Its a torsional loading to the WF beam. I guess some people might hand have that but I think its prudent to design for the eccentricity.

The beam web stiffener plates will reduce the bending of the beam flange, but do not increase the beams torsional strength.

Not sure what check you ran, it isn't clear to me where there is compression due to this? I would look at the beam shear including the torsion, as well as beam bending stresses. AISC (pardon my americanness) provides equations for member strength for combined bending and torsion. (AISC 360-10 Section H1.2)

Now if the beam is locked into a concrete deck there might be a good load path for the torque to be taken out directly.

We spent a bunch of conversation on this a few months ago. Very similar condition. If I can find it I will post it back here.

I would not ignore the eccentricity. Can you add another beam adjacent to it?

No I can’t, I don’t have enough space, (originally they welded a plate under the beam, but with the time the welds broke)

• https://files.engineering.com/getfile.aspx?folder=68942c84-d930-4fdd-a850-7

@driftLimiter this is how I ended up with the sum of the stresses in compression, its the first step of the torsion check

• https://files.engineering.com/getfile.aspx?folder=13ce2628-4c47-4d55-88ef-6

What's above that vented soffit? If there's a diaphragm up there and the beam is adequately connected, then that could be used to resolve the torsion (driftLimiter mentioned this with the concrete reference, but it could be other materials with sufficient strength, stiffness, and connections as well).

Well from the other side I can see half of the beam, so best case, only the top flange is bolted on the diaphragm, that doesn't seems enough to control the torsion..

Quote (Simlac450)

only the top flange is bolted on the diaphragm, that doesn't seems enough to control the torsion..

Can't say I agree with you. Draw a section through the beam and then create a free body diagram. See what the beam is resisting and how.

alright, I'll check that, thank you y'all, I appreciate

Can you close up the section? Add side plates so that there is a box section? It doesn't have to be on both sides.

Yes I could do that, and it could be pretty straightforward to evaluate !

Creating a box section would probably resolve the beam torsion issue, but in addition, the end connections must be able to resist that torsion.

What about something like this...?

I think it will change the look too much but if I put the hss somewhere between the flanges I think it could be good solution

Simlac450:
Pay attention to Hokie’s suggestions, they are right on the money. Additionally, you have to unload, untwist, the beam (reduce the stresses) by jacking it up on the loaded side before you do the additional work on it, in place. You shouldn’t normally be welding on a highly stressed beam, because it will just yield even more, due to the heat.

Weld a plate on the side of the W beam to turn it into a fabricated box.

W section (or to me, UBs) are often used with a plate on the underside but I can rarely ever justify them except for on very short spans (so brick arching very much in effect) and with the plate itself also bearing onto the support (so some torsion is taken out).

Edit - missed Jed's post above me

Quote (GTCE)

Weld a plate on the side of the W beam to turn it into a fabricated box.

That will only work if there is enough torsional resistance at the ends of the beam. Hard to calculate that IMHO.

To get rid of the torsion at the ends, I'd unload the beam (you might need to use some jacks) put on the side plate(s) to make it into a box, then add a new clip angle at the ends to the side plate and brick (using epoxy to anchor to the brick). Carefully unload the jacks, looking for distress.
To analyze, resolve the torsion into a couple.