Bent W8x31 column flange

[Bazinga]

Hi everyone,

I have a W8X31 H beam column which our contractor had difficulty installing so their 'solution' was to wrap a soft sling around and pull into position. This resulted in some inward localised bending of the column flanges which they then cold worked straight again with a hydraulic jack. The overall beam was not bent, just the flanges around where the sling was positioned. The issue now is we know the flanges have yielded and are now work hardened with potential for micro-cracks within the material due to repeated deformation. My question is, is there a method to return the yielded section to it's 'elastic zone' through heat treatment, or is there another suitable repair method to reinstate the column so that it performs as expected under wind/earthquake loading?

 

[dik]

Nope... if it's an issue, can you weld a plate across the flange where it is 'bent'. That's about the easiest 'fix'. Is the indentation near the top or bottom support or is it in the mid height (likely near the end)... not so much of a problem if buckling is the mode of failure.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Lomarandil]

I'm with dik, the first step is to check whether the local deformation/work hardening is actually problematic. Many times you'll find that it is not. Which is good, or tons of warehouses with columns dinged by forklift operators would be in danger.

 

[dik]

I think it would be difficult to determine that and I'd just replace the damaged material with new BAR stock, with the added material and the existing taking care of any eccentricity issues.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Lomarandil]

I figure there are a lot of cases where even if the steel in the region affected was entirely neglected, the column would still calc out OK overall.

 

[Bazinga]

The 'bent' region is basically mid span of the column. The contractor has worked it back straight now, and the concern is now microfractures within the material from being cold worked multiple times. I was wondering if we could somehow normalise the plastically deformed area by heat treating. Probably the same effort just to cut out and re-weld the plate in though. Thanks for the responses!

 

[phamENG]

the first step is to check whether the local deformation/work hardening is actually problematic

This response comes up a lot. As often as it's thrown out there, nobody ever goes into any detail on exactly how they would carry out that analysis. I could see a lower bound check of just assuming that portion of the column is gone, but short of that what are you doing to check it? Determine compressive stresses in that region and doing a plate buckling analysis to see the eccentricity is enough to cause the plate to fail at a higher column loading?

 

[JoshPlumSE]

Was there an civil / structural engineer who signed off on the column / structural design?

That's the person to ask! He might have over designed the column by a significant amount and know immediately that it's fine. Or, he might have designed it right down to the last 0.1% and know that it could be reinforced slightly.

 

[Lomarandil]

Ok, I'll bite Pham.

Like Josh says, the first step is to check the base utilization. Was the column designed to the limit, or was it oversized for some reason?

If the column was originally oversized, re-run the design assuming some reduced section, applied over the full length. If that calcs out, Bob is your uncle. Bill for an hour and call it good.

The governing limit state informs our next step. If the column is governed by flexural buckling, you can apply the same type of analysis as you would for a stepped column or partially reinforced column to modify the effective buckling length for a reduced section local to the damage (not along the full length). If the deformation is near points of lateral support, this is likely to work (also check yield on the reduced section). If the deformation is near midspan, it's less likely.

If that doesn't work, you can go to FEA buckling analysis with plate elements and imposed deformations -- but at this point, replacing the affected cross sectional area is probably just as cost effective in most cases.

 

[phamENG]

Thanks, Lo. The only times I've dealt with these, they were so minor they were okay by inspection (20% utilization in original design, etc.), client didn't care and wanted a repair to satisfy their safety department, or it was so bad the column had to be replaced (by inspection). I've never encountered an 'in between' where I had to go through those steps and do a real analysis on it. They make sense, though - thanks for sharing.

 

[BridgeSmith]

We hire specialty contractors to do heat straightening (AKA flame straightening) on bridge girders fairly regularly. They do some of the straightening without heat. If you have one in your area, you could probably get a consult from them. I know that based on research and testing, we limit the number of times a bridge girder can be heat straightened to 3. (Yes, we've had a few that have been hit in the same place by oversize loads more times than that)