Response Spectra Analysis

My first response will be related to damping:
This only affects the CQC and Gupta methods of modal combinations. These methods account for the interaction effects of closely spaced modes. The program needs to know the assumed damping ration in order to calculate these interact effects.

Regarding the Use Dominant Mode for Signage:
One of the problems with a Response Spectra Analysis is that the relative signs of the member forces, joint deflections, joint reactions and such are lost. This is related to the statistical methods of combining the modal results together.

The loss of signs for your results is okay when you're looking at a single result. The axial force in a member from the RSA may be 10 kips compression or 10 kips tension. Either one is perfectly valid. But, when you try to look at something that relies on the RELATIVE sign of multiple results (like when you plot your deflected shape) then this can be a real problem.

The "Use Dominant mode for signage" option is a way to give relative sign results to your RSA results. If you've got one mode which has 85% mass participation then it should be reasonable to use the relative signs for that mode for your overall RSA results.

I'm in the process of putting together a RISANews / blog posting about this subject.... this has some pictures and such which might help. They show how the loss of signage affect deflected shapes and such.