Unbraced Lengths

[jbuening]

I'm modeling a steel frame structure (pipe bridge) and this is both a question related to RISA as well as structural engineering. Attached is a screenshot of my model. In RISA there is four different unbraced lengths you can input for the member info, Lb yy, Lb zz, Lcomp top, and Lcomp bot. I plan on using bolted angle clips on the ends of each bracing member (like M21, M22, etc) to attach to the main girders, so they will have moment releases as shown. The braces will not have a full depth stiffener/connection plate. For unbraced lengths Lb yy and Lb zz, do you choose the largest distance between the braces (say for M10)? I assume the Lcomp top and bottom would be the full span length, since the angle clips don't technically brace the compression flange?

 

[JoshPlumSE]

Lbyy would be the weak axis buckling of the girder. Therefore, it could be set to "segment" as the bracing members (M21, M22, etc) would restrain the lateral direction or lateral buckling of the member. Lbzz is the strong axis buckling of the member and should be left blank (IMO) because there is no bracing for the strong direction.

Lcomp-top and Lcomp-bot are top flange and bottom flange bracing for beam type buckling. This isn't as straight forward. AISC 14th edition appendix section 6.3 gives some guidance that you might want to review.

If the members share the same top of steel elevation, then I have normally said that they are effective for resisting beam buckling. Meaning that they would be a brace for Lcomp-top. But, that I would not consider Lcomp-Bot to be braced unless the clip angle connection extended most of the depth of the girder.

The Lcomp stuff is a bit of an engineering judgment call, though. The commentary to section 6.3 of the AISC appendix says that the brace must control "twist" of the beam if it is to be considered a brace point. Does the clip angle really inhibit the twist of the main member? It might depend on the depth of the clip angle. Or, on whether there is any decking at the top flanges.

 

[gmailman1]

Josh,
Thanks for the reply. It appears the Lbyy and zz are more related to column type members and the Lcomp is more for beams. Correct? The members do indeed share the same top of steel elevation. For the main beams it appears I could set the Lbyy and Lcomp-top to "segment" and leave others blank. Do you agree? I have since made the cantilever beams separate members since the Lcomp will be different (comp bottom flange is now only at the columns and no Lcomp-top)

 

[jbuening]

After I posted this I realized I should have read the help file in RISA. Thanks for the contribution Josh. I don't have an answer for gmailman1 though, maybe you can assist.

 

[JoshPlumSE]

GmailMan -

Yes, Lbyy and Lbzz are related to axial force capacity, so that usually means columns. The LComp values are related to flexural capacity, which usually means beams.

For your case, I would tend to agree with the Lbyy and Lcomp-top being set to "segment" and leaving the others blank.