Dewars
There are lots of different designs. For fast filling of the test rig I had a small centrifugal LN2 pump which was lowered through the neck. We tried (sorry, would have tried) that for the air conditioning experiment, but it was (would have been) a bit noisy. So we tried (hell, I can't be bothered pretending) method two. You seal the neck with a rubber bung containing a dip tube which goes to the bottom of the flask. As heat leaks in the thing self-pressurises and drives liquid out, just like the LPG tank on a car. That's how LHe dewars work, by the way, though you sometimes have to encourage things with a bit of work on a gas-filled bladder.

Whether this works for LN2 depends on how good your dewar is. Too good and it doesn't push enough out: too bad and it empties too quickly.

You are quite right about the physiology. We weren't poisoned because nitrogen is not (at those pressures) poisonous, which is where the air-is-mostly-nitrogen comes in. We weren't asphyxiated because the room was well ventilated, we weren't putting that much more nitrogen into it, and because the nitrogen rich area would have been at the floor. Touch shit for the mice, possibly.
(ubergeekian, Wed 21 Sep 2011, 12:05, 1 reply)

Interesting.
you'd need to watch it, though, - the N2 rich air is only at the floor because it's cooler. N2 is marginally less dense than air, all else being equal, so it's really the room vents that matter.

I'm surprised by the amount in both yours and Boris's posts though - I can trigger a full 02 alarm just by mis-filling a small storage dewar, although all our labs run odd sealed air handling systems to be fair.
( the mighty badger Aphrodite, on a bar stool, by your side, Wed 21 Sep 2011, 14:05, closed)

Sure
It was the coolth what done it. It was just a normal office, so no alarms of any sort. None in the helium labs either - you knew something was leaking when your voice went funny. Seriously.
(ubergeekian, Wed 21 Sep 2011, 16:36, closed)