Seismic Analysis (Static ELF) - Which period to use? 3

[fracture_point]

Hi all,

Firstly, let me state that I am going to be undertaking both a static analysis and a dynamic analysis (response spectrum) as I want to compare forces and scale them if necessary.

For the ELF in the x-direction, the first mode has a period T = 2.4s, and excited about 16% of the seismic weight. The 6th mode has a period T = 0.4s and excites approx 20% of the seismic weight. There is discussion as to which period to use for design, as the 6th mode period develops a significantly higher base shear (3 times the amount)

I am of the opinion that if 20% of the mass is excited in the x-direction (the most of all modes), then this is the predominant mode for the x-direction and should be used for design. However another argument has been presented to me that the mode with the highest period and 16% mass participation (second highest participation factor) is the predominant mode and should be used, which develops a significantly lower base shear than the T=0.4s.

Hoping to hear yoru thoughts.

 

[WARose]

Since you will be doing a dynamic analysis as well....it's not something to get that excited about. (No pun intended.) But when I do the ELF procedure, I just use the procedure they give for hand calculations for figuring it. (See Sect. 12.8.2.1 in ASCE 7-10 for example.)

If you just had to use the computer output, the first mode for lateral excitation is typically what is meant by "fundamental".

 

[KootK]

If your 6th mode period develops 3X the base shear of your 1st mode mode period, I'm of the opinion that ELF is probably not a meaningful tool for evaluating this structure.

 

[OUe]

Any time I designed a structure with the dynamic analyis method, I obtained around 70-90 percent of the base shear in the first or second mode. Also keep in mind that FEA software like RISA, SAP2000, etc... will not behave well with eccentric masses. I would recommend forcing a joint moment if you have large cantilevers because it will alter your mass participation and make it inaccurate.

 

[JoshPlumSE]

Start with section 12.8.2 of ASCE-7 (I'm looking at the 2016 version).

12.8.2.1 (Approximate Fundamental Period).... This section is meant to be used with the ELF method and give you some limits on the period to use. This will also allow you to calculate an upper limit on the period you can use for ELF calculation.

Then look at the period (from your FEM analysis) with the highest mass participation in a given direction. That's probably your "fundamental" period. If multiple modes have similar mass participation (i.e. 3 mode all have about 20%) then use some engineering judgment to choose between those particular modes. I'd personally choose the one with the lowest period.... Because this is conservative for the ELF method.

Then you compare your FEM period to your upper limit period from 12.8.2.1. If your upper limit period is lower then the FEM, then you use the upper limit.


Note: I will point out that the period used for ELF can still be important because it can becomme the basis for scaling the response spectra results.

 

[cal91]

I'm curious what the building looks like if it has more mass participation in its 6th mode than in its 1st mode?