9 Dec
2009
9 Dec
'09
10:58 p.m.
On 12/9/09, Chris Elmquist <chrise at pobox.com> wrote:
Wow! Neat! (I almost said "Cool" ;-)
From 1983 to
1989, at ETA Systems, we ran our computers submerged in a
bath of liquid nitrogen.
:-) Another thing to consider if you bring a machine into
the warm from
a cold environment is condensation. The thermal expansion that Ethan
pointed out is important but you will also often see water condensing
on surfaces and components as the thing warms up. You definitely want
to wait for that moisture to evaporate before you apply power.
As someone else already observed, condensation isn't a problem in
Antarctica. The air is ultra-ultra dry. In the winter at Pole, the
outdoor *Relative Humidity* of a chunk of -80F or -100F air can be
well over 50%. Bring that very same air inside and warm it to +72F
and it plummets to as low as 0.5% RH.
We had massive tape drive problems with our bank of LT02 tape drives
until we brought them out of the nice shiny new IceCube Laboratory and
back into the main station building where we could hook a 4-transducer
ultrasonic humidifier to the building's water supply (the ICL has no
water except what you carry with you). Once we pumped several gallons
per day into the air near the tape drive, the RH in that rack rose to
15% or so and the tapes became much more reliable.
Condensation on machines brought from cold to warm in most of North
America or Europe (or Brazil, Japan, Oz, NZ, etc.) would be something
to watch for and be careful around. We never had those problems at
Pole.
-ethan