Steel HSS Column Buckled & Sheared...Advice and Opinions Please

[mappryan]

Ok, to start out let me describe the situation here. My firm has been contracted to perform a failure investigation on an existing HSS 6x6x1/4 column. The building was constructed in 1994 as a grocery store. The building is currently in the process of being converted to a department store. During demolition of the building this HSS column buckled and sheared right through its compression flange (see attachment). The project manager for the GC is adamant that the column was struck at about 5' from the base by the trackhoe operator. Of course the trackhoe operator says that is absolutely not the case. I am an EI with little to no experience in this type of failure and would greatly appreciate any help or insight into the possible causes. Here is the pertinent information (hopefully I dont omit anything).

- HSS 6x6x1/4
- Unbraced Length is ~26 ft
- Tributary Area supported by column ~1430sq ft
- Typical Metal Deck roof
- Typical bar joist framing.
- Top is pinned. Base appears to be fixed(no baseplate visible)
- Failure point is about ~18ft from base
- significant rain occured earlier in the week (project manager noted water flowing from the scuppers and check and verified all (4) 10" dia. roof drains were working.

I determined in my analysis that the euler buckling load is ~74kips in a pin-pin configuration. Dividing this by the 1430 sq ft would mean a load of ~52 psf would be required to fail this column. I assumed 15psf for dead load and 20.8psf for 4" Rain Load but that still only brings me to 35psf. I am unsure how the impact at 5' from the base would cause the buckling at 18'. I understand that in the fixed-pin configuration the buckling would be shifted up from midpoint. Is is common for a column to shear right through itslef when it buckles? I have attached a few pictures and a portion of the plan where this occured. Any insight or opinions would be greatly appreciated. Thanks

 

[BadgerPE]

Just a quick thought....in the lower picture of your attachment it appears that there is are marks from some sort of impact +/- 1' below the sheared location. Maybe something else struck the column much higher near the failed location?

 

[csd72]

mappryan,

this looks like a classic forklift strike to me.

The reason as follows:

The shear due to eccentric axial load is negligible compared to other effects (somewhere in the realm of 2% of axial load).

Failure of this type of column tends to produce warped/buckled members not torn ones like that shown.

The types of deflections that would be required to rupture that steel by applied stresses would be in the realm of several feet.

a few notes on your assumptions:

It is typical for the base plate to be fixed to the top of the pad footing which is usually a foot or so below finished level. The column is usually designed as pinned top and bottom though in reality it is possible for the base to behave as fixed.

You have some shoring in there to support the roof which is great, but beware of uplift on the roof if the amount of steel left at the rupture is not enough to resist this.

Is it possible that this is a welded splice?

 

[ishvaaag]

You name Euler load. Since slender might be close, but the standard check would be against the carrying capacity.

More than calculations you need inspection. It is impact what is being negated, so you search for any mark of impact. Are there any pictures of the situation prior to the failure? Also, suspiciously, and as so many times, failure occurs during subsequent works in the life of the building. Usually this means that fundamental assurances previously taken when constructing the building have been damaged, eliminated or undermined. Has in any way the structural system been deprived of any required element whilst in the demolition? If so, one can't just proceed on tributary area, but on analysis of what the new condition asks of the structure in general, and the failed or failing element in particular.

More, to buckle is buckle, but it is rare for ductile structural steel to break; yet the tube may have been there a transverse weld to form the required height of the column. Analyze if the rupture is in some heat affected zone for any such potential weld.

Coupon analysis of the broken zone (open lips) and basic steel etc could also potentially reveal (apart from weld) any anomalous structure of the steel itself; even if probabilistically rare, a poor condition of the material leaving this particular piece out of specs may have contributed to the failure.

Of all this, I venture the cause is something done during the demolition/reform that shouldn't have been done, impact or not. I would say that more than half structural failures of importance, with fatalities or not, fall in that category. So examine what elements of the structure if any were being removed, by the instructions of whom, and in what way they were being retired, and analyze for the situations that appear.

 

[MiketheEngineer]

BTW - I hope you are not relying on that scaffold to hold up the roof. Each leg is only good for about 3,000 lbs

 

[JedClampett]

What's the end game here? Are you an advocate for the GC to show that the trackhoe operator caused the failure? It's highly unlikely you're going to find an initial design problem. In seventeen years, the column has seen almost all its normal loads (I mean its rained before this) and probably a few abnormal ones. If someone drove a machine into the column, and the column failed, thats pretty much the end of the story. It comes down to who's more likely to lie? The guy who's directly at fault.
Anyway, I'd concentrate on getting a new column in there. I don't really think this is a structural issue, more of a who's going to pay issue.

 

[spats]

The column was definitely pierced by some type of equipment. I have seen and investigated a handfull of instances where a column was pierced/almost severed by a forklift tine, and you picture indicates the same kind of damage. However, it seems a little high up to be a forklift, unless it was a high-mast type. There is absolutely no way this failure could have been caused by roof loading, so save the wear and tear on your calculator.

 

[Lion06]

I was thinking the same thing spats. I don't see any way this could have happened from any type overload on the column. If a buckling failure occurred, I would expect to see localized plastic deformations in the region, but certainly not a a shear failure through practically the entire column.

 

[hokie66]

I agree that the column failed due to impact, but like csd72, I question whether this is at a splice. Closed sections welded without backing plates are notorious for weakness at that point, and that would explain the location of the failure.

 

[Ron]

I also suspect this is a weld splice. The failure is too straight, so preferential failure such a this is usually related to a pattern defect or weld line.

This is also not a typical shear failure. If you look at the buckled compression side (the side that was hit by the equipment), you see the walls are folded. On the opposite side the walls have just popped apart. If you look at the failure surface I would bet that you have a tensile failure of a partial pen weld.

Add to that the fact that the failure occurred about 18 feet above the floor and a base plate is not visible. If you get to the base plate you'll probably find that the distance from the base plate to the failure point is 20 feet....the typical mill length of structural sections, that had to be spliced to get the 26-foot clear height.

 

[OHIOMatt]

Just a thought, is there a chance the guilty party hit the column and severely bent it and then tried to cover it by pulling it straight? They may have stopped when they saw the factures open.

 

[ToadJones]

Spliced at that location or not, someone drove a high-lift (Lull, Gradall etc) right through the column, probably from the same side that the guy in the man lift is on in the lowest picture. The sides of the HSS buckled out as the "flange" side pulled inwards in tension.
To suggest this from roof loading is quite a stretch.....it would be like the roof slammed down a few feet really hard and popped right back up!

 

[SAIL3]

Agree with hokie & Ron, looks to me also that it is failure at a splice which was probably partial pen. Too localized to be a buckling only type failure.
Also, I would check all structural members that are depending on or influenced by this col not now being a point of support ie. especially the conn. of bms etc. at their far ends away from this col

 

[dik]

Any chance you can post a close-up of the failure? One that you can see the actual failed edge.

The column is slender enough that it wouldn't take much to bend it severly, in particular with a P-delta loading. Failure appears to be too 'straight' and as noted, may have occured at a weld line.

Any idea of what loading was on it at the time? For a column that high, the failure was close to the point of contraflexure and may have prevented catastrophic collapse (pin ended columns, in particular slender ones, often have a degree of fixity 'built in'.

It may not have taken much of a nudge to precipitate the failure. A heavy equipment operator may not have realised that it was nudged. Any chance something adjacent to the column (since removed) was providing some lateral support?

What work was ongoing at the time. Was the 'crack'

 

[dik]

...observed at the time.

Dik

 

[a2mfk]

The OP said it was a "track-hoe", not sure what that was doing inside of the building. Interesting how so many people morphed that into a forklift. One wrong rotation of the track hoe could clip that column and do that damage...