Tension rods in compression - Response spectrum analysis 1

[cp2013]

Hello,

I am working on the seismic design of a dome braced with tension rods using a response spectrum analysis according to ASCE 7-10.
I know that response spectrum analysis cannot handle non linear elements and that my tension rods can enter in compression.

My problem is that I have some combinations for which all the tension rods are in compression and other combinations for which all the tension rods are in tension. Is it realistic ? Does it mean that there is a problem from the analysis ? Or does it mean that actually my tension rods should be totally dummied for the design under the combinations where they are all in compression (is there a way to avoid that ??).

Thanks.

 

[glass99]

Non-linearity in linear analysis including response spectrum is always difficult. If you have symmetrical X bracing in a frame, you might justify that the global behavior is as if you had only a single diagonal which worked in both tension and compression.

Alternatively, if the tension rods are pretensioned and never go into compression, they are linear!

 

[Agent666]

I have setup a nonlinear case first which buckles the braces using an equivalent static load distribution and then used the results of this case to do a modal or response spectrum analysis so you have the appropriate braces in tension being the only ones considered for a certain direction of loading. Setup cases for both directions as required.