Determining if a beam connection will torsionally brace a steel column? 1

[abusementpark]

I sat in on an AISC seminar recently and recall the presenter discussing that a steel column needs to be "torsionally" restrained at any brace point to preclude any torsional buckling modes. So, I got to thinking about typical practices.

We often assume that steel beams framing into a column in each direction adequately brace the column for purposes of compression design. The beams, if adequately connected to the LFRS, certainly provide lateral bracing, but what about torsional bracing? Wouldn't this depend on what type of beam to column connection is provided? What about single plate shear connections? Does AISC provide any guidance on this matter?

Thanks.

 

[DST148]

AISC 360-10, Appendix 6 - Stability bracing for columns and beams - This Chapter gives brace stiffness requirements.
Bracing systems that are attached only near the centroid of the member are generally ineffective in controlling twist.

 

[hokie66]

Typical shear connections wouldn't provide much torsional resistance, but why is it required? Did this presenter explain the reasoning behind his statement?

 

[abusementpark]

AISC 360-10, Appendix 6 - Stability bracing for columns and beams - This Chapter gives brace stiffness requirements.
Bracing systems that are attached only near the centroid of the member are generally ineffective in controlling twist.

I'm looking at the same section in AISC 360-05. I see requirements for axial stiffness of braces for restricting translation, but I do not see rotational stiffness requirements for preventing twist.

Typical shear connections wouldn't provide much torsional resistance, but why is it required? Did this presenter explain the reasoning behind his statement?

He worked an example where he checked the capacity of a 20' long W14x38 column that was braced at each end laterally and torsionally, and also braced at mid-span in the weak-axis laterally, but not torsionally. He checked three buckling failures based on three effective lengths:

KLx = 20' (strong-axis flexural buckling using Section E3)
KLy = 10' (weak-axis flexural buckling using Section E3)
KLz = 20' (torisional or flexural-torsional buckling using section E4)

 

[BAretired]

A base plate with four anchor bolts would constitute a torsional brace at the bottom. How would the column be torsionally braced at the top? And if it is not torsionally braced at the top, what is KLz, 40'?

BA

 

[DST148]

@abusementpark - See if the last paragraph of the commentary given on page 16.1-503 of AISC 360-10 would help. Also try thread507-186501. Some of the posters on that thread have attended AISC seminar including the one by Yura - the author whose recommendations have been adopted by AISC for beam and column bracing requirements.

 

[abusementpark]

Ok, I recently spoke with somebody at the AISC Steel Solutions Center. He said that determining whether the members framing into a column are adequate for torsional bracing is just a matter of engineering judgement. There is no good method to quantify the requirements for torsional bracing. He said that the standard framing connections of beams into columns are adequate if the beam is attached a structural diaphragm, even the single tab plate.

However, he said there is some question as to whether or not extended end plate connections or connections involving deeply coped members are adequate as torsional bracing and that research is currently being conducted associated with this issue.