A question on pressure testing welds?

[AdrianH]

This is just an idea before you all tell me I should not as it's dangerous! But, if I made a box/cube out of 2mm steel and each edge was say 4 inch long, in one flat drilled a hole and welded a flat nut or flange so I could stick water in to it and pressurize it, what would be a good safe pressure to test that the welds were OK. How does one work it out?

I am not talking a 1,000 psi here as I think 2 mm wall would have ruptured well before with three faces trying to lull the corners apart, but say would 50 or 60 psi water pressure, be OK.

If you are doing fuel tanks say diesel or petrol, expansion tanks in vehicles etc. how would you test them?

Or do I have it wrong and do you pressurize with say 20 psi air with a gauge on and check an hour later for leaks?

Adrian

 

[Kent]

Pressure vessel work normally requires welder and proper QC testing stds. If its for yourself and there is no risk to others I think a suitable overload test might be employed, I don't think little leaks are the issue at 50 or 60 psi it is effectively a potentially damgerous bomb

 

[AdrianH]

It is for me, and I do understand the bit about being a bomb which is why most testing I have read about is with a liquid and not air. But how do I work out what would be an suitable test pressure. I am learning on my own, have tried to even pay for tuition with out success so all I can do is practice and practice, and what looks good on the outside may not be penetrating on the inside correctly, hence the idea of a test on my own work. Still only idea stage at the moment.

Adrian

 

[mike os]

water pressure test is fine.. NEVER use air or compressed gasses

 

[eSCHEn]

You could do the water test in a tank of water. Would prevent a lot of mess although 4" cubed is only about a litre so not that much really.

 

[daleyd]

When I did my appreticeship we had an introduction to welding and made a cube welding each of the sides with a different method (gas, mig, tig,arc..) finally soldering a threaded fitting on one face. This was then connected up to a hosepipe.......

The instructors was knocked up in about 15 minutes and was completely watertight. It took the rest of us about 4 days and all we succeeded in doing was making square sprinklers

 

[rtcosic]

I am not talking a 1,000 psi here as I think 2 mm wall would have ruptured well before with three faces trying to lull the corners apart, but say would 50 or 60 psi water pressure, be OK.

If you are doing fuel tanks say diesel or petrol, expansion tanks in vehicles etc. how would you test them?

Hose pipe.
Mains water supply pressure is usually around 3 bar (45 psi) - more than enough to show up leaks.
Or air via a bicycle pump to 0.2-3 bar (3-5 psi) and place in container of water - just as you did as a nipper when fixing punctures in your bicycle.

More than enough to show up leaks if that is your objective.

 

[AdrianH]

OK so doable then, picked a cube as 12 edges and 48 inches to weld up, will check my scrap pile and cut the steel tonight have a start.

Will go for the water under pressure, will have to see what gauges I have may have to turn a flange. I have got some npt taps some where.

I may just try and take it to bursting at some point, depends on if I get it to seal in the first place.

Adrian

 

[Bill Edwards]

If you want it to get some idea as to the strength of your welds then forget it.
You'll need a lot of pressure to blow apart even some of the most awful welds. You'll be far better welding test pieces then acid etching them with common household products and doing nick break tests.

If you want to see if they're water tight then filling with water and pumping up to 20 psi will quickly reveal any holes. But don't expect this to reveal defects.

 

[Bsmith]

Could you not pressure it up with a pressure washer to test for leaks? Would need to vent air probably.

 

[ben8vgte]

If you just want to test for leaks a cheap pressure tester for plumbing from ebay or such will do the job perfectly.

 

[AndyStobbs]

What is the anticipated pressure the part will see in service? This will dictate your test pressure.

 

[AdrianH]

Could you not pressure it up with a pressure washer to test for leaks? Would need to vent air probably.

Now that could be dangerous I have seen people use pressure washers to do simple hydro-forming. Even the small karcher I have says max pressure 110 bar which is like 1,500 psi.

What is the anticipated pressure the part will see in service? This will dictate your test pressure.

. It not got one it is just for me to have some practice on and then test for leaks. I really could just buy one from ebay or a plumbing shop as Bsmith says, but I think I will test to water pressure which is around the 3 bar figure as suggested.
When someone builds a fuel tank for gas oil and say it is 2 metres tall the the pressure on the base of the tank is 2.38 psi (not 10 or 11 as I originally had), and I was wondering how they test it before filling and just hoping for the best.

Adrian

 

[sardine]

Fuel tanks are tested at 3.5 psi for inland waterways boats. That's enough to find any leaks.

 

[AdrianH]

Fuel tanks are tested at 3.5 psi for inland waterways boats. That's enough to find any leaks.

Thanks for that information, I have been searching and I guess I found a similar answer in this http://www.tqc.co.uk/pdf/fueltank_leaktest.pdf
I did find this which mentioned your scenario
"A simpler test, that is also considered a hydrostatic test but can be performed by anyone who has a garden hose, is to pressurize the vessel by filling it with water and to physically examine the outside for leaks. This type of test is suitable for containers such as boat fuel tanks, which are not pressure vessels but must work under the hydrostatic pressure of the contents. A hydrostatic test head is usually specified as a height above the tank top. The tank is pressurized by filling water to the specified height through a temporary standpipe if necessary. It may be necessary to seal vents and other outlets during the test."

But it did not mention what height?

Adrian

 

[Maker]

A hosepipe will show any leaks. You could use a pressure washer and it would be safe, provided there is no air in it, all that would happen is a little pinprick hole would opun up and squirt water out and all your pressure would go instantly.

acid etching them with common household products

I don't know how well this will show up on thinner material like this but you can use Citric acid for this, I got some food grade stuff online years ago for next to nothing and it works perfectly. You do need to prep the edge really well first, starting with the cut in half weld, flapdisc on a grinder then progressively finer grits of sandpaper, either by hand or on a ROS, and finally by hand with wet 'n dry paper, the aim being to get it really polished. Now you apply the acid and leave it for an hour or so.

 

[Rig Pig]

This is just an idea before you all tell me I should not as it's dangerous! But, if I made a box/cube out of 2mm steel and each edge was say 4 inch long, in one flat drilled a hole and welded a flat nut or flange so I could stick water in to it and pressurize it, what would be a good safe pressure to test that the welds were OK. How does one work it out?

I am not talking a 1,000 psi here as I think 2 mm wall would have ruptured well before with three faces trying to lull the corners apart, but say would 50 or 60 psi water pressure, be OK.

If you are doing fuel tanks say diesel or petrol, expansion tanks in vehicles etc. how would you test them?

Or do I have it wrong and do you pressurize with say 20 psi air with a gauge on and check an hour later for leaks?

Adrian

From hard learned experience the best thing to test a diesel tank with is diesel - I repaired a 4x4 tank a few years ago looked fine, filled it with water - no leaks, brilliant so I refitted it and tipped a jerry can of diesel in, went for a cup of tea and when I came back there was a wet patch on the tank....

dye penetrant is probably the easiest way of testing a weld in the home workshop, it will find any pinholes in the weld.

 

[daleyd]

From hard learned experience the best thing to test a diesel tank with is diesel - I repaired a 4x4 tank a few years ago looked fine, filled it with water - no leaks, brilliant so I refitted it and tipped a jerry can of diesel in, went for a cup of tea and when I came back there was a wet patch on the tank....

dye penetrant is probably the easiest way of testing a weld in the home workshop, it will find any pinholes in the weld.

Don't you need access to the back of the weld for that? (I'm not sure, but I thought I heard that somewhere?)

 

[Bill Edwards]

Yes, diesel is very good at finding leaks as it penetrates, water doesn't.
But water dries off and won't set on fire or contaminate the weld if you find a hole to fill.

 

[Maker]

dye penetrant

What's that?