Standard for Glass Sealed Hermetic Connectors?

[MikeRCS]

I'll try to make this as short as possible, so please bare with me.

We make corrosion probes that are placed into the process with the back side connector out of the process for electrical hookup to read the probe. Over the lifetime of the probe, the probe element may corrode thru allowing process to enter the internals of the probe, meaning the hermetic glass seal 6-pin connector on the back side is now the pressure barrier.

Trying to get CRN's for the probes so we can sell in Canada has become somewhat of a headache. We must conform to a design standard, but I cannot figure out how? I have done wall thickness calcs on the wall of the connector, using the BPVC assuming the connector is a pressure vessel. However this does not address the glass seals. I cannot find any standards pertaining to glass seals.

According to UG-101(m), MAWP may be determined from burst tests. However, when this was presented to ABSA, they said that a burst pressure of 10X the MAWP must be obtained if glass is your pressure barrier. That would mean our burst pressure would have to be in the area of 60000psi!!!

Any ideas why the burst pressure has to be so high for glass? Any ideas on a standard for glass sealed pins in a connector?

Thank you.

 

[ApC2Kp]

MikeRCS,

There might be some different requirements for primay verus seconday seal that might make the requirements for glass seal secondary barrier more reasonable. One flow meter sensor that had a ceramic signal feedthru and glass-fused into a Kovar housing was required to be subjected to 10 thermal cycles while at test pressure. The straight hole of the housing allowed the ceramic to drift and finally leak after 3 cycles. The remedy was to design a shoulder into the straight hole to hold the ceramic insert against the pressure. The flow sensor had very thin 0.002 inch thickness stainless steel diaphragm - so leakage thru to the ceramic feedthru was very real possibility. The ceramic feedthru and housing was made by Vernitron in Cleveland, I think. The glass frit used to seal the ceramic to the KOVAR housing was selected to match the low expansion charcteristics of the ceramic and the KOVAR housing. The test pressure was the CL600 rating of flow meter body.

A brittle material is usually designed at a higher 10X factor to avoid getting anywhere near the tensile failure point. If the glass part could be shown to not be subjected to pressure tensile stresses (just shear stress?) maybe the design factor is much lower than the 10X of glass housing.