User login

Navigation

You are here

A suitable format for transfer between various FEA-packages?

pragtic's picture

Dear iMechanicians,

While developing my own fatigue FEA-post-processor called PragTic (www.pragtic.com), I have got to a crossroad. PragTic have used quite sophisticated method for import of FEA data from formatted ASCII files, which allowed import of files from ABAQUS, ANSYS or COSMOS, with not that much problems. But, well it is definitely slower and e.g. for ANSYS more cumbersome. Also, it prevents adequate work with results on shells, because the information there gets too complicated (3D) due to existence of results on bottom and top sides of the element.

In fact, the only other solution in PragTic is a complete import of MSC.Nastran files (*.dat for topology, *.pch for results).

Well, my friends, who want to use PragTic in a simpler manner than with that general ASCII import, offered me to prepare macros in Ansys' APDL and Abaqus' Python, which would convert the data in the current model to files with the structure, which I should propose.

I want it to be an ASCII file, because it allows an immediate check. I do not think, it is reasonable to made them to convert it to MSC.Nastran native files, which PragTic already knows. I have some knowledge about FEMAP neutral file, but the information there is quite hidden in the specific file format. It should be noted, that PragTic makes only fatigue post-processing and
do not work very much with element properties (only shell thicknesses
are used now).

So, I want to ask you, if you could recommend me some suitable recognized transfer format.

Thanks a lot for any advice.

Regards,

Jan

Comments

The VTK formats could be a good choice.
You'd only need to support a subset of it of course, but the whole specs are here;
http://www.vtk.org/VTK/help/documentation.html

The main selling point to choose this, is that Paraview is increasing in popularity, and if people convert their output to a VTK format they can work with their results in many other applications as well.

I'd use the XML version and let users supply a celldata for thickness, and just two sets (top and bottom) when it comes to the stresses and strains for shells. Even if the file contains more it can easily be ignored.

Subscribe to Comments for "A suitable format for transfer between various FEA-packages?"

Recent comments

More comments

Syndicate

Subscribe to Syndicate