Sealing and pressure testing steel trough girders

[bugbus]

I am working on a steel trough girder bridge. Due to its relatively small size (not possible to inspect by a person inside), there is a requirement that the girder needs to be fully airtight and confirmed by pressure testing.

Apart from that, there are no guidelines that I am aware of, so I wonder if anyone has previous experience?

Figure below is showing the typical cross-section with proposed seal plate shown in red. Thickness would be maybe 5 or 6 mm. The seal plate has not been considered in the capacity of the girder so slenderness is no issue, although I need to confirm that the thickness is OK for the pressure test. The seal plate will rest on the diagonal bracing so it is not suspended in air as it would appear in this below cross section.

For the weld along the edges of the seal plate, I would imagine some small fillet weld as shown below is likely. Due to the fact that our code does not allow a single-sided fillet weld to carry a bending moment, I would need to treat the edges as though they were pinned. But I don't think this will be an issue.

In terms of pressure testing, this is where I am a bit stuck. The fabricator will need to install a valve and pressurize the girder. I am leaning towards adopting a similar approach for PT ducts - something like 2x atmospheric pressure with no notable drop in pressure after 1 minute.

My final concern is the performance of the weld when the beam is eventually loaded. No doubt the seal plate will want to buckle (which is OK as we are not relying on its strength), but would this buckling cause any unforeseen loads on the welds that would cause them to fail?

 

[FacEngrPE]

2X atmosphere is bordering on where the girder starts acting like a pressure vessel. You likely do not need that much pressure to prove air tightness. Perhaps 5 psi (1/3 atm) held for several hours, and /or use a tracer gas to identify leaks. If you did not PT the welds there likely will be leaks.
See the attached for some things to watch out for if you are just using change in pressure to prove tightness.

 

[bugbus]

Thanks FacEngrPE, that pdf is useful.

After running some numbers, you're correct that 2x atmospheric pressure is a LOT! Even at 1.1-1.2x, it is creating some pretty large forces and deformations in the girder, which should be manageable however.

The 6mm seal plate wants to bulge out about 30 mm. I doubt that adding stiffeners to it would be desirable for fabrication, so maybe a slightly thicker plate would be needed.

 

[3DDave]

If you depend on watching the pressure make sure to correct for the temperature of the atmosphere inside the closed girder. If it is properly sealed then temperature fluctuations will create pressure fluctuations.

 

[BridgeSmith]

I'm still trying to figure out what possible advantage there could be to using a rectangular steel box as a bridge girder. I've seen trapezoidal shaped tub girders on tightly curved superstructures, but having the webs vertical would mostly negate lateral thrust resistance provided by using a closed shape.

If the drawing is to scale, and the box is about 2m wide and about 1.5m high, I'm not seeing the problem with being able to inspect it. It's not exactly roomy, but it's not that tight either.

Rod Smith, P.E., The artist formerly known as HotRod10