Effective Length of Beam with Joists

[MSOE2010]

Suppose you have a wide flange beam with joists framing into it, bearing on the top flange at 4'-0" o.c. and the beam has a compressive axial load. For the purpose of this question ignore any dead load, live load, self-weight, etc. The only load is the axial load on the beam.

Would you say that beam effective length is 4'-0" in the weak axis for the axial load?

From my point of view: the joists only brace the top flange so to say the effective length is only 4'-0" is incorrect. My thinking is that the k factor would be something similar to the k factor for plate buckling. A simple span plate with a simple support on one edge and free on the other edge (k = 0.425 I believe).

 

[Lion06]

I would say that the buckling length of the beam is the actual length. I would use k=1 with the actual length.

I don't think that it's analogous to plate buckling just because, well.... a WF isn't a plate and doesn't behave the same way because of having the web and other flange.

You also can't say that the joists brace it for weak axis buckling because the entire section isn't braced.

 

[asixth]

There is a concurrent thread running on the Australia codes forum with references that deal with classification for restraining elements. It may be worth a read for your own interest.

thread744-269173

I would be inclined to say that the floor joist does provide lateral restraint to the W-beam despite the fact they are cleated from the flange of the section.

A similar application to that described which I design regularly are braced roof frames where W-beams do experience large compression due to bracing forces. Generally the purlins which cleat from the rafter are taken as to provide lateral restraint for the W-beam buckling about it's weak axis.

The reference I will use for these statements is a design guide published by the Australian Steel Institute "Design of Portal Frame Buildings" by Woodcock, Kitipornchai and Bradford.

http://www.eabooks.com.au/epages/eab.sf/?ObjectPath=/Shops/eabooks/Products/9780909945916

I have attached the paragraph that relates to weak axis buckling.

I believe floor joists would by more capable to restrain the bearer because the floor sheet would behave more rigidily than relatively thin roof sheet.

 

[CTSeng]

see attached

 

[paddingtongreen]

If this is new, I would put bottom chord extenders on the joists on one side only, and tell the joist manufactures what axial loads to expect in addition to the vertical.

The case is interesting, even on one flange, the joists prevent complete column buckling. What is the column buckling mode if lateral displacement of one flange is prevented?

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Lion06]

asixth-

Would you say that a column with a wind girt attached to only the exterior flange braces the column for weak axis buckling? I wouldn't.

 

[Lion06]

paddington-

I remember one of our senior engineers saying that you can't count on bracing at one flange only. If you do, it will manifest itself as a torsional buckling problem in the column. This apparently came from a Yura seminar. I wish I could get my hands on the notes.

 

[WillisV]

StructuralEIT - I would say the bracing on the top flange would restrain the "beam" (column) from pure buckling in the weak axis, but it would not restrain it for torsional buckling. Generally the strength you get from torsional buckling is very similar to weak axis buckling of the member for the entire length, so its easy enough to just say its unbraced for weak axis buckling and not bother with it any more.

 

[asixth]

SEIT

Based on what was written in the Woodcock book and provided that I could prove that the girts could satisfy the requirements for lateral restraints than I would. I wouldn't agree with the book saying that thin metal sheet can be considered a "rigid" diaphragm. I think it is unethical to rely on light-gauge metal sheeting to transmit compressive loads in diaphragm action.

However for the column example you describe, I have found in my experience that the size of the columns for single-story buildings is usually governed by limiting the drift and side-sway moments. When the calculations are done on the columns for compressive forces I have found them to have capacity reserved even if a ky=2.2 was used.

Can you please post the notes if you can find them.

 

[paddingtongreen]

SEIT,
Torsional buckling usually requires the section to rotate about it's approximate polar axis, both flanges must displace laterally in opposite directions, that is why the joists only have to prevent lateral movement of the top flange for beams.

In this case, the section would have to rotate about an axis running through the restrained flange. That's what I find interesting.

In practice, I would probably use the joist spacing and a k of 2.0 or more if it would work.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Lion06]

See page 544 and 545 of the Guide to the Stability Design Criteria for Metal Buildings.

http://books.google.com/books?id=DD...age&q=column bracing one flange yura&f=false

This talks about the torsional failure mode, and also provides a detail (Fig. 12-10 a) that I've seen used (It was used on the building where my daughter does gymnastics - it's a PEMB).

 

[Gumpmaster]

StructuralEIT, I was just going to reference the Galambos version of this book (Dated 1998). The text is pretty clear, there's a load limit after which torsional bracing must be provided. If no torsional bracing is provided, the entire member length must be considered.

 

[paddingtongreen]

SEIT Thank you. Good explanation there in that guide.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[apsix]

The "Design of Portal Frame Buildings" book referenced by asixth above has similar recommendations:
For lightly loaded portal columns the effective length is the girt spacing, but if the columns are also supporting a travelling crane, the effective length is based on the spacing of restraints to the inner flange (assuming that is greater).

BTW, in the context of restraint provided to purlins/girts the cladding is considered a rigid diaphragm, that's why the design of those members assumes the cladding side flange is fully restrained.

 

[csd72]

I have tended to design these for twice the purlin spacing under axial load to allow for the inherent flexibility.

 

[msquared48]

If the WF was a flush beam, of the same depth as the joists, with wood infill between the flanges, then I would say 4 feet or whatever the joist spacing was for the weak axis bracing. The strong axis would still be the span length.

Anything different would be the full length.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask