Enormous Result File 1

[huiying]

Hi,
I'm doing a transient analysis and random analysis on a 20,000 element model with Nastran as solver. The result files (.f06 and .xdb) I obtained are both very huge, in the area of about 10GB! I wonder if that is common.
I managed to get a result for transient after reducing time-step from 500 to 100, however, I'm still unable to get a soln for random because the results says there is some run-time error. Is there anything I could do to reduce the runtime?

Thanks & Regards.

 

[Spirit]

Hi huiying.
The first thing you should do is reduce the amount of requested output. This can be done both reducing the output in term of frequency steps required (FREQi cards) and the perameters you want to trace (you probably aren't interested in having the responses at ALL GRID points, or at ALL elements). As for the solution time, the only thing you could do (assuming you are using a modal and not a direct solution) is using restarts. As for the runtime error, I couldn't say anything without knowing further details.
Hope this helps.

Regards,



Spirit


'Ability is 10% inspiration and 90% perspiration.'

 

[huiying]

What is "restarts"?

 

[jxc]

What Spirit is referring to in place of over-writing your .f06 file you can restart it from where it stopped. Info. on how to restart is in any of the NASTRAN & Patran manuals or online help.

You do not need both the .f06 & .xbd files, you can stop one or both. In NASTRAN the .xbd file allows you to read your results without adding them to you FE model file, keeping that file size manageable.

I hope this helps, everything I explained again is in the manuals or online help.

jxc