MAT #
Example
Nonlinear Viscosity
Damper
*MAT_DAMPER_NONLINEAR_VISCOUS
N/A
No
Cable
*MAT_CABLE_DISCRETE_BEAM
71
Yes
Rigid
*MAT_RIGID
20
Yes
B.2. Material Model Examples

B.2.1. Isotropic Elastic Example: High Carbon Steel
MP,ex,1,210e9
! Pa
MP,nuxy,1,.29
! No units
MP,dens,1,7850
! kg/m3
B.2.2. Orthotropic Elastic Example: Aluminum Oxide

MP,ex,1,307e9
! Pa
MP,ey,1,358.1e9
! Pa
MP,ez,1,358.1e9
! Pa
MP,gxy,126.9e9
! Pa
MP,gxz,126.9e9
! Pa
MP,gyz,126.9e9
! Pa
MP,nuxy,1,.20
! No units
MP,nuxz,1,.20
! No units
MP,nuyz,1,.20
! No units
MP,dens,1,3750
! kg/m3
B.2.3. Anisotropic Elastic Example: Cadmium

MP,dens,1,3400
! kg/m3
TB,ANEL,1
TBDATA,1,121.0e9
! C11 (Pa)
TBDATA,2,48.1e9
! C12 (Pa)
TBDATA,3,121.0e9
! C22 (Pa)
TBDATA,4,44.2e9
! C13 (Pa)
TBDATA,5,44.2e9
! C23 (Pa)
TBDATA,6,51.3e9
! C33 (Pa)
TBDATA,10,18.5
! C44 (Pa)
TBDATA,15,18.5
! C55 (Pa)
TBDATA,21,24.2
! C66 (Pa)
B.2.4. Blatz-Ko Example: Rubber

MP,dens,1,1150
! kg/m3
B3

MP,gxy,1,104e7
! Pa
B.2.5. Mooney-Rivlin Example: Rubber

MP,dens,1,.0018
! lb/in3
MP,nuxy,1,.499
! No units
TB,MOONEY,1
TBDATA,1,80
! C10 (psi)
TBDATA,2,20
! C01 (psi)
B.2.6. Viscoelastic Example: Glass

MP,dens,1,2390
! kg/m3
TB,EVISC,1
TBDATA,46,27.4e9
TBDATA,47,0.0
! Go (Pa)
! G (Pa)
TBDATA,48,60.5e9
! Bulk modulus (Pa)
TBDATA,61,.53
! 1/
B.2.7. Bilinear Isotropic Plasticity Example: Nickel Alloy

MP,ex,1,180e9
! Pa
MP,nuxy,1,.31
! No units
MP,dens,1,8490
! kg/m3
TB,BISO,1
TBDATA,1,900e6
! Yield stress (Pa)
TBDATA,2,445e6
! Tangent modulus (Pa)
B.2.8. Transversely Anisotropic Elastic Plastic Example: 1010 Steel

MP,ex,1,207e9
! Pa
MP,nuxy,1,.29
! No units
MP,dens,1,7845
! kg/m3
TB,PLAW,,,,7
TBDATA,1,128.5e6
! Yield stress (Pa)
TBDATA,2,202e5
! Initial strain at failure
TBDATA,3,1.41
! r-value
TBDATA,4,1
! Yield stress vs. plastic strain curve (see EDCURVE below)
*DIM,STRAIN,,5
*DIM,YLDSTRES,,5
Strain(1) = 0,.05,.1,.15,.2
YldStres(1)=207e6,210e6,214e6,218e6,220e6 ! yield stress
B4

EDCURVE,ADD,1,Strain (1),YldStres(1)
B.2.9. Transversely Anisotropic FLD Example: Stainless Steel

MP,ex,1,30e6
! Pa
MP,nuxy,1,.29
! No units
MP,dens,1,.00285
! kg/m3
TB,PLAW,1,,,10
TBDATA,1,20e3
! Initial yield stress (Pa)
TBDATA,2,5000
TBDATA,3,.2
! Hardening parameter
TBDATA,5,1
! Maximum yield stress curve (see EDCURVE below)
*DIM,mnstrn,,6
*DIM,mjstrn,,6
mnstrn(1) = -30,-10,0,20,40,50
mjstrn(1) = 80,40,29,39,45,44
EDCURVE,ADD,1,mnstrn (1),mjstrn(1)
B.2.10. Bilinear Kinematic Plasticity Example: Titanium Alloy

MP,ex,1,100e9
! Pa
MP,nuxy,1,.36
! No units
MP,dens,1,4650
! kg/m3
TB,BKIN,1
TBDATA,1,70e6
! Yield stress (Pa)
TBDATA,2,112e6
B.2.11. Plastic Kinematic Example: 1018 Steel

MP,ex,1,200e9
! Pa
MP,nuxy,1,.27
! No units
MP,dens,1,7865
! kg/m3
TB,PLAW,,,,1
TBDATA,1,310e6
! Yield stress (Pa)
TBDATA,2,763e6
TBDATA,4,40.0
! C (s-1)
TBDATA,5,5.0
!P
TBDATA,6,.75
! Failure strain
B.2.12. 3 Parameter Barlat Example: Aluminum 5182

MP,ex,1,76e9
! Pa
MP,nuxy,1,.34
! No units
B5

MP,dens,1,2720
! kg/m3
TB,PLAW,,,,3
TBDATA,1,1
! Hardening rule of 1 (yield stress)
TBDATA,2,25e6
TBDATA,3,145e6
! Yield stress (Pa)
TBDATA,4,0.170
! Barlat exponent, m
TBDATA,5, .73
! R00
TBDATA,6,.68
! R45
TBDATA,7,.65
! R90
TBDATA,8,0
! CSID
B.2.13. Barlat Anisotropic Plasticity Example: 2008-T4 Aluminum

MP,ex,1,76e9
! Pa
MP,nuxy,1,.34
! No units
MP,dens,1,2720
! kg/m3
TB,PLAW,,,,6
TBDATA,1,1.04
! k (MPa)
TBDATA,2,.65
! Initial strain at failure
TBDATA,3,.254
!n
TBDATA,4,11
! Barlat exponent, m
TBDATA,5, 1.017
!a
TBDATA,6,1.023
!b
TBDATA,7,.9761
!c
TBDATA,8,.9861
!f
TBDATA,9,.9861
!g
TBDATA,9,.8875
!h
B.2.14. Rate Sensitive Powerlaw Plasticity Example: A356 Aluminum

MP,ex,1,75e9
! Pa
MP,nuxy,1,.33
! No units
MP,dens,1,2750
! kg/m3
TB,PLAW,,,,4
TBDATA,1,1.002
! k (MPa)
TBDATA,2,.7
!m
TBDATA,3,.32
!n
TBDATA,4,5.0
! Initial strain rate (s-1)
B.2.15. Strain Rate Dependent Plasticity Example: 4140 Steel

MP,ex,1,209e9
! Pa
MP,nuxy,1,.29
! No units
MP,dens,1,7850
! kg/m3
B6

TB,PLAW,,,,5
TBDATA,1,1
! LCID yield stress vs. strain rate (see first EDCURVE command
below)
TBDATA,2,22e5
TBDATA,3,2
! LCID Elastic modulus vs. strain rate (see second EDCURVE

command below)
*DIM,StrnRate,,5
*DIM,YldStres,,5
*DIM,ElasMod,,5
StrnRate(1) = 0,.08,.16,.4,1.0
YldStres(1) = 207e6,250e6,275e6,290e6,300e6
ElasMod(1) = 209e9,211e9,212e9,215e9,218e9
EDCURVE,ADD,1,StrnRate(1),YldStres(1)
EDCURVE,ADD,2,StrnRate(1),ElasMod(1)
B.2.16. Piecewise Linear Plasticity Example: High Carbon Steel

MP,ex,1,207e9
! Pa
MP,nuxy,1,.30
! No units
MP,dens,1,7830
! kg/m3
TB,PLAW,,,,8
TBDATA,1,207e6
! Yield stress (Pa)
TBDATA,3,.75
! Failure strain
TBDATA,4,40.0
! C (strain rate parameter)
TBDATA,5,5.0
! P (strain rate parameter)
TBDATA,6,1
! LCID for true stress vs. true strain (see EDCURVE below)
*DIM,TruStran,,5
*DIM,TruStres,,5
TruStran(1)=0,.08,.16,.4,.75
TruStres(1)=207e6,250e6,275e6,290e6,3000e6
EDCURVE,ADD,1,TruStran (1),TruStres(1)
B.2.17. Modified Piecewise Linear Plasticity Example: PVC

MP,EX,1,443478
! psi
MP,DENS,1,.03
! lb/in3
MP,NUXY,1,.45
! No units
TB,PLAW,1,,,11
! Modified PLP
TBDATA,1,1000
! Yield strength
TBDATA,2,50000
! Tangent modulus
B7

TBDATA,8,.25
! Thinning strain at failure - 25%
TBDATA,11,1
! Rate (1/s)
TBDATA,12,1
LCID3 (stress-strain curve for above Rate )
TBDATA,13,4
! Rate (4/s)
TBDATA,14,2
LCID4 (stress-strain curve for above Rate)
TBDATA,15,45
! Rate 45/s
TBDATA,16,3
LCID5 (stress-strain curve for above Rate)
! Define stress-strain curves for each rate

*DIM,strn1,,10
*DIM,strs1,,10
*DIM,strn4,,10
*DIM,strs4,,10
*DIM,strn45,,9
*DIM,strs45,,9
strn1(1)=0,.004,.006,.008,.01,.014,.018,.022,.026,.03
strn4(1)=0,.004,.006,.008,.01,.014,.018,.022,.026,.03
strn45(1)=0,.004,.006,.008,.01,.014,.018,.022,.026
strs1(1)=0,375,520,675,745,900,1050,1200,1325,1400
strs4(1)=0,475,675,825,950,1200,1450,1675,1850,1925
strs45(1)=0,600,950,1175,1475,1775,2100,2400,2650
EDCURVE,ADD,1,strn1(1),strs1(1)
B.2.18. Powerlaw Plasticity Example: Aluminum 1100

MP,ex,1,69e9
! Pa
MP,nuxy,1,.33
! No units
MP,dens,1,2710
! kg/m3
TB,PLAW,,,,2
TBDATA,1,0.598
!k
TBDATA,2,0.216
!n
TBDATA,3,6500.0
! C (s-1)
TBDATA,4,4.0
!P
B.2.19. Elastic Viscoplastic Thermal Example

MP,dens,1,830
! kg/m3
MP,alpx,1,10e-6
! m/m-C
MP,ex,1,3.06e9
! Pa
TB,PLAW,1,,,12
! Elastic viscoplastic thermal
TBDATA,1,3.447e7
! Initial yield stress
TBDATA,12,1
! LCID1 (effective stress vs. effective plastic strain)
TBDATA,13,2
! LCID2 (elastic modulus vs. temp.)
B8

TBDATA,14,3
! LCID3 (Poisson's ratio vs. temp.)
TBDATA,15,4
! LCID4 (yield stress vs. temp.)
TBDATA,18,5
! LCID7 (coefficient of thermal expansion vs. temp.)
! Stress-strain curve
*DIM,strn1,,10
*DIM,strs1,,10
strn1(1)=0,.004,.006,.008,.01,.014,.018,.022,.026,.03
strs1(1)=0,1.224e7,1.836e7,2.455e7,3.06e7,3.54e7,3.75e7,3.90e7,4.05e7,4.12e7
! Temperature dependent curves
*DIM,temp1,,4
*DIM,nuxy1,,4
*DIM,ex1,,4
*DIM,ystrs1,,4
*DIM,alpha1,,4
temp1(1)=70,120,250,500
ex1(1)=3.06e9,1.85e9,1.12e9,6.79e8
nuxy1(1)=.445,.451,.454,.456
ystrs1(1)=3.447e7,1.85e7,1.042e7,5.627e6
alpha1(1)=1e-5,8.4e-6,7.6e-6,7.2e-6
EDCURVE,ADD,2,temp1(1),ex1(1)
EDCURVE,ADD,3,temp1(1),nuxy1(1)
EDCURVE,ADD,4,temp1(1),ystrs1(1)
EDCURVE,ADD,5,temp1(1),alpha1(1)
! Note: Be sure to define a temperature range that exceeds the actual resulting temperatures. In this particular case, LS-DYNA will expect all temperatures to be between 0
and 500 degrees C. This requirement is on the LS-DYNA side.
B.2.20. Geological Cap Example: SRI Dynamic Concrete

MP,gxy,1,1700
! ksi
MP,dens,1,2.226e-7
! kip-sec2/in4
TB,GCAP,1
TBDATA,1,2100
! K (ksi)
TBDATA,2,.7
! (ksi)
TBDATA,3,.1
TBDATA,4,.2
! (ksi)
TBDATA,5,1.473
! (ksi-1)
TBDATA,6,10.8
!R
TBDATA,7,.00154
! D (ksi-1)
TBDATA,8,.884
!W
TBDATA,9,18
! X0 (ksi)
TBDATA,10,0
!C
B9

TBDATA,11,0
!N
TBDATA,12,1
! Ftype
TBDATA,13,-0.3
! Toff
B.2.21. Johnson-Cook Linear Polynomial EOS Example: 1006 Steel

MP,ex,1,207e9
! Pa
MP,nuxy,1,.30
! No units
MP,dens,1,7850
! kg/m3
TB,EOS,1,,,1,1
TBDATA,1,350.25e6
! A (Pa)
TBDATA,2,275e6
! B (Pa)
TBDATA,3,.36
!n
TBDATA,4,.022
!c
TBDATA,5,1.0
!m
TBDATA,6,1400
! Melt temperature (C)
TBDATA,7,30
! Room temperature (C)
TBDATA,8,10
! Initial strain rate
TBDATA,9,4500
! Specific heat
TBDATA,10,240e6
! Failure stress
TBDATA,11,-.8
! Failure value D1
TBDATA,12,2.1
! Failure value D2
TBDATA,13,-.5
! Failure value D3
TBDATA,14,.0002
! Failure value D4
TBDATA,15,.61
! Failure value D5
TBDATA,17,140e9
! EOS linear polynomial term
B.2.22. Johnson-Cook Gruneisen EOS Example: OFHC Copper

MP,ex,1,138e9
! Pa
MP,nuxy,1,.35
! No units
MP,dens,1,8330
! kg/m3
TB,EOS,1,,,1,2
TBDATA,1,89.63e6
! A (Pa)
TBDATA,2,291.64e6
! B (Pa)
TBDATA,3,.31
!n
TBDATA,4,.025
!c
TBDATA,5,1.09
!m
TBDATA,6,1200
! Melt temperature (C)
TBDATA,7,30
! Room temperature (C)
TBDATA,8,10
! Initial strain rate
TBDATA,9,4400
! Specific heat
TBDATA,10,240e6
! Failure stress
TBDATA,11,-.54
! Failure value D1
B10

TBDATA,12,4.89
! Failure value D2
TBDATA,13,-3.03
! Failure value D3
TBDATA,14,.014
! Failure value D4
TBDATA,15,1.12
! Failure value D5
TBDATA,16,.394
!C
TBDATA,17,1.489
! S1
TBDATA,18,0.0
! S2
TBDATA,19,0.0
! S3
TBDATA,20,2.02
! 0
TBDATA,21,.47
!A
B.2.23. Null Material Linear Polynomial EOS Example: Brass

MP,ex,1,200e9
! Pa
MP,nuxy,1,.3
! No units
MP,dens,1,7500
! kg/m3
TB,EOS,1,,,2,1
TBDATA,1,0.0
! Pressure cut-off
TBDATA,3,1.5
! Relative volume in tension
TBDATA,4,.7
! Relative volume in compression
TBDATA,17,16e5
! EOS linear polynomial
B.2.24. Null Material Gruneisen EOS Example: Aluminum

MP,ex,1,100e9
! Pa
MP,nuxy,1,.34
! No units
MP,dens,1,2500
! kg/m3
TB,EOS,1,,,2,2
TBDATA,1,-10000
! Pressure cut-off (Pa)
TBDATA,3,2.0
! Relative volume in tension
TBDATA,4,.5
! Relative volume in compression
TBDATA,16,.5386
!C
TBDATA,17,1.339
! S1
TBDATA,18,0.0
! S2
TBDATA,19,0.0
! S3
TBDATA,20,1.97
! 0
TBDATA,21,.48
!A
B.2.25. Steinberg Gruneisen EOS Example: Stainless Steel

MP,gxy,1,11.16e6
! (Pa)
MP,dens,1,.285
! (kg/m3)
TB,EOS,1,,,5,2
TBDATA,1,49.3e3
! Initial yield stress (Pa)
B11

TBDATA,2,43
! Hardening coefficient
TBDATA,3,.35
!n
TBDATA,5,.36e6
! Maximum yield stress (Pa)
TBDATA,10,32
! Atomic weight
TBDATA,11,2380
! Absolute melting temp.
TBDATA,15,2
! Spall type
TBDATA,16,1
! Cold compression energy flag
TBDATA,17,-50
! Min. temp. parameter
TBDATA,18,200
! Max. temp. parameter
TBDATA,29,.457
!C
TBDATA,30,1.49
! S1
TBDATA,31,0.0
! S2
TBDATA,32,0.0
! S3
TBDATA,33,1.93
! 0
TBDATA,34,1.4
!A
B.2.26. Cable Material Example: Steel

MP,ex,1,207e9
! Pa
MP,nuxy,1,.3
! No units
EDMP,cable,1,1
! See EDCURVE below
*DIM,EngStran,,4
*DIM,EngStres,,4
EngStran(1) = .02,.04,.06,.08
EngStres(1) = 207e6,210e6,215e6,220e6
EDCURVE,ADD,1,EngStran (1),EngStres(1)
B.2.27. Rigid Material Example: Steel

MP,ex,1,207e9
! Pa
MP,nuxy,1,.3
! No units
MP,dens,1,7580
! kg/m3
EDMP,rigid,1,7,7
B12
Appendix C. ANSYS LS-DYNA to LS-DYNA

Command Mapping
When you initiate an ANSYS LS-DYNA solution via the SOLVE command (and also when you issue the EDWRITE
command), an LS-DYNA input file called Jobname.k is created. This file contains all of the model information
that you have entered into ANSYS. However, the information in this file is in terms of LS-DYNA keyword commands,
not ANSYS commands.
The table below contains information that will help you interpret the Jobname.k file. The left column lists ANSYS
commands that are unique to ANSYS LS-DYNA, and the right column lists the corresponding LS-DYNA commands.
ANSYS commands that perform an action (such as EDCLIST) have no corresponding LS-DYNA command and
are not included in this table. ANSYS and LS-DYNA commands that are used to define material properties are
listed in Material Model Examples.
ANSYS Command
Corresponding LS-DYNA Command
EDADAPT
*PART
EDALE
*CONTROL_ALE
EDASMP
*SET_PART_LIST
EDBOUND
*BOUNDARY_SLIDING_PLANE *BOUNDARY_CYCLIC
EDBVIS
*CONTROL_BULK_VISCOSITY
EDBX
*DEFINE_BOX
EDCADAPT
*CONTROL_ADAPTIVE
EDCGEN
*CONTACT
EDCMORE
*CONTACT
EDCNSTR
*CONSTRAINED_EXTRA_NODES_SET
*CONSTRAINED_NODAL_RIGID_BODY
*CONSTRAINED_SHELL_TO_SOLID
*CONSTRAINED_RIVET
EDCONTACT
*CONTROL_CONTACT
EDCPU
*CONTROL_CPU
EDCRB
*CONSTRAINED_RIGID_BODIES
EDCSC
*CONTROL_SUBCYCLE
EDCTS
*CONTROL_TIMESTEP
EDCURVE
*DEFINE_CURVE
EDDAMP
*DAMPING_PART_MASS *DAMPING_PART_STIFFNESS
EDDC
*DELETE_CONTACT
EDDRELAX
*CONTROL_DYNAMIC_RELAXATION
EDDUMP
*DATABASE_BINARY_D3DUMP
EDENERGY
*CONTROL_ENERGY
EDGCALE
*CONTROL_ALE
EDHGLS
*CONTROL_HOURGLASS
EDHIST
*DATABASE_HISTORY_NODE *DATABASE_HISTORY
Appendix C. ANSYS LS-DYNA to LS-DYNA Command Mapping

ANSYS Command
EDHTIME
*DATABASE
*DATABASE_BINARY_D3THDT
EDINT
*INTEGRATION_BEAM
*INTEGRATION_SHELL
EDIPART
*PART_INERTIA
EDIS
*STRESS_INITIALIZATION
*STRESS_INITIALIZATION_DISCRETE
EDLCS
*DEFINE_COORDINATE_VECTOR
*DEFINE_COORDINATE_SYSTEM
*DEFINE_VECTOR
EDLOAD, , (FX, FY, FZ,

MX, MY, MZ)
*LOAD_NODE_SET
EDLOAD, , (UX, UY, UZ, *BOUNDARY_PRESCRIBED_MOTION_SET

ROTX, ROTY, ROTZ, VX,
VY, VZ, AX, AY, AZ)
EDLOAD, , (ACLX, ACLY, *LOAD_BODY_GENERALIZED
ACLZ, OMGX, OMGY,
OMGZ)
EDLOAD, ,TEMP
*LOAD_THERMAL_VARIABLE
EDLOAD, , PRESS
*LOAD_SEGMENT
*LOAD_SEGMENT_SET
*LOAD_SHELL_SET
EDLOAD, , (RBUX, RBUY, *BOUNDARY_PRESCRIBED_MOTION_RIGID

RBUZ, +RBRX, RBRY,
RBRZ, RBVX, RBVY, RBVZ,RBOX, RBOY, RBOZ)
EDLOAD, , (RBFX, RBFY, *LOAD_RIGID_BODY
RBFZ, RBMX, RBMY, RBMZ)
EDMP,HGLS
*HOURGLASS
EDNB
*BOUNDARY_NON_REFLECTING
EDNROT
*BOUNDARY_SPC_SET
EDOPT
*DATABASE_FORMAT
EDOUT
*DATABASE_OPTION
EDPART
*PART
EDPVEL
*SET_NODE
*INITIAL_VELOCITY
*INITIAL_VELOCITY_GENERATION
*CHANGE_VELOCITY
*CHANGE_VELOCITY_ZERO
C2
Appendix C. ANSYS LS-DYNA to LS-DYNA Command Mapping

ANSYS Command
EDRC
*RIGID_DEFORMABLE_CONTROL
EDRD
*DEFORMABLE_TO_RIGID
*RIGID_DEFORMABLE_D2R
*RIGID_DEFORMABLE_R2D
EDRI
*DEFORMABLE_TO_RIGID_INERTIA
EDRST
*DATABASE_BINARY_D3PLOT
EDSHELL
*CONTROL_SHELL
EDSP
*CHANGE_SMALL_PENETRATION
EDSTART
r = d3dumpnn lsdyna command line option
EDTERM
*TERMINATION_NODE
*TERMINATION_BODY
EDVEL
*SET_NODE
*INITIAL_VELOCITY
*INITIAL_VELOCITY_GENERATION
*CHANGE_VELOCITY
*CHANGE_VELOCITY_ZERO
EDWELD
*CONSTRAINED_SPOTWELD
*CONSTRAINED_GENERALIZED_WELD_SPOT
BFUNIF or TUNIF
*LOAD_THERMAL_CONSTANT
LDREAD
*LOAD_THERMAL_CONSTANT_NODE
REXPORT
m = drelax lsdyna command line option
C3
C4
Appendix D. Thermal/Structural Preload

Example
This appendix contains an example of a combined thermal/structural preload analysis followed by a structural
transient. The example demonstrates the implicit-to-explicit procedure discussed in Chapter 15, Implicit-toExplicit Sequential Solution, with emphasis on how to handle temperature loads in this type of sequential
solution (see Section 15.1.1: Special Considerations for Thermal Loading).
The input file creates a cantilever beam structure modeled with thermal brick elements. The analysis is performed
in four separate phases. In the first phase, a steady-state heat transfer analysis is conducted using ANSYS thermal
elements. A thermal gradient is produced in which the left side of the beam is 20 degrees hotter than the right
side.
In the second phase, the thermal elements are converted into ANSYS implicit structural elements. Next, the
nonuniform temperatures from the end of the heat transfer analysis are read into the structural model with the
LDREAD command. A pressure load is also applied to the top of the beam, and one end of the beam is restrained.
A static thermal/structural preload analysis is then conducted.
The results of the second phase show that the temperature gradient causes the beam to bend to the side, and
the pressure load causes the beam to bend downward. This is the desired preload condition, which should match
the results at the beginning of the subsequent transient dynamic analysis.
In the third phase, the implicit elements are converted into explicit elements and a stress initialization to a prescribed geometry is conducted. This pseudo-dynamic relaxation analysis uses the displacements from the drelax
file (but not the temperatures) to establish the preload. The temperatures are applied directly as static thermal
loads in the Jobname.K file itself. The results of this analysis (at time = 0) match those of the ANSYS implicit
analysis very well.
In phase four, a transient dynamic analysis is conducted. The pressure load is ramped up, while the temperature
load is held constant. This is because only one type of thermal load is permitted in any given analysis.
The complete input file is shown below.
fini
/clear
/title, ANSYS LS-DYNA Thermal/Structural Preload Example
! units: kilogram, meter, second, Newton, Pascal, Joule, Watts
! = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
/filnam,thermal
/plopts,info,1
/psf,conv,hcoe,1
/view,,1,2,3
/prep7
et,1,SOLID70
r,1
! thermal brick element

! dummy real constant set
! Temperature-independent thermal material properties:

mp,dens,1,8.03e3
mp,kxx,1,16.3
mp,c,1,502.0
! kg/m^3
! W/m-C
! J/kg-C
Appendix D. Thermal/Structural Preload Example

block,0,1,0,1,0,11
esize,0.25
vmesh,all
cm,nbeam,node
cm,ebeam,elem
fini
/solu
antype,static
time,1.0
nsel,s,loc,x,0
esln
bf,all,hgen,3.0e3
esel,all
nsel,s,ext
nsel,r,loc,x,1.0
sf,all,conv,30.0,0.0
nsel,all
! steady state heat transfer analysis
! left side of beam

! 3,000 W/m^3 heat generation on left side
! exterior beam nodes

! right side of beam
! hf=30.0 W/m^2-C and Tbulk=0 degrees C
/title, ANSYS Heat Transfer Analysis

save
solve
fini
/post1
set,last
plnsol,temp
/wait,3
save
fini
! = = = = = = = = = = = = = = = = = = = = = = = = = = = =
/filnam,static
/psf,pres,norm,2
/dscale,,1.0
/prep7
bfdele,all,all
sfdele,all,all
! remove thermal body force (heat generation) loads

! remove thermal surface (convection) loads
!!!etchg,tts
!!!keyopt,1,2,1
et,2,SOLID185
emodif,all,type,2
etdele,1
numcmp,type
keyopt,1,2,1
! convert SOLID70 elements into SOLID45 elements

! similar formulation to default SOLID164
!
!
!
!
!
desired structural brick element

ETCHG,TTS creates SOLID45 elements
remove thermal bricks
compress element type numbers
uniform reduced integration to match SOLID164
! Temperature-dependent structural properties

mp,dens,1, 8.03e3
mptemp, 1, 0.0, 500.0
mpdata, ex,1,1,193.0e9,93.0e9
mpdata,nuxy,1,1,0.29,0.28
mpdata,alpx,1,1,18.0e-6,16.0e-6
! Note:
!
!
!
!
!
!
!
!
kg/m^3 (canNOT be temp-dependent)

degrees C, two temperatures
N/m^2 (temperature-dependent)
unitless (temperature-dependent)
m/m-C (temperature-dependent)
Do not define plasticity in this ANSYS implicit preload

analysis, since implicit-to-explicit sequential solutions
assume linear elastic material properties and small strain.
Plasticity can be added in the subsequent transient run
mplist,all,all
fini
/solu
antype,static
D2
! thermal strain (preload) structural analysis
time,1.0
nsubst,2,1000,2
outres,all,last
ldread,temp,last,,,,thermal,rth
tref,0.0
bflist,all,temp
nsel,s,loc,y,1
cm,nbeamtop,node
esln
cm,ebeamtop,elem
sf,all,press,1.1e5
nsel,s,loc,z,0
d,all,uz,0.0
nsel,s,loc,z,0
nsel,r,loc,y,0.5
d,all,uy,0
nsel,s,loc,z,0
nsel,r,loc,x,0.5
d,all,ux,0
! temperature body loads (overwrites TUNIF)
! reference temperature for "instantaneous" alpha
! apply pressure load (N/m^2) to top of beam
! restrain aft end of beam
nsel,all
esel,all
/title, ANSYS Implicit Static Preload Analysis
save
solve
fini
/post1
set,last
/dscale,,10
plnsol,s,eqv,2
/wait,3
! Note:
!
!
! exaggerate displacement
Make sure that the yield stress is not exceeded, since

the preload needs to have a linear response for the
subsequent dynamic relaxation analysis
save
fini
! = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
/filnam,dynamic
/psf,pres,norm,2
/dscale,,1.0
/prep7
sfdele,all,all
! remove structural surface (pressure) loads
etchg,ite
! convert SOLID185 elements into SOLID164 elements
! Add plasticity to temperature-dependent properties

tb,biso,1,2
tbtemp,0.0
tbdata,1,66.7e6
tbdata,2,1.93e9
tbtemp,500.0
tbdata,1,60.0e6
tbdata,2,0.93e9
! Note:
!
!
!
!
!
!
!
!
!
temperature-dependent BISO (2 temperatures)

first temperature
N/m^2, yield stress at 0 degrees
N/m^2, tangent modulus at 0 degrees
second temperature
N/m^2, yield stress at 500 degrees
N/m^2, tangent modulus at 500 degrees
Be sure to define a temperature range that exceeds the

actual resulting temperatures. In this particular case,
LS-DYNA will expect all temperatures to be between 0 and
D3

!
500 degrees C.
This requirement is on the LS-DYNA side.
edmp,hgls,1,5
edpart,create
! add hourglass control

! create part list
eddamp,all,,0.10
eddamp,1,,1.0e-6
! mass (alpha) damping

! stiffness (beta) damping
mplist,all,all
tblist,all,all
fini
/solu
time,0.2
! termination time (seconds)
rexport,dyna,,,,,static,rst
! Note:
!
!
!
!
!
! create drelax file
The temperatures written to the drelax file are NOT used by

LS-DYNA to establish the thermal preload condition. Instead,
the temperatures written via the *LOAD_THERMAL_CONSTANT_NODE
command are used. The latter are stored in the ANSYS database
as body force BF loads from the LDREAD command. These thermal
loads remain active unless deleted (BFDELE,ALL,TEMP).
bflist,all,temp
! no need to re-issue LDREAD command
eddrelax,ansys
! stress initialization to a prescribed geometry
! Increase the pressure load after a bit to introduce plasticity in

! the beam. The beam should remain bent in the XZ plane (due to the
! thermal load) while bending further downward in the YZ plane.
*dim,etime,,5,1,1
*dim,epress,,5,1,1
etime(1)=0.0,0.025,0.050,0.100,0.201
epress(1)=1.1e5,1.1e5,2.2e5,2.2e5,4.4e5
edload,add,press,4,ebeamtop,etime,epress
dlist,all,all
! extend curve out

! force into plasticity
! pressure (N/m^2) on beam top
! keep beam restraints from previous implicit run
edrst,100
edhtime,200
edhist,nbeamtop
edhist,ebeamtop
edenergy,1,1,1,1
edout,glstat
edopt,add,,both
edwrite,both,,k
nsel,all
esel,all
/title, ANSYS LS-DYNA Dynamic Relaxation and Transient Analyses
save
solve
fini
/post1
set,first
/dscale,,10
! exaggerate displacement
/title, ANSYS LS-DYNA Dynamic Relaxation Analysis

plnsol,s,eqv,2
/wait,3
set,last
/title, ANSYS LS-DYNA Transient Dynamic Analysis (Plastic State)
plnsol,s,eqv,2
/wait,3
D4

save
fini
/exit,nosav
D5
D6
Bibliography
[1.] F. Barlat and J. Lian. "Plastic Behavior and Stretchability of Sheet Metals. Part I: A Yield Function for Orthotropic
Sheets Under Plane Stress Conditions". Int. Journal of Plasticity, Vol. 5. pg. 51-66. 1989.
[2.] F. Barlat, D. J. Lege, and J. C. Brem. "A Six-Component Yield Function for Anistropic Materials". Int. Journal of
Plasticity, Vol. 7. pg. 693-712. 1991.
[3.] R. Hill. "A Theory of the Yielding and Plastic Flow of Anisotropic Metals". Proceedings of the Royal Society of London,
Series A., Vol. 193. pg. 281197. 1948.
[4.] F. K. Chang and K. Y. Chang. "A Progressive Damage Model for Laminated Composites Containing Stress Concentration. Journal of Composite Materials, 21. pg. 834-855. 1987.
Index
A
Action buttons
Drop Test Animate Over Results dialog box, 1717
Drop Test Graph Time-History dialog box, 1718
Drop Test List Time-History dialog box, 1718
Drop Test Set-up dialog box, 1712, 1713, 1714,
1715
Status tab, 1715
Target tab, 1714
Velocity tab, 1713
Adaptive meshing
in explicit dynamic analysis, 310
Animating Results, Postprocessing, 122
ANSYS LS-DYNA overview, 11
Applying loads
explicit dynamic analysis, 34
Array parameters, 42
B
Beam Elements, 29
BFUNIF command, 412
C
Commands used in explicit dynamic analyses, 11
Components, 42
Constraints, 48
Constraints and initial conditions, 47
Contact Surface
Definitions, 61
Contact Surfaces , 61
Contact Options , 65
Definition of Contact Types, 65
Coupling and constraint equations, 411
Creating Components, POST26, 121
D
Data curves, 45
Data Smoothing, 126
defining material models, 72
Discrete Elements, 210
Drop Test Graph Time-History Variables dialog box,
1718
Drop Test List Time-History Variables dialog box, 1718
Drop Test Pick Nodes dialog box, 1716
Drop Test Set-up dialog box
Basic tab, 1710, 1710, 1710, 1711, 1711, 1711
Drop Height, 1710
Gravity, 1710
Number of Results Output, 1711

Set Orientation, 1711
Solution Time, 1711
general information, 179
Status tab, 1715
Target tab, 1713, 1713, 1714, 1714, 1714
Contact, 1714
Dimensions, 1713
Material Properties, 1714
Orientation angle, 1714
Velocity tab, 1712, 1712, 1712
Angular Velocity, 1712
Translational Velocity, 1712
Dynamic relaxation, 413
E
EDHTIME command, 37
EDLOAD command, 412
EDNDTSD command, 36
EDRST command, 37
EDSTART Command, 131
elastic material models, 71
Element Output Data, 123
Element Solution, Postprocessing, 126
Elements
Beam and Link Elements, 29
BEAM161, 29
COMBI165 Spring-Damper, 210
Discrete Elements, 210
explicit dynamics, 21
explicit solid and shell elements, 21
General Element Capabilities, 211
LINK160, 210
LINK167, 210
MASS166, 211
PLANE162, 26
SHELL163, 23
SOLID164, 21
SOLID168, 28
Triangular Shell Formulations, 24
equation of state material models, 72
Example problems, explicit dynamics, 14
Explicit dynamic analysis
adaptive meshing, 310
applying loads, 34
commands, 11
constraints, 36
data smoothing, 36
defining contact, 33
defining parts, 37
displaying load symbols, 35
DOF coupling, 36
Index
example problems, 14
initial velocities, 35
modeling, 31
overview, 11
viewing load curves, 35
explicit dynamic analysis
output control options, 36
DOF coupling
drelax file, 151
dynamic relaxation, 151
explicit dynamic analysis, 36, 36
explicit-implicit element type pairs, 141
implicit-explicit element type pairs, 151
importing stresses, 141
thermal-explicit element type pairs, 155
explicit-to-implicit sequential solution, 141
F
foam material models, 72
H
Hourglassing , 91
I
implicit-to-explicit sequential solution, 151
Initial velocity, 410
L
LDREAD command, 412
Link Elements, 29
Load curves
displaying in explicit dynamics, 35
Load symbols
displaying in explicit dynamics, 35
Load types
Loading, 41
array parameters, 42
components, 42
constraints, 48
constraints and initial conditions, 47
coupling and constraint equations, 411
data curves, 45
dynamic relaxation, 413
general loading, 44
general loading options, 41
initial velocity, 410
nonreflecting boundary, 411
temperature, 412
Loads
Index2
LS-DYNA input file, 54

LS-DYNA solver, 55
M
mass scaling, 101
material models, 71
3-parameter Barlat, 711
anisotropic elastic, 74
Bamman plasticity model , 729
Barlat anisotropic plasticity, 713
bilinear isotropic, 77
bilinear kinematic (strain rate independent), 710
Blatz-Ko rubber, 75
cable, 737
closed cell foam, 724
composite damage, 718
concrete damage, 718
crushable foam, 726
defining, 72
descriptions, 73
elastic, 71, 73
elastic fluid, 75
elastic viscoplastic thermal, 719
elastic-plastic hydrodynamic, 721
elastoplastic spring, 736
equation of state, 72
foam, 72
general nonlinear spring, 735
Gruneisen equation of state, 728
honeycomb, 726
inelastic tension- or compression-only spring, 736
Johnson-Cook plasticity, 729
linear elastic spring, 735
linear polynomial equation of state, 727
linear viscous damper, 737
low density foam, 725
maxwell viscoelastic spring, 736
modified piecewise linear plasticity, 716
Mooney-Rivlin rubber, 76
nonlinear elastic, 71
nonlinear elastic spring, 736
nonlinear inelastic, 71
nonlinear viscous damper, 737
null, 731
orthotropic elastic, 74
others, 72
piecewise linear plasticity, 715
plastic kinematic, 711
Power Law plasticity, 719
pressure dependent plasticity, 71
rate sensitive Power Law plasticity, 714
rigid, 738
Index
spring-damper (discrete), 72
Steinberg model, 733
strain rate dependent plasticity, 714
tabulated equation of state, 729
temperature dependent bilinear isotropic, 78
transversely anisotropic FLD, 79
transversely anisotropic hardening, 78
viscoelastic, 77
viscous foam, 725
Zerilli-Armstrong model, 731
Modeling
Monitoring an LS-DYNA solution , 53
N
Nodal Solution, Postprocessing, 126
nonlinear elastic material models, 71
nonlinear inelastic material models, 71
Nonreflecting boundary, 411
O
other material models, 72
output control options
Output Controls, Postprocessing, 121
Output Data, 123
Output Files, POST26, 121
R
restarting, 131
EDSTART Command, 131
effect on Output Files, 135
Restart Dump File, 131
Rigid bodies, 81
inertia properties, 81
loading, 82
switching (D2R and R2D), 82
S
scaling, 101
Shell elements
formulations, 23
Shell Formulations, Triangular, 24
SMP, 52
Solid elements
Solution
editing the LS-DYNA input file, 54
solution control and monitoring , 53
Solution process, 51
Spring-Damper, 210
spring-damper (discrete) material models, 72
Subcycling, 111
P
Parallel processing in ANSYS LS-DYNA, 52
Parameters,array, 42
Parts
POST26, Creating Components, 121
POST26, Output Files, 121
POST26, use, 125
Postprocessing, 121
Animating Results, 122
Creating Components for POST26, 121
Data Smoothing, 126
Element Output Data, 123
Finding Additional Information, 127
Nodal and Element Solutions, 126
Output Controls, 121
Reading ASCII Files for Miscellaneous Output Data,
126
Results vs. History, 121
Using POST26 with ANSYS LS-DYNA, 125
Writing the Output Files for POST26, 121
pressure dependent plasticity material models, 71
Temperature loading, 412

Termination controls, 51
thermal-to-explicit sequential solution, 155
TIME command, 36
Triangular Shell Formulations, 24
TUNIF command, 412
Index3

Ansys Ls-Dyna Users Guide

Uploaded by

Copyright:

Available Formats

Ansys Ls-Dyna Users Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ansys Ls-Dyna Users Guide

Uploaded by

Copyright:

Available Formats

ANSYS LS-DYNA User's

ANSYS, Inc. and

ANSYS LS-DYNA User's Guide

Copyright and Trademark Information

U.S. GOVERNMENT RIGHTS

ANSYS LS-DYNA User's Guide

ANSYS LS-DYNA User's Guide

ANSYS LS-DYNA User's Guide

ANSYS LS-DYNA User's Guide

ANSYS LS-DYNA User's Guide

1.1. Overview of Steps in an Explicit Dynamic Analysis

Build the model (with the PREP7 preprocessor)

Review the results (with the POST1 and POST26 postprocessors)

ANSYS Basic Analysis Guide

ANSYS Modeling and Meshing Guide

1.2. Commands Used in an Explicit Dynamic Analysis

Activates adaptive meshing

Assigns mesh smoothing to explicit dynamic elements that use the

Creates a part assembly

Defines a boundary plane for sliding or cyclic symmetry

Specifies bulk viscosity coefficients

Creates a box shaped volume to be used in a contact definition

Specifies adaptive meshing controls

Specifies contact parameters

Lists contact entity specifications

Specifies additional contact parameters for a given contact definition

Defines various types of constraints

Specifies contact surface controls

Specifies CPU time limit

Merges two rigid bodies

Specifies whether subcycling will be used

Specifies mass scaling

Specifies data curves

Defines system damping

Selects numerical precision type

Deletes or deactivates/reactivates contact entity specifications

Activates initialization to a prescribed geometry or dynamic relaxation

Specifies output frequency for the restart file (d3dump)

Specifies energy dissipation controls

Specifies plotting of load symbols

Defines global ALE controls

Specifies the hourglass coefficient

Specifies time-history output

Specifies the time-history output interval

Specifies number of integration points for output

Specifies stress initialization in a full restart analysis

Defines inertia for rigid parts

Defines a local coordinate system

Defines material properties

Defines a nonreflecting boundary

Smooths noisy data and provides a graphical representation of the

Applies a rotated coordinate nodal constraint

Specifies the type of output, ANSYS or LS-DYNA

Specifies LS-DYNA ASCII output files

Creates, updates, or lists parts

Selects and plots contact entities

Plots time dependent load curves

Applies initial velocities to parts or part assemblies

Section 1.3: A Guide to Using this Document