Permeability Issues Relating to Sealing a Vacum 1

[mcswe]

Does anybody have any knowledge relating to Permeability issues relating to various sealing compounds. We have a problem with water ingressing into a small vacum on one of our components (Volume in the region of 1.7x10-5 m3) . Having carried out numerous investigations all evidence seems to point to ingress through the compound used to seal the volume.
I have tried to use Darcy's Law and various permeability values given by material suppliers to calculate the mass flow rate across the seal, for different materials, but i keep coming up with unrealistic answers. I was wondering if anybody had any similar experiences or advice.
Thanks

 

[kenvlach]

With values for the seal cross-sectional area & thickness, the calculation is quite easy. If not, I have used helium to back calculate the effective Area/thickness ratio: simply fill vessel with He, then record the pressure decay. Use 1 atm He for glass vessels, higher for metal.

What are your system characteristics? Vessel construction,
seal material & permeation constant (with units),
seal area & thickness, ambient atmospheric H₂O pressure, etc. At 25^oC? Any bake-out that may have damaged the compound? What vaccuum & how long does it take?
Please give your calculation. What is unrealistic?

 

[btrueblood]

Permeability rates can vary pretty dramatically for some materials, esp. two-part mix potting compounds. A lot of processing variables can affect the final seal permeability. They also vary depending on the gas you are interested in, water vapor ingress rates generally aren't predicted well by helium leak rates.

 

[IRstuff]

How do you know that it's not simply entrained water in the seal itself?

TTFN

 

[kenvlach]

btrueblood,
Helium is often the easiest means to determine effective geometric values for the equipment seal, and also for finding tiny physical leaks if a helium leak detector is available.
Agree that water vapor permeates faster through many substances, but sealant manufacturers should be able to provide permeability constants for all the atmospheric gases.

IRstuff,
That's only a problem with improper sealant or inadequate cure. Aerobic-curing silicone rubber is possibly the worst choice -- gives off acetic acid and lets water through like a sieve.

 

[btrueblood]

Kenvlach,

From experience, there is no reliable conversion from helium gas leak rates to water vapor leak rates, even in metallic materials where you are dealing with physical (capillary, not diffusion) leakage. Sorry, but a "geometric value" for an equipment seal, when you are looking at leak rates below 10^-6 scc/sec in helium, just flat out won't correlate to water vapor, or if they do, you got lucky. Similarly other polyatomic gases (i.e. more than 3) don't correlate to diatomic or monatomic leak rates. Sealant manufacturers (e.g. potting compounds) will give you water permeability rates, but will caution you that many processing/cure variables affect water permeability. Membrane and polymer manufacturers will give you rates, but caution you that processing and interface variables will affect permeability rates. The best method for finding permeability rates is a test on your own hardware.

 

[kenvlach]

btrueblood,
What's to disagree about?
Helium permeation can yield geometric values of equipment seals to aid in calculating water vapor permeation (the thread topic). Then, mcswe can determine whether the measured ingress is consistent with the seal manufacturer's water vapor permeation constant.

Of course, leakage will obscure permeation tests. In which case, helium leak detection is a useful tool.