Holding Vacuum Level

Don't know what a "drop test" is.

Pulling a vacuum on earth applies about 14.7 PSI to the outside of the part.

In all my years of using vacuum bags, we never worried about small leaks. We just let the pump run until the resin cured. Much of our work was with pre-prges which don't flow readily, I'll admit.

Visual inspection should reveal any leaks after introduction of the resin. Look for tiny bubbles.

The drop test is as described, "pull vacuum, close valve and do not lose more than 20 millibars in 20 minutes. The leaks during infusion are what we're trying to avoid, air entering the part during the infusion causes a snow flake pattern and a condition known as dry glass. Sometimes these leaks show up between the time the infusion ends and before the resin goes through exotherm.

Sealing tape may be the problem. What type tape you using?

You might have moisture content that is boiling off causing gases to be introduced in the vacuum space....

The tape is an industry standard, approved by our customer as are all of our materials. The building is environtmentally controlled and we monitor the humidity levels. This tset is done before any resin is introduced to the part so so gassing can occur. Thanks for all the suggestions so far, this is a problem I've been chasing for some time. since I ran out of ideas thought I'd throw it out on the web and maybe locate someone smarter than me!

Can you build a frame and use an O-ring in the table?

that may not work for you. I know vacuum bag material is not easy to work with.

Maybe the workers are sweating...?

Vacuum gauge measurements are a relative measurement....the ambient pressure could vary causing vacuum gauge to change as well....Do you have any large exhaust fans in the area that may be cycling...?...yeah that's a long shot.....How do you test the valves? Leaks this small can be extremely hard to find, sometimes you just have to start replacing most likely offenders...

You hit the problem I'm chasing, these are minute leaks and most of the time we keep trying things till we get a good drop test. I'm looking for a more scientific method if possible. As far as an exhaust fan there is no measurable difference in atmospheric pressure with or without the fan running and also between the inside and the outside of a building. HVAC of a building does not vary the atmospheric pressure, major changes are brought on by weather changes and elevation.

Well perhaps ultrasonic leak detecting might work...

Ultrasonic leak detector....

http://www.sdt.eu/index.php?page=applications-uld-leak_detection_on_compressed_air_gas_and_vacuum_systems_by_means_of_ultrasonic_signals

Low tech approach....

As stated in my 1st post "We have electronic leak detectors but these aren't showing a leak." Have not tried the stethoscope......YET! I'm grasping at straws here.

Ultrasonic...? ...recommended with extended distance cone for max amplification....vacuum leaks are the hardest to detect, especially the small ones...

http://www.sdt.eu/index.php?page=products-uld-sdt270-accessories-eds&hl=en

For such a minute drop, you might just having your bag stretched during a period of time; so after a while trying different things, it just stopped giving way and you believe that you somehow fix it, in an uncertain and mysterious way.

Do a good pre-vacuum time, correct as needed and then trickle the resin.

Are you pulling the vacuum from a single point? Maybe multiple monitored points or location may help.

Since you are pulling a vacuum, then the leak would be into the interior and showing at your pump exhaust. Is that where you are checking?

The suggestion about "outgassing" (described as moisture, but could be any vapour) I feel is a good one. Could be the adhesive solvent from your sealing tape, could be the mould surface itself, could be the release agent on the mould, could be outgassing from the plastic sheeting itself.

You describe the "snowflake" effect that happens when you apply resin. Maybe worth the cost of materials to run a few trials and see where these are happening. That would be proof positive of the location.

It could also be a leak at your isolation valve itself. Has that been checked?

What I'm concentrating on now is the vacuum ports. We use a steel port that is pulled down in to a pocket machined in the mold. Mold is fiberglass. I discovered that the vacuum was leaking through the fibers in the machined pocket. Painted the pocket with resin, let it cure and then it passed. Called Loctite and them trialed a 2-part silicone based sealant to seal the ports in the mold. I am doing the trials on these ports, they are small only a 2 inch diameter head with a 1 inch hollow thread to pull the port into the mold. A pipe fitting with a hose barb screws into this hollow part of the port.

The resin sealed port, with the new sealant, finally passed. The port without the resin has not passed yet but I'm still fighting variables. I've been working on this as I'm asking you people for information and appreciate the responses so far.

A manifold gauge set-up may help pinpoint the problem ports

How big is the part?

How big is the mold?

Is there more than one mold, part?

Quantity?

Cycle time?

Resin gel time?

Can you use pre-preg?

Can you bag the mold and the part?

Hi TvanD

Here's a silly idea. When testing for pressure leaks we use soapy water and the bubbles increase in size, could you use bubbles and see if they decrease? A manometer using mercury is an option for measuring vacuum, requires less than a metre.

Tony

If you can afford to sacrifice a few sets of fibre you could submerge the whole thing in a dye before doing a dummy test: the position of the leaks should be obvious when you open up the mold.