Overview

The FE input file (finite element input file) is processed in two ways during the optimization procedure. Initially, the optimization preprocessing step reads the FE input file and all necessary information is written into the optimization database. Afterwards, in each optimization iteration only the properties changing during the optimization will be modified in the FE input file.

The flow diagram below is identical for all FE solvers. The optimization preprocessor TOSCA_PREP executes the program fem2onf. This program converts the FE-input file into a neutral file format (ONF). The essential information regarding the FE-model is written in the neutral file. The first file called <jobname>_req.onf contains the result requests. These are the FE-results needed for the optimization module. The second file contains FE-model (<jobname>_fem.onf) information, that is, grid points, elements and etc. The last file contains the group definitions (<jobname>_grp.onf).

The optimization iterations begin when the preprocessor is finished. No previous optimization iterations exist for the first optimization iteration. However, FEM_MODIF modifies the FE-input files (model1.<input>, model2.<input>, ...) which ensures that requested results are present after FE-analysis.

The chosen FE solver calculates the requested results (model1.<output>, ...). The specific results required for the optimization are read by RES_2_VTM and the results are stored in the database (tosca_memory). Using these results the optimizer TOSCA_OPT generates a modification file (<jobname>_mod.onf). Afterwards, FEM_MODIF once again modifies the original FE-input file using the modification file. The standard configuration ensures that the FE-input files will be stored for each optimization iteration in the directory SAVE.<input>. The directory SAVE.<input> also contains the original input file having the name model_org.<input>. The first and last FE results files are stored in the directory SAVE.<output>.

Flow diagram for SIMULIA Tosca Structure

"input" and "output" is the FE input file extension and result output extension, respectively. The extension names correspond to the extensions given by the FE solver.

Solver type

SIMULIA Tosca Structure supported input

SIMULIA Tosca Structure supported output

Abaqus

*.inp

*.fil or *.sim (see below)

*.odb

ANSYS®

*.cdb

*.rst

*.dat

*.inp

MSC Nastran®

*.bdf

*.op2

(*.dat)

Some features in SIMULIA Tosca Structure are not available for all solver interfaces.

Solver

Sensitivity-based Optimizations with contact

Use of internal Forces

Frequency Response Optimizations

Abaqus

X

X

X

ANSYS®

X

X

X

MSC Nastran®

X

X