AISC 15th Spec - F7.3 (WLB for Box Beams) Doesn't make sense 2

[b-clark]

We design a lot of box beams where I work, so we use F7 a lot. In the AISC 15th, they added a brand new section for slender webs that wasn't there before.

It's broken down into two sub-sections:
(1) Compression flange yielding
(2) Compression flange local buckling

I'm not seeing how either of those failure modes are relevant to web local buckling, nor the equations that are given in this section. Has anyone else looked over this new section? See below.

 

[bootlegend]

I don't have a reference in front of me but I think some slender webs aren't enough to brace the compression flange from buckling, meaning the flange buckles into the softened area next to the compression flange. It looks different than regular web buckling.

 

[JoshPlumSE]

It's taken me a little while to understand it. But, reading through the commentary, I gathered that this section was based on section F5 (I shaped members with slender webs), but with a doubling of the aw value to account for their being two webs. ​

F7.3c) Applies to sections with slender webs.
Subsection (1) is correctly related to compression flange yielding. Just it applies an Rpg (a bending reduction factor related to plate girder that really only comes up when we've got a slender web. Right?

Subsection (2) then covers the other scenario where you have BOTH a slender web and a slender flange.


It's definitely confusing, but after taking some time..... I think it makes sense. I'd like to experiment with numbers or graphs about what happens to the capacity and Rpg factors as the slenderness increases or decreases. But, that would be just to see the trend as it heads towards and away from those two slenderness points.

 

[271828]

Rpg is the ancient Basler way of dealing with so-called "bend buckling." If you have a copy of Salmon & Johnson, this is covered under Plate Girders. Rpg has some serious staying power. LOL

The basic idea is that the web buckles in the compression region before there is yielding anywhere in the section. The resulting stress distribution at the ultimate strength is nonlinear through the web. There is stress redistribution because the compression portion of the web becomes more flexible and more stress is thrown to the compression flange. All of this is taken into account in Rpg. It's the same formulation as that in F5 for regular plate girders.

 

[b-clark]

These explanations make sense after some thought. Thanks to all of you for your help.