Ignoring SMF Requirements? 1

[GlulamSam]

Good morning. I have been an EIT at my firm for the last 2 years and am working toward my PE license. I have worked on a number of residential projects similar to the one shown below with very tall (20') great rooms at the rear of the home. Most of our projects are SDC D. For the lateral design on projects like this our first approach is to see if the other walls on the upper floor can handle all of the lateral loads. This approach often doesn't work (plate height differences, the other upper floor walls are cantilevered past the main floor walls below, the shear walls just aren't enough to handle the load, etc.), which results in a special moment frame (SMF) being designed at the great room.

As I gain more experience I am digging into the code books more and more, and what I'm finding in there doesn't sit right with me compared to how our firm designs these SMFs. We do not provide the required lateral bracing as required by code. We follow the prequalified RBS connection procedure for the beam-to-column connection, but we do not design the continuity plates for the "maximum probable moment" per the code. Rather, we design it for the calculated moment from our loadings on the frame. I have brought these concerns to my boss (the owner of the company). He says that we don't do the bracing because the gravity loads on the beam are small enough that it won't matter, and also that none of our competitors are doing the bracing either. He says we don't use the maximum probably moment because, again, our loadings are so small that the maximum probably moment will never happen. These explanations aren't satisfactory to me, because as I read the code these things are required no matter what. In my opinion if we're trying to design a SMF without the bracing and without following the codified procedure for the RBS connection then we aren't really designing a SMF.

I understand that the code can often be onerous, but these explanations don't sit right with me because essentially they boil down to "ignore the code." So my questions are these:

1. What is your opinion on SMFs without lateral bracing, and RBS connections without using the maximum probably moment?
2. Are there other viable lateral designs for these walls? In reading the code I don't believe that an IMF or OMF could be justified due to these frames being more than 1-story, height restrictions, load restrictions, etc. I have done some reading on the Simpson Moment Frames, but I don't believe that anyone in our market is using them.
3. Is "our competitors do that so that's what we'll do" good logic? If there is ever an issue with these frames I doubt that that argument will hold water. But maybe my boss is right and the loads are so small that these things I'm worried about will never happen in a million years.

I'd be grateful for any advice or suggestions you could give. Thank you!

 

[KootK]

I believe that you are correct on all counts and that your boss appears to not understand the principles of capacity design that govern SMF. Your boss does, however, have my sympathy for feeling pressured to pander to the median level of technical rigor common in his marketplace.

I find the bracing issue to be or particular issue since achieving that within a wood structure would be rather difficult I would think.

 

[JoshPlumSE]

1. What is your opinion on SMFs without lateral bracing, and RBS connections without using the maximum probably moment?

If you don't do that, then it's not an SMF. It might be an IMF. Certainly would be an OMF. The firm I used to work for (heavy industrial) in the post Northridge days used to design our moment frames as SMF's, but used the R value of a OMF.... So, if you're using a lower R value, then this would certainly be okay.

2. Are there other viable lateral designs for these walls? In reading the code I don't believe that an IMF or OMF could be justified due to these frames being more than 1-story, height restrictions, load restrictions, etc. I have done some reading on the Simpson Moment Frames, but I don't believe that anyone in our market is using them.

I'll leave this to someone else who works in residential. I tend to think there is a way to get an IMF or OMF to work for this situation.

3. Is "our competitors do that so that's what we'll do" good logic? If there is ever an issue with these frames I doubt that that argument will hold water. But maybe my boss is right and the loads are so small that these things I'm worried about will never happen in a million years.

The only arguments that hold water are a) This is what the code allows us to do... and b) The standard of practice in the industry is....

And, it's far better to use argument a) rather than argument b). Argument b can be used for obvious situations like where the code says something vague like "all aspects of deformation shall be considered" then you can easily say, the standard of practice is to use one of the following analysis programs, many of which do NOT include this particular form of deformation (I'm thinking of panel zone shear deformation).

 

[GlulamSam]

Thank you KootK and JoshPlumSE for your commentary.

I find the bracing issue to be or particular issue since achieving that within a wood structure would be rather difficult I would think.

Yeah, that's the main reason we don't do it, and I guess our firm justifies that by saying the loads are small. Guess everyone else around here does the same...I think I should find a new career.

I tend to think there is a way to get an IMF or OMF to work for this situation

I just can't get past the wording "Single-story steel ordinary moment frames" and "Single-story steel intermediate moment frames" (ASCE 7-16 12.2.5.6 and 12.2.5.7). On projects like the one above... am I not looking at a 3-story structure, or at least 2-story if you count a 20' tall room as 1 story?

 

[KootK]

...I think I should find a new career.

Right? It's tough. You gotta explore it enough to feel that you really did but, once you've done that, it starts to feel as though it's too late to turn back.

 

[JoshPlumSE]

Residential structural seems especially tough to me. Tighter margins, more competition, et cetera.

School and Hospital work, while challenging, seemed better to me. The quality of engineers working on it and the end product we put out seemed so much better. That's even true for the healthcare "adjacent" work like Skilled Nursing Facilities, anchorage of hospital equipment and dialysis centers.

I cut my teeth on heavy industrial work, and that was pretty forgiving (IMO). I learned a lot, especially about foundations, steel moment frame, seismic loading, dynamics, and equipment anchorage. The company always assumed the liability and I didn't have to stamp much.

 

[ARPASEVAN]

There is a lot of testing and research behind those frames and choosing what requirements to use/not use is not very "professional". Your boss is 100% wrong. You either use SMF (or any other frame) with all design requirements or you don't.

You may use a multi story OMF in most residential buildings and do away with braces. [URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1647886456/tips/OMF_ae4nxw.pdf[/url]

Simpson Moment frames may also be an option. They use R=8 and do not require a brace.

 

[dold]

Ditto to what Koot and JoshP are saying. You can't just use a R=8 of a SMF without detailing it as such and meeting all of the ductility/compactness/bracing/connection/etc/etc/etc requirements of a SMF. Using R=8 and then only designing the connection for the corresponding forces seems like it would be unconservative since you've said "ah, the frame is super ductile and will dissapate all this energy", but in reality it's relatively stiff and will probably remain elastic anyway and therefore not dissipate all that much energy internally, so, your frame/connection forces are about 8/3(ish) times higher than what you're designing for.

And yes, Koot, detailing the bracing for SMF in wood construction is damn near impossible (brace stiffness-wise).

I'm 95% sure there is an exception somewhere in IBC or IRC that allows the use of OMF for 1-2 story light framed residential/dwellings in SDC D. They acknowledge that SMF detailing for a house is juuuuuust a bit overkill. Single family homes just aren't heavy enough to truly utilize/mobilize the ductility of a SMF. And I'm pretty sure that exception allows you to still use R=6.5 (wood shearwalls) for the entire structure and just amplify your loads for design of the moment frame. (don't quote me on that though).

 

[KootK]

Sweet, thanks for that ARPASEVAN. Would have been nice if they gave us enough margin to cover some face brick. I bet a lot of wall / window averaging takes place...

 

[JoshPlumSE]

A couple of last things that I just thought about.


1) Regarding the design of the continuity plates to a lower value than the maximum probable moment:
There are ways to use a significantly lower force level:
a) You can use the "overstrength" seismic loads Omega*E to calculate your connection design strengths. Not sure if this is applicable to SMF's or not. If so, then it would be applicable to continuity plates. If not, then I've seen folks use R*E as the "over strength" factor. Essentially saying that they're using the ELASTIC seismic load as if no ductility exists in the system.
b) Limiting strength of panel zone.
c) Column yielding.
d) Foundation uplift.... etc.

I believe there is a design example in the AISC Seismic Manual that walks you through at least one of these situations.

I suspect that at some point, someone looked at these force levels and decided that something else was always going to control over the maximum probable moment. However, there should ALWAYS be something in the calculations submitted to plan checks that spells this out.

2) Regarding the bracing of the SMF connection / beam:
a) Section E3.4c, item 2 is about unbraced connections. It adds extra restrictions about the columns and such and may not apply to RBS connections specifically. But, it definitely applies to SMFs.


b) You might even be able to make an argument that this is NOT a single story moment frame!... Sounds crazy, but hear me out. The middle floor is not a real floor, right? It's just there because of the windows and such so that you can have a great room with huge vaulted ceilings. Right?

If the lower beam were a strut and didn't have a moment connection, would the system have worked? If the answer is "yes", then how can adding moment connections at that level cause this to "not work", when you know it improves the seismic behavior of the system.

It's a little convoluted. But, it's one of those things where the code has some relatively arbitrary reasons for forbidding something. So, if you can demonstrate that those arbitrary reasons don't make much sense for your situation then you should be allowed to use engineering judgement to bypass that arbitrary limit. Granted, you're taking on some more liability by doing so.

 

[GlulamSam]

And I'm pretty sure that exception allows you to still use R=6.5 (wood shearwalls) for the entire structure and just amplify your loads for design of the moment frame

I'll look around for this, because that would be amazing.

You might even be able to make an argument that this is NOT a single story moment frame!

Well I've still got the basement floor (it's daylight, not full height foundation wall) so even if the tall room counts as one story then I have another story at the basement. Although I guess that section from the code that ARPASEVAN posted seems to say that a OMF would be fine for more than 1 story.

But unless the suggestion presented by dold (R=6.5 for the whole structure) is correct, then with an OMF I'd need to design the entire back wall of the home - including the wood shear walls - for R=3, correct?

 

[ARPASEVAN]

The section that allows you to use R=6.5 for shear walls and the corresponding R for moment frames is 12.2.3.3:

https://res.cloudinary.com/engineer.../tips/Screenshot_2022-03-21_124722_r54mhs.jpg

]

 

[dold]

But unless the suggestion presented by dold (R=6.5 for the whole structure) is correct, then with an OMF I'd need to design the entire back wall of the home - including the wood shear walls - for R=3, correct?

I'd stick with what you're saying (R=3 (or 3.5 or whatever)) for the entire line as discussed in 12.2.3.3 (see ARPASEVAN's comment above) unless you can find the exact exception I seem to remember. I very well could be getting sections melted together. What I'm thinking of is more along the lines of vertical combinations. Safe bet is to pick a R-value for each line of resistance. Also note that I've spent approximately 10 minutes reading the IRC in my entire life - I just recall at one point a colleague mentioning an exception about OMF in SDC D for dwellings. I live in IBC world.

 

[kissymoose]

I live in IBC world.

Oh no, you're missing out.
Sometimes I feel like residential engineering is an oxymoron.