Interesting Break in Truss 2

Full-length sister. The demand on the webs is pretty low so it don't take much to repair it.

The low end plate connection looks shot to me and sistering alone wouldn't address that. I would be inclined to:

a) Cut the plate away where it would have originally overlapped with the web to be replaced. You'll never salvage it so it's best to just get it out of the way cleanly.

b) Put a new web in there as snugly as possible acknowledging that:

i) It's likely a compression web under gravity load and;

ii) Full contact at each end may not be realistic and is unnecessary of the connection is designed to suit.

c) Glue and screw plywood gusset connections at each end that bypass the existing plates altogether for all of the members coming into the joint.

fixmytruss.com has some samples

Thank you KootK, your comment echoes pretty much exactly what I was outlining in my original post.

Ismfse, thank you too, didn't know of that resource.

XR250, I appreciate your comment but sistering in this case doesn't seem very achievable considering that the original web is in such terrible shape.

XR, the other thing I noticed. What are all the nails coming thru the OSB securing? Is there something on the roof located over this truss?

Shingle nails presumably.

I think XR250 meant sistering the full length of the broken web member extending onto the top and bottom chords.

Nothing on the roof other than ordinary shingles. House isn't that old (25 to 30 years) so maybe at most one extra layer of shingles if the roof replacement was done on the cheap without lifting the old.

SW, yes, you're right. That would make a lot more sense now that I reread his comment. Still though, I am not a fan of leaving the broken web in that attic. "Feels" cheap... yeah yeah, I know, engineering isn't about feelings... but still.

Probably broken when it was lifted into place and the builder didn't care. It's right at the attic opening, so a big FU to the original buyer.

Is that a nail through the horizontal board that is dead center on the broken fragment? If it was nailed before the break happened the nail should be better centered on where the web should be, 2 inches or so to the left per the 4th picture.

SWComposites said:

I think XR250 meant sistering the full length of the broken web member extending onto the top and bottom chords.

Click to expand...

Such a scheme is often infeasible because it would induce a tension perpendicular to grain failure in the truss chords from the axial coming out of the web.

This is speculative but one explanation for that bizarre split is that:

1) the web tried to buckle under compression.

2) the board running perpendicular to the web attempted to brace it.

3) the brace force in the nails pulled the web apart in a version of tension perpendicular to gain failure.

This would be somewhat consistent with lsmfse's overload theory.

If it buckled under compression then all the webs the horizontal boards are attached to should still be 2 inches to the right and out of alignment with the plane of the truss. It would also have bent the plate.

There is a large offset at the base of the board, like someone dropped an anvil on it which would not happen from buckling. They tried to bash it back into place and nailed the horizontal board to keep the lower half in place.

could the roofers have piled up a bunch of shingles at the location of the broken web? is it located at a spot where a truck with conveyor could have placed the shingle bundles?

Generally speaking, yes, it has clear access from the driveway/street and this location could have had a bundle of shingles on it in the past.

I don't think we will ever know what caused this. I am not sold on the idea of this being the sole result of buckling under compression, as per KootK, because the plates are in perfect condition with no sign of deformation. Perhaps it was a combination of factors - poor wood with existing splits, construction loads or loads during reroofing, etc.

3DDave said:

If it buckled under compression then all the webs the horizontal boards are attached to should still be 2 inches to the right and out of alignment with the plane of the truss.

Click to expand...

I disagree. All of those webs, and the transverse boards, would have stayed put. It would only be the part of the one web that moved laterally when it ripped apart in tension perpendicular to grain.

3DDave said:

It would also have bent the plate.

Click to expand...

SSE said:

I am not sold on the idea of this being the sole result of buckling under compression, as per KootK, because the plates are in perfect condition with no sign of deformation

Click to expand...

I disagree with the plate bending business as well. I feel that you guys are spuriously imagining a version of web buckling where the web would move a bunch and, therefore, generate significant rotations at the ends of the web. A tension perpendicular to grain failure is a fracture mechanics rupture kind of thing and would require very little movement to generate. That, particularly in a case such as this where the split was surely exacerbated by some existing stress concentrator in the member to begin with.

You would also be surprised at how much the end of a truss web can rotate, taking the chord along with it, without distressing the plates visibly. It's not as though the roof or ceiling sheathing is going to offer any significant rotational restraint. When I worked for truss manufactures, I would handle these things every way imaginable, including tossing them around the yard with a forklift. They are altogether flex-i-licious out of plane.

Still, the buckling hypothesis is just one of may possible causes that might explain the end result.

Those webs do actually look pretty long to me for 2x3's in compression. But, then, all my truss work has been in snow country.

Thank you KootK. Your comment certainly drives discussion and gives me something to think about. I appreciate your input.

P.S. This is snow country. Ottawa, ON.

You're most welcome SSE. These forensic things are always a fun ride.

@SSE: do you see any indication of deliberate bracing on any of the webs of any of the trusses?

@KootK, Your question confuses me slightly. Bracing - Yes, plenty. In the pictures you can see plenty of bracing as is. It was present on both sides of the king posts.

Edit: I will add that the bracing was symmetrical on both sides of the roof but was only applied to the first web members (closest to the king posts) in each truss.