R=1.0 with an Importance factor of 1.5 5

[WARose]

I have a situation where I have to use a Response modification factor of 1.0. (Long story as to how I got there.)….but along with that, the client wants to use an Importance factor of 1.5. You do the math on that and that’s getting into the elastic range in a seismic event.

I can't find anything in the code that will (explicitly) get me out of it……can anyone think of a code based argument that could get me out of it? (Or perhaps just a common sense type argument.)

 

[BAretired]

Did you mean "getting beyond the elastic range in a seismic event"?

BA

 

[WARose]

Did you mean "getting beyond the elastic range in a seismic event"?

No, I mean it's amplifying the seismic load (rather than reducing it). Essentially my design load will not stress beyond the elastic level. (Assuming the design event is never exceeded.)

Kootk, if you are reading this....I'd appreciate your weigh-in.

 

[Deker]

The importance factor is meant to reduce your effective response modification factor to reduce ductility demand and the associated damage to the structure. At R of 1.0, the structure will already be expected to behave elastically under the design earthquake. One caveat to this is that the client may want the structure to behave elastically under the maximum considered earthquake, which requires R of 2/3 (equivalent to R of 1 with I of 1.5). There are some code provisions that do aim to achieve this level of safety, such as for members spanning between structures (ASCE 7-10 12.12.4).

Has the client provided a code based argument for requiring R of 1.0 and I of 1.5? If it is simply the client's prerogative, I'm not sure that any code provision can save you. Best of luck.

 

[WARose]

One caveat to this is that the client may want the structure to behave elastically under the maximum considered earthquake, which requires R of 2/3 (equivalent to R of 1 with I of 1.5).

Not sure I have heard of this before (i.e. that a R of 0.66 was considered the threshold of elastic response.) Thanks for the info.

Has the client provided a code based argument for requiring R of 1.0 and I of 1.5? If it is simply the client's prerogative, I'm not sure that any code provision can save you. Best of luck.

The R value I'm kind of doing to myself (as I said: long story). But the "I" value is something the client wants. I kind of want to arm-wrestle him out of the 1.5. (Considering the R I am using.)

 

[jittles]

What exactly do you want to get out of?

In any case, I think Deker covered things pretty well.

 

[JoshPlumSE]

I remember using an R = 1.0 for nuclear work sometime back in the late 90's. I can't remember how we got there either. But, it was a small structure, so we just dealt with it. I can't remember if we used an I = 1.5 or not. I think we got away without it because we said it was a temporary structure.

Though I remember we turned the column anchorage design over to the client (who was the owner of the nuclear site).

 

[wannabeSE]

My understanding is the I of 1.5 compensated for the reduced seismic design force, SDS = 2/3 SMS and SD1 = 2/3 SM1. So the I of 1.5 takes it back to the maximum expected loads.

 

[ihsan_]

I'm agree with Deker as well. In addition, if i were you I'd ask myself what level of earthquake should I consider?

As you know, the design earthquake: "the 475-year return period (or 10 percent probability of exceedance in 50 years)" and the maximum considered earthquake: "the 2475-year return period (or 2 percent probability of exceedance in 50 years)"

 

[KootK]

I think that we're missing two important pieces of information here so I'll assume them and let you react as needed:

1) Your client's motivation with the 1.50 importance factor. As neat and tidy as the MCE proposal is, I've not encountered many private sector clients sophisticated enough in the dark art of seismic design to make that ask. I'm guessing that you're client's real motivation is that they feel that they're doing some high stakes stuff with some expensive toys within your structure and they want a low risk that whatever environmental loads may come their way are gong to mess with that. Essentially a poor man's performance based design ask.

2) Your own motivation with the R=1.0. One reason for this may be cost as you may be trying to obviate the need for high ductility detailing which can get expensive. More likely, you're trying to respond to your client's desire for uninterrupted operation by providing a system whereby seismic resistance is not dependent upon high degrees of ductile response and the damage that implies. Again, a poor man's version of performance based design.

How'd I do with that? Assuming that I'm close...

3) Does either strategy listed above increase the margin of safety against collapse during the design seismic event? I would say no. Looking at it from a rudimentary, equal displacement theory basis, buildings designed to either strategy - or even both strategies -- collapse under the same design seismic input.

4) Does either strategy listed above help with the performance based project goal? I would say that both strategies do that since each will promote elastic, limited damage seismic response for a greater range of seismic induced displacement. In this sense, employing both stratagem concurrently represents double dipping and I suspect this may be the crux of what you're asking here. If so, I agree with your intuition here and feel that I = 1.5 is probably excessive combined with R = 1.0. Have fun explaining this to your client though. I can barely explain it to myself adequately. And my confidence interval on my ramblings here is only on the order of 81%.

 

[WARose]

Thanks Kootk. The client is wanting the 1.5 because they don't want any (significant) damage in a seismic event.....that's part of the reason I went with a R=1 for the system......BUT.....all this is combining for a level of conservatism not since William F. Buckley. So what I am trying to do is strike a balance between their desire for no damage and something not quite so excessive.

Ergo, I was thinking of justifying a R=2 on my end.....and keeping their 1.5. But that leaves me wondering (assuming this event hit): how much damage are we talking with 2? (For reinforced concrete.) I would assume we are talking some yielding here and there and some cracking....but hopefully not much beyond that.

 

[TehMightyEngineer]

OP: What is the seismic criteria for the site? Ss? S1? Site class?

I'd establish what the client wants to limit to avoid "double-dipping" as KootK put it. In my opinion a higher importance factor should be used to reduce risk of collapse, lower R factor should be used to reduce in-elastic action and resulting damage (increase durability).

It may be that there's better ways to do this depending on what your project actually is. Base isolation, tuned mass dampers, etc.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/

 

[MotorCity]

It seems like you and the client are both applying your own factor of conservatism (you by using R=1.0 and the client by using I=1.5).

Does the client know that you are using R=1.0 and what effect it may have on the cost of construction?

If so, and he/she still wants to impose the R=1.0 and I=1.5 criteria, and is willing to pay the premium then march on.

If not, I would choose R based solely on that which corresponds to the lateral system being used and apply I=1.5 per the clients request.

If you still want an additional layer of conservatism on your end, there are plenty of other ways to do it (select the next largest beam size, use one additional rebar, etc).

 

[KootK]

how much damage are we talking with 2? (For reinforced concrete.) I would assume we are talking some yielding here and there and some cracking....but hopefully not much beyond that.

I really don't know how to quantify that in any meaningful way. It's something that I've wondered about a lot myself. For "special" stuff, I usually have a pretty clear picture of what we're expecting to happen when it's go time, and where. For the low R, conventional, my understanding is murkier. Is it the same as special just less? Or is ductility a more distributed phenomenon in the conventional construction situation?

This sounds like one of those cases where you may need to educate the client a bit and somewhat forcefully steer them towards a good recommendation. Many clients aren't equipped to make good decisions of this sort and appreciate a consultant that makes a non-wishy washy recommendation. If it were me, my pitch would be something like this:

1) All permutations here involve some degree of risk, even the MCE / R=1.0. 2% in 50 etc. If this is very important to the client and they are able to articulate their tolerable level of risk, they might consider engaging your services to do a true performance based design.

2) If PBD isn't the path, then I'd recommend I = 1.5 and R = 3.0 (or whatever conventional is for your system). Nearly the same as MotorCity. This is an effective R = 2.0 and should give you a building that will sustain unquantifiably modest damage during a design earthquake that comes around once every 475 years or so. The odds of a seismic event coming along in the next fifty years and causing enough damage to interrupt operations should be pretty remote.

Is your design earthquake a New Madrid fault seismic event? If that comes to pass, it'll be complete mayhem out that way with the east coast devolving into something like a Handmaid's Tale scenario. Not even worth thinking about facility operations at that point. The priorities will be weapons, canned food, and fertile partners.

 

[WARose]

Thanks Kootk....good feedback again.

 

[azcats]

The priorities will be weapons, canned food, and fertile partners.

I think you found your new signature KootK.

 

[KootK]

I think you found your new signature KootK.

Ha! Tempting. Notice how I approached it gender neutrally.

 

[TLHS]

I feel like you're stuck in the minutia rather than the end goal here. What does the client actually want? An R of 1 or 1.3 with an Ie of 1.5 is potentially very reasonable if the end goal is ensuring elastic response for a high criticality installation. The argument would then be whether they actually need that.

These factors do different things, potentially, but they definitely overlap and the I value is used as a proxy for a couple of things.

The issue is documenting what you're promising and what the owner wants. If functionality is the goal, there's a lot more than just designing the structure to work. This is a question of overall seismic performance of the building, equipment, and other items, and a discussion regarding whether functionality in that sort of situation is even reasonable given the environment that will surround the facility. Even in critical facilities, you'll often have different design earthquakes being considered for full on structural failure and operability.

I have very much done high importance factor, low R design in high seismic areas for some rare types of installations and equipment, but there needs to be some solid design basis documentation in place explaining it.

 

[TehMightyEngineer]

how much damage are we talking with 2? (For reinforced concrete.)

KootK is definitely more of an authority than I in regards to seismic theory, but if I can add my $0.02:

Everything I've read about R factors is that for the most part they're established based on historical estimates and assumptions and have been tweaked over the years. The original design basis for their values is somewhat lost or not based on a strict criteria. Thus, converting the R value into accurate, real-world design estimates for durability and post-seismic event damage seems dubious at best.

Since the Christchurch seismic event I've noted more attention being placed on "resilient design" calculating life-cycle costs, post-event recovery costs, etc. I imagine many papers are probably being produced for the NZ market that could be utilized in a psudo-PBD analysis and help answer your questions of damage levels.

Being that I don't do any practical high-seismic design I don't have any good references on these resilient design methods. However, I imagine some googling will turn up some references from the NZ engineers. I'd also reach out to these guys

http://usrc.org/

and see if they have anything they can offer.

Is your design earthquake a New Madrid fault seismic event?... The priorities will be weapons, canned food, and fertile partners.

They're not messing around when Ss = 300, huh?

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL)
American Concrete Industries

https://americanconcrete.com/

 

[TLHS]

From a general conceptual standpoint, any R value that exceeds the presumed overstrength of the material/design (generally assumed around 1.3 for steel, and varies by material) is assuming some amount of ductile yielding that would generally be considered 'failure' under other loading conditions. Whether or not that's acceptable depends entirely on the owner's expectations, the structural system, the accessibility for assessment, and other similar things. It's also a question of what level of forces to use for equipment, since the code design methods are generally restraint only with amplifications to act as a proxy for containment on hazardous materials. There are other standards that go into this, to some degree.

TME, I'd be very careful going down a piecemealed seismic approach from high level sources with regards to performance based design. If it's something that people wanted to chase for a project, there are ATC and ASCE sources that can be used as a north American basis but they are a large investment in learning time and generally an unreasonably large investment in analytical time. There are efforts being made to generalize some calculation based on typical framing systems, but it's still early days.