my magnet is possessed
David Vander Velde (dave@kunmr.chem.ukans.edu)
Thu, 17 Dec 1998 14:55:48 -0600 (CST)
Dear AMMRL, We are having difficulties with a Varian R2-D2 magnet of a
kind I have not experienced before (and I have been dealing with its
vacuum problems for the past 11 years ...) These problems surfaced after
some asbestos abatement/construction in the room. We know that the magnet
was covered with a plastic sheet for a while, which could have allowed
helium to concentrate underneath. We don't know if any other adverse
events happened while we weren't looking.
Our symptom was sky-high helium and nitrogen boiloff. (Previously, our
problem had been high helium boiloff, low nitrogen boiloff as our vacuum
softened.) There is no sign of ice on the magnet inside our out that
might indicate a touch or an internal shift of the components. When I
hooked up the diffusion pumping system that had always worked in the past
to bring our boiloff down in 6-8 hours of pumping (with the magnet
isolated from the pump, the pump is still producing an excellent vacuum),
I found the pressure inside the magnet was ca. 20-fold higher than I had
ever seen in a cold magnet, ca. 1.5 E-5 on my Penning guage. Three days
of continuous pumping has not lowered either the indicated pressure or
more importantly, nitrogen/helium boiloff (both guages are offscale). The
only time I have seen the pressure go down was transiently, during a
helium fill. Obviously, I cannot afford to feed this insatiable magnet
very long, and I either have to get a good vacuum back or let it quench.
I know that pumping helium can be slow work, but our rate of progress
seems to be negligible.
I would appreciate any advice in general, and/or opinions on these
questions:
1. Do we need a bigger pump? This one is rated ca. 100 liters/sec ...
whatever that means.
2. Are we pumping against some kind of leak, or reservoir of gas? The
failure of the pressure to go either up or down suggests we are at an
equilibrium, but an equilibrium of what?
Wishing you all a new year containing no magnet trouble,
Dave Vander Velde, University of Kansas