KL/r > 200 (Weak Axis)

[Samwise Gamgee]

I have a moment frames which will cause axial load in the beams. The software flags that KL/r > 200 (for weak axis buckling).

Question : As deck is perpendicular to the beam, can we consider the deck would brace the beam in Y-axis and effective length in Y-direction won't be the full length of the beam ? My plan was to use the flexural torsional buckling check (E4.d and calculate the critical buckling stress for lateral bracing offset from shear center). Check what is the equivalent length to achieve this and use that as the effective length in Y-direction. Is that reasonable ?

 

[KootK]

I think that you're generally heading in the right direction here.

As deck is perpendicular to the beam, can we consider the deck would brace the beam in Y-axis and effective length in Y-direction won't be the full length of the beam ?

Yes, true weak axis buckling will be precluded and, rather, take the form of constrained axis torsional buckling. In routine design, we typically use weak axis buckling as a simplification to represent constrained axis buckling which is more complex to evaluate. Like this:

a) Check weak axis buckling based on the weak axis, unbraced length.

b) Add bracing to the beam bottom flange if we don't like the answer from [a].

This is precisely what I would do in this situation.

Check what is the equivalent length to achieve this and use that as the effective length in Y-direction. Is that reasonable ?

I'm not sure that part makes sense. The software will just take that equivalent length and then still run it through the weak axis buckling check which may still report a failure. If you're going the constrained axis buckling route, I'd handle it like this:

1) Check it by hand / spreadsheet.

2) Set your weak axis unbraced length to zero, acknowledging that you've addressed this outside of your software.

 

[phamENG]

The KL/r > 200 is not a mandatory check. Read up about it in the AISC commentary.

That said, it's usually best not to eschew the wisdom of the ancients on a whim. If you're going to go higher, make sure you have a full understanding of why it was done and what you're doing in your design to ensure the design is adequate. (If that involves something to trick a piece of software into calling it good, it's probably a bad idea.)

But when in doubt, follow KootK's advice - add bottom flange bracing to get KL/r solidly below 200.

 

[KootK]

With moment frame beams, you often require bottom flange bracing for the negative end moments anyhow. The constrained axis voodoo tends to yield tangible benefits primarily at the collectors that are not integral with the moment frames.

 

[lexpatrie]

You have to have a diaphragm of some sort, or a collector, to get the load into the moment frame beam. You may need discrete bracing for the negative moments in the beam, however, so you might be able to double count the brace if you do it reasonably.

I would expect most engineers would consider the deck to brace the beam (axially) without much thought. If you're looking for something more technical, perhaps this Helwig paper would be of use. Lateral Bracing of Bridge Deck Girders by Permanent Metal Deck Forms, Helwig, Egilmez, Jetann, 2005.

There's also the Fisher paper you might find of use.
Bracing of Steel Beams, Trusses, and Joist Girders Using Open Web Steel Joists - Fisher, Eng. Jour. 1Q2006

 

[PMR06]

AISC discusses this in the "Steel Design After College" series. Search "torsional buckling" and "Kz" in the pdf. I'm not sure which video it is in. The whole series is worth a listen if you haven't already.

https://www.aisc.org/education/continuingeducation/education-archives/steel-design-after-college/