Equiv Lat Force Procedure Modal Mass Participation Minimum

[bookowski]

Some questions about period/mass participation when using ELF:
- I've always assumed that where your analysis model gives worse (shorter) periods than the code approximate periods you use the analysis periods
- If doing the above is there a min mass participation for the mode?

In particular my question applies to something along the lines of this:

I have building that is SDC B so low seismic and no real restrictions. The lat system is essentially a big 'C' from walls so very torsional but stiff enough to make it work overall.

I can use the code approximate period but this appears to be too generous compared to the analysis model. However if I want to use the analysis period then the 1st mode is clearly torsion. The following modes are not clearly x or y so to find an "x" period for seismic loads I can look at the participation, but how much do I need? To get a good degree of x participation I get to a very short period, especially compared to the code approximate period. Hate to screw myself if I don't need to but at the same time the code approximate period seems silly to use if I know that it's way off. Is there any guidance on this? (and no I don't want to do rsa, junky little building that I don't want to go nuts with - just wondering if there's some guidance/code commentary on this).

 

[MrHershey]

If doing ELF I don't see any reason why you couldn't/wouldn't use the approximate period, it's very clearly allowed by 12.8.2 in ASCE 7. Even if actual period is lower, you're also probably not responding 100% in the first mode like ELF implicitly assumes. How do base shears compare? How tall is the building? What's the construction type?

Analysis models can also be sensitive to a bunch of assumptions regarding effective stiffnesses, element stiffnesses that get neglected/included, etc. So just because the model says something doesn't necessarily mean it's any more right than the approximate value from the code. For instance, a lot of people/firms will omit floor out of plane stiffness in lateral models because they don't want any frame action but this can artificially increase your period and thus lower your seismic demand. Can also swing the other direction if you haven't applied ultimate level crack factors to a concrete structure when analyzing for an ultimate level event like an earthquake.

 

[bookowski]

I see the argument for using the approximate, the code allows it and it helps. Just curious if this is addressed/discussed anywhere. You can double your base shear or more. Particular case right now is 8 story building, shearwalls.

Listing all of the other stuff that we assume and get wrong doesn't sell me much. You could chase that one forever, might as well just give up. I do see the argument that the code allows it and it saves the client money, fine. I'd still like to know if this is what everyone does.

The torsional building is one situation that doesn't sit well with me. The other one is that I often end up with loads of wall in one direction that I don't need because of the site layouts in my area (usually lot line construction and often get full walls for simplicity/logistics etc) and in the other direction have to fight for every inch of wall. So I might end up with 3x the stiffness in y vs x, doesn't seem right to assume same period in both for elf loads.

 

[WARose]

If doing the above is there a min mass participation for the mode?

The number I have typically heard over the years is about 90%.

Of course, if you have a FEA model......you can take all the guess work out of it by using the Modal Response Spectrum procedure in the code. It accounts for mass participating in different modes. (I think a requirement of that procedure is enough modes to get 90% of the mass participating per direction btw.)

 

[KootK]

I agree that you could use the simplified period estimate and be code compliant. In practice, however, I use the model predictions if I've bothered to do the modelling. It's like walking in on your parents having sex. It just can't be unseen.

I'm also particularly bothered by this when you've got building torsion as your first mode. It seems to me that the simplified period estimates must have been based on some simplifying assumptions. I don't know exactly what those were but I'm guessing something like:

1) Regular-ish building.
2) Torsion not as first mode.
3) Shear building lateral deformation behavior.

Where much of that is violated, so are the estimates in my estimation.

I very much agree with MrHershy's general skepticism regarding accuracy though, especially with shorter buildings. Once you get into mixed lateral systems, gross estimates of cracking, and considering the shear deformation of shear walls, it starts to feel as though you could drive a snowmobile through your wannabe result bulls-eye.

You should really consider offshoring your sexy NY lateral work to a lowly paid foreigner who lives for this stuff. I know just the guy.




I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[bookowski]

Well, I agree with the lack of accuracy as well but it's not the question at hand and could be said about any question asked on here. It's like a guying asking 'i'd like to get into cycling, should i buy bike a or bike b' and the response is that he'll always be fat anyway so it doesn't matter. I can double my base shear or 'unsee' the issue as you say since code allows it. Either way I'm probably way off from reality but we need to fool ourselves into thinking we're doing real stuff here or what's the point.

I assume that the elf has assumptions baked in as you've said, was wondering if anyone had that off the top of their head or how they've dealt with (or ignored) the above situations.

They're not sexy and I doubt your lowly paid but if you want/need/have interest in work let me know.

 

[Deker]

This may be my west-coast bias talking, but I wouldn't feel comfortable using ELF for a building with significant torsional response. Assuming you have a torsional irregularity, you would be required to do MRSA for SDC D-F. I know that your building is SCD B, but since the ELF procedure is predicated on regular buildings with negligible torsional response, I don't think it's appropriate to use ELF. If you've already modeled the building to get the actual period, is it that much more work to do MRSA?

If you want to stick with ELF, I would pay careful attention to your modeling assumptions and use the shorter of the actual and approximate periods. The intent of the approximate code equations is to provide short periods for conservative base shear estimation. If you take the time to model the building, you are rewarded by being allowed to use the actual period (assumed to be longer) as long as it doesn't exceed the upper limit of Cu x Ta. If your modeled period is shorter than the approximate period, I would interpret that as either (1) inappropriate modeling assumptions or (2) a shortcoming in the approximate code equations. The following articles might be helpful:

Link #1
Link #2

 

[KootK]

They're not sexy

Like most things, it's in the eye of the beholder. Sometimes you uber-urban environment structural engineers tend to have kind of a skewed perception of what constitutes a cool project. I see that a lot with my Toronto cohorts. In my neck of the woods, an eight story building has a pretty good chance of being in the top five of your projects for the year with respect to building height. I spent a little time in Dumbo last weekend, along with all the other tourists. I thought that there was quite a bit of interesting construction going on near by.

if you want/need/have interest in work let me know.

Put me in the want/have interest in work category. I currently have a relationship with a Midwestern US engineering firm that I'm enjoying. While I don't know all of your specifics, I wonder if we might do something similar. I'll describe what I'm doing and let you ruminate on it's applicability to you.

- The Midwest firm is a couple of buddies that I used to work with who've gone out on their own. They're doing great but they're not yet at the point where they'd be comfortable adding senior level engineers to their staff. And sometimes, you want the help of a senior engineer. They could certainly handle any of the work that they send my way themselves but they need to spend their time at project meetings and doing business development stuff.

- I cherry pick the work that I take on. I'm trying to position myself so that I have two kinds of work: 1) interesting and marginally profitable 2) boring and very profitable. The design work that I do with the Midwestern firm is the former. If they've got ten thousand wood headers that need designing, they don't come to me unless they're in a bind (thankfully). However, I'm happy to help out with all kinds of things that interest me.

1) Structural modelling, particularly multi-story lateral.
2) Design of long span systems and trusses.
3) Design of transfer elements.
4) Design and detailing of complex connections.
5) Pre-stressed / Post-tensioned anything.
6) Quality control on contract documents.
7) Contract administration to the extent that I can participate remotely. I like developing solutions to field issues etc.

Basically, all the fun stuff that they'd love to do themselves if they could spare the time.

- I do charge a healthy rate for my services on an hourly basis. That said, I charge in Canadian dollars which typically makes as US dollar go about 33% further than it does state side. That allows me to do fun design engineer stuff while still earning a decent hourly rate. Also, since I'm cherry picking the work that I find interesting, I tend to budget my hours pretty competitively. I go over budget a fair bit but don't mind eating it as I'm doing interesting work and doing it for people that I like.





I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[MrHershey]

With 8 stories I'd be a little bit more wary. I'd probably start doing modal response somewhere in that height neighborhood anyways. If you've already done the modal analysis to know the period, it's really not much work as others have noted. You're already halfway there. Would just doublecheck the crack factors (including slabs and other non-SFRS elements) to make sure they're not too high.

One other alternate if you're worried about the major changes in stiffness in each direction is ASCE 7 gives you an alternate procedure for shear wall buildings (based on the Chopra article that Deker posted above) that would yield different periods in each direction. It's a fairly long process if you've got more than a few walls though, but would give you another alternate if desired.

 

[TLHS]

Yeah, ELF is not at all set up for torsional modes. If seismic is vaguely critical and you've got a lateral force system that's obviously going to respond torsionally, I'd defer to a dynamic analysis if it were more conservative. I'm generally of the opinion that our seismic methods are heavily approximate and are intended to be risk mitigation, so I'm comfortable designing to the letter of the code generally. I still wouldn't mess with torsional irregularities, even if they're allowed in your case.

Doesn't ASCE 7 have a test for torsional irregularity? In Canada, there's a torsional stiffness test. If you fail it, your building is classified as irregular and you have further seismic analysis restrictions.

You end up not being able to use the static force method for structures that are torsionally irregular unless you're in a lower seismic area. In a lower seismic area, I'm sure the irregularity would be just as big a problem for wind and I'd probably be okay with using a more approximate seismic method, as it may not be the governing concern.

 

[MrHershey]

Yes, ASCE has tests for torsional irregularity. Very few of them actually apply to OP's case though since this is such a low level of seismicity. Think the only thing you're required to do for torsional irregularities in SDC B is actually do a 3D model with cracked sections, that model can still be ELF and I don't think you even have to amplify your accidental torsion for SDC B.

 

[RobertHale]

A few thoughts on this subject. One of the most valuable pieces of information I learned studying for the SE was, ELF is created to deal with buildings whose response is dominated by 1st mode translational behavior. All of the irregularity checks are designed to find buildings that have more complicated responses, be it contribuitions from higher translational modes or torsional modes. You have fairly assessed the building as being irregular. I have to assume that since you have been looking at mass participations and periods, you are practically at a Modal Response Spectrum Analsysis. Since most softwares that can give the info you have told us you have are capable of MRSA, why not just complete the loop and see what a slightly more sophisticated anaylsis gives you? You get to sleep good and you dont have to worry about the lawyers in the future.

Robert Hale, PE

 

[bookowski]

Kootk - Could work. I have similar setup with a couple of people now but I could use more capacity, they're also cherry picking which is difficult on my end. An 8 story is still the upper end of my work, most of it is random renovations or other misc work. If I get one or two 8 stories a year it's enough to keep the lights on. I've got something new right now that I could use help with asap. How do you exchange contact info on here?


Thanks for the opinions/input - doesn't sound like there is anything that I was missing. In my area it's sdc b 90% of the time and therefore wind almost always controls design. Seismic is usually a formality only. In the cases that I was referring to the choice of analysis can be enough to make seismic control and change the design. Everything is usually modeled in etabs so I agree that it's not much of an extra step to do rsa, but the less I have to do the better so was just trying to iron out any documentation/commentary that limits ELF or lays out the assumptions since it's explicitly permitted in the code even with torsional buildings for sdc b. (rob hale - I think i'm the oddball but I almost never worry about lawyers with this kind of stuff. I'm pretty sure that if anything gets me it will be something frivolous such as a slip and fall in one of my buildings, or it will be water infiltration- its' always water)

 

[Trenno]

Not sure how the ACI/CSA codes handle predictions of fundamental period, does building stiffness come into it anywhere?

I've been stung, when using AS codes, that the fundamental period for ELF methods only consider height and general lateral system. This particular building was much stiffer (lower period) than predicted by code equations, therefore attracting much more seismic load. I always like to compare the results of ELF and RSA, allows you to sleep better at night.

 

[KootK]

Link. In an interesting twist of fate, I've got a bead on some minor NY work that I can't sign off on. Not even quid pro quo.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[TLHS]

Is it bad that I laughed about a twist of fate in a conversation about torsion?