Date: Fri, 30 May 2003 00:34:42 -0400
From: Stefan Jeglinski <jeglin@4PI.COM>
> Turbulence kicking around dust or small dirt particles is also a
> definite issue. Often these windows (when broken) will show evidence
> of projectile puncture.
That seems like quite an issue. Defending
against fast-flying shrapnel is a pretty
tall order
In contrast, pressure without shrapnel isn't
very scary. I once made a capacitive strain
gauge using a membrane of 1/4 mil Kapton and
a capacitor gap of 1/4 mil. Cute thing.
Nanobar sensitivity. It withstood a lot of
abuse including inadvertent 1 bar steady
pressure and multi-bar transients. Understandably
it had to be recalibrated after such incidents,
but it didn't leak, so we were pleased, especially
considering that it would have taken many days to
get access to repair it. Since it had superfluid
on one side and absolute vacuum on the other side,
when I say it didn't leak I mean it really, really
didn't leak.
> ... But not all, hence the original question.
Maybe non-obvious shrapnel damage? Or perhaps
the occasional less-than-full-strength membrane?
==========
1) Bits of dust in a vacuum system are bad news.
-- They can produce huge "virtual leaks".
-- They can do bad things to the pumps if
they get carried downstream.
2) As a general rule, it is better to prevent
catastrophes than to clean up after them.
So I would look real hard at what could be done
to prevent catastrophic venting.
-- If somebody is opening a gate valve, design
an interlock... perhaps a pin held by a latching
relay, or a pin driven by a motor+leadscrew that
won't disengage until there is zero pressure
differential across the gate valve. (And conversely
the system won't pump down unless the pin is in place.)
[Or just replace the whole valve with a motorized
valve. I've made some cool things with motorized
valves.]
-- If somebody is breaking the glass on an
observation port, install a plexiglas shield in
front of it.
-- Et cetera.
The interlocks will pay for themselves quite
quickly.