Torsional Properties for I-Beam w/ a Cap Channel 1

[jochav52802]

Good Morning,

Per AISC's Design Guide-9, I've determined that an I-Beam installed in the field can't handle a torsional load that results fromm the supported joists being installed eccentrically by the contractor.

That said, I need to determine if adding a Cap-Channel to the bottom flange will be enough to provide the required torsional resistance.

I found a CISC guide that shows me how to calculate J and Cw, but it doesn't tell me how to calculate Wns, which is the "normalized warping function." I'd appreciate if someone could help guide me towards a solution.

Many thanks in advance!

Best regards,

jochav52802

 

[KootK]

Back to JAE's #2. I bet you've got room.

 

[KootK]

Keep in mind that if you go with the torsional cross section reinforcement approach, that's only as good as your capacity at the ends of the beams to deliver that torsion to the supporting columns etc.

 

[jochav52802]

Thanks everyone!

Hello KootK,

Does tying the joist to the support beam to restrain twist of the beam impact the joist? Doesn't the joist now need to handle the torsional moment that it's restraining from reaching the support beam? I'd appreciate your thoughts; our joist supplier will likely not want to support any additional loads.

Thanks for the extra comment; the current end connection to the columns are just shear tabs, which obviously aren't good for torsion, so they'd need to be reinforced if we reinforced the beam.

Many thanks again!

 

[charliealphabravo]

Adding a tube section will be the simplest method I think and involve the least amount of welding and possibly minimum steel since the tube is so efficient in torsion. This question came up a couple of weeks ago.

 

[jochav52802]

Hello charliealphabravo,

By add a tube, do you mean to replace the beam with a tube or just enclose it?

Thanks!

 

[cal91]

I was curious how the member forces would be effected by KootK's cool idea of bracing to the joist itself to take out it's own eccentricity.

You'd increase the vertical shear demand (and the reaction from the joist seat to the Beam top Chord) on the first two diagonals by 'V2'. You'd increase the compression of the top chord (only within the end panel) by 'H1'.

Solve for the three unknowns V1, V2, and H1 with the three equations gives...

V1 = R / (1-e/W)
V2 = R * e / (W - e)
H1 = R * e * W / (d*W - d*e)

 

[KootK]

does tying the joist to the support beam to restrain twist of the beam impact the joist? Doesn't the joist now need to handle the torsional moment that it's restraining from reaching the support beam?

Yes & yes. It's a very common means of retraining a beam torsionally against various things:

1) lateral torsional buckling.

2) lateral loads applied to the beam bottom flange.

3) minor load eccentricities in the beam.

As cal91 has nicely illustrated, there are both horizontal and vertical loads imposed upon the joist. That said:

4) flexural compression forces at the joist top chord (midspan) should dwarf the horizontal force here by an order of magnitude or two.

5) the vertical force should be small-ish relative to the joist end shear capacity.

That said, it kinda depends on how minor your torsion is. What's the joist span? What's the eccentricity? What is the beam span?

This scheme also requires that some joist axial be transferred at the joist seat to beam flange connection. Cal91's H1. Usually not a big deal.

 

[KootK]

You might get the combined section to be strong enough but not stiff enough. A WF with a cap channel is still an open section and very flexible torsionally.

I believe this to be a very salient point. A point which would preclude many of the solutions here straight away. Stiffness concerns are part of why I recommended the solution that I did.

 

[charliealphabravo]

@jochav52802

I was thinking to add a tube to the I beam as per your second post. That way continuous welds would not be needed.

Edit: By that I mean adding a thin-wall HSS full length, not enclosing the I-beam. It should be somewhere near the top of the beam, if that is where the joists bear, in order to avoid other mode checks.

 

[KootK]

A lot of folks clearly like the side plate idea. Depending on your weld access situation, you could achieve virtually the same thing with an angle or bent plate and only top-hand welding as shown below.

 

[KootK]

That way continuous welds would not be needed.

Continuous welds would not be required regardless. It's an old structural engineer's wive's tale that one can't stitch up a closed, torsion section.

 

[cal91]

Continuous welds would not be required regardless. It's an old structural engineer's wive's tale that one can't stitch up a closed, torsion section.

Never trust a structural engineer's wife. Also nice idea for eliminating the OH welds.

 

[KootK]

Thanks cal91. My wife actually is a structural engineer. Management though...

 

[charliealphabravo]

I know KootK. I saw that post a week or two ago. I was just thinking that an intermittently welded enclosure of the I-section raised more design and execution questions and is not as idiot-proof as tacking on a tube section...especially if the contractor is already strike one.

Edit: KootK's last detail looks good to me as well.

 

[cal91]

Ha that's awesome! You knew she was the one when she performed Fourier Transforms for you?

 

[JAE]

The box beam idea can certainly be an option - just remember that some type of torsional end restraint is still required to take the torsion out of the beam somewhere (usually at the ends) whether you have an open section or a closed boxed section.

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

 

[KootK]

@charlie: noted. I could get behind a tube under the bottom flange but, with a side mounted tube, I'd be a bit worried about weld quality on the overhead flare bevels.

 

[jochav52802]

Many thanks everyone for all of your input.

We're trying to get our joist supplier to agree to adding a kicker to restrain the bottom flange as we believe this is less work than having to reinforce the supporting beam for torsion as well as adding restraint to the shear end connections.

Unfortunately, I couldn't really get any good info on coming up with the "normalized warping function" mentioned in AISC Design Guide-9; the goal continues to be eliminating torsion from consideration if at all possible.

Thanks again for all of your help!