Wood Diaphragm Chords / Aspect Ratios

[abusementpark]

For a wood diaphragm, at what aspect ratio would you argue that the chord member requirement becomes negligible? It seems to me at some point the behavior will become largely shear dominated and that you shouldn't have to worry about designing and detailing a reliable chord member. I was thinking that maybe this is true for aspect ratios of 1:1 or better.

Does anyone know of any publication or articles which address this issue? I'm curious what others opinions are.

I've seen a lot of wood diaphragms built with no reliable chord member that still remain standing. The sheathing alone must have some ability to resolve a certain amount of tension force. It is common for the diaphragm nailing pattern required at the location of the highest shear to be used for the entire roof deck. So, maybe that is a factor since there are extra nails that can be used for tension transfer at the center of the diaphragm span, where the diaphragm shear force is very low.

 

[DaveAtkins]

Although it is not "technically" correct, I agree that the sheathing can act like a chord.

The design of a building with steel deck on wood purlins recognizes this, in a way. The design procedure allows you to use all of the purlins, not just the purlins at the edge of the roof, to resist the chord force.

I don't see why this type of approach can't be used for a plywood or OSB diaphragm.

DaveAtkins

 

[msquared48]

Man, I would really be reticent to rely solely on that, having seen failed diaphragm chord connections in concrete tilt-ups from seismic action.

Seems like using a toothpick to hold up a skyscraper to me.

I do not see, too, that it is just the aspect ratio that determines the chord force. It is also the overall size of the diaphragm and the lateral loads generated.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[Lion06]

How is a chord typically detailed on the two sides where the chord is perpendicular to the joist span?

 

[a2mfk]

What type of specific construction are we talking about? Roof trusses on wood frame or CMU walls?

Describe your situation where you don't think you will have a chord or it will be hard to design one?

How low of a chord force are you talking about?

For a tension chord you can always use a continuous flat strap, they come in coils from Simpson and others, nailed flat on top of the sheathing at the edge of the roof. For your compression chord I think you could rely on blocking or a continuous structural fascia (2x4 with angle clips to the ends of trusses)...

 

[a2mfk]

BTW- For chords I meant where a traditional top plate in wood framing or bond beam in CMU somehow would not suffice as a chord.

 

[abusementpark]

What type of specific construction are we talking about? Roof trusses on wood frame or CMU walls?

Describe your situation where you don't think you will have a chord or it will be hard to design one?

This was more of a general question that is not in reference to specific project or condition.

I've seen many instances where designing and detailing a chord member in wood structures is difficult based on architectural requirements. Sometimes the architect does not want a continuous fascia board on the roof edge. Sometimes, the truss heel height is too large to argue that the top plates can be considered as an effective chord member. Sometimes the building edge doglegs in and out slightly, which makes it difficult to have a continuous fascia board or top plates. Sometimes there are a series of dormers along the edge of the roof which cause discontinuities.

A lot of times the only way to have true chord member is to design additional blocking with tension ties at the end of each block across the length of the diaphragm. It seems pretty wasteful to me to put this blocking in when you have a diaphragm with depth to span ratio of 3 and the chord force is less than 500 lbs, for example. It sure seems to me like the sheathing alone will suffice.

 

[msquared48]

For me, it is hard to imagine the sheathing functioning as a continuous chord member with all the discontinuous seams in the plywood panels. Whatever these panels attach to, or transfer to, that is continuous, has to be the structural chord member. Anything else, I think you are just kidding yourself.

There has to be a structurally defined route for these forces by code, not a wave of the hand.

However, if you are dealing with a 4' X 8' diaphgram, that's a different story, and I would agree.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[abusementpark]

For me, it is hard to imagine the sheathing functioning as a continuous chord member with all the discontinuous seams in the plywood panels.

But they are always fastened to a common element at the seams.

Whatever these panels attach to, or transfer to, that is continuous, has to be the structural chord member.

What if they only attach to framing members that are perpendicular to the chord force?

 

[msquared48]

No, they are not.

Wood Blocking and wood Ledgers are discontinuous.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[DaveAtkins]

The sheathing is always fastened to a common truss top chord (or rafter) at a seam.

DaveAtkins

 

[msquared48]

It is true that the seams are attached at trusses, joists and at the edges, but in unblocked diaphragms, there are unnailed seams. There is also discontinuity at the diaphragm edges as I alluded to above unless specifically designed as otherwise - the chord member.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[msquared48]

I think that I see what you are saying at the edge of the diaphragm where the trusses frame normal to the wall. However, technically, there is a big problem with relying on tension normal to the truss top chord, and it's called cross-grain bending which is not allowed by the code. I would not do it. Moreover, there would have to be 10 nails or so in a small space to achieve the 500# link. This number of nails could split the top chord of the truss. Metal strapping nailed to blocking below would be better, much better.

That being said, there would still be a problem with discontinuous chord members at the transverse walls that are parallel to the trusses. Technically, this could be alleved if the 8' plywood panels were split to either side of the discontinuity and nailed off to achieve the required tension link of 500#, or whatever. I would definitely inspect this though as I do not have real good confidence in contractors to comply, let alone read the small details in the plans.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[DaveAtkins]

Yet another way to handle a chord force without a chord is to resist it with the shear wall below. You end up analyzing two three-sided diaphgrams butted against one another at the open sides.

DaveAtkins

 

[a2mfk]

If you design for proper diaphragm shear path into the walls (whether that is from blocking, diagonal bracing, etc.), then the top of your wall will act as the chord. The chord does not physically need to be attached or part of the diaphragm, though it could, it usually is not. So then you have to just design the top of your wall for this compression/tension force, which for most CMU structures is automatic with a bond beam. Double top plates with offset splices or strapping can provide the same tension continuity for wood framing.

Am I missing something here?

Now this is all fine and dandy for rectangular structures, but I think what Abusement is talking about is your average custom 2500sf home in a new neighborhood, for example. There may be multiple changes in the wall direction (an L shape with bump outs), several different ridges and hips, dormers thrown in, whatever other crap the residential "designers" think is good architecture. Then the answer is - there is no good answer (I think).

I think for compression I see what Abusement says about the plywood acting as a chord, especially for something like 500lb and that is only the worse case at the center of the diaphragm. Not so sure about that in tension, which Mike covered. But I think in residential roof truss construction there ends up being a lot of redundancy, not that you can analyze it like that, but there are reasons why there are few documented cases that I am aware of where a roof truss rolls over because it is not properly braced or blocked, and diaphragms fail in tension because of a missing chord member. The top of your wall will act as a chord whether you like it or not, or unless something more rigid intercepts that load.

Also, at least with wind (not seismic), there are so many other things that are going to fail before a diaphragm reaches its capacity, ie, UPLIFT...

Mike- not that I am doubting you for a second, can you describe your diaphragm chord failure in a tilt up building in seismic (that I can totally see happening).

Abusement- I mentioned this before, I think I have only used this as a drag strut, but could you not place a flat strap continuously along the edge of your diaphragm on top of the sheathing for a tension chord? I think this would be ideal, and be in the exact place where it should be. Relatively cheap and easy to install, though a lot of nailing (a continuous flat strap from a coil).

Using the deep beam analogy, it would be like putting a piece of rebar at the bottom of your concrete beam... Then for compression you'd only need blocking, and not even full depth, just below the diaphragm (if you have tall truss heights).

 

[msquared48]

The tilt-up cord failure involved the tension failure of the rebar that was welded to one of the embedded angles at one of the joints near the center of the edge of the diaphragm where the chord tension is greatest.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[msquared48]

Dave:

I can see how that might work if the length of the adjoining shearwall and diaphragm were the same.

However, that case would not be the norm as the length of shear wall is usually less than the length of the intersecting diaphragm, necessitating a drag strut, which, in this case, would be the chord member.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[BAretired]

I tend to agree with abusementpark that for an aspect ratio of 1:1, the tension or compression in the chord members becomes neglibible and need not be considered.

Even for an aspect ratio of 2:1, the tension chord is not required to do very much. For higher aspect ratios, the tension chord comes into play and must be capable of resisting a force of wL²/8D where w is the load per unit length, L is the span and D is the depth.

BA

 

[abusementpark]

Now this is all fine and dandy for rectangular structures, but I think what Abusement is talking about is your average custom 2500sf home in a new neighborhood, for example. There may be multiple changes in the wall direction (an L shape with bump outs), several different ridges and hips, dormers thrown in, whatever other crap the residential "designers" think is good architecture. Then the answer is - there is no good answer (I think).

Ding! Ding! Ding!

The bigger problem is that commercial architects are starting to incorporate all these "busy" features to some large commercial projects as well.

 

[hokie66]

When it gets too complicated to brace at the roof level, brace at the ceiling level. That's where the force goes first, anyway.