New version of LS-DYNA is released for all common platforms.
Version 971 R4.2.1 includes the following updates from the previous release R4.2.
General
Change the behavior of the command line option ncpu=#.
Case 1: ncpu=-n negative number of CPU's The code will always turn on the consistent option (CONST=1). It will override the consistency flag on *CONTROL_PARALLEL if it exists. Case 2: ncpu= n positive If there is no *CONTROL_PARALLEL, CONST=2 as default. If there is *CONTROL_PARALLEL, it uses the value of CONST on this card.
Add info when thermal linear solver 3 fails.
Fix MPP restart for thermal using direct solver.
Fixes for encrypted input: seatbelt materials, *KEYWORD inside the encrypted block, and proper handling of *COMMENT cards.
Airbag
Fix to *AIRBAG_HYBRID leakage in case where some parts have negative CP23 and don’t belong to airbag being considered.
Disable CPM particle (*AIRBAG_PARTICLE) data output to D3THDT.
Change of CPM particle to fabric PV work calculation by nodal velocity instead of fabric segment average velocity.
Fix for CPM with MPP pre-decomposition.
Make molar-fraction based inflator gas inflow available for *AIRBAG_HYBRID_JETTING.
Implement molar-fraction based airbag inflator input for *AIRBAG_HYBRID.
Bug fix for treating non-closed airbag volume (MPP only).
Bug fix for read in *AIRBAG_PARTICLE cards and give warning message if nonlinear cp is not a monotonic function within gas temperature of operation (MPP only).
Add particle friction for CPM.
Fixed the volume calculation for control volume airbags that have holes in the mesh (MPP only).
A bug fix for ALE ambient type 5 (*LOAD_BLAST).
*LOAD_BLAST + ALE coupling with area weighted nodal velocity.
Use area weighted nodal velocity for *LOAD_BLAST + ALE coupling to avoid edge effect.
Constrained Options
Fix for *CONSTRAINED_INTERPOLATION so that division by zero is avoided when no rotational DOFs are present.
Fix for MPP restart with *CONSTRAINED_TIED_NODES_FAILURE.
Fix errors in implicit RBE3 constraints.
Fixed affecting staged construction models containing nodal rigid bodies and shell elements.
Contact Options
*CONTACT_2D_SLIDING_ONLY does not apply to 2D shell elements. An error message is printed at termination.
Fixed a problem in 2D tied contact that cause a lack of force balance when more than one tied contact appears in a model.
MPP fix for soft=4 contact along with adaptive constraints.
Bugfix in beam-to-beam contact.
Fix for soft=4 contact and adaptive constraints so soft=4 can be used in forming calculations.
Fixed a memory error when two surface force transducers are used with MPP segment based contact. The bug could cause a segmentation fault in single precision versions.
Corrected the airbag thickness by load curve option for segment based contact. It was offsetting the contact surface by the total thickness instead of half the thickness.
Added support for interference contact with segment based contact.
Fixed thermal work due to friction in segment based contact.
MPP contact damping was only being applied when nodes are approaching the surface, instead of both ways (as SMP does). This should make the behaviours more consistent, and be more stable.
Protect against divide by zero in new segment based contact penetration check and in new DEPTH=23 option.
Fixed 2 surface force transducer contact when used with segment based contact. It was possible for segmentation violation to occur.
Fixed 2D thermal contact. It was likely to segmentation fault.
Skip reporting contact time step for segment based contacts.
Bug fix for shell edge contact (MPP only).
Fix problem in reading beam node list from wrong contact definition for single surface contact (MPP only).
Improvements in MPP groupable contact:
Changes to improve stability.
Fixed problem whereby not all nodes were being considered in the energy calculation for groupable single surface contacts, and the sliding energy in GLSTAT was wrong.
SMP tiebreak surface to surface was affected by node order.
Possible floating point exception in groupable contacts, including AUTO_TIEBREAK.
Fixes to MPP groupable contact and adaptivity.
MPP fixes for groupable AUTOMATIC_TIEBREAK contacts.
Fix for MPP groupable contact interface energy calculation.
Fixed airbag thickness by load curve option for segment based contact (MPP only).
Fix some MPP initialization problems related to interface linking and tied node pairs.
Control Options
Change the behavior of the command line option ncpu=#.
Case 1: ncpu=-n negative number of CPU's The code will always turn on the consistent option (CONST=1). It will override the consistency flag on *CONTROL_PARALLEL if it exists. Case 2: ncpu= n positive If there is no *CONTROL_PARALLEL, CONST=2 as default. If there is *CONTROL_PARALLEL, it uses the value of CONST on this card.
Disable NCPU option on *CONTROL_PARALLEL or control card 16.
Print a warning if the user defines this option. *** Warning in keyword *CONTROL_PARALLEL NCPU option will be obsolete and ignored in the next release Please use ncpu=# on the command line # or *KEYWORD card. *** Warning in 16th control card: NCPU option will be obsolete and ignored in the next release Please use ncpu=# on the command line.
Added new option “NOSOL” for *CONTROL_TERMINATION (Column 6):
Flag for a “non-solution” run, i.e. normal termination directly after initialization (transformation, mapping, trimming, mesh check, …).
Added new option for *CONTROL_ADAPTIVE.
ADPSIZE .LT.0 - absolute value defines the minimum characteristic element length to be adapted based on square root of element area. i.e. instead of comparing the shortest element edge with ADPSIZE, it compare the square root of element area with abs(ADPSIZE) if ADPSIZE is given by a negative value.
Added damping to stabilize the penalty formulation of sliding only contact.
Bug fix for element sorting. Combination of *CONTROL_SOLID: ESORT=1 and *CONTROL_SHELL: ESORT=2 was not correctly working.
Fixed PSTIFF=1 option on *CONTROL_CONTACT for SMP when mass scaling is used.
Added one control parameter in EFG fracture.
Database/Output
Fixed for output of 10-node tet nodes to D3PLOT (only SMP).
Fixed sign error in BNDOUT forces for implicit.
Fixed bug related to *DATABASE_HISTORY_DISCRETE_SET and *SET_DISCRETE.
Fixed problem with GLSTAT reports incorrect controlling element.
Write solid formulations -1 and -2 to D3HSP.
Disable CPM particle (*AIRBAG_PARTICLE) data output to D3THDT.
Added output of SWFORC file (and BINOUT) for the nugget pull-out failure function and fracture failure function of option 9 beam spot weld failure.
Minor fix to D3PROP output.
Disable NCPU option on *CONTROL_PARALLEL or control card 16.
Print a warning if the user defines this option. *** Warning in keyword *CONTROL_PARALLEL NCPU option will be obsolete and ignored in the next release Please use ncpu=# on the command line # or *KEYWORD card. *** Warning in 16th control card: NCPU option will be obsolete and ignored in the next release Please use ncpu=# on the command line.
Fix implicit’s output to RCFORC file.
Fix for crazy accelerations in NODOUT file if the interface force file is active.
Write density rather than temperature to BLSTFOR.
Fix for D3HSP output of *PARAMETER names.
MPP fix for KE inconsistencies in GLSTAT due to summation of rotational energy on shared nodes.
Improvements in MPP groupable contact:
Changes to improve stability.
Fixed problem whereby not all nodes were being considered in the energy calculation for groupable single surface contacts, and the sliding energy in GLSTAT was wrong.
Fix bug which causes discrete element forces not to be added to the D3PLOT file.
Fix for MPP full deck restart and sliprings, and force output plot state at beginning of second run for full deck restart.
Save/restore plot times for ASCII files in full deck restart file. This fixes a problem where the ACSII files were output EVERY CYCLE for a while on restart.
Define Options
Make *MAT_156 parameter ALM compatible with *DEFINE_CURVE_FUNCTION.
Allow *MAT_SPRING_MUSCLE parameter “A” (activation level) to reference *DEFINE_CURVE_FUNCTION.
Allow number of points in *DEFINE_CURVE greater than 99999.
Added error termination if SIDR in *DEFINE_CURVE has an illegal value.
Fix a restart problem with *DATABASE_BINARY_BLSTFOR.
Added *DEFINE_CURVE_DUPLICATE.
Fix related to contact generation. The error was caused by confusion between *DEFINE_BOX and *DEFINE_CONTACT_VOLUME.
*DEFINE_BOX_DRAWBEAD with CID>0 was broken in some cases.
New option “POS6P” for *DEFINE_TRANSFORMATION. This is an affine transformation (rotation and translation, no scaling, no shear) given by 3 start points and 3 target points.
Fix for restart in 3D EFG adaptivity with formulation 7.
Bug fixed in EFG local adaptivity.
Added MPP support for EFG solid restart (solid part).
Bug fixed in restart for adaptive EFG.
MPP support for EFG shell 41 restart (shell part).
Add options in data transfer for adaptive EFG method.
Fix for time step control in EFG 3D adaptivity from hex to tet.
Added pressure smoothing in adaptive EFG in 4/8noded switch.
An improved adaptive EFG in local boundary integration.
Bug fixed in irreversible unloading for EFG fracture.
Bug fixed in crack closure (EFG).
Bug fixed for fracture in elastic material (EFG).
Bug fixed in EOS for EFG 2D and 3D (hex).
Added one control parameter in EFG fracture.
Elements
Fixed for output of 10-node tet nodes to D3PLOT (SMP).
Fixed related to improper solid element deletion with auto sorting on.
Correct a bug calculating shell stress in a stress-activated *SENSOR.
Fixed 2D beam element failure for material types *MAT_3, *MAT_15, *MAT_54, *MAT_58, *MAT_158, *MAT_65, *MAT_165 and *MAT_106.
Fix for bug related to beam to hex weld conversion when only a subset of beams are changed.
Fix energy bug in *MAT_187 for shells.
*CONTACT_2D_SLIDING_ONLY does not apply to 2d shell elements. An error message is printed at termination.
Fixed bug affecting staged construction models containing nodal rigid bodies and shell elements.
Correct some minor 2D (4-node) seatbelt element issues.
Made 2D seatbelt improvements, triggered by setting COMP, the 6th column of *MAT_SEATBELT, to 2. Improvements include synchronization between 1D and 2D belt, automatic adjusting compression elimination, and seatbelt section force output.
Added seatbelt slipring friction decay constant.
Combination of *INCLUDE_TRANSFORM offset and *ELEMENT_SEATBELT_ACCELEROMETER parameters IGRAV and INTOPT.
Fixed bug related to large oscillations and instability in single precision on some machines when zero length springs are present.
Fix a potential problem with pressure loading being applied to shell element segments after they have failed. This could only have happened if there was more than one pressure load applied to a segment or if the element was a 2D continuum element with more than 1 edge loaded.
Fixes to MPP groupable contact and adaptivity.
Error terminate when scalar nodes are not defined in warped beams.
Fix typo that caused MPP full deck restart to not properly re-initialize thick shell elements when using more than 1 processor.
Fix for hexahedron remeshing.
Make user material and EOS work together in shells for 2D plane strain and axisym.
Bug fixed in EOS for EFG 2D and 3D (hex).
Implicit
Fixed sign error in BNDOUT forces for implicit.
Fix for discrete beam problem which did not converge implicitly.
Fix for restart for implicit joints.
Fix implicit’s output to RCFORC file.
Make implicit 2D-type solid, type 13 and type 15, available for *MAT_076 and *MAT_175.
Fix errors in implicit RBE3 constraints.
In implicit, added kinder, gentler logic for processing of (1) prescribed motion on globally constrained nodes and rigid bodies, (2) redundant local spc constraints
Rigorous treatment of shell-to-solid interface for implicit.
Fixed minor bug related to solid element type 18. Type 18 is for linear statics, not explicit.
Add implicit support for *CHANGE_BOUNDARY_CONDITION
Remove inappropriate warning from implicit about *INTERFACE_COMPONENT.
Enhance Implicit’s treatment of shell-to-solid constraint.
Add info when thermal linear solver 3 fails.
Implicit’s treatment of shell-to-solid constraint.
Include Options
Allowed mutilple entries for *INCLUDE and *INCLUDE_PATH.
For example, *INCLUDE include_file1 include_file2 include_file3 *INCLUDE_PATH include_path1 include_path2 include_path3
Fixes for mapping with *INCLUDE_STAMPED_PART.
*MAT_ARUP_ADHESIVE – Unit conversion of density was missed out for *INCLUDE_TRANSFORM.
Fix for mapping with *INCLUDE_STAMPED_PART.
Fix for combination of *INCLUDE_TRANSFORM offset and *ELEMENT_SEATBELT_ACCELEROMETER parameters IGRAV and INTOPT.
Fix for combination of *INCLUDE_TRANSFORM offset and *MAT_057 load curves.
Whereas material stress relaxation curves are defined in the input as stress versus time, *INCLUDE_TRANSFORM incorrectly transformed the curves as stress vs. log(time).
Initial Conditions
Permute angular rates for yaw, pitch and roll in *INITIAL_VEHICLE_KINEMATICS so they are in agreement with the user’s specification of body-fixed axis sequence. Some adjustments to rigid body mass center positioning with regard to gravity direction. Calculation of velocity field is based on the final mass center position.
Allow for gravity to point in positive global directions and maintain max coordinate values in that direction when repositioning with initial vehicle kinematics.
Fixed the reading of *INITIAL_STRESS_TSHELL and *INITIAL_STRESS_SOLID when the history data contains integers.
Fixed reading *INITIAL_STRESS_TSHELL history data in large problems.
Interface
Remove inappropriate warning from implicit about *INTERFACE_COMPONENT.
Allow printing of nodal contact forces in *INTERFACE_SSI and *INTERFACE_SSI_STATIC.
Allow *INTERFACE_SSI for purely dynamic analysis, i.e. without starting from an initial state and add tensor viscosity coefficient to *MAT_PML_ELASTIC_FLUID.
MPP fixes for INTERFACE LINKING “tied node pairs” option.
Fixed keyword read of FTENSR in *INTERFACE_SPRINGBACK.
Fix problem of SMP writing interface linking file.
Fix some MPP initialization problems related to interface linking and tied node pairs.
Loads
MPP support for NIDBO in *LOAD_BLAST_ENHANCED.
A bug fix for ALE ambient type 5. (*LOAD_BLAST)
*LOAD_BLAST + ALE coupling with area weighted nodal velocity.
Use area weighted nodal velocity for *LOAD_BLAST + ALE coupling to avoid edge effect.
Compute density in the shock front and post-shock for *LOAD_BLAST_ENHANCED.
A fix to the Mach stem incident pressure (*LOAD_BLAST_ENHANCED).
Material Models
Fix for multiple failure criteria in *MAT_ADD_EROSION applied to type 2 solids.
Fixed 2D beam element failure for material types *MAT_3, *MAT_15, *MAT_54, *MAT_58, *MAT_158, *MAT_65, *MAT_165 and *MAT_106.
Fixed shell *MAT_116 input so that it cannot segmentation fault.
Fix for *MAT_ARUP_ADHESIVE and restarts.
Fix energy bug in *MAT_187 for shells.
Modify muscle materials so that user defined curves can be used for activation levels.
Made 2D seatbelt improvements, triggered by setting COMP, the 6th column of *MAT_SEATBELT, to 2. Improvements include synchronization between 1D and 2D belt, automatic adjusting compression elimination, and seatbelt section force output.
Added output of SWFORC file (and BINOUT) for the nugget pull-out failure function and fracture failure function of option 9 beam spot weld failure.
Fix for *MAT_055 shear term in Tsai-Wu failure criterion.
Fixed thick shells when used with orthotropic user-defined materials. Material direction transformations were wrong resulting in bad stress.
Make *MAT_156 parameter ALM compatible with *DEFINE_CURVE_FUNCTION.
Allow *MAT_SPRING_MUSCLE parameter “A” (activation level) to reference *DEFINE_CURVE_FUNCTION.
Make implicit 2D-type solid, type 13 and type 15, available for *MAT_076 and *MAT_175.
Minor change for name of *MAT_224: *MAT_TABULATED_JOHNSON_COOK. Old name *MAT_TABULATED_VISCOPLASTICITY_WITH_FAILURE still works.
Fixes and minor modifications for *MAT_100_DA requested by Daimler.
Miscellaneous fixes for *MAT_224.
Fix for *MAT_153.
Fixed bug in *MAT_024 user-defined failure (could result in improper material constants being used in failure subroutine).
Fixed brick element stress initialization of *MAT_057, *MAT_073, *MAT_083.
Bug fix for user material with thermal coupling.
Fix for *MAT_077 when used with tet type 13 in single precision.
Fix bug in *MAT_083.
Fix bug in *MAT_187.
Fix for *MAT_159 related to internal energy calculation.
Fix for *MAT_081 and parameter TDEL.
*MAT_192 fixes for the initialization process; the most significant is that the density used for calculating maximum over consolidation is now the buoyant density as per 971 R3. In previous releases of LS971 R4, the full density was used.
Set *MAT_083 time step choice back to 971 R3 value.
Allowed *MAT_ADD_THERMAL for metal only.
Bug fix for *MAT_187 time step calculation (solid elements only).
Bug fix for combination of *MAT_187 and hourglass type 6 (solids).
*MAT_089: damping calculation was, in some cases, referencing an undefined variable.
SPH Update 1: Fix bug in START and DEATH time per part for SPH. 2: Fix a bug for *MAT_059 for SPH elements. 3: Implement *MAT_022 (*MAT_COMPOSITE_DAMAGE) for SPH 3D elements. 4: *SECTION_SPH_TENSOR : Write an error messag in case this option is used with material other than *MAT_NULL
Fix related to *MAT_126 related to switching solid element formulation to type 1 from 0 erroneously.
Fix for *MAT_195.
*MAT_ARUP_ADHESIVE – Unit conversion of density was missed out for *INCLUDE_TRANSFORM.
Small change for *MAT_100_DA with DG_TYP=4.
Accuracy fixes in single precision for C-S strain rate in *MAT_036 and *MAT_133.
Allow *INTERFACE_SSI for purely dynamic analysis, i.e. without starting from an initial state and add tensor viscosity coefficient to *MAT_PML_ELASTIC_FLUID.
Small modification for *MAT_106: use last point in curve lrc instead of extrapolation to avoid possible negative base in Cowper-Symonds.
Fix for *MAT_COMPOSITE, didn’t work for adaptivity run.
Combination of *INCLUDE_TRANSFORM offset and *MAT_057 load curves.
Bug fix affecting problems with highly nonlinear heat capacity such as phase change problems.
Make pore air available for solid formulation=0.
Fixed bug in pore pressure analysis that could prevent convergence of steady-state analysis type on MPP systems.
Fix bug that caused steady-state pore pressure analysis to terminate early when end of construction stage was reached.
Bug fix for pore pressure analysis stability calculation for steady-state was not always correct.
Fix bug in pore pressure analysis. If suction builds up first, then the analysis is switched to steady-state, it could fail to converge.
Fix bug in pore pressure analysis when nodal rigid bodies are present and the model contains wedges or tets with automatic sorting.
Fix initial and ambient hydrostatic pressure setup (MPP only).
Compute blast wind velocity directly from the incident wave decay parameter rather than ratio of pressures.
Fixed initialization bug for umats using both ortho + part damping.
Fixed bug in cohesive element warning message for thick shells.
Parameter
Fix for D3HSP output of *PARAMETER names.
PERTURBATION
*PERTURBATION_SHELL_THICK: fix bug for different section properties.
Rigidwall
Better fix for MPP input problem with rigid walls having stick conditions.
Section Options
SPH Update 1: Fix bug in START and DEATH time per part for SPH. 2: Fix a bug for *MAT_059 for SPH elements. 3: Implement *MAT_022 (*MAT_COMPOSITE_DAMAGE) for SPH 3D elements. 4: *SECTION_SPH_TENSOR : Write an error messag in case this option is used with material other than *MAT_NULL
Fix for large ID’s in EDGSET of *SECTION_SHELL.
Fix related to multiple failure criteria in *MAT_ADD_EROSION applied to type 2 solids.
Fix for discrete beam problem which did not converge implicitly.
MPP support for EFG shell 41 restart (shell part).
Write solid formulations -1 and -2 to D3HSP.
Added output of SWFORC file (and BINOUT) for the nugget pull-out failure function and fracture failure function of option 9 beam spot weld failure.
Fixed thick shells when used with orthotropic user-defined materials. Material direction transformations were wrong resulting in bad stress.
Solved the stability problem of nonlinear discrete elements and beams by using the max slope of the stress-strain curve even if it is the last segment of the curve.
Make implicit 2D-type solid, type 13 and type 15, available for *MAT_076 and *MAT_175.
Fixed bug in Hughes-Liu shell so strains are output if more than 10 integration points are specified.
Fix for degenerate solid type 3 into wedge elements.
Modify step calculation for discrete beam element to correct instability seen when nodal masses differ by a large amount.
Fixed minor bug related to solid element type 18. Type 18 is for linear statics, not explicit.
Fix for *MAT_077 when used with tet type 13 in single precision.
Fixed extra history data output for type 2 and 3 thick shell elements.
Fix related to *MAT_126 related to switching solid element formulation to type 1 from 0 erroneously.
Fix for discrete cable beam time step calculation.
Fixed wrong strain output for one-point integration shell.
Fix for full deck restart. This involved solid type 5.
Fixed a possible stability problem with discrete beams that have non-linear stress-strain curves and the final curve segment is stiffer than the others. This could result in such models running at a smaller time step.
Fixing reading bug for shell types 25,26,27.
Fixed a bug in type 17 shell: update thickness only after convergence.
Fixed *MAT_123 for the case that ES7 or ES8 or EPS7 or EPS8 was a non-zero value. Elements were being deleted prematurely.
Update the valid material check for forms 3 and 5 thick shells.
Fixed the calculation of angles for fully integrated thick shells to prevent seg fault.
Sensor
Correct a bug calculating shell stress in a stress-activated *SENSOR.
Set Options
Fixed for *DATABASE_HISTORY_DISCRETE_SET and *SET_DISCRETE.
Bug fix for trimming: When deleting trimmed nodes, *SET_NODE_LIST_TITLE was not correctly read.
Fix bug in 3D SPH lagrangian formulation (MPP only).
Implement 3D total lagrangian formulation for SPH elements (MPP only).
SPH Update 1: Fix bug in START and DEATH time per part for SPH. 2: Fix a bug for *MAT_059 for SPH elements. 3: Implement *MAT_022 (*MAT_COMPOSITE_DAMAGE) for SPH 3D elements. 4: *SECTION_SPH_TENSOR : Write an error messag in case this option is used with material other than *MAT_NULL
Title
Allowed the user to define parameters in *TITLE or _ID line. The code will try to translate all defined parameters. If any error encountered, no error message will be issued and it will continue.
Restart_Input_Data
Add implicit support for *CHANGE_BOUNDARY_CONDITION
Fix MPP restart for thermal using direct solver.
Fix for restart with ground motion.
MPP support for accelerometers in full deck restart, which were not properly preserving history.
Fix typo that caused MPP full deck restart to not properly re-initialize thick shell elements when using more than 1 processor.
Fix MPP full deck restart bug related to minimum timestep.
Fix for MPP full deck restart and sliprings, and force output plot state at beginning of second run for full deck restart.
Save/restore plot times for ASCII files in full deck restart file. This fixes a problem where the ACSII files were output EVERY CYCLE for a while on restart.
MPP fixes for retractors and sensors with full deck restart.
