Reinforcing a Slender Steel Pipe Column

[spats]

I am working on a project where they are adding support frames for roof-top equipment. The support frame posts are located directly above the existing pipe columns, and in some cases, add a significant amount of load beyond the capacity of the existing columns. The existing columns are 8" standard pipe (A53 Grade B, 35 ksi) and are approximately 30' tall.

Question is: how do you reinforce a steel pipe? I don't think concrete fill would help much, since the column is still fairly slender, and the process would be difficult. Adding braces/kickers down from the roof to cut down the unbraced length would likely not work because I'd be kicking to bar joists that probably don't have the stiffness to create a suitable brace point. If I reinforce with some sort of steel sections, how do I keep it from looking like Hogan's goat? Welding on t-sections or the like is all I can thing of. Would the reinforcing need to be all four sides? What length to run the reinforcing?

I need suggestions. The existing roof is steel joists and joist girders.

 

[JAE]

We usually try to first weld on two steel channels such that the channel flange tips are just touching the sides of the columns and stitch welding them down the length.
You end up with a pseudo box column with a pipe inside it.

So if your column is a 6" I.D. pipe with an outside diameter of 6.63 inches - we'd use two C6 channels. The flange tips would not touch each other - leaving room for the welds.

 

[JedClampett]

What I always try to do (admittedly, it almost never works out), is to find a larger pipe with the same ID as the pipe's OD, split it and weld it around the existing pipe.
For example, an 8 inch pipe has an OD of 8.625 inches. A Schedule 140 (yeah, it's a monster) pipe has an ID of 8.75 inches. Split it, weld it to the pipe at the ends and weld the seam back together. If you're worried about shear continuity, take out some plugs and plug weld the pipes together intermittently. Coat the pipe with some epoxy and polyurethane coating, and you're done.
It might be heavy, but it can be made to look decent.

 

[AELLC]

Equal leg angles work too.

The definition of a structural engineer: overdesign by a factor of 1.999, instead of the usual 2.

 

[spats]

Thanks all. I like the channel idea best. Do you think it's OK to run the reinforcing short top and bottom, and do some length of solid welding each end so the load can get into, then out of the channels, and only count on the pipe to handle all the load at each end? (run-on sentence). I can't practically get past the joist tails at the top, leaving about 3' unreinforced. At the bottom I could just stop the channels at the SOG.

 

[JLNJ]

Depending on how much capacity you need, you might be able to live with the reduced column section on the ends. The critical buckling isn't affected as much as you might think where the column and has just a fraction of its center section's MOI.

It's why all those warehouse columns whose flanges are completely mangled near the floor from forktruck collisions (usually) don't collapse.

 

[JAE]

I'd agree with JLNJ. Just get as close as you can to the ends. As long as there isn't any significant moment at the ends.

Typically if the majority of the column length is stiffened, the new kl/r would apply for overall, global buckling of the column.

If there IS moment at the ends - check the strength of the pipe only for the compression and bending (without buckling affect) as the channels, or whatever stiffening elements you are using, won't be engaged.