New version of LS-DYNA is released for all common platforms.
The release provides fixes and extensions. You may contact your local distributor for receiving the executable.
These notes briefly highlight changes made to LS-DYNA since the last production release of version 971, i.e., revision 7600.398.
Database/Output
Fixed in writing intfor database if the total number of words exceeded 2G.
Fixed in reading of binary dynain file, *INCLUDE_BINARY.
Zero out kinetic energy and momentum for each material for the case of implicit statics.
Fix for intfor database plotting problem.
Corrected secforc computation for 4-noded EFG solid.
Added secforc calculations for EFG solids and EFG shells.
Fix for rigid body velocity calculation in MATSUM where deformable body nodes are shared by rigid body.
Fix in reading stress initialization file (m=file) if jobid option is used.
Modify pressure output to dbfsi based on segment orientation.
Fix matsum problem with discrete spring elements related to bug report.
Fixed the output of timestep control if no element controlled the timestep. If the timestep is controlled by dt2ms of *CONTROL_TIMESTEP, report it as “dt2ms”.
Removed d3thdt at the termination time since the accceleration is incorrect.
Outputted stress Syz= and Szx=0 to d3plot for materials 36, 37, 122 and 125 if shell formulation is not 24, 25 or 26.
Fix in the momentum and rigid body velocity output for MATSUM.
Output number of nodes and elements to glstat if adaptivity is invoked.
Fixed element numbers in d3part database.
Updates to thermal and thick shell output.
Added mass center coordinates to the SSSTAT_MASS_PROPERTIES output.
Redfined failure terms in dcfail output file to include the area scale factor and also the exponent.
Added compiler informatoin to banner.
Fix spotweld assembly output problems.
Fix for SMP output of rcforc to binout/lsda in case there are TRANSDUCER type contacts.
Fix in *DATEBASE_FSI_SENSOR.
Changed pressure output to dbfsi when *AIRBAG_ALE(_ADVANCED) is used output pressure = coupling pressure + ambient pressure.
Fix inconsistency of rigid body output to d3hsp.
Fix in *DATABASE_SPH if *DATABASE_HISTORY_NODE is not defined.
Material Models
Fix for material 127 for plane strain 2d elements.
Fixed types 6, 7, and 9 hourglass control for solid elements with material 4.
Added error termination if BETA in material 6 is zero.
Fix for divide by zero during restart with materials 123 and 24 when using the dynain file for initialization.
Modified initialization of void fraction f in material 120:
f is set to history variable dam from INITIAL_STRESS_SHELL if it is greater than F0 from the material card. If smaller, f is set to F0 (as before).
Fixed 2-sheet type 9 beam spotweld failure with the rate sensative parameter option.
Fix for material 76 when bulk viscoelastic constants are specified.
Fixes for material 175 for memory clobbers & missing calls to the fitting subroutines.
Fix Toyota’s new 3 sheet weld failure criteria as well as it’s extension to the 2 sheet case.
Corrected material energy in fabric materials 13 and 14 when viscous damping is used.
Fixes to material 187 related to SMP parallel and also to initialization.
Fix for material 4 whenever the yield stress is set to zero for the thermoelastic case.
Fix for material 24, VP=1 with load curve to scale yield stress to account for strain rate effects.
Fix for material 24, VP=1 for solids that led to a negative internal energy.
Minor corrections for Toyota spotweld capability.
Minor fix to material 187 related to damage calculation.
Fix for zero divide for spotweld failure option type 1.
Fixes for Toyota spotweld type 9 option related to double storage allocation.
Fixes to material 120.
Updates for toyota new spotweld requirements.
Added load curves to account for rate effects in *DEFINE_SPOTWELD_RUPTURE_PARAMETERS.
Fix for variable viscosity in *MAT_NULL.
Small modification to fabric material forms 2 and 12 which could generate initial compressive stress when an airbag reference geometry was used and compressive stress elimination is active.
Update Mat_138 to include bug fixes already made in 971 R3.
Moved new material 34 form 14 option to scale the unload curves from R3 into R2.
Fixed rotational instability in material 22 for solid elements.
Fix for energy calculation bug in material 123.
Update material 120 to include void contraction even smaller than F0
Fixed 3-sheet Toyota spotweld beam failure options 6 and 9 whereby both had problems with transforming resultants before taking their difference at the center node.
Minor fix for material 22 for fully integrated solid elements
Fix for material 72_R3 to correct potential divide by zero problem on the first cycle.
Output effective strain for material 187 since effective plastic strain is not available and added internal energy calculation.
Fix in material 72 related to unstable response.
Updates for material 120, *MAT_GURSON.
*MAT_CONCRETE_DAMAGE and *MAT_CONCRETE_DAMAGE_REL3 (materials 72,72r3):
Internal energy now calculated and output works with multi-integration-point element types.
Fixes for material 186 and user defined equations of state.
Include material 155 for SPH elements.
Fix for material 195 – Tension cutoff.
Fixes for material 77 plane stress implementation.
Fixed solid element hourglass stiffness parameters for material 181 and 183 such that types 6, 7, and 9 hourglass controls should work well for both the compressible and incompressible models.
Updates including fix of round-off bug in material 138 cohesive model.
Correction in principal stress calculation for material 126 yield surface 2.
Minor fix for material 24 viscoplasticity rate formulation.
Moved strain rate history information to the first history variable slot for the solid material 24, VP=1 (now consistent for shells and solids).
Fix for yield 96 in the Barlat plasticity model.
Elements
Fixed sorting of EFG shell 41 for 3-node/4-node background element.
Add thick shell types 1 and 2 for linear implicit analysis.
Fix related to mass scaling and constrained beam offset option.
Fixed linear curve interpolation routine for type 9 beam spotweld failure to stop extrapolation of dynamic scale factor curves.
Fixed null section force problem in secforc for ELFORM=13.
Fixed a serious problem in the tangential moment calculation of DAIMLERCHRYSLER solid spotweld failure when the TRUE_T option was used.
Corrected shear load rate calculation for type 9 beam spotweld failure with dynamic load scale factor curves. Before the fix, the rate was much too large.
Fix for *ELEMENT_BEAM option where n2=-n1 grounds the discrete beam.
Fixes in MEMORY option on *CONTROL_SPH card.
Fix for resetting mass of spotweld beam element which is lengthened for tied contact.
Fix for spotweld beam to prevent huge energy excursions.
Fix for 3D EFG adaptivity.
Fix for 3D EFG mass scaling.
Added esort=3 option to use DKT triangular shell elements in place of degenerate quad elements.
Delete solid spotwelds if surrounding shells fail and update *CONTACT_SPOTWELD accordingly.
Added bulk viscosity to the DKT, type 17, shell element.
Added spotweld neighbor segment thinning to version R2.
Added bulk viscosity to shell type 4 (C0 triangle).
Stress initialization bug fix when the option tet4to10 is used.
Fix for penta type 15 for *MAT_ADD_EROSION.
Fix in EFG shell 42 due to erosion support.
Fix SPH bug: Particles were not deactivated with a renormalized formulation.
Fix computation of mass for superelements by removing the “mass” associated with scalar nodes.
Put in some error checking on illposed superelement problems.
Fix a problem with superelement input processing.
Fix initialization of superelements for the case of implicit/explicit switching.
Fixed consistency problem related to the type 17 tetrahedron that resulted in error terminations.
Fixed time step calculation for type 17 tet (time step was too large).
Fix related to tet4 to tet10 option.
Fix in efg shells when element failure happens.
Contacts
Fix with dynamic relaxation and contact entities.
Added support for *DEFINE_CONTACT_VOLUME for SPH elements, which was not working.
Fixed a segment based contact problem when used with 2-surface force transducers that could cause an infinite loop when one segment has nodes on both surfaces.
Fix in beam contact when there is only one beam.
Fix for contact entity problem.
Fix related to tied contacts where the master and slave thickness are specified as negative values.
Fix SMP thread problems in 2D automatic contact that could cause unpredictable results.
Added square shell edge option to segment base contact. Activate by setting SHLEDG=1 on *CONTROL_CONTACT SHLEDG is the first paramter on the new 6th card.
Fix for SPH contact when DEFINE_CONTACT_VOLUME option is used in the automatic contact generation.
Fix with segment based eroding surface to surface contact that could cause penetration of solid parts after erosion.
Fixed 2-surface force transducer contact when used with segment based contact. Before the fix, results were not too bad, but now are correct.
Fix for initialization problem of CONTACT_GUIDED_BEAM: incorrect neighbors were being computed on one end of each beam, resulting in errors as the beams moved along the guide nodes (SMP and MPP).
Fix for round-off problem related to contact boxes.
Fix to automatic_tiebreak contact due to differences between 970 and 971.
Fixed some problems with define connection spotweld failure. Now, the check for improper tied contact type works.
Added failure messages if a node is released for automatic…tiebreak OPTION=5,6,7.
Fix with MPP guided cable contact in case the number of nodes or beams is different between different interfaces.
Fixes to *CONTACT_FORCE_TRANSDUCER for slave-master definition.
Fix for *PART_CONTACT : the stiffness scale factor was being improperly applied due to a bug in some changes back in January 2005. A workaround for those broken versions is to input the square root of the desired scale factor on the PART_CONTACT card, as the code has been squaring the input.
Fix for seatbelt slipring node in contact.
Fixed new segment-based contact depth of penetration reduction factor for the case when there was initial penetration at time zero. It is supposed to not penalize that, but did penalize some of it.
Fixed energy calculation in symmetric treatment of segment-based contact in case where segments in contact have nearly colinear normals.
Fixes for OFFSET tied contacts which fail to release when attached segment is deleted.
Fix for surface to surface tied interfaces to release constraint if segment fails.
Constrained
Fix for *BOUNDARY_PRESCRIBED_MOTION for the acceleration of node of a nodal rigid body.
Fix with *CONSTRAINED_NODAL_RIGID_BODY_INERTIA. If any of the nodal rigid bodies preceeding this in the input get merged out due to common nodes, the inertia would be applied to the wrong nodal rigid body, in some cases overwriting memory with unpredictable results.
Fix related to *CONSTRAINTED_NODAL_RIGID_BODY_SPC and also updated formats.
Fixed for *CONSTRAINED_NODAL_RIGID_BODY if RRFLAG is negative.
Fixed related to deleted nodal rigid bodies which caused energy growth.
Fix associated with implicit constraint and attachment modes.
Made changes to the moment calculations for the joints to make them consistent with the revolute moment calculations.
Fix for *BOUNDARY_PRESCRIBED_MOTION DOF=-4 and VAD=1.
Minor fix for constraints on rigid bodies related to round-off.
Fix implicit constraint loading for special cases of solid only models.
Fixed implicit’s use of ale subroutine for rebar-in-solid problem.
Fix for the newly added 2nd order terms for ALE mesh rotation (should not be used for PRTYPE 0,1,2,3,4).
Fix for a coupled thermal-mechanical problem using the ALE element formulation 12 (single material and void).
Fix for ale rezoning for equipotenial and isoparametric options.
Fixes to *AIRBAG_ADVANCED_ALE.
Fix in *LOAD_ALE_CONVECTION.
Fix for *ALE_REFERENCE_SYSTEM affecting PRTYPE 9 mesh expansion center.
Enhance volume fraction filling algorithm to take care of very large dimensional model .
Relax upper bound of leakage control starting volume fraction to 0.6 to avoid unstability when both sides of the structure have leakage control.
Miscellaneous
Updated *CASE driver program.
Fix in mapping for *INCLUDE_STAMPED_PART including check of segment normals to get order of integration points right.
Made various fixes/updates to ‘*PERTURBATION’:
More of correct handling of defaults; specified as zero stays zero.
Fix with perturb_shell_thickness and adaptivity.
Fix shell thickness bug using wrong part info.
Fixed of double mass scaling when dynamic relaxation is active.
Fixed *INCLUDE_TRANSFORM bugs, including those related to:
*CONSTRAINED_JOINT_STIFFNESS
*MAT_020 (rigid)
*MAT_196 (discrete beam)
*DEFINE_COORDINATE_VECTOR
*CASE
Fix related to acceleration filtering via IACCOP=2 in *CONTROL_OUTPUT.
Saved interface force file name for restart.
Fix in adaptivity reported by Dynamore.
Add support for Dytran coupling with blockage and shell failure options.
Fix for the restart of adaptivity job if ADPTHK option is used.
Fix in 3d-adaptivity if not all parts are adapted.
Fix in *DEFINE_FUNCTION.
Fix in mapping (with history variables).
Fix treatment of implicit nonlinear print flag.
Extend array dimensions to allow 500 airbags, up from the current 100.
Added new parameter ISNAN in ‘*CONTROL_SOLUTION’:
$=============================================================================== *CONTROL_SOLUTION $ solu nlq isnan $ I I I $ 0 0 0 $=============================================================================== ISNAN - Flag to check any NaN in moments/forces after element formulation 0-no check 1-check
The command option “checknan=1” or the sense switch control “swn” can activate this option as well.
Fix related to 2D_Segment sets when multiple sets are defined.
Fix for 2d thermal adaptive problem with convection boundary conditions (2d segments generated improperly and temperatures for thermal solver were set to the wrong time).
Extend Hybrid inflator option to work with _CM jetting.
Fix with jetting related to _CM option.
Added rigid bodies to 100 step initialization option in *CONTROL_DYNAMIC_RELAXATION.
Fixed *PARAMETER if “&” is not at the first column.
Fix for *SET_PART_ADD with set offsets defined.
Some improvements to *CONTROL_ADAPTIVE_CURVE.
Fix in 2d r-adaptive where *LOAD_SEGMENT is treated as in 3d adaptivity.
Fix potential bug for *CONTROL_CHECK_SHELL when multiple parts are defined.
Fixed an adaptive bug that was added 07/11/2006 causing adaptive runs that use *LOAD_SHELL to lose their load.
Fixed in adaptivity triggered if any element fails.
Updated NASTRAN and DYTRAN reader from MSC.
Add support for MSC Dytran + Dyna coupling.
Make lscasedriver (*CASE) work for Windows platform.
Fix for *SET_NODE_GENERAL usage for full deck restarts with the PART option.
MPP Database/Output
Add echo of decomposition information to the d3hsp and mes0000 files for MPP.
MPP fixes for interface force file in case there are tied interfaces.
Modify sensor handling for MPP, to fix problem with PART_SENSOR.
Fix for ALE dbdensor — MPP only.
MPP Material Models
Fix MPP bug in material 169, *MAT_ARUP_ADHESIVE that could cause LS-DYNA to crash if input variable XEDGE = 1.
Fix beam spotwelds with OPT=9 failure which were broken when rate effects were added — MPP only.
MPP Elements
Fix with MPP beam spotweld neighbor thinning.
Fix for one noded beam — MPP only.
MPP Contacts
Fix affecting contact forces for contact entities — MPP only.
Fixed memory allocation problem in MPP related to tied contact initialization.
Add MPP constrained contact check: if a slave or master node is in more than one constrained interface, issue a warning and release the node. This check was already in SMP.
Fix for MPP 2-surface force tranducers when used with segment-based contact for the case when some processors in the contact communicator do not have any force transducer contacts defined.
Adjusted the tie criteria for spotweld — MPP only.
Fix in MPP segment-based contact when there are control volumes in the data.
Fix user defined thickness in tied contact — MPP only.
Change in MPP forming contact: honor newly documented behavior of the contact orientation flag: reorient if flag is 3, but not if flag is 4.
Fix to SINGLE_EDGE contact — MPP only.
Fix possible case of MPP eroding contact failure: if segment list was used for one of the side definitions, some contact segments might get lost.
Change to MPP contact release condition to better handle the case where both the slave and master sides have 0 thickness.
Fix in MPP single surface contact friction calculation that has been broken since 971 version 5445.
Fix for MPP TIEBREAK contact initialization, in case a slave node is near two separate slave segments, it was sometimes not taking the closest one.
Add extra logic for tie interface with user specified thickness according to manual.
t3 = 0.6*(slave_t+master_t)
t2 = 0.05*clen
t = max(t3,t2)
Add additional check which SMP code does to avoid tie to the wrong segment resulting in a short beam controlling the time step — MPP only.
Fixed a serious MPP thermal contact bug that would cause bad thermal stiffness and temperature data at elements shared by processors if there is more than one thermal contact definition in the model.
Fixes for MPP auto-tiebreak contact: was reading the tiebreak option off the wrong card, and accidentally reinitializing the tie point every cycle, resulting in no contact forces.
Fixed MPP 2-surface force transducers when used with segment-based contact (Nodal forces were put on incorrect nodes possibly causing segmentation faults.).
MPP changes for slipring node in contact.
MPP fixes for contact force based sensor: forces were not being correctly computed.
MPP Constrained
Fix treatment of rigid body forces in MPP implicit.
Fixes for *BOUNDARY_PRESCRIBED_MOTION including:
Curve scale factor applied twice.
Bug in VAD=4.
Fix of MPP not properly handling *CONSTRAINED_LINEAR in some cases.
Fix MPP implementation of implicitly constrainted joints for explicit.
Correct MPP processing of tied constraints for implicit.
MPP Miscellaneous
Fix in *DEFINE_COORDINATE_NODES, FLAG=1 — MPP only.
Fix related to MPP and adaptivity.
Fix for *CONTROL_DYNAMIC_RELAXATION — MPP only.
Fix for interface linking — MPP only.
Fix problem in MPP solid remeshing in case some processors had no part of the outer boundary.
Support *DEFINE_DEATH_TIMES — MPP only.
Volume calculation of airbag with holes made consistent with SMP.
