Leak Testing of component under high pressure?

Dear Orc General:

Create a reservoir. Connect the valve to the reservoir. Place a pressure gauge on the reservoir. Purge the system of air/ fill it with liquid of your choice. Pressurize the liquid to 800 psi. Record the pressure over an hour/day/month, or count the drops that come out of the valve.

David A. Smith

Most leak test systems don't use liquids, even though the system under test is designed for liquids in use. A compressible gas (like air) will let you pressurize the workpiece (maybe with a reservoir connected in line) and then watch the pressure decay in response to any leakage.

There are two main advantages to this method. First, with high pressures and small reservoir (or workpiece) volumes, the amount of volume leakage needed to produce a very measurable pressure drop can be quite small, but is always proportional to the volume lost. With fluids that aren't compressible, the pressure can go from maximum to zero with only a tiny little bit of volume lost. You might detect a leak; but you'll have a VERY hard time quantifying it.

With air, and with pressure being nicely proportional to volume losses (you'll want to bone up on volume/pressure equations from your old fluid dymanics text), leak rates can be calculated quickly and accurately, even if the leakage is very small. This means that leak tests don't have to take a lot of time. Also, air will leak quickly through holes that fluids might only seep from. This means you can "amplify" any leaks that exist, and get real results more quickly.

And, of course, if there ARE leaks in what you're testing, the only thing leaking is air. You don't have to deal with catching or containing fluids that might come out at high velocities and high pressures from any direction.

There's a company in Marine City, Michigan, called TestRite, that does a LOT of very creative stuff of the exact kind you seem to be looking for. It'd be worth a phone call for sure.

KG __ I'm sick of spam. The 2 in my address doesn't belong there.

Orc General wrote:

Dear Kirk Gordon:

But you are not going to compress the gas to 800 psi, are you? A word of caution on how much pressure to apply would be in order. Otherwise his next of kin will come looking for you.

...

David A. Smith

N:dlzc D:aol T:com (dlzc) wrote: >

No, 800 PSI would be a bit much in the air pressure department. I figured the next of kin would spare me for two reasons. First, I don't expect many people to have that kind of air pressure just lying around, waiting to be plugged in to the first inflatable toy that happens to become available. Second, I DID recommend a company that knows how to do this, and that has the skill and experience to fill in the details that I didn't include (and plenty that I don't even know.)

The thing about air leak testing is that it doesn't NEED the same pressure as the application will endure when it's in use. If you're looking for leaks (and NOT burst-strength results, or other nasty stuff like that), then all you really need to know is whether gasses or fluids can get out of a container. In most cases, if they can get out at high pressure, they can also get out at lower pressures, though in smaller volumes. Using air instead of fluids often increases the rate of volume loss through any given leak point, since air will frequently go through holes that fluids won't. That makes it possible to produce detectable rates of loss without gigantic pressures.

Higher pressures can cause some special problems, of course, that the design of low pressure tests needs to account for. If the workpiece under test, or it's seals, or whatever, are prone to deformation at high pressures, then leaks might exist only when the pressures are dangerously high, and might not be detectable by safer and saner methods. In that case, this is a truly demanding situation, and will need some very specialized equipment and procedures.

Even at relatively high pressures, however, air seems to me like a better and safer medium for leak testing. If air leaks out of a workpiece at, say, 200 or 300 PSI, that can be a hazzard. But reasonable caution, and maybe a lexan shield or two, can save the next of kin from needing to dust off their black suits and dresses. However, that same leak, at the same pressure, could cut holes in a brick wall, if the fine, narrow, high-velocity steam of leakage happens to be made of something like hydraulic oil. Equally important, gas can be compressed. You can pressurize a test subject, then turn off the air source and wait to see what happens. If a leak results, it only lasts as long as the contained pressure does, which can be a very short time if the internal volume is small. Hydraulic systems can't just be pressurized and then isolated. If the source is turned off, then pressure goes to zero immediately, with or without leaks. This means that if a leak starts, it won't just be a short and diminishing burst of danger, but a continuous stream that lasts until either the pressure source is turned off, or until all the fluid is gone. Not good. (An accumulator can maintain pressure when the pump is turned off; but that's one of those details that's determined by good test design, and which is way outside the scope of a newsgroup discussion.)

So, to answer your question directly: I HOPE nobody is busy hooking up 800 PSI air lines, this morning. I certainly didn't advocate that explicitly. And since gas testing offers some advantages at low pressures, I'm hoping that I've merely pointed the author of the original post in a direction that will keep him out of trouble, rather than getting him into any.

I think maybe I'm gonna grow a beard and change my phone number though. Just in case. You can never be too safe.

And thanks for jumping in with your concerns about safety. I hope your question was unnecessary; but it was absolutely appropriate.

KG __ I'm sick of spam. The 2 in my address doesn't belong there.

I'm familiar with the concept of testing high-pressure gas systems with fluids, but not the converse. I presume that at a given pressure, the energy stored in a fluid is vastly lower than in a gas, which should provide some added measure of safety when testing a system for leaks. I'm not sure I'm comfortable with the idea of testing a fluid system with gas at the rated pressure.

jim andrews

The rated pressure will be a problem no matter what you do; but gas testing has some advantages. Please see my response to a similar question, earlier this morning.

KG __ I'm sick of spam. The 2 in my address doesn't belong there.

I'm not proposing that you be legally liable. But one can rent a cylinder of compressed gas for less than $100, and it is delivered with contents at

Don't worry about a reply. Just keep this in mind. Not everyone is looking for a good or correct answer. Some are just looking for the first, easiest way out of their particular "trap". A little "disclaimer" is always in order.

I don't mean to jump down your throat...

David A. Smith

Dear jim andrews:

His recommended solution is to take it to an expert. I concur. It is always cheapest, and the have the expertise and training to evaluate perhaps even more than you require.

David A. Smith

No problem. And I understand the part about some folks not really wanting a good answer. I generally assume that they do, however, and offer the best I can when it's a topic that I think I can contribute to. It seems to me that if too many of us become too cynical about the content of online newsgroups, then the value of the groups is diminished for those who DO give and take the best information they can.

Also, it's been my experience that there are two or three lurkers for every person who actually participates in group discussions. I frequent another group where I know this is true. In response to things I post in the group, I'll sometimes get notes in my my personal e-mail that include comments, thanks, complaints, or follow up information that I find interesting or informative, even though the people sending these have NEVER posted in the open.

So, you never know who's watching, or how many of the silent types might benefit from what goes on in public. A disclaimer is a good idea, I guess; but it seems like it ought to be unnecessary. I didn't charge money for my advice, after all. And, as Huck Finn once said about a preacher: "He didn't charge nothing for all that talkin', and it was worth it!"

KG __ I'm sick of spam. The 2 in my address doesn't belong there.

You know, it's a strange world.

Except for gross and obvious leaks like missing plugs or casting porosity, I think I've seen more leaks that closed on rising pressure than leaks that opened on rising pressure.

I have used air as a medium for leak testing, but mostly at very low pressures, amounting to a few tens of inches of water column. O-ring seals in particular are self energizing, so it's not uncommon to find an o-ring assembly that will leak at 2" wc, but not at 200 psi.

Personally, I've never leak- checked with air above 100 psi shop air pressure, and the part was under a foot of water when the pressure came up. I've never asked a production worker to do a gas leak check above 15 psi.

All of the hydraulic systems on which I've worked can indeed be pressurized and isolated. The pressure stays up for a considerable time because of trapped gas and strain energy. I've even seen the pressure increase by half when the system was isolated. But releasing the pressure from such a test article, even at very high pressue, usually happens without a lot of fuss. Not so for pressurized air systems; they store a lot more energy. I disagree with your evaluation of the relative safety of pressure testing with air vs. liquid.

On an unrelated note, do not rely on polycarbonate shields to protect you from shrapnel. The stuff has some odd behaviors when subjected to chemical attack and subsequently stressed. It's chemically attacked by a lot of common substances found anywhere, including glass cleaner, Loctite, and soap.

-Mike-

Dear Mike Halloran:

Not that a spool valve "opens on rising pressure", but they do leak more at higher pressure. Spool valves are used commonly as directional control valves.

And I completely agree with the closing on rising pressure part. We use check valves to *slow* the reverse flow of water into an ozone generator. Because no ozone resistant material seals on only a couple of feet of head.

pressurized and

And the funny thing is, you can flow 1-2 wt% ozone gas through it with little or no effect for years... well it gets a little hard to see through.

David A. Smith

To me, the normal 'internal leakage' of spool valves through their clearances is distinct from the 'external leakage' for which one would test an assembly. Unless one were manufacturing spool valves.

That is funny; I'd expect polycarbonate's impact resistance to fall, a lot, with reduced transparency.

-Mike-

Dear Mike Halloran:

When used as a variable area flowmeter. there is not a lot of impact going on! ;>) Yes I know we were talking about valves...

David A. Smith