We have just come back from a failed attempt to revive an NMR magnet which
quenched last December, and as amateurs in this area we would appreciate
any ideas or opinions or suggestions from any of you who will listen to our
tale of woe.
We have a collaboration with a research institute in Hermosillo, Sonora,
Mexico, which focuses on food chemistry. We donated to them an MSL-200
solids system which we no longer had space for and in the spring of 2000 we
brought the magnet to field ourselves using a borrowed turbo pump and a
borrowed power supply. The system ran for 3 years successfully doing both
liquids and solids, and a number of masters and Ph.D. students depended on
it for their thesis work.
Last December it quenched while the person in charge was in the hospital,
and the only information available is that the magnet was discovered with
"ice in the bore" even though it was near the end but well within the He
fill schedule. We went down (a 5 hour drive) in February of this year and
hoisted the magnet, pulled out the bore tubes, cleaned and re-greased the
(3 year old) seals, and re-assembled the magnet. We did not take off the
bottom plate of the N2 cannister or check the alignment, since we had no
reason to believe that the magnet was damaged.
At the end of June we went down again with a different borrowed turbo
vacuum pump (the one we used in 2000 had since died) and our colleagues
ordered $6,000 worth of liquid N2 and He (yes, those are dollars, not pesos
- liq. He is trucked from Mexico City). We had planned to pump and chill
in a long weekend, but the vacuum went down very slowly. We returned to
Tucson and left it on the pump for a week as it crept down to a final
vacuum of 4.3 x 10-5 torr. The magnet expert of our team was unable to
return the next weekend so I, the applications NMR guy who hasn't filled a
magnet in 8 years, went down to chill the magnet alone.
I closed the vacuum valve and filled the N2 can with liquid N2. Then I
filled the He can with liquid N2. Near the end of that fill, it seemed
that we were running out of N2 and because it was 4:30 on a Friday in
Mexico, we ran to the other end of the building to try to order more N2
liquid by phone. We got put on hold and many minutes later someone came by
and told us that liquid N2 was spilling on the floor. We raced back to
find a torrent of liquid N2 falling from the dog bone center hole onto the
magnet. (violation of rule #1 - never leave the room) Anyway, the point
where the N2 was landing was on the slope of the top of the magnet and
rolling down the outside, so it didn't seem that the top seals were frozen
(a wad of Kim Wipes was stuffed in the top of the bore). Soon afterward,
however, we noticed frost forming in the inside of the bore. This was
melted with a heat gun and the bore was swabbed dry. Ice formed again,
uniformly throughout the bore. Laying on the ground and looking up through
the bore, I could feel a torrent of cold air streaming down. By the next
morning the bore was completely filled with ice and half of the liquid N2
had boiled off in the N2 can. The outer surface of the magnet was not
cold. We gave up at this point and the liquid He was wasted since it was
not possible to return it to the vendor.
My initial assessment was that the alignment of the bore tubes had been
damaged in the quench and there was a touch between the inner bore tube and
the 77K tube. The odd thing, however, was that the ice formed absolutely
uniformly throughout the inner surface of the inner bore tube.
When our magnet expert returned from vacation and heard the story, he felt
that it was not a touch but rather a failure of vacuum, possibly caused by
the spilled liquid N2. The lack of ice on the outer cannister was
explained by the larger space between it and the N2 can, and the presence
of mylar superinsulation. The inner bore tube, being very close to the 77K
tube, will ice up quickly if the vacuum fails.
Another magnet expert agreed that it was not a touch and that the vacuum
was the problem. He didn't think that spilling the liquid N2 caused the
problem. He said I should have rough-pumped the vacuum space multiple
times, venting with N2 gas to remove water before starting with the turbo
pump. He said the turbo, even a very small one, should reach the 10-5
range within a few hours. The fact that it took a week means that either
there is a leak in the vacuum space or there was water in the vacuum space
(we never saw visible water when we disassembled the magnet). I should
mention here that a very fine crack was observed in the metal at the top of
one of the helium towers several months before the quench. It's not clear
whether this is a crack in the surface chrome coating or a real structural
problem. The crack was growing steadily with each He fill. When the valve
to the magnet was closed, the turbo pump reached a vacuum in the low 10-6
range within about 5 minutes.
The questions we have are:
Do we need to re-align the bore tubes (i.e.,.was there a touch)?
Is it worth attempting again, just with liquid N2?
Is there any chance that the turbo pump was "low volume" or not
suited for an NMR magnet, and this might explain the one-week duration?
Did the N2 overfill cause the problem?
Can we check to see if the crack is a leak?
Any comments or opinions you have would be greatly appreciated. We have
all learned a lot from this project, but there are limits to what we can do
with practically no money.
Getting an NMR magnet into Mexico was not an easy task, lots of palms had
to be greased, so I'm not real excited about replacing the magnet.
Thanks again!
Neil
Neil E. Jacobsen, Ph.D.
NMR Facility Manager
Department of Chemistry
119 Old Chemistry
1306 E. University
University of Arizona
Tucson, AZ 85721
520-621-8146
FAX 520-621-8407
Received on Tue Jul 20 2004 - 12:09:53 MST