Vacuum Leak Test - gauge and rate? 1

[IceMan30]

I am trying to set up a vacuum leak test. I want to apply a vacuum to an assembled panel we produce(~1'x1'x1.5" thick, very little internal space, basically just the space between internal components and outer bezel), then lock the vacuum and measure decay over time. I am thinking ~15-30 mm HG, enough so I know I have significant vacuum, and I plan to measure decay every minute (possibly more) for 5-10 mins. Knowing the true internal volume will be very difficult, if not impossible.

Three questions:
- Does that pressure range for a vacuum test of this scale seem appropriate?
- Any suggestions on a digital gauge? I'm hoping to keep cost to <$200 for now, as this is exploratory at the moment. I need a good range,I would like <0.5% error, and 3+ decimal accuracy, I think. Maybe I can splurge on a high end (>$1000) gauge later may be an option.
- Any thoughts on what an acceptable "leak rate" or "decay rate" might be. It will be very hard for me to measure/determine volume, so a true leak rate may be out of the question. Part of the point of this test is to determine leak rate ourselves for this application, but I didn't know if there might be a standard(s) from various industries.

Thanks!
Jimmy

 

[Compositepro]

If you have flat surfaces on your panel it will not be capable of withstanding much pressure, and its volume will change with pressure.
The simplest leak test that is still very sensitive is a bubble test. Just immerse your panel in warm water and look for bubbles. The warmth will expand the air inside to create the pressure so you do not need a connection for pressure, which may itself be the leak source.

 

[handleman]

You need to know what you are trying to achieve with your "leak rate". Some sort of IP rating? Why are you wanting to measure rate? Sounds like you don't have any particular target you're shooting for or reason to need to know the leak rate. Even if you are just wanting to set yourselves a benchmark for the future... What happens if you start not hitting your benchmark? Are those parts bad? Are they not going to perform? If you don't know what you actually need for your product to function properly, you may spend a bunch of time/money/etc. later trying to hit something that was set arbitrarily.

 

[btrueblood]

I wouldn't worry about absolute accuracy of the gage, you're measuring a differential which takes some of the error out of it, a 1% gage or transducer is likely "good enough". But I would create a test volume, making something like a cake pan (thicker wall though) with a known/measurable and fairly large volume that will take a measurable amount of time to "leak down". Measuring any leak rate into a zero volume gives an infinitely small leak-down time.

 

[cdill21]

Iceman,

Pressurize it and spray it with soapy water, or Snoop if you want to be more controlled.

We spent a long time trying to standardize a similar process. Moral of the story is that we rigidly controlled the upstream processes that contributed to the seal and only used a leak test for a catastrophic failure. Fact is, if your seal has many tiny holes, air can flow, but water can't, (think Gore-Tex). On the other hand, you can have one large hole that water will definitely get through, but may have a larger "leak rate" than all of your small holes.

To your standards question, I exhaustively researched this before and as far as I can tell there are no published leak rates, you would have to determine that yourself, but I wouldn't recommend it (gory details upon request).

Find out what IP rating matches what you are trying to accomplish, or what your customer wants, and test for that.

https://www.enclosurecompany.com/ip-ratings-explained.php

The testing is very direct. "I want my enclosure to be water tight for 30 minutes at 2 meters deep." OK, find a two meter deep water hole, put your enclosure in for 30 minutes. Did it stay dry inside? Yes? Congratulations! Your enclosure meets the requirements of IPX7.

Good Luck,

Chris

 

[IceMan30]

Thank you all for your replies! Some responses and perhaps better insights:

- These parts are currently tested via a positive pressure "bubble test," somewhat as described by CompositePro. We have a requirement to perform some sort of seal check, and this has been the method for some time. However we have reason to believe this test may be causing damage (due to blowing out seals, as positive pressure is applied inside the volume). We are looking for a viable alternative.

- To handleman's point, I am absolutely trying to come up with an arbitrary benchmark. That's how many of these projects start. My point in posting was to see if there is any "tribal knowledge" I can tap into to help me along the way, so I end up at a reasonable benchmark. Obviously each situation is different and requires its own engineering judgment; I was just looking for a little insight on the chance someone may have gone down this path before.

- brueblood - I get what you're saying about a larger volume, and I agree it would help with a number of issues; unfortunately it might be difficult to employ in this particular situation.

- Chris - thanks for the info on your own "standards" search. It has been slowly dawning on me that this might be a futile effort. The more I learn about 'leak rate' the more I realize how big this can of worms really is (which touches on handleman's points as well).


Thanks for your input guys!

 

[robertib]

Speaking as someone who works in the vacuum deposition industry, quantifying leak rates depends on what you are trying to achieve.

1. What is the seal for - liquid-atmosphere, liquid-vacuum, gas-atmosphere, gas-vacuum, gas-liquid, etc?
2. What is the difference in pressure between one side of the seal & the other - 0 bar (eg atmosphere-gas at 1 bar pressure), 1 bar (eg atmosphere-vacuum), more (eg water-atmosphere at 3 bar mains pressure), etc?
3. What are the failure requirements - minute trace elements, visible droplets, flood, etc?
4. What are the consequences of the leak - cosmetic damage, minor damage to linked components, total failure of product, etc?

For example, we have the following:-

water-vacuum - 2 bar water pressure to 10-6 mbar vacuum, trace elements are failure, potential total failure of product - our He leak rate is required to be better than 10-9 mbar
water-atmosphere - 2 bar water pressure to 1 bar atmosphere, visible droplets are failure, potential minor damage to linked components - our He leak rate is required to be better than 10-6 mbar
atmosphere-vacuum seals - 2 bar water pressure to 10-6 mbar vacuum, trace elements are failure, potential total failure of product - our He leak rate is required to be better than 10-9 mbar