simply supported beam subjected to torsion with top flange unrestrained.

[Ringoo]

I am designing a simply supported beam subjected to torsion and flexure .The beam spans 24ft and is subjected to point load of 18kips at e=4” at 1/3rd and 2/3rd of the span. The beam sitting on concrete platform at the ends .The top flange and the web are not laterally restrained ,but the bottom flange is anchored down to concrete platform using the hilti bolts.
Is stiffener at the ends enough to transfer the top flange forces(due to torsion) down to the bolts? I want to make sure that beam is torsionally restrained at ends.
Am I missing something here?

 

[JAE]

Ringoo - is this a wide flange beam? If so it would have a very small torsional stiffness and you'd have lots of rotation.

For the end designs:
1. Vertical stiffener plates can be used to ensure that the end of the beam doesn't warp torsionally to any great extent.
2. The stiffener plates would be under compression on one side - thus buckling of the vertical plate under compression would need to be checked.
3. The welds of the stiffener plates to the beam would also have to be designed but may not be that big of an issue.
4. You mention Hilti bolts. With typical flanged beams the natural gage between the flange-to-concrete bolts might be small - thus the bolts might exceed the minimum bolt-to-bolt spacing. Need to check that. With close spacing you lose capacity rather quickly (ACI Appendix D - the wicked code).
5. An external system to prevent the ends of the beam from rotating might be in order for that size of load. This could be some kind of bracket or diagonals to provide lateral restraint to the top of the beam at its end to minimize force into the bolts.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[KootK]

The stiffeners will provide rotational restraint. However, they will not provide meaningful warping restraint. Torsionally, your finished product will be pinned. To achieve torsional fixity and restrain warping at the ends, you would need to connect the flange tips with vertical plates running parallel to the web.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[JAE]

KootK - I think you pointed out what I didn't do very well in my post above. The beam will rotate a lot along its length.
I was only speaking of the end rotations. So your comments are good related to the beam length.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[KootK]

Yeah, upon reread, I think that we're saying the same thing at the ends JAE. If the warping restraint plates were installed at the beam end, there would be no access to the bolts. Could one push the plates into the span a touch to address that? I would guess so.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Tomfh]

Flexure + Torsion = Hollow Section

 

[Ringoo]

Thanks JAE and Kootk for your valuable comments.