Damaged Column 6

[BPres]

We have a large assembly building. At one end we have a 10 ton bridge crane followed in series by several 5 ton bridge cranes. All have a lift height of 21 feet.

One of the columns between the 10 ton and the 5 ton next to it was damaged by a fork truck. About 6 inches off the floor. The column is a W10x77. One flange was dented in approximately 1-1/2" on one side of the web. On the other side of the web the flange bowed out approximately 1".

Aside from AISC Manual of Steel Construction Allowable Stress Design (page I-149, Permissible variations in cross section), are there any thumb rules or codes dealing with damaged columns?

 

[dik]

I'd probably look at plating the flanges on both sides by cutting out the bent portions. May require temporary support of part of the superstructure.

Dik

 

[amanpreet]

I would personally go for the bracing to that portions so that in future 'cos of creep and sustained loads and some undeterministic loads could make the situation more worse than what it is right now.
So i would go for the steel bracing in that portion. providing extra support to the column.

regards
Amanpreet

 

[boo1]

Assuning its A36, why not support the beams and heat straighten and add a welded cover plate.

 

[jen4950]

I'd say coverplate it (no cutting of existing structure btw); and don't try to straighten it out- the material has already yielded locally, and it doesn't sound like an architectural issue- so heating up the remaining cross section to bend it back straight only makes matters worse. Pretend the material isn't there and weld a piece of 1" plate that's wider than the damaged flange (0.87" tf).

 

[boo1]

Common carbon steel beams (A36,A588/A588M-05) are often straightened with heat. If the beam is unloaded, and supported, heat straightening will not add additional stresses.

The purpose is to introduce in the beam carefully planned tensile stresses to pull the beam straight. The explanation is as follows:
Stresses are due to volume changes with heating and to decreasing yield strength at elevated temperature. Metal subject to thermal expansion while heated tends to be compressed by the surrounding cool structure. The heated volume has lower yield strength at high temperature, and then it is easily upset to shorter dimensions. Upon cooling the same material tends to contract in all directions and is now stressed in tension by the attached cool structure which did not move appreciably in the process.

See procedure and examples:

http://www.eng-tips.com/viewthread.cfm?qid=106922

http://www.steelstraightening.com/companyhistory.htm

 

[BPres]

Very good references. However, this is a column with no easy way to take out of service. I will take the jen4950 advise.

 

[BPres]

As stated above, one flange was dented in approximately 1-1/2" on one side of the web. On the other side of the web the flange bowed out approximately 1". I would have to heat and bend the flange before I could weld a 1" plate over the bent area.

What if I welded a 1" thick plate between the outer edges of the flanges (1 flange bent, other flange straight) on both sides of the web. Would that do the job?

 

[boo1]

The problem we had with attaching flat plate doublers to deformed member is the fit up. You could attach two angles on inside root between the flange and the web.

The flame straightened we did, left the members in place. We added temp supports and jacked them in position, unloaded the beam. The gray beard welder made it look easy.
I have see welders do to it on buckled plate to.

 

[jen4950]

Sure- shore the area, barricade the factory floor and perform a textbook straightening procedure. And lose sleep over residual stresses.

Or.. fabricate a plate- that's quick and easy. Split it into two plates and go on the inside of the flange.

Bottom line it's pucker factor and CYA- has the bay collapsed? You just want to be sure that the loads get to the foundation. And the existing section still has capacity (obviously). Get some plate and sleep well at night in a cost effective manner IMO. It's simply Area of steel- in the words of a boss- just kill it. You can spend tens of thousands of dollars in shoring, analysis and fabrication- or you can just slap some good ole' steel on there with proper development and analysis for change in C.G. and be done with it.

This is not a swiss watch or the Ritz, it's a steel column that's still standing in a factory. Just reinforce it and be done with it.

 

[geoffdale]

Just a thought. Could you reinforce the bottom portions of the column with a reinforced concrete casing? This may stand up to forktrucks etc better. Of course, you would have extra dead load on the footings which would need to be checked? My 2 cents.

 

[apetr26542]

When welding cover plates, consider the column has compression stresses in it, and the cover plate you are welding does not, AWS recomends stresses be less than 3 ksi.
If you are over this, you must incorporate this into your repair or remove the stresses.

I would go for heat straightening, seen a few jobs where you could not even tell a member was dented or dinged.

 

[apetr26542]

Obviously the flange was torqued, after it is straigntend they could also weld in some flange stiffeners, I also like the concrete encasment idea, they could weld some studs to the column to grab, not sure how the load path would work to the foundation.

 

[BSE05]

I have been involved in a couple of these problems. Fork truck drivers love columns. Even had a hole punched in a tube column. We use the "blowout" patch technique, welding a plate of equal thickness over the bent or punched area. It makes every feel better.

Recently I saw a column (after it was fixed by the Client)that had been hit by a fork truck, three of the 4 anchor bolts had sheared and the column twisted over 90 degrees at the base and was still straight at the top.(roof up about 35 feet) The fix was to wack it with a sledge hammer and it sprung back. I'm glad I wasn't involved in that one!

So think about column protection with warehouse design.

 

[RARWOOD]

When I worked in a manufacturing plant for a short time I dealt with one damaged column. The solution adopted was to box in the column with some heavy channels and cover plates. The box column was welded to the existing column about 10' off the floor so that the box section would pick up the vertical load.

The main reason for adding the heavy channels, was to protect the column so the third time they hit the column they would due less damage.

 

[Redfish205]

My vote would be the for the jen4950 approach. Would not cut the column at all

 

[pba]

I'd do all of the recommendations! Contact local steel companies to see if the column could be economically straightened. Steel is a plastic material and the concept of residual stresses within the 'fixed' section should not influence the yield stress.

As this column supports cranes it is possible that the section is oversize for the building loads provided that the crane is ot used during the repair works.

Even with a straightened column welding on new plate/plates gives better certainty that the repair will perform adequately. I would be surprised if the re-bending actually achives the original profile. It is likely to be an approximation.

Concrete casing is a cheap way of giving extra protection to the columns but I'd be relucatant to rely on it for structural strength - As a fresh graduate, I attended site to examine a cast iron column where I'd asked for the casing to be removed for inspection. The casing turned out to be concrete but the contractor had indeed removed it and had shattered the cast iron in the process... You would know what the concrete was for but not necessarily other people.

 

[boo1]

When welding any in-service member, caution should be used so in the heated condition the load does not cause failure.....

 

[jen4950]

" As a fresh graduate, I attended site to examine a cast iron column where I'd asked for the casing to be removed for inspection. The casing turned out to be concrete but the contractor had indeed removed it and had shattered the cast iron in the process... "

I want more details on this one!

And how did the structure behave without the column?

 

[UcfSE]

I'd like to hear from the OP. BPres, how did you solve this problem, or is it still in the works?