RSA displacement

[Aslinn]

Hi,
I'm trying to run a RAS analysis on an unreinforced masonry building, but the result I got was very strange. I tried several spectra on two models, and both gave me extremely small displacements that are less than 3 mm.
What I did is the following:
- Under RAS parameters tab, Add standard to the Loading combinations;
- Under RAS parameters tab, Add a spectrum as Given accelerograms, one example I used was ChiChi.dat;
- After I ran the analysis, I reviewed structural displacement under Global Response Parameters tab for all the 4 combination outputs, both x-axis and y-axis, and all the nodes.
Doing this, the maximum displacement I got so far is 2.5mm, which is obviously wrong since it does not match the results from pushover and time-history.
I must have missed something important. Could you please help me with what I did wrong with RSA? Really appreciate that.
Best Regards,
Aslinn

[huffte]

Aslinn, are you getting displacements from the "Deformed Shape Viewer" or from the "Global Response Parameters" table? At least on my computer, on which I am running SeismoStruct 2020 Release 1 Build 1, the numbers on the plot are off by the factor specified in the response spectrum definition compared to the numbers in the table, which appear to be correct. I may need to adjust some setting to get the plot correct, but I am not sure. Hope this helps. Best of luck in your work.

[seismosoft]

Note that the response in RSA is linear elastic, contrary to the response in nonlinear dynamic analysis and pushover analysis. This could also explain the discrepancies
Seismosft Support

[Aslinn]

Aslinn, are you getting displacements from the "Deformed Shape Viewer" or from the "Global Response Parameters" table? At least on my computer, on which I am running SeismoStruct 2020 Release 1 Build 1, the numbers on the plot are off by the factor specified in the response spectrum definition compared to the numbers in the table, which appear to be correct. I may need to adjust some setting to get the plot correct, but I am not sure. Hope this helps. Best of luck in your work.

Hi,
Thank you for your reply. I'm using SeismoStruct 2020 Release 1 as well. It seems to me that the numbers under 'Deformed Shape Viewer' and 'Global Response Parameters' are the same, but after I multiply them by the factor, 9.81, the numbers are more reasonable. Is this what we should expect?
Best Regards,
Aslinn

[seismosoft]

Probably this is a coincidence. A more reasonable estimate is the (nonlinear displ)=(q-factor)*(linear displ), however this is also a big approximation.
You should not mix and compare linear and nonlinear deformations directly.
Seismosoft Support
PS: This discussion is beyond the scope of this support forum. You should better make this discussion with your supervisor.