Load combination for Seismic design (UBC 1997) 1

[engrmarks]

Hello engineers

What load combination should I use on seismic drift check? Should I use Ultimate or service level?
My college professor also told me that I can use pure earthquake load 1.0E (derived from base shear force). Is it okay to use this?

Thank you all!

 

[haynewp]

I’ve used the worst case combination in the past but looked in ASCE 7 and couldn’t find the answer. A 1.0E case isn’t realistic since the p-delta from gravity loads would also affect the drift, and therefore the stability which is what limiting EQ drift is about. So I will continue to envelope all LRFD combinations with EQ until I find a reason not to. And it’s ultimate and not service to answer that part of the question.

 

[engrmarks]

I forgot to mention that he (my college professor) only use pure earthquake load 1.0E on low rise structure two-/three storey buildings.

 

[JAE]

Look at your local code but if in the US - check out ASCE 7.
There - the seismic drift (for static type analyses) is defined as the elastic drift from 1.0E multiplied by the Cd factor, divided by the importance factor.

This Cd essentially removes the "R" reduction factor from the demand. Also the Ie importance factor is removed as it is an added level of safety demanded by the occupancy.
This gets you back to the true lateral drift that a building is estimated to experience in its design event.

In other words, there is really no "service level drift".

 

[haynewp]

I don’t understand comparing the drift to a case that doesn’t include the realistic gravity times p-delta which can be significant.

 

[haynewp]

I still don’t know why it’s not any of the realistic cases that can occur during an earthquake. The whole point is to maintain structural stability.

“The gravity load and seismic load combination used in drift analysis is stated in Load Combination 6, Section 2.3.6 of ASCE 7-16: (D+Ev+Eh+0.5L+0.2S).”

https://www.structuremag.org/?p=14696

 

[engrmarks]

I still don’t know why it’s not any of the realistic cases that can occur during an earthquake. The whole point is to maintain structural stability.

“The gravity load and seismic load combination used in drift analysis is stated in Load Combination 6, Section 2.3.6 of ASCE 7-16: (D+Ev+Eh+0.5L+0.2S).”

https://www.structuremag.org/?p=14696

That's a great article. Thanks for that! About the pure 1.0E alone, maybe that was based on his engineering judgement? As for myself, I'm fairly new to the industry.

 

[engrmarks]

Look at your local code but if in the US - check out ASCE 7.
There - the seismic drift (for static type analyses) is defined as the elastic drift from 1.0E multiplied by the Cd factor, divided by the importance factor.

This Cd essentially removes the "R" reduction factor from the demand. Also the Ie importance factor is removed as it is an added level of safety demanded by the occupancy.
This gets you back to the true lateral drift that a building is estimated to experience in its design event.

In other words, there is really no "service level drift".

My country adopts UBC '97 for the most part but also uses ASCE-7. What is generally the difference between UBC '97 and ASCE-7 in terms of seismic analysis?

Understood that seismic analysis is at strength level, I always have a hard time understanding that part.

Thank you.

 

[wannabeSE]

If you are using the 1997 UBC, look at section 1630.9. It is a bad idea to mix the 97 UBC requirements with ASCE 7. They have similar ideology with different limits and different formulas. 97 UBC uses Δ[sub]m[/sub] = 0.7 R Δ[sub]s[/sub] and the UBC uses strength (ultimate) force levels to calculate the elastic drift, Δ[sub]s[/sub] .

 

[haynewp]

Sorry, I don't know about UBC 97. But the AISC 14th examples manual refers to ASCE 7-10 section 12.4.2.3 load combinations when calculating drift. The equation I posted above is similarly derived from ASCE 7-16, but the other equation (0.9D -Ev + Eh)from that section appears to also need to be checked drift.

So the combinations that include "seismic load effects" when using 7-10 or 7-16 appear to be the ones required when checking drift, but not the set of seismic load effects equations that include omega since drift does not include omega. Then multiply the drift times Cd.

 

[JAE]

Ya I may have confused things by referencing ASCE 7 when the original post clearly indicated the UBC 96 as the code in question.

The general concept of deriving seismic drift in the UNC and ASCE7 are generally the same approach.

So I’d second what wannabeSE states above as well as haynewp.

And I’d double that taking 1.0E alone isn’t correct. You still look at load combinations with dead included to derive drift values.