Plate Girder Design with AISC 13th Edition

[Shaylon]

Hello all,

Does anyone know why the plate girder design chapter (Chapter E) in the ASD 9th edition was omitted in the 13th edition? With that being said, how would all of you go about designing a tapered plate girder with the current edition? Thanks.

Shaylon

 

[271828]

See the following sections:

E7 for computing Pn of a slender section. You'll also need E4 if flexural-torsional or torsional buckling applies.

User note at first of Ch. F for selecting the correct section for Mn. You can end up in F3, F4, or F5 depending on what's slender--the flange and/or web.

See G2 and G3 for shear strength without and with tension field action.

 

[271828]

Tapered plate girders will be covered in the upcoming AISC Design Guide 25, if I have the story correct. If you need help faster than that, I'd recommend to write to the AISC Steel Solutions Center. They might have something helpful.

 

[slickdeals]

DG 25 is out already. Check out AISC.org.

 

[firai]

One of the main features of the AISC 360-05 was that it unified the provisions for rolled and built-up members, so the only chapter you need to look at when designing for flexure is F.

With that said, you should refer to DG 25 for tapered members as others mentioned, which just came out. I read it while it was still in draft (Dr. White was my grad steel professor), and the moral of the story is mainly to use the direct analysis method (which is also advocated in 360-10).

Structural Design Engineer
New York, NY

 

[ToadJones]

DAM for designing a plate girder?
I'm confused

 

[ToadJones]

eh ?

 

[peal724]

great reference

AISC Engineering Journal
First Quarter 2009 Volume 46, No.1

"Design Aids for Built-Up I-Shaped Beams with Slender Webs" by Paul P. Nasados, Jr.

 

[271828]

DAM for designing a plate girder?

The DG's main focus is tapered built-up members as used by MBMA companies. Most of those are in moment frames that can either be designed using the effective length method or direct analysis method. The DM is very easy for those companies to implement with their software.

 

[ToadJones]

I guess I was curious because regardless of the method of analysis, you are still designing the girder for forces/moments resulting from the analysis.

 

[firai]

ToadJones, I mentioned DM because in my humble opinion, figuring out an accurate-to-conservative force/moment to design the tapered member for and correctly accounting for the stiffness in the analysis is more difficult than actually designing the plate girder.

I apologize if my comment caused some confusion.

Structural Design Engineer
New York, NY

 

[271828]

I don't quite understand that,firai. They have had their software accurately deal with tapered member analysis for many years.

I think the main reason they use the DM is to avoid dealing with Kx. It's easy for them to push the initial geometry to H/500 to one side, reduce AE and EI and run the analysis.

 

[jrw501]

I haven't looked at AISC DG25, but the WRC Bulletin (#173) "Design of Tapered Members" (G.C. Lee, M.L. Morrell and R.L. Ketter, 1972) provides some insight and is used in conjunction with AISC 3rd Ed. App. F3. It basically modifies the equations for prismatic members with taper factors.

 

[271828]

The DG25 approach is much more consistent with the 2005 and 2010 Specifications.

However, I admit that, even as a specialist in this sort of thing, I find DG25's formulations to be somewhat baffling and counterintuitive. The average design engineer who's very familiar with the AISC Spec. will find it frustrating.

Now that I think of it, the Lee et al. documents might be a lot easier for a typical design engineer to pick up and use. Not sure if one needs the Lee et al. documents. Just look at Appendix F of the 1993 AISC Specification.

 

[JoshPlumSE]

The problem (as I understand it) with the Lee et al. approach was that it was actually supposed to be relatively limited. Look at the old AISC specs. The rate of taper was limited. Flanges had to be of equal area, et cetera....

I believe most of the metal building guys had programs that extended their software beyond those limitations. But, what criteria did they use to do that? There was never a clear specification on how to do this. Therefore, every company probably did this a little differently. At least now, we should get to the point where there is some consistent guidance on how these members should be designed.

 

[jrw501]

@JoshPlum, I think that is correct. Lee et al.'s approach was limited to (what I'd guess) is the most common case where the taper is linear (and I think there's a limit to how dramatic the taper can be), flanges are equal sizes and there's at least one axis of symmetry.

 

[271828]

But, what criteria did they use to do that?

In my experience, they jettisoned the Lee et al. work and adapted the 1989 Specification themselves. Something like the following:

Overall analysis: First-order using stiffness matrices for tapered members.

KL/ry buckling: doesn't change significantly due to taper, so use the Spec. as is.

KL/rx buckling: use rx at a magic distance from the smaller end (something like 0.57L IIRC) and treat the member as prismatic. Compute Kx using analytical formulas or eigenvalue analysis within their program.

Accounting for slenderness in the axial strength calcs: evaluate at stations.

Flexural Yielding or FLB: this is a station by station check, so no mods reqd.

Flexural LTB: Evaluate at closely spaced stations and use prismatic member equations.

Interaction equations: station by station check.

Shear check: station by station check, so no mods required. I think most did/do subtract the vertical component of flange force per Blodgett's method, although this method is mentioned as unverified in DG25.

Sorry for typos and poor grammar--in a hurry.