Beam Unbraced Length

[marinaman]

Take a look at the attached simple beam layout.

If a person is working thru simple beam design at beam 1 and beam 2, what is the unbraced length that would be used to determine the allowable stress?

- BM1 and BM2 are sitting on a concrete wall and welded down (pinned, not rotationally restrained....except beams can not rotate along about their length)

- BM3 is attached to BM1 and BM2 by typical double angle bolted connections

- BM1, BM2, and BM3 are all the same size, W12 members

One side of me says the unbraced length for beam design of BM1 and BM2 is 10' because BM3 prevents the top flange (compression flange) from laterally moving under lateral torsional buckling

Another side of me says 20' because there's no diaphragm there

This is giving my mind a fit!!

 

[BAretired]

The unbraced length of BM1 and BM2 is 10' provided they are both torsionally restrained at each end and the middle.

BA

 

[JLNJ]

The brace point needs to prohibit either translation or rotation.

Your diaphragm (if you had one) would prohibit translation.

If the clips on the ends of BM3 are stiff enough they will prohibit rotation. Job done.

If BM3 is connected with true pins, then no brace point.

 

[KootK]

In general, I agree with 10'. You might google twin bridge girder buckling however. It's a real thing and, I suspect, speaks to your concern here. It can happen in the absence of lateral restraint but is unlikely to be anywhere near critical for your setup.

 

[Settingsun]

Do you adjust this for top flange loading, and restraining only the bottom flange at the supports? Assuming those are both the case? Those combine to increase the effective length by about 50% in some codes eg Australia and I think Euro.

 

[marinaman]

Thank you guys for all the comments.

steveh49, that's a good question. In the case illustrated, I was thinking of placing a stiffener between flanges at the bearing locations.....this way, even though only the bottom flange is welded down, the stiffeners unify the entire section at the bearing location and there would be no way for the top flange to move at the bearing ends.....that's my thinking anyway.

 

[Settingsun]

The full depth stiffener is all we have to do to restrain the top flange at the support, so all good there.

The Aus code requires 40% increase in effective length for top flange loading but there's some fat in that. The British code only required 20%. I had a feeling though that it's not a factor in the US code - perhaps compensated in some way. It's also not a factor when St Venant torsion is the only torsion resistance but I haven't got to the bottom of why that is, when the exacerbation of twist should still occur due to geometry.