Altair OptiStruct 2021 Release Notes

HIghlights

  • PART superelements
  • Divergence
  • Frequency domain viscoelasticity
  • Cyclic symmetry
  • Auto-Contact for Explicit Analysis (Beta feature)
  • Altair Compute Console (ACC) GUI to submit jobs

New Features

Stiffness, Strength and Stability
Continuous Sliding (CONSLI) support for Preloaded Linear Analysis
Continuous Sliding (CONSLI) is supported for Preloaded Linear Analysis
MODULUS (Long or Instantaneous) option added in MAT1, MAT9 and MATHE
Modulus specified in MAT1, MAT9 and MATHE can either be based on Long term or Instantaneous. Default is Long term and this option is only relevant when MAT1/MAT9/MATHE are used with viscoelasticity material (MATVE). For the frequency- domain viscoelasticity (MATFVE), the MODULUS option is irrelevant because the specified elastic modulus is always considered as Long-term modulus. Prior to v2021, the modulus was assumed to be instantaneous.
RESTARTR when multiple nonlinear subcases continued from the same subcase
RESTARTR is now supported for the case where each new subcase in restart model will be continued from the same subcase which had already been solved in the original run (before the restart run). This means that all new subcases in the restart run can be independent to each other. Typical use case is all the independent subcases in restart run is continued from the pretension subcase which is solved in the original run (before the restart).
Cast Iron Plasticity
Cast Iron Plasticity is used to model gray cast iron. It allows elastic-plastic behavior with different yield strengths, flow, and hardening in tension and compression.
Cast Iron Plasticity material data can be specified by using the MCIRON Bulk Data Entry which has the same Identification Number as an existing MAT1 Bulk Data Entry. The tensile and compressive stress-strain curves can be defined using the TABLE_T and TABLE_C fields on the MCIRON entry.
Cast Iron material via MCIRON is supported for Small and Large Displacement Nonlinear Static Analysis. It is supported for CHEXA, CTETRA, CPENTA, and CPYRA elements (both first and second order).
Static Stabilization for Axisymmetry and Plane Strain
Static Stabilization is available for axisymmery and plane strain elements. Static Stabilization is activated by STABILIZ on NLADAPT Bulk Data Entry.
2nd order element support for Axisymmetry and Plane Strain
Higher order elements (2nd order) are now supported for axisymmetry and plane strain.
Contact Results for both sides (Positive or Negative) of shell elements
Contact related results such as Contact pressure, status and so on are now available for both sides of shells. The positive or negative sides are determined based on the normal direction of shells.
Contact Status output for Stick and Slip
Closed contact status in h3d file has “Close-Stick” and “Close-Slip” as new status outputs.
Frequency Domain Viscoelasticity (MATFVE)
Frequency domain viscoelasticity material is available with MATFVE Bulk Data Entry. There are several ways to specify the material properties in MATFVE.
  • FORMULA
  • TABLE
  • PRONY
  • PRELOAD
PRELOAD option allows the input of storage and loss modulus from uniaxial and volumetric tests. For information regarding each option, refer to MATFVE Bulk Data Entry. MATFVE can be combined with MAT1, MAT9 or MATHE.

Enhancements

Enhanced Cohesive element output
Additional outputs such as Cohesive energy by mode and Cohesive energy per area by modes are now available.
Explicit Dynamic Analysis:
Auto-Contact (Beta Feature)
Auto-Contact for Explicit Dynamic Analysis is now available. TYPE field on CONTACT Bulk Data Entry should be “AUTO” to activate auto-contact. Using ACTIVA/DEACTIVA continuation line, the particular surfaces specified on that line will be only considered (ACTIVA) or excluded (DEACTIVA) from auto-contact generation. Edge-to-Edge contact can be considered with PSURF entry.
Enhanced 2nd order Tetra elements (10-noded CTETRA)
Enhanced 10-noded CTETRA allows the same time step as first-order tetra (4-noded CTETRA). This enhanced 10-noded CTETRA is now turned on by default for explicit analysis. Regular 2nd order tetra can be activated through HGHOR=REGULAR on the EXPLICIT continuation line on PSOLID.
Pin Flag support for Beam and Bar elements
Pin Flag (releasing the dofs) is supported for CBEAM and CBAR elements for explicit analysis.
Force output for CBUSH
CBUSH force output is now available in h3d file for explicit analysis.
MAT2 and MAT8 support
MAT2 and MAT8 material properties are now supported for explicit analysis.
Composite Support
PCOMP(G), PCOMPP/STACK are now supported for Explicit Dynamic Analysis.
Aeroelasticity:
Static Aeroelastic Divergence Analysis
Divergence can occur when deflection of lifting surfaces of an aircraft leads to additional lift, which in turn leads to further deflection in the same direction. A Divergence analysis determines divergence dynamic pressures using a direct complex eigenvalue analysis. The lowest eigenvalue correlates with the critical divergence dynamic pressure.
Aeroelastic Divergence Analysis Input
Divergence analysis determines the divergence dynamic pressures which are the eigenvalues from a complex eigenvalue analysis. The analysis is activated by a DIVERG Subcase Entry pointing to a corresponding DIVERG Bulk Data Entry. The DIVERG Bulk Data Entry contains information regarding the number of eigenvalues to be extracted and the Mach numbers for which these eigenvalues are to be extracted. A CMETHOD case-control entry referencing a EIGC Bulk Data Entry should be specified to activate complex eigenvalue extraction.
Heat Transfer:
DM support for Nonlinear Steady-State and Nonlinear Transient
Domain Decomposition Method (DDM) is supported for parallelization of Nonlinear Steady-State and Nonlinear Transient Heat Transfer analysis.
MUMPS as default for Nonlinear Steady-State
MUMPS is now the default solver for Nonlinear Steady-State Heat Transfer analysis.
User Material for Nonlinear Transient Thermal Analysis
The MATUSHT Bulk Data Entry, in combination with the LOADLIB I/O Option Entry, allows for the definition of thermal material through user-defined external functions. The external functions may be written in Fortran or C. MATUSHT is currently supported only for Nonlinear Transient Thermal analysis.
Fatigue:
Surface Damage
Since fatigue is a surface phenomenon, it is a common practice to assess damage only at the surface of a structure. Damage is assessed on the surface of a structure modeled with solid elements. Two options are available to assess surface damage of a structure (surface damage using the membrane stress or grid point stress). The surface damage calculation is automatically turned on when multiaxial fatigue analysis is carried out.
Surface damage using membrane stress
When requested in FATPARM, OptiStruct automatically creates membrane elements on the surface of the structure to assess surface damage.
The membrane fatigue method is available in fatigue analysis with linear static analysis, linear transient analysis, random response analysis, and frequency response analysis. It is supported for SN and EN Fatigue.
Surface damage using grid point stress
When requested in FATPARM, OptiStruct uses grid point stress to calculate damage on the surface of the structure.
Grid-point stress fatigue method is available in fatigue analysis with linear static analysis and nonlinear static analysis. It is supported for SN, EN, and FOS. Pseudo damage is not supported with grid point stress-based damage. Grid-point stress-based Fatigue is not supported for Weld Fatigue, Solder Fatigue, Vibration/Dynamic fatigue, Transient Fatigue, or Pseudo Damage analysis.
Input
The SURFSTS field can be set to MBRN (membrane stress) or GP (grid point stress) for surface stress after sub-keyword STRESS in FATPARM Bulk Data Entry. Membrane stress is calculated in multiaxial fatigue analysis by default, unless grid point stress is chosen.
Output
No additional output request is required. Damage/Life/FOS output of an element set will automatically output surface damage. If membrane stress is chosen, the worst damage caused by the membrane stress will assigned to the original solid element. If grid point stress is chosen, damage of surface nodes will be output.
Stress Gradient Effect
Stress gradient effect can be taken into consideration through either FKM guideline method or Critical Distance method. It is supported for both shells and solid elements. For solid elements, the stress gradient effect is only available with grid point stress in fatigue analysis using results of static analysis. For solid elements, SURFSTS field on FATPARM is automatically set to GP when Stress Gradient effect is activated.
The Stress Gradient method is supported for Uniaxial and Multiaxial SN, EN and FOS Fatigue. It is not supported for Weld, Vibration, and Transient Fatigue analyses.
Nonlinear Analysis with EN
Small Displacement Nonlinear Static Analysis results can be used to assess the fatigue characteristics for both SN and EN fatigue. PARAM,NLFAT,YES should be turned on for this situation.
When EN fatigue analysis is carried out with nonlinear analysis results, cyclic strength coefficient (K’) and cyclic strain hardening exponent (n’) are not required. No plasticity correction such as Neuber correction, Hoffmann-Seeger correction, and Jiang-Sehitoglu plasticity model is involved in damage calculation as stresses and strains are already elasto-plastic stresses and strains.
Composite:
Max Stress Criteria, STRS, for Composite Shells
New composite failure type, Max Stress, is now available for Composite shells. This is activated by setting the CRITERIA field to STRS on the MATF Bulk Data Entry.
Support of Direct Coupling input for Tsai-Wu Failure Criterion
For TSAI3D and TSAI.
Transverse Shear Stress for PCOMPLS
Redistribution of transverse shear stress with PARAM,COMPSHST is now also supported for PCOMPLS.
Optimization:
Nodal thickness as design variables with DVCLRE1/2
In case the nodal thickness is defined on the shell element, the average nodal thickness of the element can be defined as design variables using DVCREL1/2. If the nodal thickness is not uniform within the element, the ratio of nodal thickness will be maintained during the optimization. Item code for DVCRLE1/2 is “T”.
Loading frequencies as arguments for DRESP2/DRESP3
Loading frequencies associated with DRESP1 response will be passed down to DRESP2/DRESP3 through DFREQ1, DFREQ1L, DFREQ1V, and DFREQ1LV.
GRID ID input for GRIDCON (GRID-based Free-Shape)
Direct input of GRID IDs are supported for GRIDCON continuation line for Grid-based free-shape optimization (TYPE = VERTEXM on DSHAPE Bulk Data.
Enhanced LEVELSET
LEVELSET Topology Optimization has been enhanced to improve its robustness. LEVELSET Topology optimization can be activated by LEVELSET continuation line on DTPL Bulk Data Entry.
  • Minimum member size and Draw direction constraints are supported.
  • Maximum Member size control is currently not supported with Levelset Topology Optimization. If present, then the setting will be ignored and a message is printed in the .out file.
  • Pattern Grouping, Pattern Repetition, and Extrusion Constraints are not supported by Levelset Optimization. If present, the run will error out.
  • PARAM,TOPDISC,YES is not supported in conjunction with Levelset Topology Optimization. If present, then this parameter is ignored and a message is printed in the .out file.
  • Multi-Model Optimization (MMO) and Fail Safe Optimization (FSO) are not supported with Levelset Topology and the run will error out if present.
  • Draw Direction constraint is supported. SINGLE and/or SPLIT constraint types are only supported. All DTPL entries in the model should have or should not have the Draw direction constraint defined, if Levelset Topology is specified. If some DTPL entries in the model have draw direction and others do not, then the run will error out.
General
PART Superelement
PART Superelement allows the definition of each PART by its own partitioned bulk entries. PART can be defined, not only by a superelement but also by FE data. Each PART is self-contained and consists of grids, elements, properties, materials and loading specific to that PART. Supported Superelement format for PART superelement is op4 and punch.
  • Each PART can have its own ids for grids, elements and properties and so on.
  • A PART can consist of Superelements (op4, punch) or FE data (Grids, elements, properties, materials).
  • A PART can be moved and connected to the residual part.
  • A PART can be repeated (REPEAT).
New Bulk Data and I/O Options for PART superelements.
SEBULK
Definition of the type of superelement (TYPE field) and the superelement boundary search options.
SECONCT
Definition of GRIDs/SPOINTs that are used to connect the PART.
SELOC
Defines a partitioned superelement relocation by listing three non-collinear points in the superelement and three corresponding points not belonging to the superelement.
BEGIN SUPER
Definition of PART (Superelement or FE data).
ASSIGN,INPUTT4
Assignment of .op4 for the analysis with PART superelement.
Cyclic Symmetry
Cyclic symmetry is available for Linear static Analysis and Normal mode analysis. New Bulk Data Entries for cyclic symmetry.
CYJOIN
Defines grids on the segment boundaries.
CYAX
This entry is used to list the grids that lie on the axis of symmetry.
CYSYM
Specifies the number of segments.
LOADCYN
This entry is used to define the loading.
LOADCYH
This entry is used to define the harmonic coefficients of loading New Case Control entries for cyclic symmetry.
HARMONIC
This option is used to specify the solution harmonics to be used.
NOUTPUT
This entry is used to specify the segments for which results must be recovered and output.
Skip Case Control Section entries
SKIPON, SKIPOFF can be used to skip entries defined in case control section. SKIPON turns on the skipping of lines. SKIPOFF turns off the skipping of lines. If SKIPON is defined but there is no SKIPOFF, the all the entries after SKIPON and until BEING BULK will be skipped.
SKIPON, SKIPOFF are only relevant for case control section of data. They will be ignored, if defined in BULK entry.
Monitor Point Output - MONPNTi
MONPNT1, MONPNT2 and MONPNT3 are supported for Static analysis, as well as Aeroelasticity.
MONPNT1
Defines an integrated load monitor point at a point (x,y,z) in a user defined coordinate system. The integrated loads about this point over the associated nodes will be computed and printed.
MONPNT2
Element output monitor. Stress, Strain and Forces are supported. CBAR, CBEAM, CELAS1,CONROD, CBUSH, CWELD, CQUAD4, CSHEAR, CHEXA, and CTAXI elements are supported.
MONPNT3
Sum of grid point forces with respect to a user-defined point for a user-defined section.
Skip Degrees-of-freedom from AUTOSPC SPCOFF/SPCOFF1
SPCOFF/SPCOFF1 defines degrees of freedom which will be skipped from AUTOSPC.
HDF5 Enhancements:
CSHEAR force output
CSHEAR element force output is now available.
Frequency Response and Random response
Complex displacement and SPCFORCE, PFPANEL results are supported for Frequency Response. PSD/RMS Disp/Velo/Accel/SPCF and RCROSS results are supported for Random Response.
Graphical User Interface (GUI):
Altair Compute Console (ACC) GUI to submit jobs
With this release, a new utility is included called Altair Compute Console (ACC). It replaces individual menu entries for a group of Altair solvers (including OptiStruct, Radioss, MotionSolve, AcuSolve, HyperXtrude and several more). It is the easiest way to launch a solver on a local host or submit simple job to a remote Linux server/cluster or PBS system. It includes an interactive GUI for selecting input files, defining run options, submit multiple solver runs using a queue, schedule a delay, monitor solution progress, kill/pause a job, and provides easy way to execute Fluid-Structure Interaction (FSI) solution sequence for AcuSolve with OptiStruct and MotionSolve.

Resolved Issues

  • Applied Power calculation for Thermal convection is corrected.
  • MPC force output with LGDISP is inaccurate in text file output such as .pch or .mpcf.
  • Some Nonlinear Analysis restart job no longer fails with programing error when temperature loading is present.
  • Interlaminate Shear Stress for continuum shells (PCOMPLS) is correct.
  • GROUP option in DSIZE for MMO job no longer fails with programing error.
  • Stress results for normal modes analysis is correct when JOINTG is present in input file.
  • Sliding distance no longer is output as zero for continuous sliding (CONSLI).
  • .dens file no longer is created for topology optimization (even without user’s request).
  • Pretension force output in .pret file is corrected for DDM runs.
  • For asynchronous rotordynamics model with modal complex eigenvalue analysis, a programming error does not occur when there is more than one rotor speed and the number of desired complex eigenvalues (ND0 in the EIGC card) is left blank.
  • When there are multiple MATMDS Bulk Data Entries and dynamic analysis is performed (i.e. normal mode analysis), density was only used from the first MATMDS.
  • Reading issue no longer occurs when using op4 DMIG in residual run.
  • Slow performance of nonlinear transient calculation has been improved.
  • SPCF output is no longer incorrect, if CNTNLSUB points to the subcase ID.
  • The accuracy of Node-to-Surface (N2S) TIE/FREEZE for Linear analysis and the small displacement Nonlinear Analysis has been improved.
  • Advanced restart with overhang constraints or extraction constraints are enhanced to handle such use cases properly.
  • A programing error may occur when advanced restart with mode tracking is performed. This has been fixed.