Literature for diaphragm action through a gable roof. 1

[StrEng007]

Can anyone point me to literature that discusses diaphragm action for gable roofs?
Most of my resources talk about diaphragms for flat roofs, which also tends to lend itself to monoslope roofs.

I've tried to argue (or convince myself) of the manner by which loads travel through a gable diaphragm but due to the lack of information out there, I'm wondering if ANYONE actually has it down pat.
Sorry to those of you who are tired of this subject. I was comfortable with it in the past, but now it's got me questioning things again (does that happen to any of you?).
thread507-430158

I have a sneaking suspicion that when any sort of ceiling is present, the lateral loads are taking a more direct path, whether we want it to happen or not. So whose to say if all these assumptions we make about pitched roof diaphragms are actually true.

 

[Celt83]

If i recall correctly this has a chapter on diaphragms authored by Diekmann and has a small section on gable roofs, Link

 

[phamENG]

There's a company out there with an article from an engineer looking at shear transfer at ridge vents and the mechanics of that transfer. It gets into this question a little as I recall. I'm on my phone so can't post it directly, but a search for "shear transfer at ridge vents in high seismic regions" should get you there.

But yeah... occasional what have I been doing!? moments are way too frequent to be good for my health.

 

[StrEng007]

phamENG, is this what you're talking about?
Ridge Vents in Heigh Seismic Regions

Believe it or not, I still get stuck on how the friggin load gets into the diaphragm in the first place. Especially with the CMU wall/tie-beam setup. Only the truss connection can resist the F2 load (bottom of truss bearing at top of wall), so I believe WW and LW lateral loads must enter the diaphragm via that connection. From there, a local portion of the truss top chord must have some strut action that delivers the load into the diaphragm. My buildings get so much wind load that I'm often using (2) types of truss clips (one for uplift, one for F2) and some sort of roof boundary clip to hold the F1 in the other direction. It's insane.

There you go, I'm having another one of those moments now.

 

[StrEng007]

If i recall correctly this has a chapter on diaphragms authored by Diekmann

Do you recall it being worth the additional 600 pages of dead load? Bookshelves are sagging over here. I tried finding a digital version without much luck. Too often I've ordered a book just to find the couple sheets I was looking for barely justified the carbon footprint required to get the book to my door.

 

[SWComposites]

If you have high wind loads, why are you questioning needing multiple clips?

 

[Eng16080]

In my mind, the gable roof diaphragm acts similar to if the roof was flat, provided that edge nailing of the sheathing is satisfied at the ridge. If it isn't, per the ridge vent detail, I suspect it can still be viewed the same way, with perhaps some reduction in strength. I'm not sure how the reduced strength would be accounted for or if it's even significant. I suppose at the extreme, you could look at it conservatively as two separate diaphragms.

I still get stuck on how the friggin load gets into the diaphragm in the first place. Especially with the CMU wall/tie-beam setup. Only the truss connection can resist the F2 load (bottom of truss bearing at top of wall), so I believe WW and LW lateral loads must enter the diaphragm via that connection. From there, a local portion of the truss top chord must have some strut action that delivers the load into the diaphragm. My buildings get so much wind load that I'm often using (2) types of truss clips (one for uplift, one for F2) and some sort of roof boundary clip to hold the F1 in the other direction.

This seems like a good description to me. I assume you're using blocking between trusses, and the F1 force you mention is between the blocking and wall.

"The Analysis of Irregular Shaped Structures" is a great book that gets into diaphragm and shear wall design in detail. I'm not sure it addresses your specific question, although gable roof diaphragms are clearly shown in some of the problems. The book weights 2.6 lbs. Each pound is worth it.

 

[phamENG]

StrEng007 - that's the one.

I agree with you - where ceiling framing is perpendicular to the applied lateral load, that's more than likely the load path. Especially when you have a ceiling that can brace them out of plane. I still detail the roof as a diaphragm for a couple reasons, though:

1) Same reason I don't like gyp shear walls. Get a tree limb through the roof during a hurricane and your lateral force resisting system has a way of dissolving right when you need it most.

2) It's good practice. Houses with nice, neat ceiling framing that all runs in the same direction went out of style in the mid 90s. Most of the stuff I design has framing running in at least 3 directions at every level by necessity with ridge beams in the roof. So continuous ties at that level are rare.

3) How does the load get out of the gyp diaphragm and into the end walls? Those connections aren't often engineered, and you'd probably find that getting the load transfer there to be robust enough you'd be dealing with an atypical connection. One that the drywall guys will ignore if they're even away that structural engineers exist.

If you have a truss that needs more than the IRC prescriptive connection to get the load from the walls to the roof, then it would be wise to inform the truss manufacturer so the bottom chord to top chord connector plate can be sized accordingly.

 

[StrEng007]

If you have high wind loads, why are you questioning needing multiple clips

I was never questioning needing multiple clips. Please refer back to my OP. This is to do with the behavior of a diaphragm.

I assume you're using blocking between trusses, and the F1 force you mention is between the blocking and wall

Yes, that's correct. If my F1 is low, I've taken this load "out through the truss connection" before.
However, IMO that creates a concentrated cross-grain loading and potential roll-over for the actual truss. Logically, the diaphragm is happiest when it has a line of continuous support. I'm dealing with 175MPH winds at Exposure D.

So the options were are:
1. Take F1 with connector (not preferred)
2. Provide continuous blocking with clip for shear transfer to wall.

3. If you want to get really fancy, push your lateral load to the fascia, bring it back through soffit sheathing and into the wall. Simpson has a discussion on this.

Why a Structural Boundary Member Between a Truss/Rafter is Not Optional

a great book that gets into diaphragm and shear wall design in detail

Great book. I thumb through it every couple months looking things up. It doesn't cover this discussion in particular.

If you have a truss that needs more than the IRC prescriptive connection to get the load from the walls to the roof, then it would be wise to inform the truss manufacturer so the bottom chord to top chord connector plate can be sized accordingly.

This is a really great point and I'll have to look into it. Is this a rule of thumb for you, or does the Code say something like "hey, get your truss designer involved if you can't make prescriptive stuff work"?

You know, maybe it's a local thing (ie, truss designers here already anticipate this concern) but I don't recall ever having this discussion with any other engineers here in Miami-Dade. Much of what I see now is already a lot more informative than those 80's and 90's designs.

From what I can tell, truss designers only consider loading in the transverse plane of the truss and aren't concerned much with axial/load transfer unless a truss is designated as a drag and a load is provide. By that standard, pretty much a majority of common truss are all acting as pseudo drags, especially in high wind areas.

StrEng007 - that's the one.

I had a chance to get into that article and two things the author mentioned made me go, "huhhhh?"

1. Shown below, the author is implying that unblocked diaphragms don't require formal chords. To me this an error. Is this explained in the Code?

2. What on earth is this moment force to diaphragm action he's showing here. Maybe I'm just interpreting this incorrectly, but is this analogous to a flat bar in bending about it's strong axis, without the presence of any flanges sort of deal? I understand the concept, but where does the NDS, AWC, etc. account for this?

I can't imagine this would ever work in a area with a decent amount of lateral load.

 

[Celt83]

For $40 probably not worth it, if you can find it for $20 I found the book worth that for just the Diaphragm chapter as it contains some good info on sub-diaphragms for wall anchorage and openings in shearwalls/diaphragms which Malone references in his book.

 

[ChorasDen]

How does the load get out of the gyp diaphragm and into the end walls

Careful on this one. To my knowledge, there is no design information regarding capacities or requirements for wood framed diaphragms that develop diaphragm action from gypsum wallboard. Diaphragm behavior and shear wall behavior are distinctly different, and I would approach the consideration of using gypsum for developing forces with caution.

 

[phamENG]

ChorasDen - agreed. I don't like the idea at all, but I've heard more than a few people talk about it. Breaks at intermediate walls are also a bit of an issue. Lots of problems with it. The shear connection at the boundaries is just one of many.

 

[lexpatrie]

Haven't read the full thread, but reality is the gypsum board has similar stiffness to the OSB/Plywood sheathing so in reality they share load, the gypsum tends to get damaged at earlier force levels and by code we're supposed to ignore the gypsum ceiling diaphragm if there's a "legit" wood diaphragm, and it's not to be used at all in seismic, as I recall, but it does absorb some load due to the stiffness of the gypsum board. You can see that pretty clearly in some of the NEHRP and California/FEMA seismic research where the ceilings get torn up by the seismic loads.

@pham - Keep in mind friction and bearing work in reality but are not always considered in engineering design, so there are load paths "out" that may not be explicitly considered in an engineering design.

As to the folded plate, as I recall, the strength is similar, but the stiffness is increased. As one does typically use a flexible diaphragm the stiffness of the diaphragm doesn't do much to the force distribution (unless you're super into reality precision versus normal design procedures that produce acceptably safe designs, though I'd be surprised the stiffness dramatically changes the design of the walls). From a design standpoint I've seen APA or AWC seminars where they say treat as a flat diaphragm. I don't have that at hand, if somebody else finds it feel free to post a link, otherwise I'll try to dredge it up.

This looks kind of good - Diaphragm seminar from Peter Von Buelow, Michingan.

As a side note, Dolan, is pretty big researcher in light frame construction. He and Frank Woeste do a lot of interesting research a lot of other parties probably consider "beneath" studying.