Pressure testing with compressed air 2

[Grindy]

I was tasked with designing and manufacturing a test rig to be used to ensure that aluminium castings are not porous. The customer for these aluminium castings had been testing the castings previously, however we have now been tasked with carrying out the testing.

The customers test procedure was as follows:

- Seal/cover the casting ends to create an enclosed vessel
- Using compressed air, pressurise the casting to 10 PSI
- Close off the compressed air supply and using the attached pressure gauge ensure that the casting maintains 10 PSI for 10 minutes
- Should the casting maintain the pressure, the compressed air will be exhausted and the casting will pass the test

The reason the castings are pressure tested is that they have been known to leak. This is a problem as when the casting is in service it will be used beneath water and the cavity which requires the pressure testing houses electronic/electrical components. The castings are double impregnated as the leaks are due to micro porosity.

The internal volume being pressurised is approximately 1.5 litres.

I have created a test rig which securely clamps the casting to a steel plate, thus covering one end of the casting. A separate plate is secured to the other end of the casting, creating a sealed vessel. Compressed air is then fed into the casting via a hole in the steel base plate. Once 10 PSI is showing on the gauge the ball valve is closed sealing the casting. The casting must then hold the 10 PSI to prove it has no micro porosity.

My question is this:

I understand that using compressed air to pressure test is not as safe as hydrotesting.

However using water/fluid will not be practical due to water marks left on the casting and the mess created by filling/emptying the casting.

If I reduce the volume inside the casting using a solid steel billet so that less compressed air is required am I making it safer?

Do I need to place a shield or guard around the casting during testing?

Does anyone have any other suggestions as to how I can make this test as safe as possible while still using compressed air?

 

[handleman]

thread794-26767

-handleman, CSWP (The new, easy test)

 

[btrueblood]

Yes, reducing the volume of air will reduce the danger.

Can you test with the casting held under water? Doing so allows the water (open tank) to absorb a great deal of any potential blast should a rupture occur, and is a quicker test (no bubbles = good part). If not, then a shield would be strongly recommended.

A blow-out plug or rupture disk, that will relieve pressure in the test device if it exceeds some limit (say 15 psig) is also a good fail-safe device (for when the regulator goes south, or somebody fiddles with it).

 

[MikeHalloran]

Also, work with your customer on finding a casting sealant that's dependable enough to not need testing. Call your local Henkel Loctite expert for assistance.



Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

adding to Mike's post: and work with your foundry to ensure porosity controlled to a size that is dependably sealed by the sealant.

 

[Ron]

Put a guard around the test specimen! In the US you would be required to do so under OSHA regulations. In the UK, I understand you have similar regulations.

 

[Pud]

Why not get a proper leak tester?

e.g.

http://www.tqc.co.uk/leaktesting/series10_standard_leak_test_fixture.htm

(This is in the UK, but I'm sure there is similar stuff wherever you are.)

Used to use this sort of kit to test gas domestic cookers - fast and accurate.

H

http://www.tynevalleyplastics.co.uk

 

[Cockroach]

Yeah, be careful. I blew up a valve like this, caused over quarter million dollars damage to the building. Nobody hurt, but was knee deep in OH&S personnel for months.

I would put 1 atmosphere internal pressure to your unit, but then submerge it in a water bath, swimming pool if you like, and look for bubbles. You're not at an incredible high pressure, so elastic expansion of the gas providing the necessary "explosive decompression" along the pressure containing envelope is not an issue here. Still, 14 psig can do damage if a small component breaks free and comes at you.

The water surrounding will dampen any projectile motion, in addition to being a great medium for visual failure in the pressure boundary.

But be careful.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[btrueblood]

Ron, I think OSHA only gets involved if the pressure goes over 15 psig, or above a size limit well in excess of the OP's stated 1.5 liters. But agree, adding a guard "just in case" is never a bad idea.

The beauty of a submerged test, is that the water is acting as a guard, and is also part of the functional test being performed, thus it is unlikely that an operator would remove or defeat the "guard" for convenience.

Water spots on the aluminum parts can be minimized or eliminated by suitable treatment of the test water (e.g. rinse aids, like you put in home dishwashers).

 

[Yagonyonok]

Could you possibly pull a vacuum instead of pressurizing? At only 10spi you could get the same differential pressure, and just watch to make sure that you are holding the vacuum instead of holding the pressure.

Engineering is the art of modelling materials we do not wholly understand, into shapes we cannot precisely analyse so as to withstand forces we cannot properly assess, in such a way that the public has no reason to suspect the extent of our ignorance.
-A R Dykes

 

[Grindy]

Thank you all for your responses.

I like the idea of using the submerged test. However due to the size and weight of the casting and fixture I have no way of lowering the fixture into and out of a tank. (The guys on the shop floor would not be able to lift the test rig and casting safely by hand and we have no lifting equipment that could do it.)

I did give the wrong volume in my original question. The correct volume is 3.4 litres.

I have used the equation from the link Handleman gave me:

E=P1*V*(ln(P1/P2))

Where E=Stored Energy in kilojoules
P1=Pressure of tank or piping in kPa (absolute)
V=volume of system in cubic meters
P2=Atmospheric pressure in kPa (absolute)

We then divide the stored energy by 5788 kJ which is the amount of energy in one pound of TNT and evacuate the area accordingly.

(Reichertc)

I calculated the stored energy inside the casting to be 0.306kJ. This was based on a volume of 3.4 litres and a test pressure of 10 PSI.

I am able to part fill the cavity with very small sand bags which will reduce the volume (I was going to use a steel block but because of the shape of the casting I am unable to fill a large enough volume, unless I put lots of small peices inside).

As I can't presently submerge the test rig in water, would you all agree that 0.306kJ is a low enough value to make pressure testing with compressed air safe in this instance?

Some of my colleagues think I'm being over cautious and it will be fine as its 'only' 10 PSI. From your responses I appear to be right in being cautious?

 

[zeusfaber]

Have you calculated the likely burst pressure of this casting? How does it compare with your proposed 10 psi?

Although not free of risk, what you're trying to achieve here is not the same as the classic "Take it to 1.5 x WP" and see if it ends up the same shape and size as when you started" exercise.

Before working out how much effort you want to expend making the test safer, it would be a good idea to understand how likely you are to burst the system (or blow your fixture apart) and hence how much risk you are attempting to remove.

A.

 

[ColonelSanders83]

The concern with Air test is pressure and volume. You have niether. If I were hanlding a test like this I wouldn't even worry about it.

Nobody thinks twice about the 40 psi and 16 liter pressure vessles used to maintain contact between your car and the road (tires) nor does anyone worry about an air compressor in their garage routinly around 8 liters and 100 psi with no proper maintence.

If you want a formal risk evaluation procedure I would recomend ASME PCC-2, it has an exellent section on energy calcuations and exclusion distances. I have used thins many times for large volume pnuematic tests up to 880 psi.

Just my two cents worth.

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[dhengr]

ColonelS:

You took the words right out of my mouth. I have watched this thread with considerable trepidation and haven’t driven in the last two days, out of fear, because I have to walk past the compressor in the garage, and past some tires to get near my car. Your two cents worth...., just a little logic and common sense; shows that simplifying a problem to its basics, or comparing it to a concept we can grasp, the water pressure in your house piping, sometimes makes it become much less relevant and scary. Wooten’s Third Law states: “The acquisition of uncommon knowledge inhibits the application of common sense.”

Your two cents worth... has far more value, at least equal in real terms, to what the various employers and companies have spent, over the last two days, in lost productive time to underwrite and facilitate this intellectual exchange. Maybe a few of the posts weren’t made from the work place.

Why not run this pressure test before the first impregnation has cured, thus driving the sealant further into the porous material. You might not need the second impregnation, or apply that to the outside of the casting since if this is operated underwater you might get an inward pressure of more than 10psi. By the way, if you don’t test underwater, for fear of water spots, when this is going to be used underwater, how do you know it is the casting leaking instead of the gaskets on one of your end cover plates?

 

[monkeydog]

Grindy,

It sounds as if this is going to be an on-going production test.

My first choice would build a "burst chamber" with tubing or flex lines going in and out of the chamber for quick set-up and test. But cost is probably a driver. You have access to sandbags, why not put sandbags around casting during test?

An argument for the burst chamber with existing valves, gauges and lines, is if there is a leak, the first thing you will do is check if the tooling is leaking. If your existing test set-up is a "loose" collection of hoses, gages and valves, that would be suspect if there is a leak.

 

[EngAddict]

10 PSI will not do damage, wear proper safety gear and you will be fine. It isn't like you are pneumatically testing a pressure vessel to say 1.25xDP.

You could also do a soap sud test to see if it bubbles out (if it won't mark the casting). A lot of tanks are tested to 100 kPa w/o safety screens.

Alternatively get it tested in China. A client recently got some vessels we designed, based on burst test, fabricated and tested in china. There is a whole group of workers standing around watching this burst test with no safety gear or screens when it popped at about 6 MPa. Luckily it was a hydrotest otherwise it might have ended up in the next province. Dodgy.

 

[Artisi]

Have you considered a vacuum test, certainly a lot safer but probably harder to locate any leak.

 

[Grindy]

Thank you all for your help and advice. We didn't consider a vacuum test. However the test procedure was dictated to us by the customer so it was for this reason we pressure tested the castings.

So far the test rig is proving successfull.