Bracing of wood members 1

[EngDM]

For rectangular wood members in bending, I am curious if the lateral bracing for unbraced length is required to be at the compression flange. An application would be for a wood roof joist, maybe a 2x12, for uplift. Consider in this case that the joists are only restrained by a T&G deck at the top, and no blocking or cross bracing between joists.

I'm wondering what value is used for Le in the Cb equation (CSA O86) for uplift. Table 2.7 of the Wood Design Manual 2020 has 'a' being the distance between compression edge purlin bracing, but it doesn't indicate or hint at an uplift case.

 

[KootK]

I would have to think it would be the compression edge.

 

[jayrod12]

Agreed with KootK's answer. There should be blocking if there's going to be uplift. Otherwise you need to use the full length if unbraced.

 

[XR250]

I suppose it depends on the depth. I believe 2x4 wall studs only need sheathing on one side.
My guess is that you could look at 99.9% of beach houses and this has not been addressed.

 

[SE2607]

I suppose it depends on the depth. I believe 2x4 wall studs only need sheathing on one side.
My guess is that you could look at 99.9% of beach houses and this has not been addressed.

According to the NDS spec, you don't have to worry about it if l/d<4 which is a 2x6. If it's deeper than that, you have to consider distance between lateral supports. The reason "99.9%" of the beach houses have not failed is that they have not been subjected to the design loads. Lack of failures is no reason to ignore the code, but it might be a reason to modify the code, which, based on my experience dealing with the 200 pound guardrail load for residential construction, is a total waste of time.

 

[EngDM]

I suppose LTB (or whatever the wood equivalent is, not sure if a wood rectangle fails torsionally) is taken account of in the KL factor, it just so happens that you can get some pretty large unbraced lengths before needing to use anything other than unity for KL.

 

[Celt83]

I suppose LTB (or whatever the wood equivalent is, not sure if a wood rectangle fails torsionally) is taken account of in the KL factor, it just so happens that you can get some pretty large unbraced lengths before needing to use anything other than unity for KL.

Take a look at TR14: https://awc.org/wp-content/uploads/2021/12/AWC-TR14-0312.pdf

 

[XR250]

99.9%" of the beach houses have not failed is that they have not been subjected to the design loads. Lack of failures is no reason to ignore the code, but it might be a reason to modify the code, which, based on my experience dealing with the 200 pound guardrail load for residential construction, is a total waste of time.

Fair, but no one is dying because of an LTB uplift failure of a rafter. The roof will likely blow off prior to that.
Guardrails are so sketch to begin with I would not want them to drop the loads.

 

[KootK]

not sure if a wood rectangle fails torsionally

Yeah, rectangles fail in torsional buckling. In the extreme, high aspect ratio rectangles eventually become "plates in bending". Rectangles are pretty great torsionally at low to moderate aspect ratios. You'll see this in action in the generous LTB limits on HSS, precast spandrel beams etc.

Wood in any kind of torsion is a nebulous thing. Yeah, sure, St. Venant on a rectangle is pretty straight forward. But, then, the limits on checking for wood beams are very liberal. A lot of timber beams in service really look more like two stacked beams torsionally.

 

[SE2607]

Guardrails are so sketch to begin with...

Not if they are designed properly, and, to resist 200 lbs at a height > 3.5 feet, as you know, can become quite onerous. If it were an impact load instead of a live load, I would be ok with that.