Venting holes when welding to sealed HSS 1

[StructMB]

I was asked to comment on a deformation observed on a truss's bottom chord fabricated with an HSS. Indeed, the two side walls of the HSS are bulged out for approximately 3/8" for the bottom chord's entire length. It does not seem to be an overstress problem since this member is most likely under a tension load. The truss is 40' long and the damaged bottom chord was made with a HSS 8x6x1/4. I've noted that a reinforcing plate was added to this sealed hollow steel sections without the use of a venting hole. Would this field welding of a reinforcing plate (1/4" fillet weld, 6' long) be enough to increase the internal pressure in the HSS to cause this bulging of the side walls?

 

[Hokie93]

What process did the fabricator use to achieve the camber? Damage to the bottom chord during the cambering process comes to mind as a possibility. My second guess would be damage caused by freezing/expanding water. I have seen HSS beams that have been deformed by freezing water (the webs looked like a pot-belly pig) but they were in an outdoor environment, completely unprotected from the elements.

 

[SAIL3]

freezing seems to be the most likely...but why no evidence of bulging on the top and bot of the member?....maybe only 1/2 or 2/3 full of liquid....

 

[bridgebuster]

Like all problems, no one noticed anything during construction.

@StructMB: Is the section 8x6 or 6x4? 40' long or 50' long?

My initial thought: There's a fixed condition at the bottom chord inducing compression. Hokie93 brings up a good point about the camber. Distortion can occur if not done correctly. However 3/8" over 40 to 50 feet isn't much. If the bottom of the bottom chord is flat, probably not a camber issue. If it were due to the camber I'd expect to see a concave surface on the bottom of the HSS.

You could drill a hole in the chord to see if there is water; that may solve the mystery.

Here's my dark horse theory: If it's not compression in the bottom chord the offset of the gusset plates is creating an eccentric load - what do you think KootK? On a bridge truss we would never do that.

 

[KootK]

Here's my dark horse theory: If it's not compression in the bottom chord the offset of the gusset plates is creating an eccentric load - what do you think KootK? On a bridge truss we would never do that.

What do I think? I think that you may be my brother from another mother. My dark horse theory is identical to yours. That's where I was headed with these questions:

1) What's the average spacing between bottom chord panel points?
6) Does the degree of the bowing fluctuate at all between panel points? Or is it utterly constant?

I envisioned a scenario where you'd be getting a compression strut in the side walls between successive panel points due to the eccentricity that you've described. The joints are concentric in terms of the member center lines but still require that there be a moment where the gussets meet the chords. With the distance between panel points being 10', I deemed that too far for this effect to produce uniform buckling along an 8" deep HSS face. That's just instinct however.

3) Is the truss top chord flat or pitched?
4) What kind of section is used for the top chord and do you see any of the same phenomena there?

I figured that, if it were a flat truss with similar top chord panel point spacing and member size, the phenomenon would manifest itself at the top chord as well. That is, apparently, not the case.

5) Any chance the bowing is greater near the supports and somewhat less near mid-span? I'm talking overall here, not between panel points.

I figured that, if gusset eccentricity were the culprit, the effect would be less apparent at locations of low truss shear. Apparently that is not the case either.




I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[racookpe1978]

Its up against the wall piece, but do you know if both sides of the lower HSS are buckled, or only the interior (visible) side of the HSS?

 

[calvinandhobbes10]

Any serviceability issues? Ceiling or partition cracks in the space below? Roof ponding above? Uneven doors or floors? It looks like precast exterior, any deflection visible on that? What alerted the owner to the issue? If nothing, I'd be inclined to simply install intermittent stiffeners to prevent further sidewall buckling, and let it go at that. Sounds like a construction problem as others have noted. The service loads simply don't relate to the buckling observed (7 ft deep truss spanning 50' sounds like it was driven by a roof step geometry, not required depth). If it ain't broke don't fix it. It looks dry and conditioned, so if it gives you more peace of mind, also drill some intermittent vent holes in the underside of the chord to ensure temperature and water don't collect.

 

[bones206]

It looks like there is no flaking or cracking of the paint/primer on the distorted area. That could be an indication that the bulging occurred sometime during fabrication. My guess is that the reinforcing plates were also added during fabrication, since the plate looks to have the same uniform coating as the rest of the truss.

 

[dhengr]

StructMB:
If the truss had been stood up in a vert. position, and standing on a conc. floor slab on its bottom chord, and then loaded from above on the top chord, for some reason, or maybe dropped in the fab. shop...; the vert. load from above would have been transmitted down the diag. web members to their connection gussets, and then on down to the slab. This would cause the bot. chord HSS webs to have a bending moment from the diag. web connection gussets, plus a sizeable compression load on the bot. chord webs. The sum of these loads might have caused the kind of web bending/buckling I think I see in your photos.

 

[Compositepro]

I have seen square tubing like this split open by freezing water. Freezing, alone, is not the problem. First ice must form a plug that traps a fairly large volume of liquid water. Then the trapped water progressively freezes and builds pressure in the remaining liquid water. Ice will tend to form faster at the top because ice and zero degree water are lower density than warmer water. Cold drafts and nearby thermal masses can completely change the pattern caused by freeze damage. I imagine that structural loads would also affect how the internal pressure might bulge a tube. But progressively greater deformation will be evident as you approach the last point that water was freezing before it burst.

 

[IFRs]

Why does it matter? If the member will be in tension for its entire life, nothing has really changed - the cross sectional area is the same. If it goes into compression, isn't it a bit stronger because more metal is further from the neutral axis?

 

[bigmig]

It looks like it was damaged prior to installation....happens all the time.
It is good to inspect your job (and photograph) frequently for this very reason.