Transverse Stiffeners - Torsion 6

[ToadJones]

Do transverse stiffeners do anything for the torsional capacity of a wideflange?
Intuitively I would think they would help locally with warping.

 

[271828]

The stiffeners are flexible out-of-plane so wouldn't do much, if anything to help with warping. I don't see how stiffeners would help with pure torsion either, so I don't think they'd do anything for torsion. This assumes a member with regular proportions. If the member has very high b/t, then the stiffeners might prevent distortion.

 

[Teguci]

I don't know. But, since the beam will be restrained from warping at the stiffeners, I can picture the beam rotating more to compensate. As 271828 notes, the stiffenrs add nothing to pure torsion resistance. My hypothesis is that the stiffeners will cause the beam to rotate more but not add anything to its torsional resistance (perhaps a worse condition).

 

[271828]

Teguci, how will the beam be restrained from warping at the stiffeners? Warping not the same as "distortion," which I think is what you're referring to. See the AISC Design Guide 9 Figure 2.2. Stiffeners would do next to zero to restrain this type of deformation.

A plate added to restrain warping would have to be rotated 90 deg. to the usual orientation of a stiffener. See Figure 3.3(a). These plates are rotated in the direction that would restrain warping deformations.

Stiffeners = no help for torsion of regular rolled shapes, as far as I know. It would be cool if they did, so if someone has a reference, I'd be very interested in seeing it.

 

[rb1957]

i think they'd help in making the entire section work as one, stiffening the joint between the flanges and the web.

 

[271828]

In other words, controlling distortion, right? LOL

 

[rb1957]

i think the weak point for an open section in torsion is the joint between the flanges and the web. Stiffening that would improve the torsional stiffness of the section and the torsional capacity of the beam.

LOL?

 

[271828]

I just thought it was funny because I thought we were talking about the same thing. Sorry.

The OP was talking about regular steel beams because he said "wideflange." Design Guide 9 shows how to compute stresses and deformations for those, and stiffeners wouldn't come into the picture. For that particular type of section, I don't think they'd do anything.

I could be wrong, though, so if someone has a reference, I'd like to see it.

 

[jtx]

I think the issue is trying to resolve the shear flow in the flanges that's induced by torsion. The shear flow in flanges flow in opposite directions creating a couple acting about the shear center of the beam. In order to resist the torsion, you need to resist the shear flow couple that develops.

The stiffeners would not be able to resist that couple. Once the shear flow goes into the stiffener, it would creates a couple within the stiffener, where the stiffener is connected to the flanges, and since the stiffener is unsupported at one edge, a resisive couple would not develop.

The only thing the stiffener might do is resist any possible distorsions in the flange as 271828 notes.

That's my thought...Let me know if anyone agrees.

 

[jimstructures]

The kind of stffeners that would reduce warping are the kind that convert the W section into a closed shape. i.e. welding plates from flange to flange or welding the open side of channels to the web. If I remeber correctly AISC DG 9 has an example of attaching a channel to the web (open side against the web and running parrallel to the axis of the beam for a short distance).

I am not aware of any research that supports this but maybe DG 9 has more information. I don't have a copy available.

Jim

 

[msquared48]

That's why the shapes bestr recognized to resist torsion are tubes and cylinders, not I sections.

Mike McCann
MMC Engineering

 

[ToadJones]

And here I thought I had a silly, almost embarrassing question.
I find myself agreeing with 271828 as far as the out-of-plane stiffness of the stiffeners being inadequate as far as reducing warping.

 

[JAE]

jtx - I agree totally with your post. Stiffeners don't do much of anything for torsion.

 

[slickdeals]

I agree that full depth stiffeners don't offer much in terms of torsional resistance, but intuitively I think that if a number of these stiffeners are provided along the length (at a close enough spacing), then it should have some benefit.

 

[ToadJones]

slick-
i thought the same and i am now trying to talk myself out of it

 

[slickdeals]

jtx: are you referring to full depth or partial depth stiffeners?

 

[jtx]

Slickdeals: Full depth stiffeners assuming they are connected to both flanges and the web, leaving one side unsupported.

 

[a2mfk]

Toad- can you explain your exact situation?

 

[SAIL3]

I think Teguci has come closest to the resulting behavior.
Without the stiffeners the WF will both distort and rotate.
By adding the stiffeners, this prevents distortion to occur in the vicinity of the pair of stiffeners without causing a corresponding deflection in the stiffeners.
The added stiffeness would be in the resistance of the stiffeners to bending in their strong axis.
The top fla would try and deflect laterally in one direction while the bottom fla would try to deflect laterally in the other
putting the stiffeners in double curvature, assuming stiffeners are welded into both top and bottom flanges.
It would be difficult to quantify the added resistance.

 

[271828]

If anybody's interested, the following journal paper is relevant to this discussion.

http://ascelibrary.org/sto/resource/1/jsendh/v121/i7/p1119_s1

The author refers to a 1983 ASCE Jnl of SE paper by Szewczak et al. On Page 1639, 3rd Paragraph, the researchers state:

As noted by previous researchers and shown in Fig. 3(b), the transverse stiffener has very little torsional stiffening abilities and shows rotations and warping normal stresses approximating those of the unstiffened member.

Note however that this section of the paper only addresses stiffeners at the ends. The remainder of the paper looks at other types of stiffeners--see the first full sentence on Page 1641.

In these two papers, the authors go through quite a lot, looking at relatively exotic and very expensive options. Why would they be looking at these if all one has to do is add transverse stiffeners?

Blodgett shows a funny little experiment on Page 2.10-18, Figure 33. The twisting angles are compared for a web plate, a web plus flanges plus closely spaced transverse stiffeners, and diagonal stiffeners. The the web plate twisted 10 deg. The web+flanges+stiffeners twisted 9 deg. The diagonal stiffeners helped a great deal, and it twisted only 0.25 deg. Take that experiment for what it's worth, though, because there is no warping stress generated using those boundary conditions.

Unless someone produces a reference that shows that transverse stiffeners add significant torsional strength or stiffness, I'm convinced that they don't help with torsion of wide flange beams.