Elastic Seismic Design

[3doorsdwn]

Interesting question: if someone actually wanted an elastic seismic design (i.e. where there wouldn't be any damage in a seismic event).....how would you go about it? The first thing that comes to mind (if you are using the ELF method of load calculation) is to NOT divide your load by R (i.e. the response modification factor).

Another interesting question is: if you are doing an elastic design, would the building system height limitations of Chapter 12 in ASCE 7 still apply? After all, they are built around controlling damage (and locations of said damage).

This thing came up in my mind because a client asked about the feasibility of a damage free seismic event. (Obviously cost prohibitive.) And it got me to thinking that if someone actually asked for it: legally I'm not sure how it should be done.

 

[InDepth]

You hit the nail on the head....R = economics. I think everything should be design close to an R of 1 in order to help communities recover from disasters and minimize disaster waste (sustainability).

First off...."damage free seismic event" is like saying concrete doesn't crack. You really got to define what your client means by a damage free seismic event....does it mean 1/16" wide cracks or hairline cracks...or even equipment damage inside the building.

Try looking at some of Englekirk's work on precast moment frame systems.

Also take a closer look at the ELF method assumptions with regards to distributed yielding (Read NEHRP)....I don't think it applies if you have an elastic system.

If I remember correctly if your system is to remain "elastic" you would have to use modal analysis with an unreduced design spectrum. You could even modify the spectrum for damping provided by soil-structure interaction at the foundation. Furthermore, you could run a few time histories.

Also remember that your floor acceleration levels for your equipment and non-structural items will be much higher and could damage them. The contents of your building may be worth more than the building itself.

Also think base-isolation....look at the work of Forell-Elsesser of San Francisco.

 

[JStephen]

One thing to consider is that the loads are probabilistic, so even if you have perfect data and perfect designs, you're assuming a certain probability of overload occurring. So you could design your "no damage" structure, do everything right, then see it damaged in an earthquake the next year.

A second thing to consider on the economic end of it is that it makes a big difference where you are. In an area where seismic loads are pretty low anyway, it might not be that big a deal to do your design, versus, say, California.

(A couple of years ago at a seminar put on by FM, the speaker mentioned that the way the seismic factors in ASCE were calculated, the probability of a overload/failure wasn't necessarily uniform across the country like you might expect. Essentially, the factors are tweaked so that seismic design is less conservative in high seismic area. I forget the details, though.)

Lastly, consider that the success of designs has to be confirmed somwhat by experience. If there is a large body of experience showing that buildings built to certain standards don't fall down, that's taken to mean the method is essentially correct. On the other hand, if nobody bothers tabulating data on zero-damage buildings, you might be missing the historical background to confirm your analysis.

 

[irizarry]

Actually, I don't think codes will permit not having a height limit even with R=1. If you check IBC 2006 section 1613.6.2, it says that you can you can take the height limit for OMFs and OCBFs to 160ft in seismic categories D,E or F, provided the following conditions are met:
1. The value of R as defined in Chapter 17 is taken as 1.
2. For OMFs and OCBFs, design is in accordance
with AISC 341.

So, you can probably get a taller limit, but nonetheless there will be one.

Also, remember R=1 gets an elastic design for assumed earthquake spectra of code, but it doesn't necessarily mean thats the actual earthquake that will happen.

 

[3doorsdwn]

Excellent responses all.

irizarry, that's an excellent point you made. If I had remembered that part of the code, I could have answered part of my own question.

 

[sandman21]

1613.6.2 is for isolated systems using OMF or OCBF

 

[slickdeals]

I think the recent damages to buildings in earthquakes will provide more drive to Performance Based Design (PBD). I know recently a 57 story building was designed in San Fransisco using PBD.

http://www.structuremag.org/article.aspx?articleID=707

We are Virginia Tech
Go HOKIES