time history analysis: warning about file size

03-Analytical/modelling capabilities
Post Reply
hania.kkari
Posts: 3
Joined: 28 Jan 2012, 03:02

time history analysis: warning about file size

Post by hania.kkari »

Dear support/huffte

I'm running the dynamic THA for a braced frame about 15 floors, I m getting a warning at the start that the file of the results will be estimated to e more than 4 gb, and that seismostruct does not support output files more than 2 gb, but I m still can see the graphical results and getting what I need. My question is : is this message only for if I want to open the output file? Or what I'm getting as results will not be correct as I crossed the size limits.?

Thanks a lot
huffte
Posts: 1005
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: time history analysis: warning about file size

Post by huffte »

Well, my guess is that the results you do get are still fine. However, it would seem to be a wise thing to trim down the size, both for the sake of ensuring that the results are correct, that no unknown bug exists when excessive file sizes are attempted, and to make your time spent in post-processing the results more productive.

Are you outputting stress/strain peaks and curvatures? If so, can you eliminate this part of the output in the options, or at least, select only those critical elements?

Are you carrying out stress recovery in the options? If so, could this be eliminated?

In the Analysis Output dialogue, what is your frequency of output? Could you adjust it to decrease the output? Are you requesting results for all nodes and elements? Could you identify the ones of utmost import and thus reduce the file size?

In my experience, it is a wise choice to narrow down your results as much as possible without adversely affecting your conclusions, and to simplify models as much as possible, again without sacrificing an accurate solution to the physical problem.

Here is a great opportunity to be an engineer, to use your logic and training to identify 'what to leave in and what to leave out' as Bob Seger once said.

Best of luck, hania.kkari. I'll be interested in any worthwhile findings on this issue you might be able to report in the forum.
Tim Huff
Post Reply

Return to “03-Analytical/modelling capabilities”