Square tube butt-weld hydrotest 1

[CaliSolarMechE]

We're having trouble with weld quality on our long, continuous 4"x4"x0.11" (11ga) galvanized structural members. We're using a field, full-penetration, convex, backed, groove weld with a gap width the same as the material thickness (3mm). The joint must be as strong as the tube to resist torsion. There are 5 to 10 ASTM A500 tube sections welded end-to-end.

Right now our only remedy is to X-ray each individual weld, which seems like overkill. Can we take this down to 5% or 25% spot-checking? There is no code governance, only AWS D1.1.

Also, I thought of using a hydrotest as an alternative to X-ray. Will this tell us if the weld is good? By using SolidWorks FEA to check the impact of the hydrotest pressure with regards to stress on the tube, I think we can use up to 40 psig for the hydrotest.

I would appreciate your thoughts!

Best regards,

Elliot S.
Richmond, CA

 

[racookpe1978]

No - the hydro test - if to a high enough pressure to "test" the weld against faulure, would "round out" the square tube dur to internal pressure against the four "square" being resisted only by the bends at each corner.

3 mm wall, 3 mm gap. How are you backing the weld?

 

[CaliSolarMechE]

So, you don't see a real benefit from leak checking via hydrotest; is this because a superficial weld would form a seal but still have poor strength?

For the backing, we're using a galvanized sheet metal tube, bent into a rectangle that loosely fits within the "ID" of the square tube. Because of the weld bead on the square tube interior wall, the backing is not too firm or tight against the steel tubing.

 

[paddingtongreen]

If the tubes are galv'd inside and out, that is a lot of zinc. was the weld prequalified? is the weld metal appropriate?

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[weldtek]

Cali...
I concur that the hydro test isn't likely to guarantee anything other than a leak tight joint.
Also as noted above, welding galvanized surfaces can result in porosity. Is this the quality issue you're experiencing?
If so, remove the coating by grinding, prior to welding.
There are some electrodes that are somewhat less susceptible to contamination, but, you didn't state which process you're using.

 

[CaliSolarMechE]

racookpe, paddingtongreen, and weldtek,

So what I can take from this is the following:

1) A hydrotest is not likely to give us meaningful information.

2) Galvanized metal must be prepared correctly if you want a reliable weld.

3) Using prequalified welds improves quality by ensuring that we have seen a particular process work, and by ensuring that the welder has the information necessary to make the right choice of application.

So, if I understand this right, if we have more confidence in the weld process and welder, then we can pull back our inspections from 100% radiography to 100% visual + 5% radiography. (5% arbitrary)

Is there anything else that I'm missing here?

I really appreciate your input. Thanks!

Elliot S.

 

[weldtek]

Cali...,
Without specific instructions from a governing code or specification to guide you, you'll have to evaluate the options for welding and inspection, weigh the risks, and make the appropriate choice. D1.1 is written in such a manner that the onus is on the 'engineer' to determine the extent of inspections necessary.
Spot RT is a viable inspection tool and, is in fact, used regularly, even for critical fabrications such as pressure vessels. Having said that, it's important to realize that there are limitations to any inspection method and, with spot RT, you're not getting the whole picture. Recognize too, that most codes impose further inspection, up to 100%, when defective welds are found.
From your comments, I think you're on the right track.

 

[metengr]

Yes, I would agree with weldtek. Hydrotesting is only used to find through-wall defects in welds and base material, it is not appropriate for this type of application.

If it were me, I would perform a 10% RT sampling of production welds and increase coverage as necessary until performance issues or technique are handled.

 

[racookpe1978]

I think you're mis-interpreting/mis-using the backing plate. Particularly by using galvanized steel as the backing plate.

A few basics first. For any weld, the backing plate is used to substitute for the "back of the weld" on the root pass. After the root pass and first two cover passes, the "back of the weld" is the new metal already deposited, so there is no use nor purpose to a backing plate or purge gas in the pipe. The new metal is being deposited as melted steel/weld filler, so the melted and cooling steel need a purge gas/backing plate to ensure there is no "crap" (oxidized metal, dirt, contaminants, partially melted or partially covered slag, etc.) deposited inside the melt volume where you can't see it or grind it out.

Most important thing is to ensure the unprotected back of the weld is covered by something pure and uncontaminted and unaffectwed by heat while it is being deposited, while it is melted and before it completely freezes to a solid mass. Make sense?

Notice that the "front" of the weld is already protected by the SMAW (stick) welding slag or the TIG gas or the wire electrode coating and gas. A backing plate or back purge isn't going to do any good against contaminates or inclusions coming in from the front!

For regular hollow pipes using TIG or SMAW (stick) welding, we purge the back of the pipe with Ar or He gas or a mix or Ar-He gasses (because they are non-conductive, non-water-bearing, and non-reactive with the melted/cooling metal), then qualify the welder to fill in the gap with is root pass and cover passes without burning through new holes or the pipe walls. And without gobbing up the inside of the pipe with driblets of steel or unmelted weld wire.

A backing plate or backing ring can be used inside a pipe to fill across that gap and make welding easier, and to reduce the amount of purge gas and time of purging. But the backing plate is NOT part of the new pipe or new weld - is is just to allow the weld metal something to "melt against" to form the pressure barrier.

A loose backing plate doesn't do you any good: It still lets the contaminating air to get to the back of the weld while it is still melted. And, if the current loose backing plate is partially consumed while welding, then it can't slide out.

A dirty backing plate does you no good (and much harm) because it passes the dirt into the metal you're trying to keep clean, so I am very suspicious of using the galvanized plate. I understand the sq tube is NOT galvanized while being welded ??? True? (You can't weld galvanized steel without grinding off the galv coating first anyway. And even then the weld fumes are dangerous. Very dangerous if no grinding is done.)

SO, my recommendation?

Use a thick copper plate (1/4 to 7/16 inch thick) just under the width of the inside of your sq tube. It should be 10 to 20 inches long to absorb heat. Attach with rivets countersunk screws springs/spring steel/a bent plate to the back of the copper to force it up firmly against the inside of the sq tube walls.

Test by pulling the copper backing plate through your bent sq tube like a mandrell. When it goes through smoothly and several times, you know your fixture works. Then pull the copper plate through while welding right above the copper.