Offset chord load paths in residential structures

[YoungGunner]

It's common for residential structures to have multiple jogs in the same plane and I've never had a plan reviewer concern themselves with the lack of chord continuity until now. I've drawn up a representation of the issue they have. They are calling us out because there is both a vertical offset in the diaphragm and a horizontal offset of the chords. Again, this happens ALL the time in home design, and this is the smost simplistic situation of it and there are far worse scenarios, and I've never seen or heard a reviewer call it out. Curious if anyone thinks this is an issue and what would be done to mitigate it. The shorter truss will be nailed to the taller truss and will be detailed so there is some measure of lateral continuity between the two diaphragms.

 

[Celt83]

Get yourself a copy of this: Analysis Irregular Shaped Structures Diaphragms

and this: Wood: Engineering Design Concepts; Volume IV

second link contains some of the source material referenced by Malone in their book.



I'm making a thing:

http://www.thestructuraltoolbox.com

(It's no Kootware and it will probably break but it's alive!)

 

[ChorasDen]

As with everything else, it depends. You haven't provided much info on your sketches, is this a 3in horizontal offset? A 6ft horizontal offset? The plan reviewer is asking you to consider this as a reentrant corner, and to analyze and detail accordingly. This is a very common request and requirement in high seismic zones such as out west, and I believe it to be a justifiable request from your examiner if they feel like local design conditions warrant it.

The vertical offset is a bit easier to resolve, assuming each truss is designed as a drag truss, and are connected together appropriately. My only concern might be how to consider shear defection effects between the two, as each truss has a different design depth, as it appears they have equivalent bearing heights.

 

[JLNJ]

When you offset the chords, where do the axial forces go if they cannot transmit directly?

 

[KootK]

Curious if anyone thinks this is an issue and what would be done to mitigate it.

I've been hearing a fair bit about AHJ callouts on this kind of thing lately. Someone must be hosting seminars for code officials where the impression is being given that it's time to start taking light frame structural lateral design seriously. Too seriously in my opinion.

Curious if anyone thinks this is an issue and what would be done to mitigate it.

1) Shear across the step is easy enough by nailing together the trusses on either side of the joint as you suggested. Chord continuity is obviously the big problem.

2) By the book, what you'd ideally want at the plan jog is collectors and a little transfer diaphragm. In your situation, that would have to occur at the ceiling plane which would be a detailing nightmare.

3) I suspect that what would really transpire here is that the walls below the chords would transfer the chord forces down to the floor level below where the transfer diaphragm stuff is viable because everything is at the same elevation. In effect, the chords at the level below pull double duty and become, also, the chords for the roof. Unfortunately, I'm pretty sure there's a code limitation somewhere that precludes large vertical offsets between diaphragms and their chords.

So yeah, you're pretty much hooped even though this kind of thing has been common practice forever. Hence forth though shalt consider buildings to require separate lateral systems on either side of a combination vertical / plan offset. Architects beware the cost of attempting to create visually interesting buildings!!

 

[Aesur]

At the reentrant corner, I would specify the longer truss to take an axial load and act as a chord of sorts, then I would use a CS strap attached to the top of the wall top plates and dragged into the higher diaphragm by use of shear blocking panels for a length required based on diaphragm capacity. This is fairly common for the higher end residential we design. Additionally, I would required the taller truss to be sheathed down to a ledger at the low roof with out of plane attachment to the truss.

I agree with KootK's comment about the AHJ's..

 

[KootK]

At the reentrant corner, I would specify the longer truss to take an axial load and act as a chord of sorts, then I would use a CS strap attached to the top of the wall top plates and dragged into the higher diaphragm by use of shear blocking panels for a length required based on diaphragm capacity.

That looks good on paper and would probably slip by AHJ. However, for sport, let's consider that strategy in it's full complexity:

1) You need two drag trusses, one at each end of the newly created transfer diaphragm.

2) Obviously, the transfer diaphragm area itself needs special fastening.

3) The coil strap needs to be fastened to the top plate before the trusses are installed. Doable, if mildly annoying.

3) The coil strap needs to be overhead nailed, by hand, into the shear blocking (likely 2x). In my head, this sounds difficult and likely to produce pretty beat up shear blocking.

4) On a pitched roof, I feel as though the top of the shear blocking ought to be mitered to match the slope of the roof. I can also envision the sheathing nailing beating up the shear blocking pretty good.

5) You also need shear blocking at the heels of the longer trusses.

6) The shear blocking attached to the coil straps must be continuous rather than intermittent because it has to transfer compression as well as tension. In my area, this gets architects up in arms because, depending upon the detailing, it can disrupt attic ventilation.

7) Each bit of shear blocking needs to transmit vertical shear to each truss that it hits which can take several forms:

a) Clip hardware into truss chords. It would be nice to have truss verticals at these locations but that can be tough to make happen.

b) Toe nailing into the truss chords from both sides which will be a fair bit of nailing into the truss chords at one place.

c) Using staggered, two ply shear blocking and end nailing it to the truss chords.

... probably a bunch of other options that I've not thought of.

8) At each of the drag trusses bounding the transfer diaphragm, the shear blocking will deliver an uplift load into the truss that may require hold downs into the wall system.

With all of that in play, one starts to wonder if such a fantabulous kit of parts actually possesses the stiffness to do the job that we're assigning to it. Moreover, in a lot of the buildings that I deal with, we'll have jogs like this every 30' or so. So you'd wind up doing this chord splicing voodoo for much of the length of the building.

 

[phamENG]

That is a gable roof shown in elevation, right? For the stated example, I might try to do a pair of three sided diaphragms. Seems easier.

I agree with Aesur's shear blocking at the horizontal offset, and I don't think it's quite as hard as KootK is thinking it is.

1) I wouldn't be too worried about this. But see my note below about wind/seismic. May be more important where the earth moves. So long as the diaphragm has the shear capacity that the second drag truss would usually need to take, I let it go.

2) Maybe, maybe not. My average roof diaphragm is waaaaay below capacity. Standard attachment is enough to do quite a bit where I am.

3) Not necessarily. For one thing, Simpson has their new nail gun coil straps with dimples to help set a nail properly with a gun. They also have straps that can bypass the truss and be nailed to the face of the blocking. H2A turned on its side, for example.

4) This would be a nice detail, probably never built scenario. But who knows, maybe you'll get a carpenter that pays attention to drawings. I don't know that it's that critical, though, unless we're talking 8:12 or steeper. So for folks in Canada, probably matters. Here and south, less so.

5) Probably need that anyway? Else how is your shear getting to the shear walls in the long direction (not shown in OP's sketch).

6) This doesn't bother me. Intermittent blocking can still transfer compression. Depending on workmanship it won't be quite as efficient, but it still works. The coil strap is there because the blocking is intermittent and can't transfer tension. Or am I misunderstanding your meaning?

7) I can see this causing some localized issues. Offset makes continuous strapping a challenge without the need for extra blocking and waste, and I agree all those toe nails would be problematic unless you can ensure truss verticals and space out the nails. Otherwise, the clips are probably the best bet.

8) Easy enough, but added cost.

I think it's strong enough and probably stiff enough to prevent collapse in most scenarios. When I start doing this kind of detailing, I'm rarely worried about serviceability (unless it's a really unique situation). This is the sort of detail that prevents that initial movement of the proverbial zipper. If the zipper doesn't come undone, the house doesn't come apart. The sort of storm that would cause this detail to matter (based on observation of similar houses with absolutely zero comparable detailing and what they have survived) falls squarely in the "I don't care about your drywall cracks" and "what do you mean you're not evacuating?" areas.

For the general case, I tend to fall back on the IRC for some 'lower bound' solutions. As long as I'm not getting into high capacity, blocked diaphragms, I use the IRC's braced wall rule for offsets. So as long as the jog in the wall isn't more than 4', I don't usually do anything too crazy. I'll run blocking and a strap to overlap the offset, but I won't get into serious sub and transfer diaphragm detailing. But then I'm mostly concerned with wind, and under 7-16 I'm actually getting quite a few houses coming in as Seismic Design Category A.

 

[KootK]

1) I wouldn't be too worried about this. But see my note below about wind/seismic. May be more important where the earth moves. So long as the diaphragm has the shear capacity that the second drag truss would usually need to take, I let it go.

If you let it go, it's an incomplete load path unless you're willing to lay claim to using the sheathing for in plane flexure. I feel that the whole thing is "a bit much" but, at the same time, if we're going to do this at all, then surely we ought to do it as a complete load path.

2) Maybe, maybe not. My average roof diaphragm is waaaaay below capacity. Standard attachment is enough to do quite a bit where I am.

I would say definitely for something of the scale that OP has shown. To an extent, this is a matter of choice. If you're willing to block ten truss spaces to transfer the load then, yeah, normal sheathing fastening probably gets the job done. If you plan to block two or three, which seems to be the norm, you'll likely need transfer diaphragm fastening.

3) Not necessarily. For one thing, Simpson has their new nail gun coil straps with dimples to help set a nail properly with a gun. They also have straps that can bypass the truss and be nailed to the face of the blocking. H2A turned on its side, for example.

Nifty tech. Do the dimples mess with the ceiling drywall install at all?

4) This would be a nice detail, probably never built scenario. But who knows, maybe you'll get a carpenter that pays attention to drawings. I don't know that it's that critical, though, unless we're talking 8:12 or steeper. So for folks in Canada, probably matters. Here and south, less so.

This may simply be ignorance on my part. Is there an NDS provision that provides for shear nailing across an air gap if it's sufficiently small? I've been assuming that there's not. I'm not suggesting that the miter is practical but, rather, that the whole setup is impractical because of things like the miter.

5) Probably need that anyway?

1) It is my experience that, in most areas of north America, that blocking will not normally be present.

2) At the transfer diaphragm, the blocking will likely be needed a greater density than normal. None of that "every third" truss business etc.

Else how is your shear getting to the shear walls in the long direction (not shown in OP's sketch).

Unblocked heel rollover capacity of low heel trusses or, more generally, whatever sketchy practice pervades in whatever market one finds themselves.

6) This doesn't bother me. Intermittent blocking can still transfer compression. Depending on workmanship it won't be quite as efficient, but it still works. The coil strap is there because the blocking is intermittent and can't transfer tension. Or am I misunderstanding your meaning?

You must be missing my meaning. The shear blocking that's not over the top plates is basically a segmented column in compression, not shear lug-ish as over the top plates. The only way to transfer compression across intermittent gaps in the blocking would be by using the tension straps or the sheathing in compression, both of which sound sketchy to me.

8) Easy enough, but added cost.

Added cost, added complexity, and added design attention. I don't recall having seen this detailed on anyone's drawings ever.

I think it's strong enough and probably stiff enough to prevent collapse in most scenarios.

I disagree and feel that the chord restraint provided by the shear walls below would usually dwarf the stiffness of one of these chord slice details that needs to mobilize the slip in a thousand heavily loaded fasteners in tension and shear before it fully engages. By the time that one of these splices actually performed any useful function, I suspect that your low windows would all be smashed and your walls beat to heck.

This is the sort of detail that prevents that initial movement of the proverbial zipper.

What is it that we're preventing from being unzipped? The diaphragm starts its life off with a big crack along the joint between the high and low trusses. It's born "unzipped". Ditto for the offset top plates.

 

[XR250]

In my market, the chance of this getting installed properly or even at all is slim to none. Most of these framers cannot read plans and have little if any oversight from the GC. I feel like the KISS method is always best in residential. My go to would be try to use 3-sided building principals. If it means sheathing some interior walls, then so be it. Much easier to get that done correctly than some piecey blocking detail.

 

[KootK]

If one were serious about the getting the chords spliced, I'd think that you'd need to do something like I've shown below. But, then:

1) You'd need adequate depth for that girder and;

2) You'd likely be shot. Folks don't take kindly to "sport girders" from a gravity perspective.

 

[phamENG]

If you let it go, it's an incomplete load path unless you're willing to lay claim to using the sheathing for in plane flexure. I feel that the whole thing is "a bit much" but, at the same time, if we're going to do this at all, then surely we ought to do it as a complete load path.

Theoretical phamENG agrees with you. Practitioner phamENG is quietly snickering with the contractors to endear myself and win your work. (Just kidding...I'd never do that...(or would I?...))

I would say definitely for something of the scale that OP has shown. To an extent, this is a matter of choice. If you're willing to block ten truss spaces to transfer the load then, yeah, normal sheathing fastening probably gets the job done. If you plan to block two or three, which seems to be the norm, you'll likely need transfer diaphragm fastening.

Adding extra spaces simplifies design, installation, and inspection. Trying to keep track of roof nailing is a nightmare. It also doesn't get inspected by DOR here, just the city inspector. And they won't be looking for it. So design alternatives that maximize typical nailing patterns and use unique elements that are bold and noticeable are where I have the most success with implementation.

Nifty tech. Do the dimples mess with the ceiling drywall install at all?

Probably. But I doubt it would be more than nail heads on a regular CS.

The only way to transfer compression across intermittent gaps

Are we talking about leaving sections of blocking out or just the 1/8" gap at the end because the day laborer put the kerf on the wrong side of the cut line?

I don't recall having seen this detailed on anyone's drawings ever.

There's nothing unique here. As long as those loads are provided to the truss manufacturer, they'll show up as an uplift end reaction, and they'll get a hold down that is (hopefully) sized to resist it. The rest of that load path is no different than a standard uplift load path. Leaving it out or including it is hardly any worse or better than typical design.

I disagree...your low windows would all be smashed and your walls beat to heck.

These things have a lot of moving parts, so attempts to assign an actual stiffness value are just this side of futile, but I think it would be easy enough to approximate the lateral deflection in the wall below as a result of this setup and consider P-Delta on jambs. I suspect the results will be favorable in many cases.

What is it that we're preventing from being unzipped? The diaphragm starts its life off with a big crack along the joint between the high and low trusses. It's born "unzipped". Ditto for the offset top plates.

Nah, they don't start life being unzipped. They start life as a zipper, a weak point. That's the place the diaphragm is most likely to separate. So we need keep it shut.

 

[KootK]

Practitioner phamENG is quietly snickering with the contractors to endear myself and win your work.

Ha! Could be tough as I wouldn't normally be doing any of this stuff.

Adding extra spaces simplifies design, installation, and inspection.

No doubt. But, then:

1) The longer you make the thing the less effective it gets from a stiffness & effectiveness perspective.

2) A ten space blocked drag strut is exactly the kind of thing likely to lose me work in my market.

Are we talking about leaving sections of blocking out or just the 1/8" gap at the end because the day laborer put the kerf on the wrong side of the cut line?

Leaving sections of blocking out. Is that not what is normally meant by "intermittent blocking"? The 1/8" gaps also bother me a bit in terms of the compression slop that they'd add but, given all the other issues, I wouldn't sweat that.

There's nothing unique here.

I say that there is something unique here. In Florida, I'm sure that every truss would have serious hold downs and that adding some tie down capacity to the drag trusses would be opaque. Where I'm at, the situation would be more like I've shown below where I'd be attempting to add one lone hold down in a field of other trusses that don't have them.

As long as those loads are provided to the truss manufacturer, they'll show up as an uplift end reaction, and they'll get a hold down that is (hopefully) sized to resist it.

In 20+ yrs of practice, including several as a truss manufacturer and industry rep, I've not once ever seen anyone provide these uplift loads to a truss manufacturer. Not ever. Have you?

...but I think it would be easy enough to approximate the lateral deflection in the wall below as a result of this setup and consider P-Delta on jambs.

Not lateral deflection, in plane wall deflection.

Nah, they don't start life being unzipped. They start life as a zipper, a weak point. That's the place the diaphragm is most likely to separate. So we need keep it shut.

At the truss step, the roof sheathing will start its life off with a full width, full depth crack right down the line of the trusses. What further discontinuity is required in your mind to properly unzip that? Pulling apart of the sheathing nailing laterally?

 

[phamENG]

Leaving sections of blocking out. Is that not what is normally meant by "intermittent blocking"? The gaps also bother me a bit in terms of the compression slop that they'd add but, given all the other issues, I wouldn't sweat that.

Okay, we're on the same page about the blocking.

I say that there is something unique here. In Florida, I'm sure that every truss would have serious hold downs an adding some tie down capacity to the drag trusses would be opaque. Where I'm at, the situation would be more like I've shown below where I'd be attempting to add one lone hold down in a field of other trusses that don't have them.

Fair enough. Probably another case of a webinar for city inspectors, uplift load path became a big thing here about...10 years ago? Something like that. EVERYTHING at the roof gets tied down. It's just a matter of identifying the size/capacity of the connector. Below the roof, however...well let's just say the roof will now take the house with it.

In 20+ yrs of practice, including several as a truss manufacturer and industry rep, I've not once ever seen anyone provide these loads to a truss manufacturer. Not ever. Have you?

Yep. I've gotten some confused calls from truss manufacturers. One of my first designs after starting my solo operation I almost got burned because the truss manufacturer missed a line load on a floor truss system. I was only saved by an architectural design change that clued me in to construction progress, and I was able to redesign the roof structure along with the architectural change to move the load elsewhere. Since then I've been pretty particular about listing everything that might load a truss very clearly on the drawings.

Not lateral deflection, in plane deflection.

Tomato/Tomato (Point taken, though)

At the truss step, the roof sheathing will start its life off with a full width, full depth crack right down the line of the trusses. What further discontinuity is required in your mind to properly unzip that? Pulling apart of the sheathing nailing laterally?

More or less. By the logic of your first statement, the entire diaphragm is a system of cracks that may or may not be staggered throughout the entire diaphragm. We depend on the nailing to hold it all together. In this case, the discontinuity (and especially the double discontinuity of vertical and horizontal offsets) creates a concentration of loading that needs to be transferred from one diaphragm to the other. That concentration is what will initiate the separation.

In any case, I think we agree that if there's a problem with the diaphragm, this is where it will be, right? So it's important to either a) move the load away from here (facing 3-sided lateral systems with a tie designed for deflection compatibility) or b) reinforce it to prevent the failure at this location.

 

[KootK]

Yep...Since then I've been pretty particular about listing everything that might load a truss very clearly on the drawings.

I wasn't asking "have you specified some unconventional truss loads?". That's no big deal. Rather, what I was asking was whether or not you've specified truss uplift forces originating from the demand for overturning resistance in shear blocking systems? If you have then, by all means, please post some snips of the areas in plan and how you've handled the specification of the loads. That way I can have them framed and hung next to the pictures that I have of the unicorn that I ride into work and the sasquatch that sneaks into my hot tub during the winter months.

 

[phamENG]

Touche. I haven't needed to specify that particular load. But if it comes up, I'll send you a copy of the drawing. The architects I'm working with are doing a lot of hip and flat roofs, so I'm not getting steps like this very often. And the one that does have it on my desk now is a stick framed roof, so no loads to spec to a truss engineer.

Where you are, wouldn't that be a Yeti?

 

[KootK]

In any case, I think we agree that if there's a problem with the diaphragm, this is where it will be, right?

Yes, however I don't believe that there is an actual problem in need of solving in most cases. Some time ago we were discussing something similar here and, if I recall correctly, XR250 challenged anyone to produce an example a real world diaphragm failure. What we got back was:

1) A whole lot of tumbleweeds blowing by;

2) Some pics from diaphragm testing done by Simpson, no doubt at the lab down the road from where they mass produce coil straps and;

3) A single pic of what may have been some extreme hurricane damage that was still only localized to the reentrant corner and nothing that I would classify as "unzipping".

So it's important to either a) move the load away from here (facing 3-sided lateral systems with a tie designed for deflection compatibility) or b) reinforce it to prevent the failure at this location.

Or...

c) Recognize that, in many situations, the walls running parallel to the top plate are performing the "chord" function rather than the top plates.

Tomato/Tomato

Not tomato/tomato but, rather a paradigm shift in the conceptual understanding the likely mechanism of resistance.

 

[KootK]

Where you are, wouldn't that be a Yeti?

It would. I was playing to a general audience form a range of latitudes that might not possess your cultural sensitives.

 

[XR250]

Some pics from diaphragm testing done by Simpson, no doubt at the lab down the road from where they mass produce coil straps

That made me laugh a lot.