Modal Response Spectrum Analysis

[sticksandtriangles]

Hello,

I've got a question regarding 12.9.4.1 of ASCE 7-10:

For example lets, lets say i've got my building whose first 2 modes are translational x and y modes that are in the range of 2 seconds. I go ahead an calculate my capped max period for ELFP, Ta = 0.8s per 12.8.7.

Then reading 12.9.4.1, with CuTa = 0.8*(1.6) = 1.28s, it states "where the calculated fundamental period exceeds CuTa in a given direction, CuTa shall be used in liue of T in that direction". Does this mean that I have to fake my response spectrum curve in my analysis to be "capped" at the acceleration that would occur at a period of 1.28s? see below for the curve I would anticipate to input "faked" curve (replace Cs with a%g) forcing a higher MRSA acceleration at longer periods.

This is how I read this code snip, but examples from FEMA do not seem to agree with this conclusion.

https://cdn.ymaws.com/

http://www.nibs.org/resource/resmgr/BSSC/p751_ch4.pdf

S&T

 

[JoshPlumSE]

No.

Per 12.9.1.4.1 (Scaling of forces) the CuTa is used in the procedures of 12.8 (equivalent lateral force method). It doesn't affect Vt (the modal base shears), rather it affects the V value (ELF base shears). Then you use V/Vt to scale up the modal results where necessary.

 

[JoshPlumSE]

Stated another way, this provision is a way to scale the response spectra results so that the base shear cannot be less than you'd get from the equivalent lateral force method.

 

[sticksandtriangles]

Josh,

What you said was my old way of thinking as well.
It seems odd to me to even include the line of:

where the calculated fundamental period exceeds CuTa in a given direction, CuTa shall be used in liue of T in that direction

Why not just leave it at saying scale the forces to 85% of ELFP?

S&T

 

[JoshPlumSE]

I thought the code provisions were simpler and more rational in the 1997 UBC. Where irregular structures couldn't get a response spectra base shear less than 100% the ELF. Regular structures could get down to 90% of ELF. Regular structures with a site specific spectra could get down to 80%.

Unfortunately, the code / dynamics gurus over the last decade or two have determined that the collapse probability of structures designed via Response Spectra is higher than for structures designed via ELF. And, they have taken action to trim back some of the advantages of response spectra.

Another way to think about it is the following:
Back in the 1970's response spectra was considered state of the art. A cutting edge analysis tool for seismic design. But, that's not the case anymore. It's now considered pretty basic analysis. The more "state of the art" / cutting edge analysis techniques are the ones that predict non-linear behavior.... static pushover analysis or non-linear time history. Therefore, these newer analysis techniques get some real benefits, but the older / elastic techniques do not.