Plywood shearwall foundations in cheap hotels 3

[Labs763]

This is my first endeavor into the design of very low cost chain hotels (plenty of experience in unique, steel/concrete framed hotels). I have received a structural engineering template from this hotel chain, which at first glance looked like it would really streamline my design process - it even designates framing orientation and shearwall locations (this is all wood framed except for a few steel girders).

My question/concern is that their foundation plan has no continuous strip foundations below interior transverse shearwalls. Since the framing is orientated parallel to transverse walls (i.e. framing bears on corridor walls and exterior walls), all transverse walls are non-load bearing. All that they have specified is a 5" slab on grade, and they are placing control joints at each wall. I have never attempted something like this before. I have used isolated foundations at each end of a shearwall to pick up chord forces, but I would still always provide a tie beam/cont. foundation below the wall.

Typical free body diagrams obviously just show a tension and compression force at each chord of the shearwall, but isn't there a compression block at the compression side of the shearwall that extends in from the edge? And outside of this issue, by transferring shear to the slab on grade, it becomes a structural diaphragm that unfortunately has a ton of control joints running through it. Neither are desireable issues, but it would appear they are standard designs used everywhere. I am in a high wind/seismic zone, so I can argue that and probably get a little leeway.

See attached for general floor plan.

 

[msquared48]

I would not settle for anything less than a thickened slab condition under the transverse shear walls either.

Also, I seem to recall a provision in the code limiting cantilever diaphragms to 20 feet, at least pertaining to decks. I know this layout is used a lot with interior shear walls only in apartments and condos, but never have liked it. This layout is prone to seeing higher torsional forces from unbalanced loading than one with exterior walls too.

Stick to your guns, regardless of the client here.

Mike McCann, PE, SE (WA, HI)

 

[JAE]

A shear wall on a 5” slab is stupid.

 

[Labs763]

Thank you, glad I am not alone on this. Will provide foundations under all shear walls and see if I get kicked off the job...

msquared48: Exterior walls will have shearwalls as well, but they will have foundations below so I just did not designate them to simplify the question. I just wanted to verify that only providing foundations under the tension/compression chords of the shearwall was still not a sufficient foundation system.

 

[JAE]

With the wall taking small amounts of gravity (it is not fully non-load bearing) the lateral forces would create uplift on the windward side of the wall end - you need something more than a 5" slab to hold it down at the least. And at the leeward end it would be driven downward onto the thin slab - not a good idea.

 

[KootK]

1) Like msquared48, I would typically have a thickened slab beneath the wall at a minimum to facilitate anchorage of the fasteners delivering shear from the wall to the slab on grade. That said, there are plenty of fasteners that can deliver shear properly into a 5" SOG.

2) If detailed properly, it seems to me that your shear wall chord forces may land right on top of foundation elements that run perpendicular to the shear walls (continuous perimeter and corridor footings). If so, that's certainly a less dire situation than dealing with the wall chord forces on the slab on grade.

but isn't there a compression block at the compression side of the shearwall that extends in from the edge?

3) That's not the case with wood shear wall systems. The entire chord is envisions as the discrete stud packs at the ends of the walls.

And outside of this issue, by transferring shear to the slab on grade, it becomes a structural diaphragm that unfortunately has a ton of control joints running through it.

4) Delivering shear wall loads to a slab on grade is a commonly used method in modest buildings subjected to modest shear wall demands. In a situation like this, I don't see that the control joints compromise this mechanism. You can still, easily, transfer lots of shear across those joints with even very modest levels of SOG reinforcement.

 

[Labs763]

KootK: You are hitting on my points of concern (or points I will have to argue when I choose to avoid the provided system).

The ends of the transverse shearwalls will absolutely fall over the corridor and exterior continuous foundations, so I can technically justify a concentrated tension/compression force at each end and resolve my forces. My concern however is that although we do idealize and design the chord forces as discrete stud packs, the reality is that they are not. Although the tension load will predominantly be forced to go through the hold down, the compression can be distributed along the length of wall (i.e. it will not sit entirely over the corridor foundation). See attached FBD.

In terms of shear transfer - agree there is no problem transferring shear into a 5" slab - multitude of attachment options. However, at most I would assume you get 1/2 of your welded wire bars to cross a control joint, depending on how you detail it. Some people stop it entirely. I would not be comfortable using this as a shear mechanism though, especially when the joint is directly below the wall (or maybe to the side of it).

One other point I am reviewing is if this even meets code. I cannot use a plain concrete slab due to my seismic zone and cannot use ACI 360 slab on grade design since this is supporting the building superstructure (and primary lateral force resisting elements). Even if I am proven wrong on the other two points, I do not feel there is wiggle room on this issue.

I have attached a cross section sketch of the shearwall to clarify the condition (each chord will fall on a foundation). The second sheet illustrates the proposed loading of studs in a plywood shearwall I am discussing (I work in the U.S., but yes the document is from NZ).

 

[msquared48]

From the standpoint of the vertical framing scheme and the lateral resisting system, I have two thoughts:

1. The transverse walls are screaming to be both bearing and shear walls, utilizing the floor dead load to resist overturning, and also requiring footings under the shear walls. These are the shorter walls and need the dead load to minimize the uplift.

2. The longer corridor walls see some dead and live load from the transverse corridor framing, but, due to their length, will, more likely than not, not require hold downs at the ends of the two walls. The transverse walls will actually transfer some additional dead load resistance to the corridor walls and this additional dead load can be taken into account for overturning resistance too.

Just thinking out loud here - Mechanical and plumbing requirements may drive the structural though.

Mike McCann, PE, SE (WA, HI)

 

[Labs763]

msquared48: 100% agree with you on rotating framing to force transverse walls to be load bearing. However, that is not the structural template provided by the hotel. The contractor has apparently done plenty of these and said they are essentially all built the same. Regarding your usage of dead load to resist overturning - I would design the corridor and exterior cont. ftg for negative bending moment so that it can pick up the dead load from the perpendicular walls when overturning.

So to clarify - I do not support the proposed system, I am only trying to justify it (or justify it with minor changes, like thickened slabs at transverse walls). I may still try to discuss other options later with the architect, but for now I am just trying to establish load paths and typical structural practice.

 

[JAE]

This is a structural wall. It is not a simple partition. According to code it needs a footing. According to ACI the “footing” needs to be at least 6” thick.

While you can probably engineer this thing to “work” with the intersecting Cross footings at the end, I would personally lose sleep over it if I did.

 

[KootK]

I would not be comfortable using this as a shear mechanism though, especially when the joint is directly below the wall (or maybe to the side of it).

Short of turning the framing, you may not be left with many alternatives. You don't have much dead load on this so, even if you provide the thickened slab footings, your shear path still probably runs through the slab on grade. It seems unlikely that you'll be able to make your foundations heavy enough to make a go of it with friction based on the footing weight alone.

However, at most I would assume you get 1/2 of your welded wire bars to cross a control joint, depending on how you detail it. Some people stop it entirely.

As with the thickened slab footing, this is probably an opportunity to sell the team on something reasonable without straying too far off reservation. Insist that only 1/2 of the mesh stops at the joints which should leave plenty for a competent shear friction mechanism to develop. I wouldn't rely on it but one could make a pretty good argument that the reinforcing isn't even required for the shear friction mechanism to develop. In this context, sub grade drag on the slab would perform a similar function to the shear friction reinforcing. And yeah, one does need to give some consideration to vapor barriers etc.

My concern however is that although we do idealize and design the chord forces as discrete stud packs, the reality is that they are not.

I get where you're coming from but disagree. Normally we're dealing with materials that are shear stiff and whose deformations are dominated by flexural response. Viewed in that light, I would agree with your conclusion about interior studs drawing axial loads from the flexural response. Most wood panel shear walls are the reverse however: shear soft and having deformations dominated by shear response. This changes the picture a fair bit. A useful model to have in your head is to think of a shear wall as a bunch of rigid, rectangular panels that generate wall deformation primarily via each panel undergoing a rigid body rotation that generates some racking. This tends to be somewhat true given that:

1) Much of wood panel shear wall response is dominated by fastener slip and;

2) Testing has been done that confirms this to a large degree. I'm at a loss to produce that testing quickly for you here (WSU?) but the gist of it is that they looked at a shear wall with a large point load in the middle of it to see how that point load affected uplift forces. And the result was that the concentrated load had hardly any effect at all on the uplift forces. In my opinion, this suggests that:

a) axial loads on shear walls should not be used to resist overturning other than when those loads are applied directly to the boundary studs and;

b) shear panel flexibility results in flexural wall stress being absorbed by interior studs to a far smaller degree than a classical flexural analysis would suggest.

If you visualize the shear wall panels as I've suggested, you will find that the shears on either side of each stud oppose one another and, thus, generate no axial force in the studs. That, except for at the boundary stud packs where the shear is delivered on only on side and, thus, axial chord forces accrue. Obviously, this model of behavior is not 100% true just as the classical flexural model is not 100% true. That said, this model is a large part of the wood panel shear wall story, both in terms of real world behavior and the story that we tell ourselves in practice.

One other point I am reviewing is if this even meets code.

Hopefully it doesn't meet code. I love it when prescriptive code requirements force us to do what we already though was best.

 

[JAE]

I love it when prescriptive code requirements force us to do what we already though was best.

Amen!

 

[msquared48]

JAE: As a deep beam sheathed EACH SIDE the short transverse walls could work, BUT, there is still the problem of shear transfer to the slab from the plate. Considering cover and embedment, 8” MINIMUM!

Mike McCann, PE, SE (WA, HI)

 

[KootK]

How many stories is this thing anyhow? At some point, I'd want a foundation just on account of the accumulated partition wall dead load.

100% agree with you on rotating framing to force transverse walls to be load bearing. However, that is not the structural template provided by the hotel.

Sometimes the front to back framing will be a part of the strategy for efficiently running mech from the corridor in.

 

[Labs763]

I ended up with the following code interpretation (as JAE referenced):
-Slab on grade would be supporting shearwall building elements, so it has to be considered "structural" per ACI 360
-I am SDC D, so the only plain concrete elements I could potentially support are CMU/conc walls (or stuff in residential), so ACI 318 will not permit plain concrete slab support (in addition to thickness requirements). It would have to be designed and reinforced as a mat slab.

Architect has complied (unhappily) and framing is now rotated to span between transverse walls, so all walls will be load bearing and have foundation support.

KookK: I would be interested in that research paper you referenced if you remember what it was. While I can understand shorter shearwall having all overturning T/C in the chords, I would not agree as the walls become longer (like the my attached image shows). If you were to have an infinitely long shearwall, would you still expect to only have compression in the extreme chords? At some point there is a transition to distributed forces.

Thanks for your responses. I still do not understand how other engineers allowed this design philosophy, but it does make sense that more and more of my designs are being considered "over designed" when I see this kind of thing. I will not expect to be called back to design more of these.

 

[msquared48]

And you should not want to be either.

Mike McCann, PE, SE (WA, HI)

 

[KootK]

KookK: I would be interested in that research paper you referenced if you remember what it was.

Tried and failed. The best that I could do was that I believe that I've found the thread, the poster, and the professor that led me to the paper originally.
Link. It's this cadair fellow and Dr. Dan Dolan.

If you were to have an infinitely long shearwall, would you still expect to only have compression in the extreme chords?

Yes, I would. Or, at the least, the longer that a wall gets, the more I feel that its behavior is that of a true shear panel, and the more cozy I get with the boundary members only business. This all ties back to the business that I mentioned previously where things are different with shear weak panel assemblies. It's different from, say, a wide flange steel beam with WT reinforcement top and bottom. There, the interior flanges will pick up flexuarally induced axial stress per Mc/I precisely because the web is very strong and stiff. With wood shear walls, it's the opposite.

Separating what is mechanically correct from what is common practice, I can tell you the following:

1) I'd be astounded to hear JAE or MSquared48 chime in to say that they've been designing their many, many wood shear walls using a distributed boundary member methodology.

2) My wife, up here in Canada, manages a structural team that does a lot of these kinds of buildings. She's never used a distributed boundary member methodology.

3) I design precast part time these days. Each year, I'll do about half a dozen hollow core decks with timber shear wall buildings stacked above. Tracking and dealing with the the shear wall boundary member forces is important part of that exercise. Of the many EOR's that I've worked with on these projects, not one has ever shown me shear wall boundary member forces consistent with a distributed boundary member methodology.

4) Any half-assed Googling effort will turn up about a zillion AWC shear wall design examples. If you can find a single one that uses this distributed boundary member methodology, I'll buy you a hot sandwhich.

Whether or not you/we agree with the common practice is most definitely up for debate. However, what is common practice in North America is not in my opinion.

Curious, when you design a bare steel roof deck on an ICI project, do you take the entire chord force at perimeter members or deal with it in a distributed fashion that assume that each suitably oriented framing member within the deck absorbs flexurally induced axial load?

 

[KootK]

FYI: I've summoned our NZ friends to this to, hopefully, broaden the discussion a bit: Link

 

[Labs763]

Kootk: Thank you for your response. I feel like you are not following the point of my original question, and are instead more concerned with how I design shearwalls and what I feel is common practice in the US, so please see the following clarification.

To be clear, I have never designed any wood/cold formed shearwall assuming a distributed compression or tension zone. As you stated, all United States based codes and literature point towards designing extreme chord members to resist 100% of overturning (plus whatever gravity is required). Distributing the load over multiple studs would not be conservative (for shearwall design) and I do not have any interest in testing limits such as these to save a few dollars with one less stud.

The reason I have brought this topic up is (or was, since my issue has been resolved) due to the slab on grade design and whether it is technically supporting perpendicular shearwall overturning forces. I was provided with a set of structural template drawings that I did not feel were adequate; what I was attempting to do was play devils advocate, and determine what the original engineer was designing for and how he would defend himself if I were to tell him his design did not meet code.

Although designing the chord members for 100% overturning load is conservative for wall design (vs a distributed zone), it it could potentially be unconservative for slab design if I was only designing the slab for partition dead load, when in reality I have a large, multi story overturning force being applied over a distributed compression zone. This issue was my concern - I was looking for reasons to explain to the owner/architect why this design should not be used. While you seem certain the studs over the slab on grade would not see overturning forces, I am not.

Now that I won the battle w/ O/A (but probably lost the war in that I will not get more work from them), this topic has become purely academic. To answer your question on untopped metal deck diaphragms: Do I design the extreme chords for 100% flexure? - Yes. Do I also believe that interior purlins could see axial forces due to diaphragm flexural stresses? - Yes. I believe this concept has been discussed in other threads, where w/ untopped metal deck w/ wood purlins, the purlins would see some distribution of chord forces:

https://www.eng-tips.com/viewthread.cfm?qid=436092.

I am going to reach out to AWC and see if they have any info or testing, because it really seems like there should be some data on this, even if it was not the intent of the research - especially when you have figures like the one I attached from my time in NZ floating around.

 

[KootK]

Thank you for your response.

You are most welcome. It's been an interesting discussion.

I feel like you are not following the point of my original question, and are instead more concerned with how I design shearwalls and what I feel is common practice in the US, so please see the following clarification.

I disagree and feel as though you've given short shrift to the extent to which I've stayed on track in my attempt to help you here. In my opinion, the discussion unfolded like this:

1) You were working on a project type that is not your norm and objecting to what you were seeing in sample work done by other engineers. And you wanted a second opinion as to whether or not your objections were valid.

2) It was my opinion that, for the most part, your objections were not valid. Yeah, there should a thickened slab but that's where the issues end in my opinion. And I was/am of the opinion the insisting that the framing be turned was a poor choice. That, for several reasons:

a) Front to back framing does nice things for mechanical which save people money.

b) I see no technical benefit in reversing the framing to put DL on the shear walls. All signs point towards our industry coming to the realization that interior loads on shear walls don't actually reduce uplift demands at the boundary members. So nothing gained.

c) Now you've not got DL at the perimeter of the building where you've probably got tons of windows and very high aspect ratio shear walls that could benefit from the DL.

d) You've "lost the war" as you put it with respect to the business development aspect of this assignment.

I saw this unfolding and thought that I could best serve your interests by being the angel's advocate to your devil's. And the only two ways that I knew of to do that were:

a) Point to what I know to be common practice as I believe that 90% of sound engineering is our profession's dogma and 10% is number crunching.

b) Provide you with as much theoretical backup to justify [2a] as I could.

So that's what I did. And I consider it all to have been on point.

While you seem certain the studs over the slab on grade would not see overturning forces

This is not accurate. I don't doubt that some interior studs will see boundary member style axial forces. And for just the reasons that you mentioned. A better expression of my position would be this:

a) I think that such forces may be of limited magnitude owing to the factors that we've discussed previously.

b) I accept that, as with most things, the answer lies is that grey, in between space.

c) Like I said, I'm a big believer in following the pack professionally. I have a great deal of faith in the engineers around me and the engineers that came before me. My judgement takes a back seat to no one's opinion but I always start from the assumption that I am the student. In the space that we're calling "cheap buildings", I believe that there are a lot of talented, gutsy engineers digging deep for economical solutions in a way that ICI space engineers are seldom forced to. I see creativity more so than negligence or ignorance.

To answer your question on untopped metal deck diaphragms: Do I design the extreme chords for 100% flexure? - Yes. Do I also believe that interior purlins could see axial forces due to diaphragm flexural stresses?

Right. But I'm guessing that you do not design your non-chord OWSJ all for some level of axial force even though you know that force likely exists. And that's probably because you find yourself in Rome doing as the Romans do. My point, as you've probably guessed, is that I see this thread's issues in a very similar light.