Lessons Learned About Helium-Cooled Winter-Over South Pole Instrumentation
Giles Novak, January 15, 1996
This is a summary of debriefings that I conducted with three CARA
winter-overs during 1995 and early 1996. One purpose of the
debriefings, and the entire purpose of this document, is to understand
the problems associated with running liquid-Helium-cooled instruments
at Pole during Winter, so we can maximize our chances for success in
this endevour.
The winter-over scientists that I debriefed are John Kovak (COBRA,
Austral Winter 1994), Mike Masterman (COBRA, 1995), and Richard
Chamberlin (AST/RO, 1995). Chamberlin and Kovak used liquid Helium
in the Winter, and Masterman used it in the Summer but didn't use it in
the Winter. In the Winter, he ran a cryogenic compressed-Helium
refrigerator. Chamberlin is a postdoc, Kovac an undergraduate research
assistant, and Masterman is an engineer. Chamberlin got data, Kovak
and Masterman gathered useful information on Winter operations but
did not get significant amounts of data.
I interviewed Kovak twice: The first time was a meeting, also with
Steve Platt, in Summer 1995. This meeting resulted in E-mail to the
entire COBRA team regarding telescope design. The second was a phone
call in December 1995. I interviewed Masterman and Chamberlin before
they left Pole after their 1995 Winter-over, and followed up with
Chamberlin during a phone conversation in January 1996.
The only aspects of these interviews that I will discuss are those
related to running dewars.
I divide this report into four categories:
- (A) Reliability
- (B) Serviceability
- (C) Ease-of-maintenance
- (D) Cryogen Consumption
(A) Reliability:
- Kovak advises that all pieces of equipment that are heated need to have
their temperatures electronically monitored continuously.
- Kovak ran a dewar that has elastometer O-rings and was heated by
electrical power to keep these from leaking. Problems with this power
caused dewar leaks for Kovac.
- Kovac felt that the single biggest problem that he had was dewar
leaks. Major failures of this kind are very difficult to deal with
under Winter-over conditions.
- Chamberlin felt that the best way to avoid dewar leaks and other
dewar failures was to use a reliable, leak-free, working cryostat based
on a well-tested design. Chamberlin's dewars were run at room
temperature (in a sort-of coude focus), so it was possible to use
existing dewar designs without the need to modify for cold weather.
However, his success is a testimony to the advantages of using
equipment that has been through a long shake-down period. His dewars
had been in use for years.
- AST/RO had a very scientifically productive winter. When I asked
Chamberlin "what broke?" in the AST/RO experiment, his answer was
"nothing". After some prompting, he remembered that two items had
broken: A 100 MHz amplifier on one of the IF down converters, and the
disk drive on one of the spare computers. (In both cases it was fairly
easy to replace the broken component with a spare.) Thus AST/RO had
very high reliability.
(B) Serviceability:
- The dewar must be easy to remove (Masterman got multiple frostbite
injuries while removing the COBRA 45 GHz receiver from the telescope).
This is an issue of human engineering.
- Spares are important. Masterman only had two servo amplifiers for the
AZ drive of the Python telescope. This component had a past history of
unreliability. Both amplifiers failed for Masterman, leaving him with
no way to run the telescope before mid-winter airdrop. This could happen
to a dewar component too.
- Documentation is important. When I arrived at Pole the new servo
amplifier that had been airdropped in to Masterman was wired up
incorrectly. There was no documentation at Pole that could have
been used by Masterman to check his wiring.
- A warm, well-equipped lab near the telescope is important (Kovac).
The Viper design implements this suggestion.
- As mentioned above, the one critical component failure of AST/RO was
fairly easy to fix (it took a few hours). Note that the part that
failed (an amplifier) was operated in a heated lab.
(C) Ease-of-Maintenace:
- Kovac said that when the Penguin dewar was cold it would "run his
life". It had 24 hour He-4 hold, 60-hour He-3 hold, and the cycling
was many hours long because it was an adiabatic demagnetization
system. Kovac said a 35-hour hold time would be a big improvement.
- Kovac thinks that a 2-day hold would be an "important psychological
advantage", but that daily transfers were "not that big a deal". He
thinks that a 2-day hold is not a critical requirement. He said that
it is more important to make the transfers more "painless", by avoiding
the need to move or hoist He-4 storage dewars vertically. He suggested
a long transfer line, but noted that the transfer losses would be
greater. An electrical hoist alone might not help unless it was
integrated into a well-thought-out system for moving the storage
dewars.
- Chamberlin was indoors when he transferred He-4 to instruments, but
every three weeks he had to transfer from the 12,000 liter dewar into
several 100 liter dewars. This he had to do in the cargo arch
(outdoors, but shielded from the wind). At first he would dread the
day when he had to do this chore. One reason for this was the need to
move 100 liter dewars (about 100 lbs.) up the dark slippery steps of
the AST/RO building. (AST/RO is now installing a handcrank-operated
hoist.)
- Besides transporting the 100-liter dewars, Chamberlin's two main
problems with the transfers from the 12,000 liter dewars were: (1)
Standing in the cold for 45 min. (cold feet), and (2) Occasional
problems with the fittings (broken O-rings - but see Kovac's solution
below; He-4 spurting out of a full dewar due to a fitting malfunction;
transfer tube frozen into the 12,000 liter dewar due to fitting
malfunction).
- Chamberlin said that some problems with the fittings would presumably
be even harder to deal with for the smaller fittings used when
transferring from the 100 liter dewars to an instrument out in the
cold.
- Chamberlin said that several solutions would be easy to implement: An
insulated platform to stand on to avoid cold feet when transferring, a
scale in the dome to see how full the dewar is to avoid He-4 spurting
everywhere.
- Phil Sargent of ASA told Chamberlin that he wants to put a heated tent
over the area where the transfers from the 12,000 liter dewar are being
made.
- Chamberlin said that seeing the plume is no problem. He had a scuba
mask that allowed the breath from his nose to keep the inside of the
scuba mask warm thus avoiding frost. Other winter-overs had other
solutions for visibility problems in general.
- Kovac used teflon O-rings at the top of the 100 liter storage
dewars during transfers. These worked fine.
- Kovac mentioned that at one point he got a plug in a dewar during a
transfer.
- Masterman pointed out that the weather was at times so unpleasant that
one REALLY did not want to go out to MAPO. Mark Hereld tells me that
Jamie Lloyd, the SPIREX winter-over, went out to MAPO every day to keep
the SPIREX dewar filled with nitrogen. Hereld would like to have a 2-day
hold and thinks that perhaps for the 1997 Winter-over he can get this.
His interest in doing this would imply to me that he is not completely
satisfied with a one-day hold.
- Masterman pointed out that when he had "house-mouse" duties, he
had to invest a significant amount of time cleaning up in the kitchen.
(Something like four hours.) Also, one day he spend a very significant
amount of time repairing the main fan of the heater of MAPO. On such
days, he told me that he would find it extremely incovenient to have
have to carry out a Helium-transfer outdoors.
- Requiring daily transfers of liquid Helium outdoors implies asking
someone to go to work for at least a few hours, every day for many
months, and perform a difficult task. The possibilities of illness, bad
weather, minor injury, as well as psychological factors such as the
occasional need for a day off and the other duties of station personnel
mentioned above all come into play in deciding on the hold-time
requirements.
- On some days Chamberlin never went out to MAPO but ran the experiment
from the dome. His dewars did not need daily transfers.
- Kovac and I discussed the the use of a pumped He-4 pot in a He-3
system, which makes it possible to cycle the dewar without the use (by
the Winter-over) of vacuum techniques. He felt that this improvement,
like the 2-day hold, was not a very important requirement for success.
He thinks that winter-over scientists should be very knowledgeable
about vacuum techniques and all other relevant experimental
techniques.
- Kovac thought that fully automated He-3 cycling would not have been a
significant improvement. The process of reversing heat swithces was
trivial. The Huntington feedthroughs worked just fine, but should have
the grease replaced for easier turning in the cold. Note, however,
that the Penguin was heated (although apparently not always,
unfortunately). Kovac DID feel that remote monitoring of pressures and
temperatures needed for cycling would be a significant improvement.
(D) Cryogen Consumption:
- transfer losses could be as low as 10% (Kovac).
- Chamberlin used 4 liters a day all Winter. There were no active dewars
in 1995. The 12,000 liter dewar was the main source of Helium.
- John used the badly-out-of-spec active dewars in 1994 and via
heroic efforts, was able to make the Helium last.
- The Nitrogen liquifier was cranky in 1995. Masterman spent lots of
time on it. This means that it is in principle possible to run
out of Nitrogen and have plenty of Helium. This seems unlikely,
but freedom from Nitrogen-use would provide one less thing to
look into. SPIREX will probabaly always need nitrogen.
Last revision January 15, 1996. Contact novak@belmont.astro.nwu.edu
or jdotson@nwu.edu