Torsional support at the two loose ends of a S shape

Stiffeners and welding

Look in the attachment. Check stiffeners for torsional bucklinkg caused by the reaction of the beam

Live long and prosper!

Also see page 2-13 and 2-14 in the AISC manual for requirements for end plates on unframed ends of steel beams.

Maine Professional and Structural Engineer.
(Just passed the 16-hour SE exam, woohoo!)

Why are you applying torsion to an S-shaped beam? They are not very good in torsion.

BA

I had assumed that we were dealing with LTB end restraint only rather than a true torsional load. Hopefully OP can clear that up.

My detail would be web stiffeners and a stiff base plate bolted to the support girder top flange[pre][/pre]

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

Correction: I like an end plate and welded connections for this.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

Dear JeanLucPicard,
Thank you for your reply and drawing the sketch.
One question came to my mind is shouldnt the side stiffener plates go all the way to the end plate? and if not whta's the proper length of them?

Is your detail considers as fixed torsional support or pinned in terms of analysis?

Shall the end of beam kind of anchor to the underneath beam?
I attached a detail of that was claimed is good trosional detail claimed from a colleague which I don't agree is good. Do you see any problem with that detail?

Thank you.

BAretired,
I am not appliying torsion just for fun. The S beam is a hoist monorail beam which the load applies with ecentricity.

Without side plates, the beam is torsionally pinned. That's the typical way to go about things.

With the side plates, the beam approximates torsional fixity. This is rarely done. The side plate tend to rotate in plane under the applied shear so connection to the end plate is prudent. Among other concerns, the end plate needs to extend far enough that, when it tries to rotate, it doesn't yield the beam flanges in bending. Alternately, you can terminate the side plates at web stiffeners.

Your colleague's detail would work. In fact, It's probably overkill.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

From what I've seen most monorails that aren't brand new (and even some that are) use S shapes for the local flange bending (which is a fairly common failure mode). The OP is correct that the lateral component of the monorail load is applied at the bottom flange and hence the torsional load. I've run into this exact problem quite a few times.

Maine Professional and Structural Engineer.
(Just passed the 16-hour SE exam, woohoo!)

Dear SKJ25POL,

You should look at the connection as torsional fixed, just try to imagine foam beam with those stiffeners under torsion. It will resist!

Stiffeners flanges don't need to go to the end, on the other hand, they might have to. It depends, on the loading. You could make a thick shorter flange, or you could have a situation without flanges (just two stiffeners as webs). Those flanges are there to increase stiffeners inertial, polar and torsional moments. Things that resist torsion . The principle is simple: by having flanges with given area and with given moment arm, geometrical characteristics will increase by squaring arm (Steiner’s rule)!

Torsion on S (I H) beams is common on the bridges. I have attached a drawing with stiffeners for a composite steel-concrete bridge 31meters, drawing is in mm, and bit messy... Note that the stiffeners don't go to the end of the flange (so that form work may slide on the flange).

Your sketch is a common procedure to minimize distortion on the beams end due to LBT. In case of torsion, I would go with two stiffeners . Little note, when designing connection tend to design two merging elements to connect with bolts. In order to minimize weld work on site. All the weld work should be done in workshop, on site just with bolts (utopia).


Live long and prosper!

Dear SKJ25POL,

Another way to understand torsion stiffeners is to observe torsion as bending moment acting on the stiffeners. Look at the attachment. You know that I beams are best for bending, now you can see why such a design .

The drawing represents the beam that I drew before, cut in half of its height.

Live long and prosper!

@JLP: nice details. What you’re describing and sketching represents robust torsional pinned-ness, however, not torsional fixity. The side plates are not of a proportion that would make them effective for restraining flange warping.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

JeanLucPicard:
Nice explanation, very clean and simple. That’s exactly what some of these OP’ers. need to get the picture. Sketches are almost always helpful. You would do well to attach your sketches as PDF files, they are much easier for us, at this end, to deal with. Welcome to E-Tips, it appears you will be a constructive contributor.

JeanLucPicard,
You are an angel. Thank you so much. That sketch openned my eyes. Excellent.
Thank you very much. You truly have understood the concept. Real engineer

JeanLucPicard,
I love to have your opninion on the sketch that I attached earlier today.
How that detail can be corrected.
I collected that that detail is a pin torsional connection and not a fix torsional comnnection, am I right?

What difference it makes having a fixed torional connection rather than pinned torional connection?

KootK (Structural),
Thank you for your comments but could you please sketch or show us a true fix torsional support. I have hard time puting your words in a picture.
Could please sketch us a sample of fixed torsion support that you believe is correctly detailed. I appreciate it.

Thank you all

@dhengr
Thank you! I'll make sure to do my sketches in .pdf . I always try to visualize problems, that's why I do sketches as to explain my thoughts. Much easier. I just have a trouble with the fact that I'm from Europe... The theory is the same; however, we do have different practices, codes, measuring systems, etc... On the other hand, I couldn't find a good European site, one that is active as yours, so I joined Eng-Tips (great forum)!

@KootK
I tend to disagree. The stiffeners have a way to deal with the moment trough moment arm, plus the beam is places onto a beam underneath. Therefore, the torsional rotation, warping, is not free, hence fixed support...

Live long and prosper!

SKJ25POL said:

Thank you for your comments but could you please sketch or show us a true fix torsional support. I have hard time puting your words in a picture.
Could please sketch us a sample of fixed torsion support that you believe is correctly detailed.

Click to expand...

See the sketches from AISC's Torsion Design Guide below. In summary:

1) JLP's detail, Jayrod's stiffeners, and your colleague's end plate all accomplish the same thing: rotational restraint (torsional pinning). None of these details accomplishes torsional fixing, however, in my opinion. JLP's detail undoubtedly provides the most torsional pinning. However, unless you're dealing with a monstrous torque, it's likely to be overkill. And, if your S-beam is really delivering that much torque, I've no doubt that your beam design will fail in torsion anyhow as BA has implied.

2) The key to torsional fixity is the side plates running parallel with the beam. Workable proportions usually have the length of those side plates being on the order of 2-3 times the depth of the beam. It's important to understand the forces that are required on the side plates to maintain equilibrium. Restraining flange warping requires developing large shears at the top and bottom edges of the side plates which need to be equilibrated rotationally by forces pushing up and down on the plates. That's what makes channels so clever in the sketch at the bottom. The channel flanges provide the equilibrating shears at the edges of the side plate.

SKJ25POL said:

What difference it makes having a fixed torional connection rather than pinned torional connection?

Click to expand...

Torsional fixity means lower warping stress and less beam rotation. It almost never makes sense to pursue torsional fixity. The return on investment just isn't there. Certainly, I would not think it warranted in this situation. Basically, detailing for torsional fixity amounts to turning some of your beam into an HSS. Often, it's best to just take the hint and turn your whole beam into an HSS.





I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

Messed up one of the sketches. This one shows the nature of the warping that we're trying to restrain with torsional fixity:



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

KootK (Structural,

Thank you very much. I greatly appreciate your time and valuable information.
Now I have a better sense on torional support.
Last nigth I was able to see the images but this morning from my work PC just see some x marks in blocks, dont know if is someyhing wrong with my computer or you and other people have same problem seeing the images

Thank you
Have a good day

The images look okay to me.

BA