Express Leak Detection

What is the gas? Air? Nitrogen?

He?

"hermetically (O-ring) sealed" is an oxymoron.

Dunking it in hot inert fluid an option?

What is the gas? Air? Nitrogen?

He?

The most practical that could be used.

"hermetically (O-ring) sealed" is an oxymoron.

O-ring used to designate the type of seal.

Dunking it in hot inert fluid an option?

Looking for the least amount of time invested.

OK, what is the DUT? Is it filled with air? Inert gas. Under pressure/vacuum?

I've done He leak testing on true herm. sealed (glass frit) and O-ring sealed modules.

Also did gross tests by dunking in hot fluids. Maybe a glass jar and a vacuum pump with a beaker of juice?

Not sure what your goal is here, but never used the chambers of which you speak.

stop with the games, what is this thing and its purpose?

Down at the gas station they put soapy water on the tire to see where the leak is.

Seriously, I can't tell from your description, but would immersion in a hot liquid cause gas inside to bubble out indicating leakage?

If I had my druther's, I'd prefer vacuum....but is the hermetically sealed vessel to be operated under pressure, if so how high.....?

The quickest seal test I know of is the testing of the seals of a shock absorber. They just measure the dampening coefficient. It took the shock absorber fabricators years to make a good and quick inspection design.

You seem to be much too focused on your desired goals and not sensing a misapplied gasket. Remember that NASA suffered a real O-ring disaster from looking to the goal and not the problem. I would stop worrying about how quickly or how much volume will spill until you find a reliable method to detect a poorly mounted O-ring.

If this was my task, I would start by tracking how often these O-rings are misaligned from your history of failures and successes. Then I would take a random sample of these O-ring assembled parts and measure every little thing that I can from the outside to try and predict which was miss assembled. Don't limit your measurements to external dimensions of the assembly, include time of day, period between machinery maintenance, etc. Only once you can correlate something with a failure should you worry how to measure that quickly.

I dont think he's repairng, my bet is proof of concept

I don't think he's that far along. He's asking us for a concept to meet his desires.

In an ideal world, the person or persons involved in installing components during assembly of a product would have what has been called "Pride in Workmanship".

If this were the case now, tests like these would not be necessary. My theory for testing in plain view of the person or persons responsible, may be at very least, "inspirational".

An example of this, was the "reject station" on a press line for an auto manufacturing plant, that ejected a defective part from the production path so violently that the noise was noted by the persons responsible for the defect.

I've got it.

Hook the assemblers up to an electrode that shocks them every time one of their assemblies fails the leak test.

Just kidding.

Some of the smart guys here will have some help for you.

Hah!

Like the "Honda" story.

The Boss, H. Honda, visiting the motorcycle assembly line, noticed a loose fastener. When the person responsible was pointed out, Mr. Honda asked for the wrench and thumped the guy on the head.

I vote for shock collars and "trunk monkeys".

The trunk monkeys are issued a hammer and placed inside all control cabinets and rooms that have "Do Not Enter - Authorized Personnel Only" posted on their doors.

If anyone other than those that are authorized open the door, the monkey punishes them severely with the hammer.

You could always start here:

https://asnt.org/Store/ProductDetail?productKey=42a1e772-ef52-464e-a20c-baded9ec19ad

... Learn a thing or two...

And pick-up a useful certification along the way...

PS : I, for one, do not see one iota of useful information yet {from OP}, from which to string together a useful suggestion...

Garbage-in / Garbage-out

Helium mass spectrometer an instrument commonly used to detect and locate small leaks.
http://en.wikipedia.org/wiki/Helium_mass_spectrometer

If you need high speed testing I suspect that you are designing a plant to make these in high quantity. I have never designed test equipment for anything as small as your application. For a manufacturing line making 45 gallon drums we clamped shut the two openings, injected air at 10psi gauge pressure and listened for leaks with ultrasonic sound sensors. Test time was about 12 seconds including loading and unloading the drum along a conveyor. Of that the largest portion was the time taken to fill 45 gallons to 10psi. Study a carousel type bottle filling machine. Each test may take 45 seconds but if you have 45 units in the machine, you production line can run at one unit per second.

Ultrasonics will not detect a missing seal. Concentrate on making sure that the seals are applied correctly. Looking for faults addresses the symptoms of units badly assembled, not the problem of assembling units badly.

Some remarks:

- the pressure change versus time is proportional to pressure thus it is better to pressurize the closed volumes (reference and sample) and measure the dp/dt difference between the 2. You eliminate the error due to set pressures errors if you have only one volume. The dp/dt evolution can be measured with 2 pressure transducers calibrated and with same ADC. The higher the pressure you can load the faster the test.

- the problem is much more complex since you have in fact for each volume the own leak and the leak of the closing element which can wear and present even higher pressure decay rates then due to the O-ring defects.

Instead of trying to find a way to check quality I would prefer to search a way to avoid errors and following malfunction introducing in the process the steps or elements which guaranty quality since all checks are added cost how short they are.

How is the assembly you want to check ? Is the O-ring a radial or an axial seal ? Is assembly automated or by hand ? Do you use specific tools to place the seal ?

Which kind of assembly problems have you found ?

If you can give answers to above questions I could give you some hints from own experience in sealing of hydraulic and pneumatic systems.

Tiny gas leaks produce ultra sonic frequencies, I was using electronic leak detectos around 1966-7 in the RN with huge success!

Have you looked around for something more modern and possibly better nowadays in this area? Quicker/Cleaner/Safer?

Best of luck.

The good news is that there are solutions to your problem. However, in order to obtain these solutions you must first describe the test sensitivity you require. If you are looking for sensitivity of 1E-3 std cc He/sec the test method you will use will be entirely different than if you require sensitivity of 1E-6 or less. The reasons you are not getting good answers to your inquiry are that you are not asking the right questions.

Do you really believe OP is looking for a sensitivity of 1E-6 or less, using O-ring seals?

I don't.

That certainly isn't an "Express Leak Detection" test.

But, we really don't know what OP is doing. Must be a secret.

I totally agree with you this is the reason I suggested pressure decay rate as criterium, the volume is small so that reaction will be sensitive and over alla cost less than for other methods more stringent.

What you or I believe is not the issue here. OP should cough up some details on his requirements. Lacking that, I can suggest reviewing the Nondestructive Testing Handbook Vol.1 - Leak Testing.

Agreed.

Here are some interesting articles:-

http://ultrasonic-leak-detectors.com/

http://de.images.search.yahoo.com/search/images;_ylt=A9mSs2zJ5e9Tyj8Aqk4zCQx.;_ylu=X3oDMTB1NHEyMmtrBHNlYwNzYwRjb2xvA2lyMgR2dGlkA1VJREVDMDFfMQ--?_adv_prop=image&fr=linkury-tb&va=ultrasonic+leak+detection

I don't know what you want to test or what its rating or standard it is to be tested to is.

We use IP tests for our products.

http://en.wikipedia.org/wiki/IP_Code

Does that help?

And, NEMA rating in the USA. That's not where I think OP is going.

"Leak detection" refers more to air tight than water/finger intrusion proof.

From what you described you are looking at a pressure decay tester. I have seen many of these used in testing items in similar size to what you describe. (Automotive side turn lamps and such) and they are highly reliable and extrememly repeatable.

Calibration once per year revealed negligible drift in instruments.

The major issues were with the integrity of the "master" unit and the maintenance of the interface seals where you mount the Device Under Test. (DUT)

The duration of the test depends on the precission required for your results and can be calculated from the total volume, error calculations and the allowable pressure loss.

Pressure is preferred to vacuum as the pressure will place strain onto any joints and tend to not "pull them closed" and so give a false "good" result.

We had at least five such units operating, with master artificats for each design/model of component. Test times were typically as low as 10 seconds.

We were testing ultrasonic vibration welded plastic parts, using the lamp socket as the interface. One line was running over a million (our million 1,000,000) part per year while most others were running around 500,000 tests.

Sounds like a good choice. Also provides indication of the magnitude of the leak for analysis of process creep.

I am glad that you share my point of view. I appreciate.

The OP does not give the acceptance level for leak rate, but he gives a good index: "sealed product to detect the lack of, or damage to, an O-ring after assembly".

I think he do not need a very sensitive method like He mass spectrometer. Surely with a pressure decay method as already said is enough. The leak rate level for the case of O-ring lack must be rather high and easy to determine.

The case of O-ring damage should result on lower leak rates but in a one minute decay test must be evident the O-ring damage (if any)

Kind regards

First, let me apologize to everyone for my delayed response.

To elaborate, the device under test is a submersible electric pump, divided up into sections that include the electrical control, stator housing and mechanical seal section. All dependent upon O-rings to maintain seal against static and developed pressure.

The test I searched for and found described the "chamber tests" mentioned above. This test appeared to me to lend itself well to rapid testing of an assembled product, but very little detail was given.

With the mention of the differential method, I imagined a rotary "clamshell" set of chambers at the end of the assembly line. What ever method that could detect a displace or damaged O-ring as quickly and efficiently as possible would serve at the very least to call attention to workmanship at the production/assembly level.

These pumps are classed IP68.

Thanks in advance for the help.

That is the reason I presented to this forum.

I'm back to a bucket-o-hot water/dunk test.

I agree this would be the ideal test, but the drawback here would be the containment of the water and drying of the product in an assembly line environment.

There is also the issue of hygroscopic absorption for the power cable end.

I'm confused. You don't want to get an IP68 device wet?

OK, sometimes KISS isn't always that simple.

There are a bunch of dielectric liquids available to perform these tests besides water. We often use Flourorinert liquids such as FC-77, FC-40 etc.. They are inert leak testing fluids designed for these applications.

As already posted, a pressure decay type leak test instrument from any of several suppliers is probably your best cost alternative. Search "Leak test instrument". Cycle time is dependent on total volume and desired sensitivity,with larger parts and higher sensitivity taking more time obviously. Most of these suppliers will help you with that and will estimate the time needed for a specific test limit. You'll need a reliable, repeatable, and leak free means of connecting your product to the instrument for pressurization (vacuum takes too long). You will also need a "0-leak" master of the same volume for tooling verification.

For more $ you could go with a helium leak tester with either a manual wand or a sealed chamber. Add "helium" to your search to find an instrument supplier. You'll still need a reliable, repeatable, and leak free means of connecting your product to the instrument to pressurize the interior. The medium is usually some inert gas with a small % of helium so it's not too expensive. The wand lets you locate the leak as well, a nice feature.

I've also built dunk tank testers - you'd be surprised how may automotive parts (ex. valve covers) are tested that way. You'll still need a reliable, repeatable, and leak free means of connecting your product to the instrument for pressurization. Drying the pump is another opportunity.

Or, you could consult a company that builds factory automation equipment and tooling, like the ones I've worked for (and continue to), to design and build this thing for you. Naturally I like that solution for you, but you can't hide the cost of your labor when you contract the work out.

The general consensus appears to be of the Time/differential pressure test.

My thanks to everyone for leading me in the right direction.

Neutron Radiography.