Eurocode 8 displacement analysis

[PSR_1]

The displacement for both damage limitation verification and in order to classify the system as non-sway,sway or unstable as per eurocode 8, if linear elastic analysis is used it should be multiplied by behavior factor(q)in order to get the value ds(see attached picture). section 4.3.4 of Eurocode 8 states that this value ( ds=dr*q) "need not be larger than the value derived from elastic spectrum". my question is the value derived from elastic spectrum( or say if the value of q=1 for design spectrum)is very small and when used it used it usually yields "non-sway" for frame classification,which would be unstable if ds=q*dr is used. anyone with experience with eurocode please help me verify if using elastic response spectrum to calculate displacement( which would be compared with the result from design spectrum multiplied by q) is a safe and acceptable approach even if the code clearly states it.

 

[quasiblu]

At ultimate limit state, the design spectrum is divided by "q", while the elastic spectrum is not.
You should assess the displacements (for example to prevent structural pounding, which is an ULS thing) using the elastic spectrum, or multiplying the displacement from the design spectrum by q, which is the same thing.
Equivalent lateral force analysis and modal spectrum response analysis are both linear elastic analysis. Ductility is considered using the fictive coefficient "q".
I am not aware of prescriptions stating which method you should use to assess the displacements to check if it is sway or non-sway. Does it go with ultimate limit state or damage limitation state? Damage limitation state is a different thing, Eurocode checks the interstory drifts with that, and it should not have q (q=1) and should have lower spectrum values.
Anyway, if it is non-sway for elastic spectrum at ultimate limit state, then it should really be non-sway...
Hope this will help.

 

[PSR_1]

thanks quasiblu!