LS-DYNA V971 R3.1 (R3.43919) released
New version of LS-DYNA is released for all common platforms.
The release provides fixes and extensions.
Features and enhancements included in V971 R3 that were not available in 971 R2 include the following:
New Material Models:
*MAT_138
(COHESIVE_MIXED_MODE)*MAT_153
(DAMAGE_3)*MAT_165
(PLASTIC_NONLINEAR_KINEMATIC*MAT_169
(ARUP_ADHESIVE)*MAT_171
(STEEL_CONCENTRIC_BRACE)*MAT_172
(CONCRETE_EC2)*MAT_173
(MOHR_COULOMB)*MAT_174
(RC_BEAM)
Enhanced Material Models:
*MAT_030
(SHAPE_MEMORY; now for Hughes-Liu beams)*MAT_071
(CABLE_DISCRETE_BEAM)*MAT_079
(HYSTERETIC_SOIL)*MAT_191
(SEISMIC_BEAM)*MAT_197
(SEISMIC_ISOLATOR)
- Material orientation by CID number for all orthotropic material models
Boundary Conditions:
*BOUNDARY_PRESCRIBED_ORIENTATION_RIGID
- Input sequence of rigid body rotations about 3 unique axes*BOUNDARY_PRESCRIBED_ACCELEROMETER_RIGID
- Input accelerometer data to drive rigid bodies
Control Options:
- ESORT=2 in
*CONTROL_SHELL
to select DKT formulation for all triangles - IMSCL and DT2MS in
*CONTROL_TIMESTEP
to invoke selective mass scaling - LCINT in
*CONTROL_SOLUTION
to control the rediscretization of material curves
Elements:
- PID in
*INTEGRATION_BEAM
allows mixing of different materials in a beam element *ELEMENT_MASS_NODE_SET
- Distribute a mass over a set of nodes*ELEMENT_MASS_PART
- SET option to distribute non-structural mass over a part set
- Parameter FINMASS to specify total mass (structural + non-structural mass)- Solid tet ELFORM 13 implemented for materials 81, 82, 120, 123, 124
Initial Conditions:
*INITIAL_STRAIN_SOLID
- Can now initialize strain in solid elements*INITIAL_STRESS_DEPTH
- Automatically initialize stress in solids (stress assumed linear with depth)
Loads:
*LOAD_BLAST_ENHANCED
for pressure from multiple and non-spherical charges
-*DATABASE_BINARY_BLSTFOR
to create binary output of blast pressure*LOAD_BODY_GENERALIZED
- SET_NODE option to select node set for body loads and optional local coordinate system
- SET_PART option to select part set for body loads and optional local coordinate system
Output:
- JNTFORC now includes energy for each penalty-based joint
- SLEOUT now includes frictional energy for each contact
- EIGOUT now includes participation factors for each mode in an eigenvalue analysis
General:
- Staged Construction
- Breaks building construction process, including excavation, into stages. - Particle method for modelling inflation of airbags
*PART_DUPLICATE
to automatically duplicate and transform parts- Tolerable inertia limits for rigid bodies are extended when using a double precision executable
The remainder of this release notes describes many other enhancements, additions, and bug fixes made since the release of LS-DYNA version 971 R2, rev. 7600.1224.
General
- Fixed errors in Genoa keyword reader.
- Made this version compatible with new Genoa library.
- Fixed the inconsistently of analysis time for LS-DYNA and Madymo.
- Fix bug in staged construction - could prevent DYNAIN file being written for final stage.
- Fix an MPP full deck restart problem that only occurs if the new and previous models had a different number of rigid shell elements.
- GM filter data alignment problem on HP UX (MPP only).
- Preserved the case for
*KEYWORD_ID
. - Removed
*TRANSLATE
keyword. - Bug fix for Madymo coupling (MPP only).
- Fix for GM filter to take care of memory clobber.
- Bug fix for full deck restart (MPP only).
Airbag
- Fixed an MPP bug in obsolete DEPTH=4 segment based airbag contact option.
- Fix for problem with HCONV when defined by a load curve ID.
- Fix for new option in airbag linear fluid.
- Add new DECOMP option based on airbag reference geometry instead of folded geometry. It gives better load balancing while the bag is fully deployed.
- Added pressure limit to airbag LINEAR_FLUID option.
ALE
- Fixed bug for DBSENSOR - memory of sensor location has been reused by ALE sorting routine (MPP only).
- Bug fix in ALE tracer.
- Fix an undefined variable in ALE edge coupling.
- Bug fix for ALE edge coupling.
- Support more than 10 AMMGs in ALE.
- Fix for ALE mass scaling.
- Fix for calculating new velocities and temperature after advection for ALE single material with void.
- A bug fix in
*AMBIENT_HYDROSTATIC_ALE
. For the uppermost AMMG there is no need to determine mixed element or not. - A bug in
*INITIAL_HYDROSTATIC_ALE
. e0 in*EOS
was calculated wrong. - Fixed segment based eroding contact when used with an ALE solution. The contact either crashed or failed to detect some contact after erosion.
- Fix time step drop for ALE problem (MPP only).
- Fix memory problem while using ALE coupling CTYPE=5 (MPP only).
- Fix for multiple ALE ELFORM 5 parts.
- A small bug fix for ALE small restart.
Boundary Conditions
- Fix bug associated with a double precision array overrunning boundaries in
*BOUNDARY_PRESCRIBED_ACCELEROMETER
. - Fix bug: Angle shift variable was being converted to a load curve in
*BOUNDARY_PRESCRIBED_ORIENTATION_RIGID_ANGLES
. - Retain more digits for birth and death times of
*BOUNDARY_PRESCRIBED_MOTION_RIGID
to avoid truncation error. - Fix treatment of scaling factor for prescribed motion on rigid bodies with option 4 and 8.
- Fix SPC force reporting error for implicit when mass scaling is active.
- Allow more than one
*BOUNDARY_PESCRIBED_MOTION_RIGID
to be applied to the same rigid material if at least one of them is initially turned off by sensors. - Output first state of SPCFORC at t=0 rather that than t=dt.
- Resolve
*MAT_ACOUSTIC
,*BOUNDARY_PART_ACOUSTIC
discrepancies with two-sided coupling. - Fix a load curve check for
*BOUNDARY_PRESCRIBED_ORIENTATION_RIGID_DIRCOS
. Add load curve check (consistent number of points) for _ANGLES and _EULERP. - Fix roundoff problem in
*BOUNDARY_PRESCRIBED_MOTION_RIGID
. - Fix treatment of prescribed motion RIGID_LOCAL option.
- Fix for prescribed acceleration problem with scale factor being applied twice (MPP only).
- Add use of
*DEFINE_CURVE_FUNCTION
for prescribed motion and implicit.
Component
- Set default generalized coordinates for GEBOD dummies in the event that the positioning file is absent.
Constrained Options
- Implement
*CONSTRAINED_SHELL_TO_SOLID
(MPP only). - Fixes for Bug where
*ELEMENT_INERTIA
is ignored by*CONSTRAINED_INTERPOLATION
. - Added the IFLAG option in
*CONSTRAINED_EXTRA_NODES
to the source code. - Change MPP handling of
*CONSTRAINED_JOINT_TRACTION_TRANSLATIONAL
, which was not working correctly. - Fixed a typo in
*ELEMENT_SUPER
,*CONTROL_SOLID
and*CONSTRAINED_ADAPTIVITY
. - Fix for KE calculation with
*CONSTRAINED_INTERPOLATION
.
Contact Options
- Fix MPP automatic tiebreak contact to avoid causing memory clobber and segfault.
- Fix MPP contact types 3/5/10 that caused some nodes in contact with solid elements to be missed.
- Force orientation of MPP contact interfaces that include solid elements to the orientation implied by those solid elements.
- Fix uninitialized variable that broke most MPP implicit tied contacts.
- Bypass soft=4 contact option in MPP. Any such contacts are switched to soft=1, and a warning issued.
- Fix bug (problem with MPP AUTOMATIC_TIEBREAK option 6).
- Offer the same output of 'contact gap' (INTFOR file) for automatic tiebreak options 10 and 11 as already available for option 8.
- Added tied penalty contact support to MPP implicit.
- Fixed an MPP bug in obsolete DEPTH=4 segment based airbag contact option.
- MPP fix for auto tiebreak type 9 and 11, extra damping parameters added.
- Add warning that implicit does not support
*CONTACT_ENTITY
. - Fixed restart bug if tiebreak is used.
- Bug fix for single edge contact (MPP only).
- MPP support for soft=1/implicit time step fixes.
- Check for redundant tied constraints (implicit).
- Fixed a problem with large forces in segment based contact caused by tied contacts moving nodes in the first cycle.
- Fixed a segmentation fault problem with binary interface force output of force-transducer contact.
- Fixed 2-surface force transusers with MPP segment-based contact. It was possible for a job to hang.
- Fixes bug: MPP smooth contact was not offsetting the control points for some non-forming contacts.
- Fix for MPP implicit contact/stiffness scale factor for nodes not in contact and for solid elements. Avoids divide by 0.
- Enhance MPP beam-to-beam contact to agree with SMP results (MPP only).
- Fixed bug in
*CONTACT_TIEBREAK_SURFACE_TO_SURFACE
with failure values set incorrectly. - Fix for SPH problem during contact initialization.
- Support type 26 beam contact force transducer (MPP only).
- Fixed the OPTT option on
*PART_CONTACT
for segment based contact. It was using the input value as the added thickness rather than half of it. - Bug fixed for smooth contact output information.
- Fixes for MPP contact in implicit.
- Modification for parameter usage of automatic tiebreak contact (dycoss9). If CT2CN and CN are > 0, then normal stiffness CN and tangential stiffness
CT=CN*CT2CN
are used, otherwise penalty stiffness is used. - Fixed segment based eroding contact when used with an ALE solution. The contact either crashed or failed to detect some contact after erosion.
- Fix computation of separation distance for tied contact after an adaptive step.
- Fixes to triangular thick shell contact.
- Fixed a bug that could cause MPP segment based contact to hang if both single surface and surface eroding contact were used in the same model.
- Some MPP single surface contact improvements.
- Fix MPP non-single surface contact behaviour when viscous damping is given as a negative value.
- Fix for unloading curve in rigid body to rigid body contact.
- Fix for MPP implicit contact convergence.
- Change MPP single surface and surface to surface contacts to do calculations in double precision, pad the initial penetration (by roundoff amount) and do edge projection (was missing in single surface for triangles).
- Fix some possible cases of MPP hanging in problems with automatic-tiebreak contacts.
- Change to the way MPP handles initial penetrations, so the full penetration amount is compensated for at first contact (no tolerance). The fixes some cases of nodes moving in models with no loads or movement.
- Added Contact ID to contact message "Penetrating node eliminated ...".
Control Options
- Properly set implicit drilling rotation control for tetra 4 elements.
- Add GEN (LSTC_REDUCE) to PFILE and keyword
*CONTROL_MPP_IO_LSTC_REDUCE
to use special REDUCE function in LSTC source to get consistent summation across processors (MPP only). - Fixed 3 new items related to
*CONTROL_DEBUG
output for spot welds that use*MAT_100_DA
. The first item is to add the output for weld assemblies. The second is to output parts ID's instead of zeros when default connection data is used. The third is to output strength values without scaling by the strain rate curves. - Add decomp option
*CONTROL_MPP_DECOMPOSITION_ELCOST
2 or, in PFILE,
decomp { elcost 1 }
option : 1 fjvpp, sun
2 ia 64, AMD
3 Intel xeon 64, xeon 32
4 other element costs for decomposition calculation. Users should be able to use different platforms for pre-decomposition to get consistent decomp (MPP only). - Added option to
*CONTROL_ACCURACY
. If INN=-2 triangular elements do not use the invarient system. - Fix for ESORT=2 problem when mixed with 1 point elements.
- If a segment fails which is attached to 1 node of a spotweld brick then delete the brick.
- Support RUNRSF NR option (MPP only).
- Fix related to shell element sorting and bulk viscosity. Bulk viscosity was not considered, but now is.
- Fixed a typo in
*ELEMENT_SUPER
,*CONTROL_SOLID
and*CONSTRAINED_ADAPTIVITY
. - Fix bug if
*CONTROL_ACCURACY
2nd order stress update is on and fully integrated solid elements are used with EOS. - If implicit controls are present in the structured input deck but no implicit analysis is to be performed set MTHSOL = 0 so that explicit element computations are not comprimised by sloppy user input.
- Fix for 3D EFG adaptivity with using TDEATH in adaptive control.
- Allow part set to be specified for
*CONTROL_TIMESTEP
erode. - Add new DECOMP option based on airbag reference geometry instead of folded geometry. It gives better load balancing while the bag is fully deployed.
- Added some useful output for LPRINT .ge. 1 (implicit only).
- Added LCINT to input for
*CONTROL_SOLUTION
to allow the number of intervals to be defined in the load curve discretization. - Fixed fabric (
*MAT_034
) triangle elements when ESORT=2 on*CONTROL_SHELL
. A major history variable bug caused two-thirds of triangular elements to have zero stress.
Damping Options
- Minor fix for accelerations in NODOUT whenever global damping is active.
*DAMPING_FREQUENCY_RANGE
energy now included in System Damping Energy in GLSTAT file. Previously, the energy was counted as negative external work - this is no longer the case.- Fix MPP non-single surface contact behaviour when viscous damping is given as a negative value.
Database/Output
- Fix some MPP problems in the INTFOR DATABASE.
- Fixed the effective strain rate calc in spot weld assemblies that use Daimlerchrysler failure.
- Fixed thick shell output with CMPFLG=1 on
*DATABASE_EXTENT_BINARY
. Also, fixed*MAT_002
with thick shell element type 3. - Fixed stress output for type 1 thick shell elements in D3PLOT and ELOUT files.
- BINOUT support for energy in LSDA based JNTFORC file.
- Print decomposition file and # of processors info to the screen, D3HSP, and MESSAGE files when MPP execution starts.
- Offer the same output of 'contact gap' (INTFOR file) for automatic tiebreak options 10 and 11 as already available for option 8.
- Fix bug for modal superposition using existing D3EIGV.
- Set all other components to zero the in the BLSTFOR buffer array.
- Bug fix for INTFOR file (MPP only).
- Bug fix for D3HSP output of
*MAT_083
parameters HU and SHAPE. - Fixed output of
*MAT_221
to D3HSP. - Add echo to D3HSP of control data for pore fluid and staged construction.
- Fixed bug in writing D3PART (MPP only).
- Bug fix for D3PART output (MPP only).
- Fix for rigid wall ID in the BINOUT files.
- Fix GLSTAT system damping energy calculation (MPP only).
- MPP support for node mass info output to D3PLOT.
- Fix to write all integration points to the ELOUT file for 2D elements.
- Fix for segmentation fault when writing the AVSFLT file.
- Fixed confusing output problem in D3HSP when CON2 in
*MAT_020
is set to 011111. Also, illegal values of CMO are not detected in the input. - Error termination will now occur.
- Fix printout for
*MAT_018
. Bug: coefficient "n" is printed as "m". - Add MPP support for INTFOR file larger than 4GB.
- Minor fix for accelerations in NODOUT whenever global damping is active.
- Support RUNRSF NR option (MPP only).
- Fix NCFORC output in LSDA/BINOUT file for interfaces with IDs of more than 5 digits.
- Two minor fixes in D3HSP output of joint stiffnesses.
- Fix bug in pore pressure output - on MPP systems the keyword
*DATABASE_PWP_OUTPUT
could cause LS-DYNA to hang with certain large models. - Change to keyword processing of solid/ beam/ shell/ thickshell element labels/ids related to time history blocks: previously they were all written together in the structured file. This caused the D3HSP file to be wrong (all were identified as being SOLID elements), as well as the legends being written to the wrong places in the BINOUT file.
- Fixed bug for adaptivity and
*DATABASE_BINARY_D3PART
. - Avoid deadlock for NODOUTHSF (MPP only).
- Fix for LSDA MATSUM output: the material numbers were mixed up.
- MPP fix for incorrect mass reported in RCFORC file.
- Fixed hex spotweld assembly output to SWFORC.
- Fix for bug related to SECFORC file and null beams.
- Bug fix for rb-velocity output to D3PLOT, MATSUM, GLSTAT (MPP only).
- MPP fix for RBDOUT output of rigid bodies that are shared among processors.
- Fixed thick shell type 2 strain output.
- Fixed output of
*DEFINE_CONNECTION
properties to MESSAGE file to output parts instead of material ID's. - Recoded thick shell output to ELOUT, D3PLOT, D3PART, D3THDT and D3HSP. Due to missing info, the CMPFLG=1 output is not available for
*MAT_002
or*MAT_086
or for any material with thick shell type 3. - Multi flexible bodies with KMFLAG in MPP version (for D3EIGV and for NASTRAN.pch).
- Added local coordinate system option for RCFORC file.
- Fix for
*INCLUDE_TRANSFORM
related to ASCII DATABASE. - Fixed thick shell strain output for case of more than 10 through thickness integration points or 3 or 5 integration points with user defined integration or Labatto quadrature.
- Added Contact ID to contact message "Penetrating node eliminated ...".
- Fix for RBDOUT file in the LSDA/BINOUT file, when rigid deformable switching is active, the number of rigid bodies can be different every cycle.
Define Options
- Added damage type DG_TYP=3 to
*DEFINE_CONNECTION_PROPERTIES
. - New _VECTOR option for prescribed orientation.
- Fix for
*DEFINE_COORDINATE_..._TITLE
when used with*INCLUDE_TRANSFORM
. The title was lost causing major problem when using the dyna.inc file. - Fix sign error in evaluation of
*DEFINE_FUNCTION_TABULATED
. - Bug fix for MPP using
*DEFINE_FUNCTION
with BEAM() function (MPP only). - Fix bug in MPP treatment of
*DEFINE_CURVE_FUNCTION
. - Fixed output of
*DEFINE_CONNECTION
properties to MESSAGE file to output parts instead of material ID's. - Add use of
*DEFINE_CURVE_FUNCTION
for prescribed motion and implicit.
Deformable
- Transform stresses for solid elements after switching from rigid to deformable.
EFG
- Bug fixed in semi-Lagrangian kernel in stabilized EFG.
- Fix for machine dependency problem in 3D EFG adaptivity.
- Bug fixed in axisymetric EFG - replace the generalized disp by nodal value.
- Fix 3D EFG adaptivity for
*MAT_024
. - Bug fixed in one default setting relating to new EFG formulation.
- Fix several bugs with 3D EFG adaptivity due to new implementation.
- Make 3D EFG adaptivity work for hex elements at the start.
- Disable shape function reconstruction for 8 nodes solid in EFG adaptivity.
- Bug fixed related to IEBT=4 in stabilized EFG.
- MPP support for EFG solid element new feature (MAXENT).
- Bug fixed in restart adaptive EFG. Added global control in Semi-Lagrangian kernel for stabilized EFG (8-noded cell).
- Added stabilization terms in EFG MAXENT approximation in 8-noded cells.
- Bug fixed for MPP EFG shell type 41.
- Included 4-noded integration cell in MAXENT approximation for EFG 3D.
- Included 6-noded integration cell in MAXENT approximation for EFG 3D.
- Considered the diffusive term in the MAXENT approximation for EFG 3D.
- Bug fixed in MAXENT to allow semi-Lagrangian kernel to be activated.
- Normalized the distance function in MAXENT for EFG. This is to reduce the conditioning number in Hessain matrix.
- Added a modified maximum entropy approximation for 3D EFG.
- Bug fixed in EFG MPP (3D).
- Fixed a single precision bug in EFG solid 41.
- Fix for 3D EFG adaptivity with using TDEATH in adaptive control.
- Fixed a bug in EFG shell 42 single precision.
- Added two options in implicit 2D EFG.
1. 2nd order basis function.
2. Higher-order integration rule. - Added an implicit flag in EFG 2D.
- Modified the stiffness in 2D plain strain EFG.
Elements
- Force orientation of MPP contact interfaces that include solid elements to the orientation implied by those solid elements.
- Fix for bug related to lack of convergence for tet type 4.
- Fixes for Bug where
*ELEMENT_INERTIA
is ignored by*CONSTRAINED_INTERPOLATION
. - Fixed thick shell output with CMPFLG=1 on
*DATABASE_EXTENT_BINARY
. Also, fixed*MAT_002
with thick shell element type 3. - Fixed shell elements with user integration that mix materials. Before the fix, if materials had non-zero history data such as
*MAT_002
, the result was likely bad. - Added an error check for mix materials with laminated shell theory with other material types.
- Fixed laminated shell theory when for elements with mixed material models of the same type.
- Fix for spring type 119 which showed too much mass scaling.
- Disable shape function reconstruction for 8 nodes solid in EFG adaptivity.
- Bug fix for
*INITIAL_STRESS_SOLID
with ELFORM 11. - MPP support for EFG solid element new feature (MAXENT).
- Major bug fix for ELFORM 11 advection that should help with stability.
- Fix for
*MAT_120
for nodal pressure tet (ELFORM 13). - Bug fix for initial stresses in solid elements: Element sorting to tets and pentas was not supported. Pentas were not supported at all. Writing of binary DYNAIN was also broken.
- Bug fixes for warped element type 11.
- Fixed memory expansion bug related to shell seatbelts.
- Fix bug in dummy argument for
*INITIAL_VELOCITY_GENERATION
bug with SPH elements. - Rotational mass bug fixed for shell 24.
- Fix bug for forces exchanging for shell type 25 and 26 (MPP only).
- Add new input processing capability for superelements.
- Bug fixed for MPP EFG shell type 41.
- MPP support for eroding/failing cohesive elements.
- Fix for wrong results with type 3 solid and NCPU=-number.
- Fix for cohesive element deletion to correctly handle both tied and merged cohesive nodal points.
- Added option to
*CONTROL_ACCURACY
. If INN=-2 triangular elements do not use the invarient system. - Transform stresses for solid elements after switching from rigid to deformable.
- Time step calculation problem with solid element type 12 and
*MAT_106
. - Fixed a bug in type 20 triangles for buckling analysis.
- Added MPP code so that solid weld elements will be deleted if any of their nodes are attached to a shell that erodes. This matches the SMP behaviour.
- Bug fix for parallel consistency with shells 25+26.
- Fix to change solid element type 18 to type 1 for explicit calculations. Type 18 is for linear static problems only.
- Fix for timestep calculation for tet 17 + correct connectivity output for tet 17 in the DYNAIN file.
- Fix to write all integration points to the ELOUT file for 2D elements.
- Fix to terminate calculation if
*ELEMENT_MASS_PART
is used to add mass to solids or beams. Also, minor change to avoid including added mass distribution specified on the*SECTION_SHELL
in the determination of the rotational mass at the nodal points. - Fix to prevent segmentation fault if
*SECTION_SOLID
has ELFORM=11 which is unused. - Add
*MAT_169
,*MAT_173
,*MAT_198
to wedge element (ELFORM 15) - without this, the automatic sorting would not work with these materials. - Fix for segmentation fault when NPLANE=4 with shell elements types 1 and 6.
- If a segment fails which is attached to 1 node of a spotweld brick then delete the brick.
- Support type 26 beam contact force transducer (MPP only).
- Support correct reading and writing initial stresses for element type 3.
- Add
*MAT_084
for wedges (ELFORM 15) and add missing calls to rstrss for tet elements (ELFORM=10) for*MAT_084
. - Fix for 2D adaptive remeshing to include initialization of 2D beam elements. Also fixed problem related to 2D rigid bodies sharing nodes with deformable bodies that are remeshed.
- Fixed bug in Belytschko-Schwer beam related to offset vectors not being accounted for in the calculation of the beam fiber vectors. This caused large forces to develop with no applied loads.
- Enable type 3 and 17 triangle shells for
*MAT_058
. - Fixed collapsed type 6 and 7 shell elements that are sorted to type 4 triangles for use with anisotropic materials. The stress was wrong if angles per layer was used.
- Fixed shell element types 6 and 7 when used with aniostopic material. The angle cosines and sines were not being initialized correctly. Also, fixed the output of stress and strain in shell elements that use different material models in different layers using either
*INTEGRATION_SHELL
or*PART_COMPOSITE
. - Fix bug that may cause error in reading of integration point dat for user defined element.
- A bug fix in
*AMBIENT_HYDROSTATIC_ALE
. For the uppermost AMMG there is no need to determine mixed element or not. - Fix for beam elements whose internal energy is not kept as eroded internal energy when the beam fails.
- Fix bug for SPH elements during initialization of EOSs type
*EOS_IGNITION_AND_GROWTH_OF_REACTION_IN_HE
and*EOS_PROPELLANT_DEFLAGRATION
. - Small mod to slipring convergence: if the user input 0.0 for the material "birth size" (min element size), then the code would not converge for sliprings, because this value was used when computing the initial guess. Now if they input 0.0, a small fraction of the element size is used.
- Change to keyword processing of solid/ beam/ shell/ thickshell element labels/ids related to time history blocks: previously they were all written together in the structured file. This caused the D3HSP file to be wrong (all were identified as being SOLID elements), as well as the legends being written to the wrong places in the BINOUT file.
- Fixed a typo in
*ELEMENT_SUPER
,*CONTROL_SOLID
and*CONSTRAINED_ADAPTIVITY
. - Fixed bug in restart job if deleted element set is specified.
- Fixed a single precision bug in EFG solid 41.
- Fix to stop auto sorting of degenerate solids for
*MAT_100
,*MAT_126
. - Fix bug if
*CONTROL_ACCURACY
2nd order stress update is on and fully integrated solid elements are used with EOS. - Fix for zero length discrete beams with SCOOR set to 2.
- Fix for
*MAT_123
for solid elements. - Fix for flexible rigid body with only solid elements.
- Add solid element connectivity array to loadud.
- Add effective plastic strain failure to
*MAT_106
shells. - Fix for bug related to SECFORC file and null beams.
- Fixes to triangular thick shell contact.
- Add fix for implicit MPP with superelements and inertia relief.
- Fixed a bug in EFG shell 42 single precision.
- Fixed thick shell type 2 strain output.
- Recoded thick shell output to ELOUT, D3PLOT, D3PART, D3THDT and D3HSP. Due to missing info, the CMPFLG=1 output is not available for
*MAT_002
or*MAT_086
or for any material with thick shell type 3. - Support the missing option in
*ELEMENT_MASS_PART_SET
(MPP only). - Fix for load curve option in
*ELEMENT_MASS_PART
keyword. - Added time dependent added mass option by load curve for keyword
*ELEMENT_MASS_PART
. - Fixed thick shell strain output for case of more than 10 through thickness integration points or 3 or 5 integration points with user defined integration or Labatto quadrature.
- Fix for multiple ALE ELFORM 5 parts.
- Fixed a bug in
*INITIAL_VELOCITY_GENERATION
for solid type 17. - Changes for beam types 2 and 12 with unsymmetrical cross-section. Already done for R4, now for R3.
- Bug fix for one node deformable spotweld.
- Fix bug, initialization of 1 noded beam/spotwelds, to better handle the case where there are parallel layers of material.
EOS
- A bug in
*INITIAL_HYDROSTATIC_ALE
. e0 in*EOS
was calculated wrong. - Slight modifications to prevent a nonzero pressure state in the reference configuration for parts with equation of state.
- Fix bug for SPH elements during initialization of EOSs type
*EOS_IGNITION_AND_GROWTH_OF_REACTION_IN_HE
and*EOS_PROPELLANT_DEFLAGRATION
. - Fix bug if
*CONTROL_ACCURACY
2nd order stress update is on and fully integrated solid elements are used with EOS.
Hourglass Options
- Fix for Puso hourglass control with a negative coefficient. Internally the negative coefficient caused Puso control to get confused with Lee
- Bindeman's type 7 which quickly caused the problem to go unstable. Bug #549.
Implicit
- Reset stress initialization for when implicit rejects a time step.
- Fix uninitialized variable that broke most MPP implicit tied contacts.
- Properly set implicit drilling rotation control for tetra 4 elements.
- Revision to
*MAT_106
to select properties based on ending time of implicit step. - Fix for tangent stiffness (implicit) of cohesive
*MAT_138
and*MAT_186
. Implicit had to get activated for*MAT_186
in the first place. - Added tied penalty contact support to MPP implicit.
- Modified
*MAT_106
to optionally use the total alpha for computing incremental thermal strains instead of the instantaneous alpha. Required for implicit. - Fix SPC force reporting error for implicit when mass scaling is active.
- Add warning that implicit does not support
*CONTACT_ENTITY
. - MPP support for soft=1/implicit time step fixes.
- Fix for soft=1 option for implicit use. Use element time step calculation for dt2 rather that the input value from the user which can create big problems.
- Check for redundant tied constraints (implicit).
- Fix for MPP implicit contact/stiffness scale factor for nodes not in contact and for solid elements. Avoids divide by 0.
- If implicit controls are present in the structured input deck but no implicit analysis is to be performed set MTHSOL = 0 so that explicit element
- computations are not comprimised by sloppy user input.
- Fixes for MPP contact in implicit.
- Fix implicit's treatment of interface linking and a further enhancement to the shell-to-solid interface treatment.
- Add fix for implicit MPP with superelements and inertia relief.
- Revision 2 of fix for implicit's treatment of shell-to-solid interface.
- Added some useful output for LPRINT .ge. 1 (implicit only).
- Added two options in implicit 2D EFG.
1. 2nd order basis function.
2. Higher-order integration rule. - Fix interplay between implicit and explicit when inertia relief is involved.
- Fix implicit's treatment of shell-to-solid interface constraint.
- Add implicit sense switches to screen output.
- Fix for MPP implicit contact convergence.
- Added an implicit flag in EFG 2D.
- Add use of
*DEFINE_CURVE_FUNCTION
for prescribed motion and implicit. - An bug related to
*MAT_060
in implicit is now fixed. - Fix for
*MAT_002
with initial stress in implicit shell.
Include Options
- Fixed bug in generated keyword file for
*INCLUDE_TRANSFORM
when Genoa material is used. - Bug fix for
*MAT_WINFRITH_REINFORCEMENT
+*INCLUDE_TRANSFORM
. - Add option STRESS=-3 for
*INCLUDE_STAMPED_PART_BINARY
, which activates long format for history variables and NO mapping of stresses. - Fix for
*DEFINE_COORDINATE_..._TITLE
when used with*INCLUDE_TRANSFORM
. The title was lost causing major problem when using the dyna.inc file. - Fixed
*INCLUDE_TRANSFORM
bug related to shells where _THICKNESS was left off of dyna.inc keyword. - Fix for
*INCLUDE_TRANSFORM
related to ASCII DATABASE.
Initial Options
- Bug fix for
*INITIAL_STRESS_SOLID
with ELFORM 11. - Fix bug in dummy argument for
*INITIAL_VELOCITY_GENERATION
bug with SPH elements. - Fixed bug that caused the wrong beam stresses to be written to DYNAIN file (
*INITIAL_STRESS_BEAM
). - Implement
*INITIAL_VELOCITY_GENERATION
for SPH elements defined by part or set part. - Remove the limitation on number of geometries<=100 in
*INITIAL_VOLUME_FRACTION_GEOMETRY
. - Fixed bug in
*INITIAL_STRESS_BEAM
- did not work if beam has 2 x 2 integration points. - Fix for
*INITIAL_STRAIN_SHELL
which now works even if there is no*INITIAL_STRESS_SHELL
input. - Fix for segmentation fault when NPLANE=4 with shell elements types 1 and 6.
- Support correct reading and writing initial stresses for element type 3.
- A bug in
*INITIAL_HYDROSTATIC_ALE
. e0 in*EOS
was calculated wrong. - Add MPP/idtrans support for
*INITIAL_LOAD_AXIAL_BEAM
. - Added new option to initialize bolt stress:
*INITIAL_AXIAL_FORCE_BEAM
. - Fixed a bug in
*INITIAL_VELOCITY_GENERATION
for solid type 17.
Interface
- Fix MPP support for INTERFACE_FORCE file reading, when segment linking is used and there are more than 1 segment interfaces.
- Fix NCFORC output in LSDA/BINOUT file for interfaces with IDs of more than 5 digits.
- Fix implicit's treatment of interface linking and a further enhancement to the shell-to-solid interface treatment.
- Revision 2 of fix for implicit's treatment of shell-to-solid interface.
- Fix implicit's treatment of shell-to-solid interface constraint.
Loads
- Treat enhanced blast loading as an unsupported feature (MPP only).
- Added DYNAIN output for SPF. Fixed a bug in SPF with adaptivity.
- Fix bug in
*LOAD_THERMAL_VARIABLE
if >99999 load curves. - Bug fix for
*LOAD_BEAM_ELEMENT
idtrans translation (MPP only). - Restrict unit system to IUNIT=3 for enhanced blast model type=3.
- Fix load curve ID output for nonuniform load option.
Material Models
- Fixed
*MAT_034
output error that caused a crash on SGI IRIX systems. - Fixed the effective strain rate calc in spot weld assemblies that use Daimlerchrysler failure.
- Update
*MAT_115
shell subroutine to fix single precision bug. - Fixed thick shell output with CMPFLG=1 on
*DATABASE_EXTENT_BINARY
. Also, fixed*MAT_002
with thick shell element type 3. - Fix 3D EFG adaptivity for
*MAT_024
. - Fixed shell elements with user integration that mix materials. Before the fix, if materials had non-zero history data such as
*MAT_002
, the result was likely bad. - Added an error check for mix materials with laminated shell theory with other material types.
- Fixed laminated shell theory when for elements with mixed material models of the same type.
- A small fix to avoid an error for
*MAT_135
for alpha and sgi platforms. - Fix for
*MAT_020
input issue. - Fix for
*MAT_108
related to bug. The orthotropic transformations were missing. - Added an error check for spot weld material that is used by both individual hex spot welds and also hex weld assemblies. The individual welds are not checked for failure if the material is shared.
- Fix for local coordinate systems with large ID's in
*MAT_020
. - Added spot weld failure to thermal.
- Revision to
*MAT_106
to select properties based on ending time of implicit step. - Bring the element information to the stiffness construction in user material.
- Fix for tangent stiffness (implicit) of cohesive
*MAT_138
and*MAT_186
. Implicit had to get activated for*MAT_186
in the first place. - Fixed 3 new items related to
*CONTROL_DEBUG
output for spot welds that use*MAT_100_DA
. The first item is to add the output for weld assemblies. The second is to output parts ID's instead of zeros when default connection data is used. The third is to output strength values without scaling by the strain rate curves. - Enable new damage type DG_TYP=3 also for hex weld assemblies.
- Fix for Gurson
*MAT_120
. With initial stresses from DYNAIN several elements failed at t=0 due to a bad starting value for the plasticity algorithm. Now, a better starting value is used for this case. - Fix for
*MAT_100_DA
with spotweld assemblies. The "no failure" option (negative CON_ID) was not working before. - Fix for
*MAT_120
for nodal pressure tet (ELFORM 13). - Bug fix for
*MAT_WINFRITH_REINFORCEMENT
+*INCLUDE_TRANSFORM
. - Modified
*MAT_106
to optionally use the total alpha for computing incremental thermal strains instead of the instantaneous alpha. Required for implicit. - Fix to
*MAT_106
due to initialization problem when restarting with a DYNAIN file. - Load curve input added to Gurson material
*MAT_120_JC
(card 5, row 8). This load curve optionally replaces expression with D1, D2, and D3 (see manual). - Fixed RATE option of
*MAT_123
. - Small modifications for
*MAT_187
parameters. - Fix for
*MAT_123
to solve element deletion problem. - Bug fix for D3HSP output of
*MAT_083
parameters HU and SHAPE. - Fixed output of
*MAT_221
to D3HSP. - Added
*MAT_221
. - Fix for calculating new velocities and temperature after advection for ALE single material with void.
- Increased minimum number of history variables from 11 to 14 for
*MAT_187
. - Fix for TITLE option for
*MAT_ADD_EROSION
and*MAT_ADD_THERMAL
. - Time step calculation problem with solid element type 12 and
*MAT_106
. - Fix bug for explosive materials with SPH. Elements with
*MAT_HIGH_EXPLOSIVE_BURN
were not detonating. - Change pressure calculation in
*MAT_084
to incremental. This fixes a bug that caused pressure jumps when the algorithm switched between crack-open and crack-closed modes. Results from existing models are likely to change. - Fix principal stress direction calculation in
*MAT_173
. - Fixed bug in
*MAT_024
viscoplastic, vp=1.0. - Fixed confusing output problem in D3HSP when CON2 in
*MAT_020
is set to 011111. Also, illegal values of CMO are not detected in the input. - Error termination will now occur.
- Fix printout for
*MAT_018
. Bug: coefficient "n" is printed as "m". - Fix for bug related to FIR,
*MAT_143
. Core dump occurs rather than clean error termination. - Fixed minor bug related to improper input into
*MAT_036
which caused error termination. - Add
*MAT_169
,*MAT_173
,*MAT_198
to wedge element (ELFORM 15) - without this, the automatic sorting would not work with these materials. - Fix for
*MAT_002
with unreasonable large stress output for small strain values. - Correct left C-G tensor update in
*MAT_112
. - Bugfix in RCDC option of
*MAT_082
. - Resolve
*MAT_ACOUSTIC
,*BOUNDARY_PART_ACOUSTIC
discrepancies with two-sided coupling. - Fix bug in
*MAT_ARUP_ADHESIVE
- for XEDGE=0, element failure could lead to a memory corruption with unpredictable consequences. - Add
*MAT_084
for wedges (ELFORM 15) and add missing calls to rstrss for tet elements (ELFORM=10) for*MAT_084
. - Enable type 3 and 17 triangle shells for
*MAT_058
. - Fixed collapsed type 6 and 7 shell elements that are sorted to type 4 triangles for use with anisotropic materials. The stress was wrong if angles per layer was used.
- Small modification in
*MAT_169
: With new input parameter THKDIR (card 2, column 7), the determination of thru-thickness direction can be changed. THKDIR=0.0: smallest element dimension (default), THKDIR=1.0: direction face 1-2-3-4 to 5-6-7-8. - Fixed shell element types 6 and 7 when used with aniostopic material. The angle cosines and sines were not being initialized correctly. Also, fixed the output of stress and strain in shell elements that use different material models in different layers using either
*INTEGRATION_SHELL
or*PART_COMPOSITE
. - Fix bug that could lead to thermal material data being missed during Keyword input. This occurred occasionally in large models, that contained lots of other data before the thermal material data.
- Small mod to slipring convergence: if the user input 0.0 for the material "birth size" (min element size), then the code would not converge for sliprings, because this value was used when computing the initial guess. Now if they input 0.0, a small fraction of the element size is used.
- Fixed timestep calculation for
*MAT_089
. Previously the material could go unstable if the elastic modulus increased or if the elements gained volume during elastic stretching. - Fix read problem in
*MAT_135
to ensure that is consistent with the new 971 manual. - Fix to stop auto sorting of degenerate solids for
*MAT_100
,*MAT_126
. - Fix for
*MAT_017
to correct the stress reduction after failure. - Fix for
*MAT_123
for solid elements. - Add effective plastic strain failure to
*MAT_106
shells. - Fix for
*MAT_172
(CONCRETE_EC2) beams - FRACA was treated as 1.0 if set to any non-zero value. - Fix for round-off error related to deformation gradient calculations for
*MAT_057
foam. - Added
*MAT_236
based on Boeing user material for silicon carbide coating on reinforced carbon composite. - Fixed
*MAT_021
and*MAT_023
thermal strain so that good for large strain. - Minor bug fix for shell element failure in
*MAT_120
. Resurrection of failed IP was possible in very special cases. - Small modification of
*MAT_138
: more conservative "ek" (stiffness/area) for time step calculation. - Modified cohesive
*MAT_138
to be more robust. Only one damage variable instead of two is used. The input is backwards compatible, i.e. old input files run. - Recoded thick shell output to ELOUT, D3PLOT, D3PART, D3THDT and D3HSP. Due to missing info, the CMPFLG=1 output is not available for
*MAT_002
or*MAT_086
or for any material with thick shell type 3. - Fix for shell
*MAT_120
: element deletion after NUMINT IPs failed did not work correctly. - Fix for
*MAT_DAMAGE_3
,*MAT_153
to store plastic strain. - Fix related to
*MAT_153
where a segmentation fault occurs. - Fix
*MAT_015
to take the absolute value of D3 since many users input this constant as a negative number when it should be positive. - Bug fix for shell
*MAT_123
with VP=1.0. Added upper bound for strain rates if table is used. - Bug fix for keyword input for
*MAT_120
(*MAT_GURSON_JC
): One external load curve number (LCDAM) was not converted to internal. - Fix for bug in
*MAT_187
. - Update to
*MAT_019
,*MAT_031
,*MAT_036
,*MAT_039
,*MAT_040
,*MAT_078
,*MAT_079
,*MAT_089
,*MAT_101
,*MAT_106
,*MAT_127
,*MAT_129
,*MAT_144
,*MAT_177
,*MAT_178
,*MAT_192
(and others) to eliminate load curve ID round-off in single precision. - Modified new history variable #11 of
*MAT_120
and*MAT_120_JC
. - An bug related to
*MAT_060
in implicit is now fixed. - Fixed fabric (
*MAT_034
) triangle elements when ESORT=2 on*CONTROL_SHELL
. A major history variable bug caused two-thirds of triangular elements to have zero stress. - Error stop if initial yield stress is zero for
*MAT_142
. - Fix for
*MAT_002
with initial stress in implicit shell. - Added new history variable (history var#11 in LSPP) to Gurson material (
*MAT_120
and*MAT_120_JC
). It is a dimensionless value showing material damage: gurdam=(f-f0)/(ff-f0). Only for postprocessing.
Node
- Fixed bug for wrong node id number.
Parameter
- Fixed bug in
*PARAMETER
if more than one definitions are specified. - If any
*PARAMETER
definition is duplicated, we will issue an error message and terminate the code after input phase.
Part Options
- Fix for bug related to
*PART_COMPOSITE
and*RIGIDWALL_PLANAR
. - Fix for
*PART_DUPLICATE
line with commas. - Fixed the OPTT option on
*PART_CONTACT
for segment based contact. It was using the input value as the added thickness rather than half of it. - Multi flexible bodies with KMFLAG in MPP version (for D3EIGV and for NASTRAN.pch).
Rigidwall
- Fix for bug related to
*PART_COMPOSITE
and*RIGIDWALL_PLANAR
. - Fix for rigid wall ID in the BINOUT files.
Section Options
- Fix to terminate calculation if
*ELEMENT_MASS_PART
is used to add mass to solids or beams. Also, minor change to avoid including added mass distribution specified on the*SECTION_SHELL
in the determination of the rotational mass at the nodal points. - Fix to prevent segmentation fault if
*SECTION_SOLID
has ELFORM=11 which is unused.
Set Options
- Fix problem related to
*SET_NODE_GENERAL
whenever the part ID is an SPH part. - Fix for
*SET_NODE_ADD_ADVANCED
where memory is clobbered if all node sets are defined with the advanced option.
SPH
- Fix extra file created by tracer particle for SPH.
- Fix bug in dummy argument for
*INITIAL_VELOCITY_GENERATION
bug with SPH elements. - Implement
*INITIAL_VELOCITY_GENERATION
for SPH elements defined by part or set part. - Fix bug for explosive materials with SPH. Elements with
*MAT_HIGH_EXPLOSIVE_BURN
were not detonating. - Fix for SPH problem during contact initialization.
- Fix problem related to
*SET_NODE_GENERAL
whenever the part ID is an SPH part. - Fix bug when multiple symmetry planes with SPH were used.
- Fix bug in SPH initialization of material data.
Termination
- Add support for
*TERMINATION_NODE
and*TERMINATION_BODY
(MPP only).