Contact TYPE24:
| • | Possible overcosts due to abnormal number of impact candidates identified. |
| • | Elastic part of contact energy not computed in TH |
| • | Wrong force output for shells in case of node sliding in corners of a T or X connection |
| • | Single precision + AMS + HMPP: possible segmentation violation in Starter |
| • | Possible segmentation violation in SPMD with Inacti=0 and 5. |
Contact TYPE7:
| • | Interface TYPE7 + Igap=2 + Irem_gap=2: wrong computation gap whith shells and solids (RD-4775) |
| • | Interface TYPE7 counter of initial penetrations in RADIOSS starter output message is wrong |
Contact TYPE11:
| • | Possible segmentation violation or possible wrong results with INTER/TYPE11 using /PARITH/OFF |
Tied Contact TYPE2:
| • | Penalty formulation (spotflag=25): |
| - | wrong behavior with the option Idel |
| - | Incorrect contact energy |
| - | No error message is output when a master node is also slave of another tied interface |
| • | Penalty formulation (spotflag=26) |
| - | Not currently recommended due to possible overconstrained master surface. RADIOSS Starter will print a warning and recommend kinematic formulation for shell/shell and shell/spring connections and switch to penalty formulation (spotflag=25) for solid/solid and solid/shell connections. |
| • | Possible negative mass error |
| • | Incorrect contact forces in Anim file when using solid (hexa) elements |
| • | Possible non repeatability issue (parith/on) with Ignore=2, Spotflag=1, Idel2=1. It appears when a node is projected at the border of 2 segments |
Multi-Domain:
| • | Possible starter error with /SPH + /SECT + /FRAME/MOVE |
| • | Possible errors with /FRAME orientation computation related to /SPHBCS when applied to SPH particles belonging to the subdomain |
| • | Possible time-step drop in ALE domain if INTER/TYPE11 in subdomain. |
SPH:
| • | In /SPH/INOUT if Fscaler is not defined, no SPH elements were created. Fscaler default value is now set to 1.0 |
| • | Checkpoint restart (CHKPT) not working correctly with SPH |
| • | Possible corrupted anim files when /ANIM/BRICK/DAMA output was requested for SPH |
| • | Orthotropy directions for SPH particles are now correctly updated and more stable. |
Input/Output:
| • | /STATE/NO_DEL: Writes in sta file connectivity and all variables only for non-deleted elements. This is useful when several load-cases are run sequentially on the same model: we want to use the deformed geometry and the plastic strains coming from the previous run, but the deleted elements. |
| • | /STATE/BRICK/STRESS/FULL for tshell16 is not working |
| • | Extrapolated values of /ANIM/GPS/TENS are incorrect when SPMD is used; similarly stress values for bricks and thick shells (for example SXijk value) in TH output are incorrect |
| • | Possible engine failure when /STATE/BRICK/FAIL + /PROP/TYPE20 (thick shells) are used |
| • | /LOAD/PFLUID & /ANIM/VECT/FEXT: external forces output in anim file is incorrect |
| • | Possible engine segmentation violation error if scalar values output on 1D elements are requested (for example: /ANIM/TRUSS) |
| • | Stress tensor output for coating shells (skin of brick elements) in animation file is zero |
| • | /ANIM/SHELL/EPSP/ALL: wrong plastic strain value in animation output |
| • | Starter outputs an unexpected warning (temperature is not computed) when DEF is requested in /TH/NODE |
| • | If no title in /TH/NODE, the run goes through, but there is no output in TH; error message has been added in the Starter |
| • | Encrypted functions are printed in the starter listing file with /MOVE_FUNCT |
| • | Possible segmentation error in engine due to empty slave node group in /INTER/TYPE2 |
| • | Contact TYPE7 with Iremgap=2: Possible Starter segmentation violation |
| • | Initial strain defined in /INIBRI/STRA_F is not displayed in animation |
| • | Sensor type Gauge - Starter reads Pmin and Tmin in the wrong position |
| • | element thickness (which defined in shell element of Submodel) is ignored if use ITHICK=1 |
| • | /ANIM/VECT/CONT2 : tied contact force vector direction display issue |
| • | Possible segmentation error in starter if section defined only with solid in /SECT |
| • | /SECT: improved colinearity check of nodes N1, N2, N3 which determine local section coordinate system |
| • | Starter prints a wrong element identifier in output file in case of null thickness for SHELL, /PROP/SHELL and /PART |
| • | Printout message for LAW66 improved: c value is output for LAW66 if use Iyld_rate=1,2 |
| • | Kinematic energy of part is not output in time history file |
| • | /INIBRI/STRS_F line duplicated in *.str file |
| • | Segmentation error in engine when using /INIV/AXIS combined with Y translation initial velocity |
| • | Result repeatability not respected when /ADMAS/PART is used |
| • | Possible starter segmentation violation if the model contains unused functions (affects 14.0 and 13.0.214 only) |
| • | Possible engine segmentation violation on win64 with /IMPVEL/LAGMUL |
| • | Starter may output non-ascii characters |
| - | If the number of warnings is too high |
| • | Orthotropy initialization in version 14.0 (/INISHE/ORTHO & /PROP/TYPE9): starter stops with ERROR ID : 26. |
| • | Possible Starter failure if /PROP/SPH cards are encrypted |
| • | Possible Engine failure if MONVOL/GAS with venting is used |
| • | /TH/FRAME output is null if SPMD is used (correct in SMP) |
| • | Transformations and rotations of submodels: sequences are not respected (rotations are applied first) |
| • | Possible engine failure (segmentation violation) when |
| - | /OUTP + SPMD + IMPI are used |
| - | /OUTP/VECT/ROT is requested for solid elements; /OUTP/VECT/ROT is not available for solids |
| - | MAT/LAW32 + PROP/TYPE1 are used |
| - | AMS + PROP/KJOINT2 are used |
| • | Warning 542 (rigid body inertia): inertia components mixed in the warning message |
| • | When a spring is defined by 2 identical nodes, the error message reports a wrong Element_ID |
| • | /INTER/TYPE18: Possible false memory allocation error in Starter (increase MULTIMP) |
| • | Possible engine crash after few cycles when TSHEL16 or BRICK20 results are initialized from .sta file |
| • | Possible Starter segementation error when using KJOINT2 |
| • | Possible Engine segmentation violation when using /RBE2 |
User Subroutines:
| • | Advanced user laws are not working in multithread. |
| • | Wrong user values in animation output if user law has high number of uvar (like number 99). |
| • | Results repeatability (PARITH/ON) is not ensured for solid with user defined property (igtyp > 29). |
| • | Possible segmentation error in engine if interface TYPE2 with failure + using user dynamic library. |
| • | Material User Law in dynamic Library for Shell - Wrong Integration point number |
| • | User sensor cannot be deactivated. |
Implicit:
| • | /ANIM/VECT/DISP output is incorrect (by a factor 2x) wrt displacement computed from nodes coordinates |
| • | /IMPL/NONLIN: Possible engine segmentation violation error if models contains RBE2 or RBE3 |
| • | Results cannot be output for decreasing load in post buckling analysis |
| • | Starter may take too much time to run and hang |
| • | Possible engine segmentation violation when using /IMPL/SOLVER/2 (MUMPS) |
| • | Isolid = 14 not valid for /BRIC20 |
| • | implicit fails if using composite material |
| • | Implicit MUMPS - possible convergence issue on Windows (same model converges on Linux) |
| • | Possible Starter failure in implicit nonlinear if RBE3 are used |
| • | Inconsistent max contact force output in Animation files between SMP and SPMD run. |
General:
| • | /PROP/KJOINT2 + /DT/NODA/CST: in case F(0)<>0, the joint is locked |
| • | /SENSOR based on force does not work as described in manual |
| • | /SENSOR/INTER compatibility with /INTER/SUB ID: possible Starter Error ID 339 (undefined Interface ID) |
| • | Poor scalability in SPMD parallel mode: |
| - | Possible load unbalance with consequent poor scalability when parabolic tetra4 (Irot=1) are used (domain decomposition issue in the Starter) |
| - | Poor scalability with high number of SPMD domains in case of SPH + TYPE7 contact |
| • | /SENSOR/INTER and /SENSOR/SECT compatibility with SPMD: when changing the number of SPMD domains, the sensor is not activated at the same time (impulse instead of force was used) |
| • | Possible time step under-estimation for parabolic tetra4 (Itet=1): time step smaller than tetra10 |
| • | AMS affects rotation of large rigid-bodies |
| • | /TETRA10 : non correct behavior in case of small strain and large rotation |
| • | Isolid=14 and Isolid=17 (HA8, H8C) give bad results with Single Precision version |
|
