Deviation of results from version 5 to 6

[fakharifar.mostafa]

Hi,

First, sorry for this long comment.

I was going through some of my analytical frame models that I had implemented in Seismostruct about two to three years ago using Displacement Adaptive Push Over (DAP) analyses.

Although I had saved the result files from the past, however, software required to do new analyses in version 6 and didn't bring up the results from past analyses.

Surprisingly, results for the capacity curve of the studied frame was different in the inelastic range from the one given in version 5. While ver. 5 showed a such a yielding plateau behavior, in contrary version 6 exhibited instability in strength with progressive drop in strength.

I didn't change anything between the two models (material models, elements, geometry and limit state definitions), and don't think if I had changed any of the program defaults in ver. 5.

Is such observation expected? Much thanks.

[seismosoft]

No, this is not reasonable.
Can you please send us the spf file for the version 5 at support@seismosoft.com, so that we check things further?

SeismoSoft Support

[fakharifar.mostafa]

I'll email the spf file. However, I need to mention I have opened my old file in the current version 6.5.

[seismosoft]

Thanks Mostafa,

Unfortunately, without the original file there is not way to check the possible differences between v5 and v6.5 results (older SStruct releases do not open newer spf files).

Note however that one possible source of instability in the v6.5 version of the file is the fact that you use very short FB elements. 4 or 5 elements per members is a typical discretisation for the DB elements, with FB elements one element per member is enough.
Further, try to reduce your load step by increasing the number of step.

SeismoSoft Support