Make New Entry,
Make Followup Entry
User name R. Michaels
Log entry time 16:27:44 on April 12, 2010
Entry number 316930
A failed O-ring at the beam-exit flange on the target chamber caused
the downtime of PREX. Today it was replaced by Ed Folts and the techs.
Since I get a lot of questions about O-rings, here is my best shot
at answering. If anyone knows better, feel free to follow up.
Q: Is that the only leak ? What about the rectangular O-rings ?
A: It's thought likely to be the only leak. We'll find out after
we pump down, probably by Wednesday Apr 14.
Q: What happens if we need to replace the other O-rings, e.g. the
rectangular ones on the sieve box ?
A: First the techs may try to "goop" it with silicon, a standard
practice. Or try to torque the bolts. But if it's leaking, then
replacing those O-rings is potentially a major downtime (~1 week
plus whatever time to wait for suitable RADCON). This is why we
didn't simply replace all O-rings.
Q: Will the situation improve or are we doomed to cook another O-ring ?
A: We put back the massive tungsten shielding that had been in during
HAPPEX-2. This happens to shield the O-ring in question, the one at
the smallest angle. Taking it out was a mistake, I think. We've recently
estimated the power that was going into this region and I believe
we were heating the O-ring and/or irradiating it before but will
not now. I don't know how to estimate the lifetime of an O-ring, but
it seems likely that the situation will be much improved and similar
Q: Why not use metal O-rings, etc ?
A: My understanding is that, first of all, for all "regular" expts that
don't use Septum magnets, they always had metal conflat seals which
are robust. For the Septum connection an O-ring was necessary
and the metal O-rings were foreseen and the design called for them;
but they did not work because the target flange is aluminum and
the groove was wrong (too deep or too shallow, not sure which).
The point is, metal O-rings WERE thought about and do NOT work.