From MAILER-DAEMON@otter.mbari.org Wed Dec 8 12:27 PST 1993 Received: by otter.mbari.org (1.36.108.4/15.6) id AA08751; Wed, 8 Dec 1993 12:27:00 -0800 Date: Wed, 8 Dec 93 12:27:00 PST From: Mail Delivery Subsystem Return-Path: Subject: Returned mail: User unknown To: kiwi@otter.mbari.org Status: RO ----- Transcript of session follows ----- 550 chmaxxxxchma... User unknown ----- Unsent message follows ----- Received: by otter.mbari.org (1.36.108.4/15.6) id AA08749; Wed, 8 Dec 1993 12:27:00 -0800 From: William Kirkwood Return-Path: Subject: DCon Leaks To: neji@mbari.org (Jim Newman), chmaxxxxchma@otter.mbari.org Date: Wed, 8 Dec 93 12:27:00 PST Cc: kiwi@mbari.org (William Kirkwood) Mailer: Elm [revision: 66.25] Mark - The Datacon lids are the problem not pressure bending them. All of the lids have a compound wave in them from machining the 'O'ring surface. That is why the first one leaked at an odd location instead of directly in the middle. The wave is a taper from inside to out of over .003" and a .010" delta in a 90 deg span around the edge. These are averages with some individual points even higher. I'm sure when the pressure is applied it makes the problem even worse. A bore seal may not work either. The tool marks in the plastic are to deep to be removed and leave enough wall in my opinion. If you look at it over life, the wall would be about .050 or less and it is right where tools are used. A nick or chip would make any given box doubtful for use. I suppose if you feel this is how you would like to go it could be done. As for the latches, I find them difficult to use with that little lock and oil on your hands. They also pull down to far away from the seal area and cause oil pressure to create a moment arm that tries to lift up the seal in the middle. The plastic Lexan is thick enough that this problem shouldn't bite us later and it is probably to late to really change the latches. My test design of the "claw" relieves some of that moment and eliminates the steel ring so you can see the seal. Except for the high point in the part that I modified I saw a very nice witness mark of the seal around 80% of the perimeter. This design achieves what the ring does, it keeps screws from being put into the plastic. I think the shops mod of the latches needs two more screws also. I checked with the shop about fixing the lids and it isn't as easy as I thought. The way they have to chuck up the lids and machine them means flat can't be guaranteed. My suggestions to fix this (if you agree) would be to change the durometer of the 'O'ring to 50 from 70 and to go with the "claw" added to your latch. This eliminates most or all of the moment on the lid and increases surface seal area. Extra screws will help take up the load and reduce the chance of a latch creeping over time. It's about the most graceful and simple fix I can think of. I've got info coming from Motion Ind. about the 'O'rings and the shop has no problem making the parts and mods. A gasket is also a possibility instead of the 'O'ring. I'll pursue this further and see what else I can come with. Actually just the 'O'ring change might fix it, but I prefer the ability to see the seal. Let me know what you think - Bill