Can a connecting member perpendicular to a beam provide twist restraint to the beam? 1

[FLR84]

(picture below)

Hi everyone,

Specific example --

I have a portal frame with a knee brace. The knee brace is coming in at quite an angle to the rafter. Could the knee brace provide twist restraint to the rafter, given the web of the knee brace is connected to the web of the rafter using three bolts?

Obviously it does provide SOME level of restraint, but would it be enough to prevent lateral torsional buckling in the rafter if needed? What about if the angle of the knee brace was increased even further so that it was almost perpendicular to the rafter?

Thanks!

 

[JoshPlumSE]

I don't believe this would be considered either torsional bracing or lateral bracing. In order to do so, the bracing must control the relative displacement of the top and bottom flanges. I don't see how this provides that type of restraint. Refer to AISC appendix 6 and the commentary for Appendix 6 for more information.

However, you will likely get some benefit based on the Cb factor based on the moment variation over the length of the beam.

 

[FLR84]

So in the code here it says that the lateral restraining member needs to be able to resist 2.5% of the force in the flange that its restraining. I did some modelling and calculations and the knee brace actually just passes in being able to resist the force (modelled as a cantilever with the point load on the end).

So I think in this case, unless I'm mistaken, it will actually help restrain the flange enough to count.

 

[BAretired]

The frame members appear to be channels whereas the bracing member appears to be a double (back to back)channel . In order to provide torsional bracing to the horizontal member, the vertical member is being asked to contribute by supplying moment about its weak axis. I suspect the calculation required to determine whether or not the horizontal member is torsionally braced would be complicated. The brace is not a cantilever unless it has a fixed end.

BA

 

[KootK]

I did some modelling and calculations and the knee brace actually just passes in being able to resist the force (modelled as a cantilever with the point load on the end).

I'll buy that the brace could be an LTB brace under the right circumstances but:

a) That's a pretty unconventional choice and, as BA said, difficult to evaluate properly.

b) Your cantilever evaluation should consider both the lateral column flexibility that BA mentioned AND the torsional column flexibility.

c) Keep in mind that there may be a significant axial load in the brace and that will reduce the braces apparent flexural stiffness.

d) Your brace ties in pretty close to your column which would normally count as LTB bracing. Surely it's easier just to design the rafter to span that extra little bit?

 

[hoshang]

Hi
Why not using a member between the rafter and the HSS?

 

[JoshPlumSE]

So in the code here it says that the lateral restraining member needs to be able to resist 2.5% of the force in the flange that its restraining. I did some modelling and calculations and the knee brace actually just passes in being able to resist the force (modelled as a cantilever with the point load on the end).

For columns there is both a strength and stiffness requirement for bracing. I'm not sure if there is a true stiffness requirement for LTB bracing (and I'm too lazy to look it up now). But, if there is (or if there should be) then I don't think this connection would satisfy it.

 

[Liam Nesson]

For haunched connections at the eaves of a portal frame you usually have restraints from the purlin at standard locations per attached screenshot. Under gravity load cases (and others) both the rafter bottom flange and inside flange of col are in compression and might buckle. Not sure if bracing from a col flange that may buckle is going to brace the rafter btm flange.Check out a design guide specifically for portal frame design.

 

[Tomfh]

I say yes, provided it is designed as beam spanning between purlins and column, and assuming the columns and purlins are fine.

As noted above it's unconventional. Normal practice is a fly brace in that position, if required.