torsion on cold formed steel built-up box members

[gte447f]

Does the AISI contain any provisions for designing for torsion loads on cold formed steel built-up box members?

My situation is a built-up window sill member. The sill supports a large two-story aluminum window. The sill receives vertical dead load and horizontal wind loads primarily from three window mullions, one near each end of the sill and one at mid-span of the sill. I have designed the sill for the vertical and horizontal loads, however, the architectural detail has the window framing outside the plane of the metal stud wall, supported on wood blocking that is in-turn attached to the outside face of the cold formed steel box member. I am concerned about torsion on the box member from the vertical dead load as a result of this loading eccentricity.

I am inclined to suggest a revision to the window detail to frame the window in the plane of the wall, so that the sill receives the vertical loads directly from the window without any (or with minimal) eccentricity, but I don't think that is going to fly with the architect. Otherwise I am inclined to suggest a different framing method for the sill (and probably the jambs), such as a hot rolled steel HSS member instead of cold formed steel box member.

Any thoughts or suggestions?

 

[BAretired]

Eliminating the torsion would be my first choice. Failing that, structural steel jambs and sill would seem to be a satisfactory alternative.


BA

 

[structSU10]

I agree with BA. I would not even attempt torsion on a cold form section.

 

[gte447f]

BA and structSU10,

The sill has short cripple studs below it. What do you think of this load path?: 1) Call for wood blocking to be at least as deep as the box member (in this case 6"). 2) Attach the blocking to the outside vertical web of the the box member with enough screws to transmit the vertical load to the web of the outside vertical member of the box via screw shear. 3) Assume the vertical load is then evenly distributed to the cripple studs in compression.

There would still be an eccentricity and thus a moment applied to the cripple studs since they would be loaded near their outside flanges only, but I could check for that and possibly design a stud to resist the moment.

I do not want to go this route, but it's a thought. Do you think it is a viable load path to transmit the vertical load to the cripple studs?

 

[BAretired]

Could you provide a cross section through the window?

BA

 

[dhengr]

Gte447f:
A sketch showing all the loads, details and dimensions, might save you ten thousand words which don’t really get the picture across. And, it might get you some good ideas to help you solve your problem. Supporting a two story window with a middle mullion, which is supported by a built-up light gage steel window sill, which is cantilevered out over some cripple studs below, and is finally supported on wooden blocking btwn. the sill and the cripple studs... is that about right? Sounds about like supporting a brick wall on a 2" thick layer of sponges and wondering why the wall is moving and cracking.

 

[DaveAtkins]

I would try to transfer the torsion into the box shaped sill member. It will have a continuous track top and bottom, spanning the width of the window opening. If you can get the torsion to the box shaped sill member (through wood blocking, for example), then you can resolve the torsion into a couple which is resisted by the two tracks. The cripple studs won't see any bending.

DaveAtkins

 

[hokie66]

A box shape is excellent in torsion, but you still have to design it. Why use a built up section when there are cold and hot rolled box sections available? Remember, if you use the member in torsion, the connection to the jambs, and the jambs themselves must take the reactions.

 

[XR250]

I would forget the boxed beam and use your latter idea of attaching the blocking to the face of the cripples. They can likely resist the torsion as it converts to strong axis bending. This will also put an outwards bending load on your track between the jambs - so it need to be designed. I would use 12ga cripples so you can get a wood to metal screw with wings to properly work and not have to worry about localized flange bending under suction loads.

FWIW, I do a lot of CFS design and Arch's are always showing boxed beams where they should not be used. They also always come up with goof details like this as well.
How tall and wide is the window?

 

[DaveAtkins]

The problem with using the cripple studs as cantilevers is the connection to the foundation wall. The special connection required at each stud will be expensive. I still think spanning to the king studs at each jamb makes more sense.

DaveAtkins

 

[XR250]

@Dave,

I was not intending to "flagpole" the cripples of the foundation wall (i agree, a bad idea). Rather, introduce the moment into the cripple and let it be resolved with a couple from the light gage sill (spanning between kings) and the floor connnection (in shear)

 

[kingnero]

I agree with BA. I would not even attempt torsion on a cold form section.

Would you care to elaborate why torsion should not be combined with cold formed sections? And why hot rolled sections + torsion are OK ?

 

[gte447f]

kingnero, I don't beleive that torsion on cold formed sections should be absolutely avoided necessarily. My concern in this particular case is that the cold formed section in question is a built-up box section as opposed to a closed section. I have decided not to attempt to design the built-up box section for torsion. If the window framing detail cannot be revised to load the box beam in the plane of the wall, I will likely design a hot rolled HSS sill and jambs. As a last resort, I may try to transfer the eccentric vertical load into the outside flange of the cripple studs. In this case I would design the cripples as pin-pin spanning between the floor and a horizontal sill member which would be designed for the out-of-plane loading from the cripple stud end reaction.

p.s. Sorry, I tried to upload a sketch yesterday, but was not successful.