A customer reads .sub files to access reduced matrices, load vectors and DOF definition. When reading .sub files generated by ANSYS 8.1, they get incorrect data. Was the .sub file structure changed? Can you provide details on the new .sub file structure ?


I don't think there have been any changes to the .sub file that are not documented. Documentation for the .sub file is contained in the FDSUB.INC file that can be found in the customize/include directory in the ANSYS installation directory.

The changes that I see in the format are as follows:

1. The file now uses 64 bit pointers for the records. For the change to 64 bit pointers, if your substructure files are less than 2 gigabytes, use of the 32 bit reader routines should continue to work. The 32 bit logic uses the lower part of the address, which is that same as before. Notice that for most of the pointers, only the lower part is stored in the header record. Where the high bits may be important, such as for the load vectors, the high bits are also stored. The low bits and the high bits may be combined using the Longloc=LargeIntGet(lowerbits,higherbits) function. Details for using the 64 bit addressing may be found in the Release 8.1 Guide to Interfacing with ANSYS.

2. Length of the sub file header record (was 50, now 60). New data was required for the component mode synthesis method currently under development. For the change in the length of the sub file header record (the second record on the file), all of the previous data items remain the same and will likely be sufficient for your use.

3. The addition of a second transformation record for substructures that were created by transformation. This second transformation record, the fourth record on the file, is 250 double precision items long. This is likely the source of your problems.

4. The degree of freedom record now includes numbers up to 64. DOF(33) through DOF(64) are the temperature degrees of freedom for composite thermal shells.





Show Form
No comments yet. Be the first to add a comment!