Personal tools
 

LS-DYNA V970 Revision 6763

Filed under:

Release Notes for Version 970/Revision 6763.169 November 28, 2005 (Summary of modifications made since release of revision 5434a)

NOTE:

This set of release notes supercedes ReleaseNotes-970-6763-86.txt and ReleaseNotes-970-6763-169.txt. (The superceded "ReleaseNotes-970-6763-86.txt" inadvertently omitted some notes pertaining to revision 6763.86.) Changes made to version 970 from the release of revision 5434a to the release of revision 6763.169 are summarized herein.

Material models

  • Updated *mat_151 for solid elements.
  • Fixed *mat_77 for shells.
  • Fixed bug in *mat_183.
  • Fixed double-single precision problem in *mat_106.
  • Made fixes to *mat_040 shells and solids.
  • Made modification to bypass deletion of solid element when using *mat_126 with the reference geometry option.
  • Fixed problems in rubber material *mat_027 and *mat_181.
  • Added upper bound for strain rates in *mat_024 for shells with viscoplastic rate formulation.
  • Added table of temperature-dependent yield curve for *mat_106.
  • Made minor fix to high explosive model
  • Fixed unit system 1 in wood material mat_143 (cm changed to mm).
  • Corrected pressure term in *mat_147. NPLOT=5 in *mat_147 outputs pore water pressure to "plastic strain" location.
  • Made minor fix to *mat_add_erosion.
  • Fixed read error of LCSR in *mat_112.
  • Fixed strain rate parameter in *mat_112.
  • Corrected stress calculation in 2D solids for *mat_098.
  • Added updates to *mat_101.
  • Fixed bug in *mat_181.
  • Modified *mat_119 to use small strains rather than displacements.
  • Added modifications to *mat_151 recieved from Sandia.
  • Added *mat_151 (mat_emmi = Evolving Microstructural Model of Inelasticity).
  • Added deletion of solid element spotwelds if surrounding shell elements fail.
  • Fixed problem with *mat_024 combined with tet type 10.
  • Fix for density scaling in *mat_147.
  • Added improved formulations of FORM 2,3,4 for mat_fabric. These are FORM 12, 13, and 14, respectively.
  • Fixed Toyota's spotweld failure model to handle shared spotweld nodes.
  • Fixed *MAT_58 input problem (details not available).
  • Activated crash front calculation in composite materials.
  • In mat_15, log is now taken of a number not less than 1.0. Thus (1+c*ln(epsdot)) and (1+D4*ln(epsdot)) are not less than 1.0.
  • Added viscoplastic option to *mat_123.
  • Fixed problem with *mat_181 for shell elements in single precision.
  • Added failure model to *mat_181.
  • Fixed minor bug in *mat_024 shells with VP=1 where rate parameters C & p were used in addition to the table definition.
  • Added viscoplastic strain rate to *mat_024 as the first history variable.
  • Fixed *mat_082 for VP=0. Failure did not occur for a simple test case using the C and p rate constants.
  • Fixed *mat_024 for VP=1. Error related to interpolation between curves when table definition is used.
  • Fixed *mat_098 for solid elements related to stress initialization using the dynain file. The volumetric strain must be initialialized.
  • Fixed bug in solid element *mat_103 when cmpflg=1 and aopt=0.0.
  • Fixed timestep problem related to mat_054 and shell formulation type 8.
  • Fixed barlat_yld96 material model. The wrong load curve is used if the curve option is active.
  • Fixed bug in seismic beam, *mat_191.
  • Fixed bugs related to spotweld material option opt=1. Rate effects were not working correctly.
  • Added support for non-Newtonian viscosity variations via load curve or by power law in *MAT_NULL.
  • Fixed problem with *mat_183 shell elements related to single precision.
  • Fixed problem with *mat_156 to handle large load curve ID's in single precision.
  • Added muscle model for truss element.
  • Added mat_183, a variation of *mat_181 but with a damage curve.
  • Fixed timestep calculation for *mat_106.
  • Fixed problem in *mat_023 related to user integration rules that reference Part IDs.
  • Made minor fix related to large load curve numbers in mat_083.
  • Added *mat_155 (ice model).
  • Added support for types 6 and 7 hourglass control for solid material types 85, 102, 109, 110, 111, 112, 115, 120, 131, 132, 141, 147, 151, 153, 154, 159, 164, 169, 176, 180, 192, 198, and 200.
  • Added reference geometry option to *mat_083.
  • Fixed bug in 2D solid element related to *mat_002 and full integration.
  • Added cutoffs for the green strain tensor in fabric material for options 2,3, and 4. Round-off error can cause instabilities.
  • Fixed problem with viscoelastic *mat_065.
  • Fixed bugs in *mat_059 for solid elements affecting transverse tensile failure and reversal of transverse shear strengths.
  • Added new failure criteria for 2D mat_010 elements.
  • Fixed bug affecting strain output of mat_058 when EC=0.
  • Corrected a bug that could cause spotweld failure option 7 to use the default failure values rather than the failure values from the table.
  • Fix to slope in type 4 spring material.
  • Made changes to *mat_124 failure model.
  • Updated *mat_075 per Paul Du Bois.
  • Fixed a minor bug in *mat_126 in which the wrong load curve ID was printed into d3hsp.
  • Fixed a minor bug in *mat_176 affecting reading of viscoelastic constants.
  • Made minor change to calculation of fracture strain in *mat_015 to prevent numerical overflow.
  • Modified Toyota beam weld failure criteria to only increment the failure counter (to compare with FVAL) at the beam end with higer plastic strain.
  • Fixed bug in *mat_119 for discrete beam to prevent infinite time step size.
  • Fixed type 6 hourglass control for solid element material mat_181.
  • Fixed type 6, 7, and 9 hourglass control for solid element materials mat_110 and mat_111.
  • Added support for hourglass control types 6, 7, and 9 for solid element materials mat_155 and mat_183.
  • Made minor fix to mat_181's reference geometry stresses.
  • Fixed problem related to rate effects in mat_181.
  • Fixed a bug affecting mat_002 when used with user-defined shell integration rule which changed properties through the thickness.
  • Added modifications to mat_162 from MSC.
  • Extended type 6 hourglass control to type 1 solid element materials 13, 16, 72, 163, 198, and 200.
  • Fixed bug affecting use of large load curve IDs for mat_120.
  • Made a bug fix to mat_158.
  • Fixed format error for *MAT_THERMAL.
  • Made fix to mat_057 related to large load curve IDs.
  • Made minor fix to mat_036 related to initial stress option.
  • Fixed variable initialization bug in mat_181.

Element formulations

  • Shell formulations 4, 11, 16, and 17 can now model rubber. Corrections made to shell formulations 2 and 7 for rubber.
  • Fixed a bug in shell formulation 18.
  • Updated shell formulation 15 for consistency, accuracy, and speed.
  • Corrected mass calculation in solid formulation 12.
  • Corrected time step calculation for cable elements.
  • Fixed incorrect thickness in 2D plane stress solid.
  • Corrected seatbelt bug caused by incorrect index used in the retractor array.
  • Fixed a bug in shell formulation 21: yz- and zx-stress/strain components were misplaced.
  • Fixed formulation 15 solid (prism) to correct NaN problem triggered by stiffness damping.
  • Fixed possible problem in stiffness damping in Hughes-Liu beam elements.
  • Fixed problem with stiffness damping in shells and solids. The implementation was incorrect for the type 2 and 3 solids and the energy calculation was wrong. For shells, the damping stresses were sometimes overwritten before energy was calculated.
  • Fixed inertia calculation for formulation 16 tetrahedron.
  • Corrected 2D hourglass stiffness for axisymmetric formulation 15.
  • Fixed input problem when number of element_inertia exceeds 99999.
  • Fixed beam end release problem.
  • Fixed problem with tet formulation 16 combined with mat_004.
  • Fixed problem with local stresses in 2D solid elements.
  • Fixed problem related to massless nodes in discrete beams.
  • Added new pretensioner type 7 in which the pretensioner and retractor forces are calculated independently and added.
  • Fixed NFAIL option related to rubber material and fully integrated elements. All elements were deleted in the first time step.
  • Fixed bug in truss element stiffness and added geometric stiffness.
  • Modified the type 13 nodal pressure tet element to use a smoothed pressure that conserves energy better during 3D adaptive remeshing.
  • Fixed for cross-section force calculations for type 16, 10-node tetrahedron.
  • Fixed component interface linking problem with tetrahedral elements.
  • Made minor fix to time step calculation for solid formulations 4, 16, and 17.
  • Updated damping related to the Belytschko-Schwer beam.
  • Added support for EOS 12 for Lagrangian solids.
  • Fixed a bug in shell type 16 for linear analysis.
  • Fixed a bug in shell type 21.

Contact

  • Fixed problem with contact_auto...tiebreak option=5 related to cutoff of decay curve.
  • Fixed problem with contactauto..tiebreak contact for option=1.
  • Fixed problem with contactauto..tiebreak contact related to searching and offsets.
  • Fixed forming contact in SMP where the shell thickness is greater than the edge length.
  • Added option to automatically switch from type 9 tiebreak surface to surface contact to automatic tiebreak.
  • Added a check for improperly tied solid spotweld elements that lack table data for spotweld failure OPT=7.
  • If soft=2 contact is requested for a keyword that is not supported, it now switches to soft=1 instead of soft=0.
  • Added revision 3858's node-to-surface eroding contact back as an option.
  • Fixed segment-based contact energy and rcforc output when elements erode.
  • Fixed segment-based contact problems. One problem occured with incorrect detection of penetration when there was extreme compression of solid elements (foam).
  • Added options to control birth and death time of contact when dynamic relaxation is active.
  • Added scaling of spotweld failure (*control_spotweld_beam).
  • Added error termination when requesting constraint-based contact with rigid bodies in explicit computations.
  • Made change to eroding surface-to-surface contact to make it behave more like revision 3858.
  • Modified area check of contact segments to allow smaller segments to be used.
  • Fixed 2D automatic contact to prevent likely penetration from occuring after eroding solid element expose new segments.
  • Fixed a potiential memory allocation error in segment-based contact that could cause a segmentation fault during a run. The error is only known to have occured with solid elements in combination with eroding contact.
  • Added guided cable contact.
  • Add consideration of true out-of-plane thickess of 3D shells when used with 2D automatic contact. Previously, a unit thickness was assumed because the use of 3D elements with 2D contact was not anticipated.
  • Fixed the sliding option of segment-based contact when used with triangluar segments.
  • Changed the release criteria for forming contact to prevent premature release.
  • Added option whereby negative contact damping coefficient becomes a flag for damping all velocity components, not just the normal velocity component.
  • Fixed a problem with forming contact.
  • Fixed a problem that could cause 2D contact failure.
  • Fixed sliding option of segment-based (SOFT=2) contact to reduce the likelihood of penetration.
  • Fixed automatic tiebreak contact related to cure energy growth problems and other strange behavior.
  • Made improvements to contact_auto_move.
  • Fixed damping option used in rigid body contact.
  • Fixed bug in sliding option in segment-based (SOFT=2) contact. 970/6763.086 had a totally broken sliding option.

Adaptivity

  • Fixed bug in *SET_GENERATE... options for adaptivity.
  • Fixed bug in look-forward h-adaptivity (too many elements refined).
  • Fixed d3plot and restart problems for 3D r-adaptivity.
  • Fixed bug that resulted in missing d3plot files in adaptive runs.
  • Increased some local array sizes to handle larger 2D r-adaptive problems.
  • Made many improvements to the triangular surface remesher in support of 3D r-adaptivity.
  • Made minor tolerance changes for symmetry planes and 3D adaptivity.
  • Fixed volume conservation and mass loss issues with 3D adaptivity.
  • Added feature to control h-adaptivity level by load curve.
  • Fixed 2D adaptive dynain corruption problem.
  • Fixed problem with stiffness damping used in combination with adaptivity.
  • Fixed problem of failed elements reappearring after mesh is adapted.
  • Added new option for ADPLVL in *CONTROL_ADAPTIVE. If ADPLVL is negative, we split the main element only.
  • Added curve to specify 2D adaptive mesh size.
  • Added *DEFINE_SET_ADAPTIVE so that the user can define the adaptivity refineement level by part or element set.
  • Added adpsize option to *DEFINE_SET_ADAPTIVE.
  • Made fix to 2D adaptivity related to missing lumped masses after first adaptive step.
  • Fixed 3D adaptive remeshing with SPC constraints.
  • Added another option for adaptive angle calculation.
  • Fixed error in adaptivity refinement.
  • Fixed bug in adaptivity if the material included failure.

ALE

  • Added ALE birth time option.
  • Updated coupling type 5 (coupling to eroding parts).
  • Added mat_025 for ALE.
  • Added option in *INITIAL_VOLUME_FRACTION_GEOMETRY to control degree of refinement used in each ALE element to calculate initial volume fraction of each material.
  • Added a new coupling stiffness option to *constrained_lagrange_in_solid. (uses slave bulk modulus, mass-weighted penalty).
  • Added a soft constraint coupling option.
  • Added an option to terminate the run if the critical time step calculated from ALE is less than dt2ms.
  • Fixed a bug in *initial_volume_fraction_geometry in which the initial guess on which segmental normal to use was wrong in some cases.
  • Fixed a bug that would cause ale_reference_system_group to function abnormally.
  • Modified leakage control type 2 for ALE coupling to adjust velocities in order to better conserve energy.
  • Eliminated mass scaling for ALE elements to avoid problems.
  • Added option for considering shell thickness in constrained_lagrange_in_solid.
  • Modified codes so that all gas species of mat_gas_mixture are considered in estimating ALE point source inlet gas velocity.
  • Fixed bug in ALE interface reconstruction if number of materials is greater than three in an element.
  • Added alternate names for all materials specific to ALE:
    • *MAT_VACUUM -> *MAT_ALE_VACUUM or *MAT_ALE_01
    • *MAT_GAS_MIXTURE -> *MAT_ALE_GAS_MIXTURE or *MAT_ALE_02
    • *MAT_NULL -> *MAT_ALE_VISCOUS or *MAT_ALE_03
    • *MAT_MIXING_LENGTH -> *MAT_ALE_MIXING_LENGTH or *MAT_ALE_04
  • Fixed the ALE sensor option to use only the segment center as the sensing point.
  • Added mass output to d3plot and d3thdt for each multimaterial group.
  • Disabled ALE coupling with nodes and always default to element quad points.
  • Implemented ALE multi-material switching (*ALE_FSI_SWITCH_MMG). This option will allow user to switch the material group ID after it flows through a defined section set. Using this option, coupling can occur from either side of a Lagrangian surface. Report mass of switched material in the porosity field of dbfsi.
  • Updated ALE coupling to accommodate all element shapes including tets.
  • Fixed a bug affecting ALE reference system 4.
  • Added ability to assign initial velocities using *INITIAL_VOLUME_FRACTION_GEOMETRY.
  • Added a new algorithm for *INITIAL_VOLUME_FRACTION_GEOMETRY that accounts for more than just convex shell/segment surfaces.
  • Fixed bug in ALE reference system 9.
  • Implemented *load_ale_convection for simulating heat convection from ALE fluid to Lagrangian structure. Write temperature change to dbfsi.
  • Fixed bug for lvel=0 in AIRBAG_ALE
  • Fixed bug affecting ALE FSI sensor.
  • Added capability to allow NORM in CONSTRAINED_LAGRANGE_IN_SOLID to take a value between 0 to 1. The normal of the penalty springs can now be:
    • a. segment-based normal
    • b. nodal average of the segment-based normal (weighted by segment area)
    • c. a blend of (a) and (b)

    Option c avoids the need to define multiple coupling commands in some cases.

  • Added support for ALE reference pressure in element formulations 5,6,7.
  • Fixed problem in EOS_IDEAL_GAS used in element formulations 5,6,7.
  • Added translational and expansion constraints for ALE_REFRENCE_SYSTEM_GROUP type 5.
  • Changed default gas mixture option for AIRBAG_ALE to non-linear cv, cp model.
  • Added pop pressure and venting curve in AIRBAG_ALE option.
  • Added DATABASE_FSI_SENSOR option to get ALE element pressure at the sensor location and output to "dbsensor" file.
  • Reinstated ALE solid 7 ambient type 3 condition which was disabled accidentally.
  • Implemented new ALE FSI interface forces file to get binary database similar to contact's INTFOR database. To activate the option:
    • (1) add h=filename to execution line
    • (2) set flag in *CONSTRAINED_LAGRANGE_IN_SOLID

    *DATABASE_BINARY_FSIFOR defines the output frequency.

  • Added new option to POINT_SOURCE to evenly distribute momentum among nodes.
  • Implemented _ID and _TITLE option card for CONSTRAINED_LAGRANGE_IN_SOLID.
  • Fixed advection for wedge elements.
  • Fixed bug in which kinetic energy of the inlet gas was not properly compensated for in the internal energy
  • Fixed bug affecting zero load curve velocity for Point Source.
  • Added ALE initial penetration option in CONSTRAINED_LAGRANGE_IN_SOLID.
  • Made correction to heat convection in ALE coupling while using mat_148 with nonlinear cv,cp formulation.
  • Added ALE reference system 9 - new option to create smoother expansion of meshes that includes a fixed part.
  • Implemented exclude part list for ALE_REFERENCE_SYSTEM type 9 to avoid point source from moving outward from the critical region.
  • Added option for ALE_REFERENCE system 9 to only allow expansion but not shrinking of ALE mesh.
  • Added a box option for CONSTRAINED_LAGRANGE_IN_SOLID for slave segments so that only segments inside the box get coupled with fluid.
  • Added option triggered by negative porosity pressure curve ID in CONSTRAINED_LAGRANGE_IN_SOLID whereby porosity comes from MAT_FABRIC.
  • Added a new advection option (type 3) in which total energy is advected rather than internal energy.
  • Improved algorithm for initial assignment of volume fractions in ALE.
  • Fixed bug in ALE porosity option in mat_034 curves.
  • Added more options for initial volume fraction (ALE).
  • Added extra option for ALE mass scaling.

SPH

  • Added mass of SPH elements to d3plot files.
  • Fixed problem with tied SPH particles.
  • Added a test on mass of SPH element during initialization.
  • Fixed bulk viscosity computation for all the SPH formulation.
  • Fixed a bug with *define_box_sph. Box was not moving even with the definition of the load curve.
  • Fixed bug in initialization of mat_008 SPH elements. All the SPH elements were detonating at time=0.
  • Fixed bug in SPHOUT . An Error Termination occured when *DATABASE_SPHOUT was defined before the *DATABASE_HISTORY_SPH keyword in the input deck.
  • Implemented *MAT_ISOTROPIC_ELASTIC_FAILURE (mat_013) for SPH elements.
  • Fixed bug in *BOUNDARY_SPH_SYMMETRY_PLANE. Lagrangian nodes (of shells and solids) were constrained by this boundary condition.
  • Fixed bug in SPH non-reflecting boundary.
  • Improved efficiency of SPH initialization.

EFG

  • Added 2nd order objective stress update to 3D EFG. This is controlled by *CONTROL_ACCURACY
  • Added 6-noded and 4-noded elements for EFG.
  • Corrected EFG 2D contact with beam elements.

Output

  • Added version number to lstc binary files: d3plot, d3thdt, d3part, etc.
  • Changed banner, binout, and ascii file headers to include version number, revision number, mpp/smp, and single/double indicators. Header format will now match new 971 format.
  • Changed xtfile to 32-bit format if LSTC_BINARY=32ieee.
  • Removed extra data in rbdout file.
  • Fixed bad write to d3plot when a tie-break is used and some nodes are moved.
  • Added write to all ASCII files at termination time.
  • Fixed d3crack output if LSTC_BINARY=32ieee is set.
  • Fixed missing header for nodout in a full deck restart.
  • Disable dbfsi and dbsensor if there is no FSI coupling defined.
  • Corrected temperatures written to the drdisp.sif file for itemp=-2 and itemp>0.
  • Fixed error in nodfor file. All forces were zero if node is attached to a rigid body.
  • Fixed calculation of strain tensor that is output for type 4 tetra element.
  • Fixed a bug in d3plot output when DECOMP=3 or 4 (*DATABASE_EXTENT_BINARY) in the case of a thermal problem.
  • Added output of all ASCII and intfor data at the termination time.
  • Fixed jntforc binary output to be in agreement with the ASCII output.
  • Changed default scale factor for binary file size from x=7 to x=70.
  • Fixed bug in beam d3plot output to correct viewing of beams as prisms in LS-Prepost.
  • Fixed bug in *COMPONENT_GEBOD segment acceleration output. (Affected reported accelerations of lower legs in d3plot, nodout, and rbdout. Motion of the dummies was unaffected.)
  • Made change to binary ncforc output for eroding surfaces. Since the number of nodes can change each state, write the ids array in each output state, but only for eroding contacts.
  • Fixed bug in writing D3PART if any ale elements is requested.
  • Fixed bug in writing D3PART if DCOMP=3 or 4 is defined in *DATABASE_EXTENT_BINARY.
  • Updated format for all ASCII files.
  • Fixed problem of IDs in swforc file in binary output: user ids were not being output for nodal constraints or generalized welds.
  • Added sleout output for 2D contact and added binout output of 2D contact rcforc, ncforc, and sleout.
  • Fixed ALE tracer particle output problem.
  • Changed signs on type 4 beam stress resultants to be consistent with types 1 and 2.
  • Fixed a bug in d3plot for linear shell elements 18, 20, and 21.
  • Fixed initial rotational velocities written to nodout files. Nodes belonging to rigid bodies had their t=0 velocities written as 0.0.
  • Fixed beam resultants in elout file so type 1 and 2 beams agree.
  • Corrected sign of beam moments in secforc file.
  • Fixed bug in elout for linear shell formulations 18, 20, 21.
  • Corrected strain output to elout file for material 34 forms 3 and 4.
  • Fixed bug in binary swforc output for generalized welds in MPP.
  • Fixed secforc file for 1-pt solid elements for implicit computations.
  • Cleaned up thick shell output.
  • Fix force output for Lagrange Multiplier joints.
  • Fixed output of shell resultants for materials 22, 54, and 55 when CMPFLG=1. Added option CMPFLG=2 which overwrites resultants as requested by NASA.
  • Fixed output to swforc for implicit.
  • Added resultant moments to swforc file for Toyota weld option with table.
  • Fixed binout output for solid type=3.
  • Correct reporting of joint forces to JNTFORC file for implicit simulations.
  • Added *DATABASE_BINARY_D3PROP for writing the material and property data to the first d3plot file or to a new database (d3prop).
  • Corrected output of reaction forces to bndout for rigid bodies in implicit.
  • Removed warning messages if the number of history blocks exceeds 2000.
  • Added MASS_PROPERTIES option to include mass and inertial properties in the glstat and ssstat files.
  • Fixed bug in writing d3thdt.
  • Fixed ASCII file SBTOUT in restart run.
  • Fixed cross_section problem.
  • Fixed bug in dynain related to output of undefined beam elements.
  • Fixed problem with accelerometer output.
  • Added option in *CONTROL_CPU to output cpu/elapsed time in glstat file.
  • Corrected the writing of beam reference nodes to d3eigv.
  • Added output of material fit data to messag file (convenient for plotting) for mat_077 (Ogden) when N=3 or N=4. Previously, data was written only when N=2. Fit data is also written for mat_027.
  • Fixed 2D automatic contact pressure calculation for ncforc output. Before fix, in-plane rather than out-of-plane thickness was used to evalutate the tributary area so pressure magnitudes were wrong.
  • Fixed bug for d3thdt output to prevent creation of many small files.
  • Added extra history variables in d3thdt file for ALE solid 11, 12 elements to show correct rho, volume fraction in the history file.
  • Limited the time for the next d3plot output to be endtim for implicit.
  • Added beam elements to the dynain file.
  • Fixed bug in binout swforc data in the case where there are both solid and beam spotwelds.
  • Fixed output of thick shell data to both d3iter and d3eigv databases for implicit analysis.
  • Fixed bug in writing dynain after restart.
  • Fixed d=NODUMP option.
  • Fixed bug affecting ID on DATABASE_HISTORY_NODE_ID.
  • Added beam element moment resultant to swforc file.
  • Fixed problem with secforc and spcforc.
  • Fixed cross-section bug.
  • Fixed code so that the number of cycles between running restart files can exceed 99999.
  • Added write statements to echo EOS 12 input to d3hsp.
  • Fixed sometimes incorrect header in NODOUT and SPCFORC files.
  • Fixed bug in D3THDT when the model contains any thick shell elements.
  • Fixed problem of missing control volumes in abstat file.
  • Added write to d3hsp of memory usage at the end of input phase.
  • Fixed section force problem.
  • Fixed printing of incorrect load curve ID in d3hsp for mat_126.

Restarts

  • Fixed bug related to restart and user-defined materials.
  • Fixed full deck restart with SPH elements.
  • Made various fixed for full deck restart.
  • Fixed 2 problems related to LSDA/binout and restarts:
  • Fixed bug affecting tied contact in full deck restart to correct problem where length of spotweld beams increased after each restart if the beam was offset from the surface.
  • Fixed full deck restart for changing initial velocity.
  • Fixed bug in full deck restart in which adaptivity was used in combination with *INTERFACE_SPRINGBACK_LSDYNA.
  • Fixed full deck restart bug involving failed beam elements. The deleted beam elements reappeared again after full deck restart.
  • Fixed restart problem affecting deletion of ALE parts and deletion of ALE coupling.
  • Fix restart problem involving rubber material.
  • Added hourglass forces to the full deck restart file to correct differences observed between runs that used the full deck restart versus those that do not. This is important for calculations that use solids with the stiffness form of the hourglass control.
  • Implemented DELETE_FSI command for small restart to activate or delete Lagrange_in_solid coupling. It is implemented similiar to DELETE_CONTACT. A negative ID activates coupling.
  • Fixed a bug affecting the implicit step size in a small restart.
  • Fixed bug affecting full deck restarts of models with _OFFSET contact.

Rigid Bodies

  • Fixed output bug for locking joints and translational joints.
  • Fixed rigid body stoppers when rigid body is constrained in local coordinate system. Constraint was being violated.
  • Fixed problem with initial velocities and rigid bodies with prescribed accelerations.
  • Fixed bug in *CONSTRAINED_NODAL_RIGID_BODY if any negative number is used for displacement release flags.
  • Fixed a load curve ID problem for curves defining joint friction.
  • Added error termination if rigid body is slave to contact_entity.
  • Fixed calculation of joint energy.
  • Fixed bug related to Nodal_rigid_body_spc_inertia.
  • Fixed bug affecting right ankle of *COMPONENT_HYBRIDIII.
  • Fixed locking joint when Lagrange multipliers are used.
  • Fixed bug related to qlo and qhi activation angles for shoulder abduction in all *COMPONENT_HYBRIDIII dummies.
  • Made minor fix to automatic rigid body constraint determination.
  • Change friction law so Hybrid III dummy joints are not as sticky and integration requires less effort.
  • Added updates for coupling with MADYMO 6.3.

Implicit

  • Corrected initialization of tied contact constraints for implicit when explicit/implicit switching is activated.
  • Adjusted implicit logic for artifical stablization in a forming simulation with automatic time step control.
  • Made minor fix to implicit eigenvalue solution.
  • Enabled Interface_springback_seamless after implicit transient analysis.
  • Fixed bug in prescribing x-rotational b.c. in implicit analysis.
  • Adjusted default setting when using full newton with implicit nonlinear. Deactivate line search by setting lstol = -1.0.
  • Fixed problem with beam stresses when switching from implicit to explicit.
  • Implemented mat_98 for solids in implicit.
  • Fixed switch controls via d3kil file to work with implicit.
  • Made change to retain internal energy data when switching from explicit to implicit.
  • Enhanced implicit mechanics to correctly handle spotweld failures.
  • Fixed implicit mat_003. Bad tangent modulus resulted in slow or no convergence.
  • Enabled *control_cpu in implicit.
  • Fixed bug in glstat energy balance when running implicit.
  • Corrected treatment of joints when switching from implicit to explicit. Penalty factor needed to be computed correctly.
  • Fixed implicit's loading of rigid body constraints in local coordinates for seamless springback problems.
  • Improved convergence of implicit 2D automatic contact when the the sliding option is active.
  • Turned off damping in spotweld material if implicit static analysis is active.
  • Fixed problem of active dampers during implicit static analysis.
  • Fixed 2D axisymmetric shell to correct frequency problem.
  • Installed 971 treatment of constrained joints for implicit mechanics.
  • Fixed bugs in implicit dynamics and disabled Rayleigh damping since not properly implemented for implicit solutions.
  • Made change to deactivate damping during static implicit run.
  • Made change to automatically switch to type 10 shell if type 8 is requested for implicit.
  • Fixed bug in shell-brick interface for implicit.
  • Fixed problem of no end time specification for implicit eigenvalue problems.

Miscellaneous

  • Made 970 structured input file compatable with version 971.
  • Fixed bug if *INCLUDE_NASTRAN is used.
  • Fixed an error in checking VDA binary file.
  • Fixed bug if title contains an ampersand.
  • Fixed round-off error related to initial and prescribed velocities.
  • Fixed load curve ID round-off problem in *LOAD_THERMAL_option.
  • Fixed load curve interpolation problem.
  • Fixed bug for *define_box_drawbead.
  • Fixed mass scaling bug affecting solid elements.
  • Made fixes to define_transformation and include_transform affecting:
    • initial velocities
    • mat_084, mat_076
    • airbag_simple_pressure
    • rigid body stoppers
    • MAREA in section_shell
    • damping coefficient
    • rigidwalls
    • inertia when bodies are rotated
    • nodal rigid bodies
    • cross-sections
  • Added option POINT for *define_transformation to provide a means of defining rotations about axes that are reoriented due to previous transformations.
  • Fixed energy consevation problem related to constrained_interpolation.
  • Increased the default memory size from 8500000 to 20000000.
  • Fixed error in *CONSTRAINED_LINEAR.
  • Removed obsolete keywords:
    • *CONTROL_OPTIMIZATION_TAGUCHI
    • *CONTROL_OPTIMIZATION_DESIGNVARIABLE
    • *CONTROL_OPTIMIZATION_PROPERTY
  • Added new keyword *INCLUDE_PATH that defines a directory in which to look for include files.
  • Fixed dynamic relaxation when used with the Wang-Nefske Airbag model.
  • Fixed a bug in superplastic forming simulation so that an arbitrary load curve id can be used for SPF.
  • Added an error termination for bad cyclic symmetry definition.
  • Added active part set option for nfail1 and nfail2 on *CONTROL_SHELL.
  • Centered hourglass control type 6 correctly when 2nd order stress updates are used. This can help meshes that spin.
  • Added *TERMINATION_DELETED_SHELLS to terminate when a specificed number of shells have been deleted from a part or part set.
  • Fixed bug in internal rediscretization of load curve.
  • Fixed negative volume calculation to erode solid elements.
  • Activated new features for superplastic forming simulation.
  • Fixed bugs in *PARAMETER.
  • Airbag_hybrid_chemkin can model convection energy loss from the gas in the airbag to the environment. Added this feature to:
    • airbag_wang_nefske
    • airbag_hybrid
  • Fixed bug related to rigid beams impacting rigid walls.
  • Fixed bug affecting tied_nodes_with_failure used with triangular segments.
  • Made change so that multiple SPC cards for one node are now respected unless more than one coordinate system is referenced in which case an error termination will occur. Previously, only the last of the multiple SPCs was respected.
  • Installed updates from MSC for NASTRAN/DYTRAN reader.
  • Reinstated Lagrange to Lagrange coupling which had inadvertently been disabled in *constrained_lagrange_in_solid.
  • Fixed kinetic energy calculation when RBE3'S (constrained_interpolation) are used.
  • Added *PART_SET_ADD for combining part sets.
  • Made fixes to Suzuki spot weld and added more output, including Part ID. Moved Suzuki spot weld output from d3hsp to messag.
  • Fixed bug for automemory if airbag is defined.
  • Fixed a problem with *NODE_TRANSFORM.
  • Fixed a bug in rigid body switching to prevent local system from being overwritten.
  • Fixed bug affecting *PARAMETER using in an include file.
  • Fixed bug in *initial_stress_section for mat_003 and mat_024.
  • Fixed energy growth and other problems related to constrained_shell_to_solid.

MPP contact

  • Made fix for rigid body boundary conditions in case some processors do not have the rigid body. Noise was being introduced into the solution in this case (MPP).
  • Fixed recent hang in MPP if there are contact interfaces but some processor(s) end up with none.
  • Made modification to forming contact for MPP to handle slave surfaces from solid elements. The "don't allow contacts from below" logic was causing these to be ignored (since they have no thickness and were never seen until they were below the surface....).
  • Fixed bug affecting interface forces file output. The segment set was not output correctly and corrupted the file (MPP only).
  • Fixed bug that could cause incorrect initial penetration results for some nodes in MPP surface_to_surface and node_to_surface contacts.
  • Added damping for beam-to-beam contact (MPP only).
  • Fixed rigid/rigid contact option 1 without friction (MPP only).
  • Fixed MPP segment-based contact error that could occur if contact involved eroding solid elements but the ERODING contact option was not used.
  • Added read of PENMAX parameter for surface to surface eroding contacts in MPP. This allows the release penetration distance to vary from the default value of 0.5*segment thickness.
  • Fixed bug in normal failure calculation in tiebreaks (MPP only).
  • Fixed OFFSET tied contact in MPP if the slave node is tied at the corner point of a triangular shell -- the normal vector was computed as 0.
  • Added guided cable contact (MPP).
  • Added MPP support for automatic tiebreak option 5. Options 1 to 5 are now supported.
  • Fixed beam-to-beam contact problem (MPP only).
  • Made fix to MPP eroding contact in case where no materials were specified. Contact was ignored on any processor that contained no solid or thick shell elements. Now including check for shell elements.
  • Fixed N2S eroding contact (MPP only)
  • Fixed problem of MPP occasionally including internal segments in the active contact surface for eroding contact.
  • Fixed typo that caused failure of type 23 contact (MPP only).
  • Fixed calculation of master energies in tied contacts. Shared nodes were being counted more than once. (MPP only)
  • Fixed a sign error in the force application for MPP drawbeads.
  • Fiaedx bug in MPP drawbead initialization that only showed up in some double precision versions.
  • Removed eroded segments and deleted nodes from the tiebreak interface to get proper response (MPP only).
  • Added MPP support for user input PARMAX and increased the contact searching area.
  • Fixed possible failure of MPP contact when *DAMPING_PART_MASS was used.
  • Changed behavior of MPP drawbead contact to re-interpolate the drawbead nodes each cycle. This avoids cumulative roundoff problems that were causing strange drawbead forces when then tooling was not stationary.
  • Added MPP support for new contact viscous damping option (VDC<0) to contacts type 3/4/5/10/13/26.
  • Fixed MPP initialization of tied contacts when PART_CONTACT thickness is specified.
  • Fixed force transducer problem that occurred during initialization.
  • Fixed a problem of MPP ignoring MPP_DECOMPOSITION_CONTACT_* keywords.
  • Fixed bug in MPP tied contact affecting moved nodes shared by neighboring processors.
  • Fixed problem in computing contact normal vectors for nodes shared by neighboring processors (MPP).
  • Fixed bug affecting translating contact entities (MPP only).
  • Fixed bug in MPP contact types 3/5/10 that could cause some nodes to be missed during bucketsort against triangular segments.
  • Fixed bug in friction work calculation (MPP only).

MPP Only (Miscellaneous)

  • Corrected numbering of d3hsp file in a restart (MPP).
  • Fixed bug affecting output of interface forces file INTFOR (MPP).
  • Added support of *INITIAL_VEHICLE_KINEMATICS (MPP only).
  • Made bug fix for jntforc translation (MPP only).
  • Implemented stress initialization by parts in full deck restart (MPP only).
  • Implemented CHANGE_VELOCITY_ZERO option for full deck restart run (MPP only).
  • Made bug fix to secforc coordinates system (MPP only).
  • Corrected MPP handling for airbag interaction with control volume types 10 and 11.
  • Added support for *damping_frequency_range option (MPP).
  • Fixed bug in MPP implementation of linear constraint equations and added support for LOCAL option of linear constraints.
  • Added options to control birth and death time when dynamic relaxation is active. (MPP)
  • Added 3D adaptivity (MPP).
  • Added MPP support for *control_spotweld_beam.
  • Fixed MPP initialization of rigid body initial velocity.
  • Added MPP support for solid spotweld failure.
  • Added support of ALE tracer particle (MPP).
  • Fixed MPP Madymo coupling problem.
  • Implemented component_hybridIII (MPP).
  • Fixed MPP memory error that occured when *CONSTRAINED_LAGRANGE_IN_SOLID coupling types 1 to 3 were used and there were no ALE parts.
  • Fixed bug in ALE coupling when shells are deleted (MPP).
  • Added basic support for ALE eroding coupling ctype=5 (MPP only). (Not as evolved as SMP implementation.)
  • Made Madymo coupling run more efficiently with MPP.
  • Changed MPP to allow total size of d3thdt files to exceed 2GB.
  • Added recognition of LSTC_BINARY=32ieee to MPP so that 64 bit executables can output 32 bit output.
  • Made change to write error and terminate run if user tries to restart from a dump on a different number of processors than the original run. Previously, job would hang.
  • Added MPP support for nfail (termination on shell element failure) option.
  • Implemented *load_mask in MPP.
  • Implemented tetrahedron formulation 13 in MPP.
  • Fixed problem with *DELETE_CONTACT option. In some cases, the contact to be deleted is not in the processor 0 and then the delete info was not output to the d3hsp01 file. (MPP only)
  • Implemented coordinated termination in MPP for some errors (mass increase, energy ratio, negative volume) so d3plot state is output and termination is clean.
  • Fixed bug affecting porosity leakage through fabric when using ALE switching (MPP only).
  • Fixed bug in boundary_prescribed_motion_node (MPP only).
  • Made some fixes to MPP affecting shell history variable remapping during adaptivity.
  • Fixed bug in Madymo coupling (MPP only)
  • Removed duplicate CONSTRAINED_ADAPTIVTY in dynain file to avoid problem during springback run (MPP only)
  • Made several improvements related to MPP handling of stress initialization/adaptivity.
  • Implemented new ALE FSI interface forces file for MPP.
  • Fixed two bugs in MPP adaptivity: if initial penetration tracking was on, the tracked penetrations were lost at each adaptive step, resulting in creeping penetrations. Also, some contact history was lost entirely when there was more than one node in contact with any master segment. These both resulted in dimples and lumps in the final stamped surface.
  • Fixed bug in nodal interface linking (MPP only).
  • Fixed SPH neighbor search (MPP only).
  • Improved memory allocation for SPH (MPP only).
  • Fixed SPH for 2D problems (MPP only).
  • Added Madymo 6.1 support (MPP).
  • Fixed bug for ALE reference system 9 (MPP).
  • Added support for SPOTSTP option for spotweld connected to rigid part (MPP only).
  • Fixed bug in dbout swforc data of solid spotwelds (MPP).
  • Fixed a bug in ncforc file (MPP only).
  • Fixed a bug affecting read of cross-sectional data (MPP only).
  • Fixed problem of MPP treating NFAIL/NFAIL4 = 2 as NFAIL/NFAIL4 = 1.
  • Fixed possible loss of precision when computing adaptive curvatures in MPP.
  • Fixed inconsistent decomposition when running job with/without pre-decomposition (MPP).
  • Fixed initialization problem for MPP *INTERFACE_LINKING_EDGE.
  • Fixed bug affecting airbag interaction (MPP only).
  • Disabled auto_move for MPP.
  • Fixed bug in interface forces file (MPP).
  • Added header to MPP message files.
  • Fixed sometimes spurious error message of beam "k node on beam x-axis" (MPP).
  • Fixed bug potentially affecting ALE massless nodes with applied loads (MPP only).
  • Implemented implicit joint formulation for MPP.
  • Fixed bug in ncforc (MPP only).
Filed under: