Stiffeners in Large HSS Columns

[CBEngi]

I have a custom HSS section that is approximately 5' x 10", it is being used to support large overhead doors.

My question is, the section needs to be stiffened at the top where the pulley assembly mounts to the column. How do I determine how far the stiffeners should run below the mounting bracket.

I'm worried about creating a point where the stiffness changes drastically and creates a "soft spot" in the column.
The load the stiffeners are holding is about 1/8 of the wind load (Which acts in the strong direction of the column and opposite to the dead load of the door).
So if the stiffeners run to far down the column I would have the same problem from the wind force which is much greater.

any input would be greatly appreciated.

 

[BAretired]

A sketch might help.

BA

 

[KootK]

5 foot x 10 inches is a mistype, right?

This sounds like a basic stiffened seat design from the AISC manual but I'll reserve judgment until we get some more info.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CBEngi]

Please see attached sketch.

The initial wasn't a mistake, the modifications are required to an existing section as the pulleys and sheaves being mounted are changing location.

Also I am working with the Canadian steel code, if the AISC covers the topic better I would greatly appreciate some input in a direction to look.

Thanks

 

[KootK]

Ok, that's cool. And nothing like I imagined. I see your bracket dumping moment directly into the stiffener behind it. That stiffener will need to be sized for flexure and shear etc. The stiffener will then dump the moment as a vertical force couple in the walls of the column. Size the welds accordingly.

Are you using the stiffeners to bring down the b/t ratio on the wide face of the column? If so, they'd need to run the full length of the column.

The localized stiffening doesn't bother me any.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[CBEngi]

No the stiffeners are not needed to reduce the b/t ratio. the column as a whole can withstand the loading. I was more worried about having some type of local buckling where the stiffener ends.

 

[KootK]

I see your point. You'd probably have to turn to Roark's to evaluate that. Fundamentally, it's not that different from the single plate shear tab connection that we do all the time without checking for buckling under the tab.

Out of curiosity, how are you dealing with the b/t issue? I've got that at 152 for the wide face. S16 table 4-2 limits you something around 38 depending on your steel grade. I did something similar for a free standing canopy once. We had to weld angle stiffeners to the inside of the wide faces to keep b/t from getting out of hand. Maybe since the column is existing, it's already stiffened.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[Bagman2524]

I'd make the bottom edge of the stiffener on a 45 deg angle.

Faith is taking the first step even when you can't see the whole staircase. -MLK

 

[BAretired]

I'm curious too. How can such a large b/t ratio be justified by S16?

BA

 

[CBEngi]

S16 Table 4-2 (Look to take Table 2, pg 1-127) which given more limits.
basically if your b/t ratio is larger then the limits provided then you have either have a class 3 or 4 section. So for a Beam-Column analysis you can turn the section into an equivalent size that fits within the limit.

So take the b/t ratio, back solve for "b" then calculate your section strength with this property, realizing that that portion of the section is only able to fully develop without localized failures.

so specifically for the calculation that 5' long wall in theory became something like 1'-4".

 

[SAIL3]

not knowing the thickeness or the actual loads...vital info. However, it looks like the only effective portion of the col for axial loads are the two end segments which would give you two vertical channel-shaped sections. There is no effective stiffening for the brkt & stiffener located @ the brkt.I would add horiz stiffeners at the top of col and at the bottom of the vert stiffeners. What this leaves you with is a quasi-horizontal bm in the top portion of the col eventually supported by the two effective vertical channel portions at the ends. If you did not have an applied mom at the brkt and just dealing with vert loads, then the center portion of the 5ft length can still transfer in shear this vert load to the end sections even though this portion does not have an axial load capacity of any useable magnitude and as a result would not need that quasi-horiz bm at the top.

 

[KootK]

I think that the best way to deal with this is to use an effective Fy that satisfies b/t rather than an effective section. In your case, Fy would about 7 ksi assuming the real Fy is 36 ksi. In theory, this would obviate the need to follow Sail's recommendations. I'd be tempted to follow Sail's recommendations anyhow though as they just sound like thoughtful detailing to me. This is a good situation for belt & suspenders thinking.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[KootK]

Correction: I think you'll need to scale to SQRT(Fy). I'll double check when I get back to the office.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[BAretired]

If we are using Table 2, a Class 3 section permits b/t to be 670/√Fy, so for b/t = 160, Fy = (670/160)² = 17.5 MPa or about 2540 psi.

Welding continuous stiffener plates inside an existing hollow section is not possible, but it may be possible to attach the two five foot wide plates together at discrete points from the exterior in order to provide some stiffening.

BA

 

[KootK]

Thanks for doing the math BA. 17.5 MPa is pretty lousy. Aaron's proposal may produce more advantageous results.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[SAIL3]

as BA pointed out welding inside the HSS may be difficult.
I would ,if possible, close off the top of the col with a flat pl and put cutouts in it if reg'd. For the bottom fla of the quasi-top bm I would add a horiz external stiff if it is not possible to add one on the inside. With a top bm of depth up to 10" , it may be possible to get a weld rod in the...10x10" opening clearance.
Rather than do all that fab, the other possibility is to weld a ,say HSS10x10 to the top of the col , close off the ends with a pl that extends dn the ends of the col a certain length or extend the ends of the col up to use as a closure pl for the HSS10x10. Then detail a brkt conn to this top bm. I always try to close off the ends of HSS members...open ends on HSS members , not good. Ofcourse, none of this may be applicable as the info supplied is not complete.

 

[BAretired]

If there is access from above, could insert a full height C9x13.4, in line with the eccentric pulley load. The C9 could be welded front and back from the top as far as the welder can reach, then spot welded to each 60" x 3/8" plate periodically through pre-drilled holes in each plate.

BA