For each finite element model there will be a FEMGEN database file and/or a FEMVIEW database file with a FEMVIEW index file.

Finite element data is generated in FEMGEN and may be output in a variety of formats compatible with many of the major finite element analysis packages. The transformation to various output formats is performed by a series of in-built interface subroutines.

After analysis, the finite element results are usually presented in a text file, and stored in a results file. A post-interface program is then used to convert the results file into a character file compatible with FEMVIEW; the data is then transferred to the FEMVIEW database using the command 'UTILITY READ VIEWDATA'. Post-interface programs exist as separate executables, and many of the major finite element analysis packages are supported. The structure of the character file for input to FEMGV is illustrated in figure 1 overleaf.

The character file may contain more than one model. The first model header record doubles as the header record for the whole file. The file is delimited by a special record used only for this purpose.

Each model is preceded by a model header record. Each model is delimited by the next model header record, or the file delimiter record. Within each model, a number of data sets will be present (e.g. nodal co-ordinates or results).

Each data set is preceded by a data set header and delimited by a data set delimiter. Additional headers may follow the data set header for some types of data set where additional information is required (e.g. User-Defined Results).

Figure 1