Non-linear seismic response
Non-linear seismic response
(OP)
Was talking to a client today and he floated the idea (during very loose talk) that a economic advantage could be obtained by me taking a nonlinear approach with seismic design (i.e. something like Section 16.2 in ASCE 7-10). When it comes to structural, I take this guy fairly seriously (unlike a lot of industrial clients).
I personally haven't done any non-linear work in seismic........for anyone who has: have you saved any construction costs by doing it? Is it overly complicated? Is it really for cases where damping devices are used?
Thanks in advance.
I personally haven't done any non-linear work in seismic........for anyone who has: have you saved any construction costs by doing it? Is it overly complicated? Is it really for cases where damping devices are used?
Thanks in advance.






RE: Non-linear seismic response
To answer the three questions:
1. Have you saved any construction costs by doing it? - Our savings was more because we did full-blown performance-based design. We were able to use a special shear wall building in SDC D taller than 240 feet and didn't have to use moment frames, so that certainly saved a ton (and likely made the project happen). Would imagine there's savings just going with 16.2 as well, but have no experience with it.
2. Is it overly complicated? - It's a lot more complicated than normal design. Normally we only care about stiffness, load, and yield strength and stay in the elastic range for all of our elements, then do prescriptive detailing to ensure adequate ductility. Going nonlinear introduces a lot more variables as you can't ignore strain hardening, ultimate strength, strength loss, residual strength, etc, all of which vary based on how you've detailed your structure and can also vary based on level of load (e.g. columns with high axial load hit strength degradation for moment a lot earlier). Also since you're past the yield point, you're not looking at DCRs anymore for yielding elements. You'll need to actually look at member deformations and make sure you're clear enough from the strength degradation portion of the material/component curve or meet other prescribed acceptance criteria. Also since it's time history, you don't have one force and one deformation for each item anymore. You'll have a force and deformation for however many time steps you look at for each ground motion history. Amount of data adds up real quick, even for small structures, as does the required computing power.
3. Is it really for cases when damping devices are used? - No. Our case didn't have a damping device and would think most cases actually don't. Just because you have a damping device doesn't necessarily mean you have to go nonlinear either, Chapter 18 of ASCE 7 lets you use both ELF and modal response analysis as long as you meet certain requirements.
RE: Non-linear seismic response