Etabs Equivalent Lateral Force vs Response Spectrum analysis Base Shear 1

[WayneRuin10]

Hello Fellow Engineers.

I had a simple question. I performed Seismic analysis on a 30 story building using 2 methods mentioned in ASCE 7-10 using ETABS; Equivalent Lateral Force (ELF) and Response Spectrum Analysis (RSA). I was expecting the base shear from RSA to be lower than ELF base shear, however, RSA base shear was 15 percent higher than that of ELF base shear.

Am I doing something wrong? Or is it possible for this to happen? Any thought/comments is appreciated! thanks.

 

[rscassar]

It's not impossible, the greater your stiffness than generally the higher the base shear from RSA. Check your primary modes, for a 30-storey building you should be around 3 to 4 seconds for both primary modes in x and y-translational directions. The lower the periods than generally higher the base shear for RSA.

Other than that, check the way your masses are applied. If you have selected self-weight and load-patterns and one of those load patterns includes the self-weight than that self-mass will be applied twice.

When you calculate the ELF base shear are you allowing for 30% in the orthogonal direction?

 

[WayneRuin10]

Thanks for your response! I made sure that self load isn’t counted twice. My concrete structure is only moment frame and has no shear walls that’s why I am struggling to bring the period down to 3-4 seconds. At the moment the period is approx. 5.5 seconds.

I wasn’t aware that I had to allow 30 percent in orthogonal direction...Do I have to do this even if my building is in SDC B and has Cs value of 0.13

Thank you again!

 

[Trenno]

At 5.5 seconds, you're a fair bit off the rule of thumb: period = storeys/10.

Also, a longer period should yield a lower base shear with RSA, as your building is more flexible and therefore should pick up less seismic load.

Sounds like you need deeper beams and wider columns. Don't forget, "strong column, weak beam."

 

[JoshPlumSE]

Also, a longer period should yield a lower base shear with RSA, as your building is more flexible and therefore should pick up less seismic load.

Not sure exactly what the justification of this comment is. For ASCE / IBC spectra, I think of base shear for long period buildings this way:

1) ELF and RSA for a building with a single mode should be basically the same regardless of the period. Why?... because they come from the same spot on the spectral equations (which are based on period).
2) Let's say you have a structure with a 4 second period with 70% mass participation and another mode with a 0.5
second period with 30 % mass participation. The acceleration for the 4 second period is going to be based on Sd1*I / (RT).
But, the acceleration for the 0.5 second period will probably be based on Sds*I/R.
a) The 4 second equation tends to impart a LOWER acceleration. Let's say 0.2g
b) The 0.5 second equation will tend to be a higher acceleration. Let's say 0.4g
c) Therefore, the 30% of the mass associated with 0.5 second period will tend to INCREASE the total base shear when compared to the ELF method. This would be an example where the base shear would be about 15% higher for a RSA than the same structure analyzed per ELF.

 

[Trenno]

Sorry to clarify, I was more thinking of the comparison between code based period and analysis based period. ESA typically uses code based equations to estimate the building's fundamental period and most of the time these are conservative and thus lower than analysis based periods. Not always though, so you need to check what yields the lowest period.

 

[JoshPlumSE]

Trenno -

I understand your comment now. I just wasn't thinking of the cases where the code may not allow you to use your more accurate period for an ESA analysis. Most of the time, I find that the code (at least ASCE) allows you to use the theoretically derived period for your static analysis.