AISC 360-10 Buckling Failure Clarification 1

[Rocks1]

Can anybody provide me with a clear narrative understanding of the difference between 360-10 CHAPTER E COMPRESSION E.4 "Torsional and Flexural Torsional Buckling" and 360-10 CHAPTER F - DESIGN OF MEMBERS FOR FLECTURE "Lateral Torsional Buckling"? What is the difference and is there redundancy? What would be an example of the two types of failure and what would initiate each?

 

[JoshPlumSE]

The first one (E.4 Flexural Torsional Bucklilng) is due primarily to the presence of axial loading. The member buckles not perfectly about the strong or weak axis, but includes some twist / torsion to it.

The 2nd one (F - Lateral - Torsional Buckling) is due solely to the presence of flexure, without any axial loading. I look at it like this. The compression flange of a wide flange beam can be viewed kind of like a column in axial compression. It's restrained from buckling about it's weak axis by the presence of the web. So, then wants to buckle about it's strong axis. But, it's restrained by the web again. But, the web is kind of weak with it's restraint. Therefore, it buckles laterally with some torsion movement induced by the web restraint.

 

[Rocks1]

Thanks. Makes Sense

 

[KootK]

The kindergarten level answer would be:

A) Lateral torsional buckling (LTB) is an instability response to a flexural demand and;

B) Torsional / Flexural Torsional Buckling (FTB) is an instability response to a compressive, axial demand;

C) These are largely separate phenomenon and, therefore, do not represent much redundancy for the case of beam columns.

The black/brown belt level answer requires drilling down to the fundamentals and a considerably more nuanced understanding. For that, we'll have to go for a little stroll...

1) Consider that one of the fundamental ways to approach structural stability is to view it as a minimization of potential energy. Things tend to deform in ways that exacerbate deformation and, consequently, move applied loads closer to mother earth.

2) If a member loaded flexurally about its strong axis is able to flop over such that it instead resists flexure about it's weak axis, that will exacerbate deflection, move the applied load closer to the earth, and minimize potential energy. This is LTB.

3) Recognize that, for the most part, LTB deformation would not do much to reduce potential energy in a member loaded only with axial compression. Most columns are governed by weak axis buckling to begin with so, in flopping over, nothing is really accomplished that would exacerbate axial deformation and move the applied load closer to the earth. This is why LTB doesn't really apply to pure columns. I will add some nuance to this in what follows.

4) So how then does torsional buckling of a column move the applied load closer to the earth and reduce potential energy? I find it most instructive to imagine a stocky WF column for this. As such a column twists, it effectively starts to morph into a pair of WT's, each undergoing what is essentially flexural bucking in opposing directions. This is, in effect, buckling by way of sectional warping. As each faux WT assumes curvature, the applied load moves closer to the ground in what quite similar to ordinary flexural column buckling. So, somewhat contrary to #3, you can twist-buckle and axial loaded column. There are important differences though:

a) The strain energy that needs to be induced to torsionally buckle a compression member is often in excess of an order of magnitude higher than the energy that needs to be induced to flexurally buckle that same member. This is why it takes a pretty extreme cross section to initiate torsional buckling. Things like cruciforms, plates, and high Ix/Iy wide flanges and channels.

b) As one can imagine, it really does take a crap ton of twist to move an axial load appreciably closer to the earth in a pure compression member. This, in contrast to LTB of a flexurally loaded beam, where initial buckling quickly sets a member off down a path of rapidly diminishing flexural stiffness and increased deflection at the point of load application.