Transferring Liquid Helium to Another Container

[Davephaelon]

I have a small 'cryostat' which I would like to fill with liquid helium to conduct an experiment. It consists of a circular plastic kitchen measuring cup 3 inches wide at the top, 2 and 1/4 inches wide at the bottom, which is flat, and 1 and 1/2 inches deep. The plastic wall/floor is about 1/8th inch thick. The cup is surrounded by rigid foam used to insulate houses, which varies from 1/8th inch to about 1 inch in thickness. I've used this homemade 'cryostat' in earlier experiments with liquid nitrogen, and never had any problems.

I have a 1/4 inch by 2 inch niobium-titanium alloy rod, drilled and tapped at both ends to accept 6-32 machine screws. Currently 22 gauge wires are secured to both ends via metal grommets with tabs, to which the wires are soldered, and the grommets are tightly secured with 6-32 screws. I'm currently designing a bracket that will suspend the rod slightly above the bottom of the cup. The plan is to fill the 'cryostat' with liquid helium and pass a large current through the rod. The whole experiment would probably take 15 or 20 minutes.

I intend to take this setup to a laboratory that works with liquid helium. So my question is would it be possible to transfer liquid helium from a dewar into the cup, and how exactly would it be done? I was just reading in a book "Dark Sun" by Richard Rhodes, about the making of the hydrogen bomb, that special techniques were required to move liquid helium from one container to another, but there was no elaboration.

Thanks for any answers.

 

[kishlaysingh]

you could not just transfer liqiud helium like the apparatus you've made... handaling a liquid at 4K is just not a simple task... is so volatile that the moment you see it it will vaporise

 

[Q_Goest]

What kishlaysingh says is absolutely correct. I've worked with LHe for almost 20 years. What's most surprising is how low the latent heat of boiling is. You can transfer liquid helium using vacuum jacketed piping and flexible hoses, but even with that, a portion of the liquid will boil off (on the order of 10%). You won't be able to cool down the apparatus and have it sitting in the cup like you can with liquid nitrogen or oxygen.

 

[Davephaelon]

Thanks for the replies Kishlaysingh and Q Goest. As I mentioned, I've had no problems cooling my earlier YBCO superconductors to liquid nitrogen temperature, as confirmed by the Meissner effect. It actually took quite a while for the LN2 to boil off. Liquid nitrogen is about 3/4ths of the way down from room temperature to the temperature of LHe. So, might it be possible to pre-cool my cryostat with LN2, quickly pour it out, and then feed LHe via the special vacuum jacketed tubing into the cryostat?

However, from what I'm hearing, from experienced workers in the field, it sounds like attempting to use my existing home-made cryostat is like beating a dead horse - it's just not viable. So that leaves the only option to use the purpose-built LHe containers to hold my 1/4 inch by 2 inch niobium-titanium (Nb-Ti) rod, along with support beam and wires to bring in the high voltage. I don't know anything about such containers, such as dimensions, or penetration at the top for wires, etc, or if there is some way to mechanically secure something from brackets, or screw holes, on the inside of the top lid. But my engineering challenge is to precisely align the long axis of my Nb-Ti rod with a very small cross section accelerometer chip that measures 1.5 mm by 4.0 mm, which would be located outside the LHe container, say 5 or 6 inches away. The exact separation between rod and accelerometer chip isn't critical, but alignment might be easier, the closer it is. I anticipate there will be issues with shrinkage of the support beam as it cools down from room temperature to the LHe temperature, which could throw the alignment off.

 

[Q_Goest]

I've not heard of anyone maintaining an open container of liquid helium but open containers of liquid nitrogen are relatively easy. To be honest, you don't even need to insulate them. I think the difference is how air interacts above the bath. As nitrogen boils off, it carrys air up and away from the sample. The air doesn't easily change phase so there's only some mild heat transfer in the layer of gas above the nitrogen. For helium, I suspect the problem will be with air liquifying into droplets, dropping into the liquid helium and solidifying. All that phase change with direct mass transfer is going to pump a tremendous amount of heat into the liquid. And the liquid has such a low latent heat of boiling that it won't take much solid air dropping into boil it off.

If your sample is only 1/4" in diameter and a few inches long, you might be able to insert it into the end of the transfer hose. Another option is to use a small vacuum jacketed vessel. One with a large top that can come off but can be used to keep air separated from the helium.

Liquid helium vessels are different than LN2 vessels though. They have a shield in the annular vacuum to help reduce heat transfer. If a conventional LN2 dewar is used for LHe, the helium will boil off in just a few days. The shielding inside the vessel reduces heat transfer by a factor of roughly 50.

So the question is, how long do you need to maintain your sample at LHe temperature?

 

[Davephaelon]

Q Goest:

Before going back to nap some more this morning, I had much the same idea - placing the rod inside the transfer tube, allowing the LHe to circulate around it, assuming it was big enough. That would also reduce the alignment problem, as the rod is now in a much smaller containment area. Indeed, the end of the rod might actually be flush with the end of the tube/hose, making visual alignment possible. To keep the rod, with its wires attached to either end, from popping out, maybe some simple bracket could be used. In this scenario some LHe would be lost. However, the actual operating time of the experiment would literally be only a few seconds, or several minutes if I chose to make more than one 'run'. A 'run' consists of discharging four 560 mfd capacitors in series, charged to 600 volts, taking only a fraction of a second, while monitoring the accelerometer output on an oscilloscope. Charging the capacitor bank back up to 600 volts takes something like a minute, or less; haven't run the system for a while.

My vacuum lined thermos has a 1 liter capacity and uses a loosely fitting foam top. Actually, with just a little LN2 left in the bottom it usually survives overnight. As I was napping I also thought about the fact that LN2 consists of two nitrogen atoms bound together for an atomic mass of 28. In contrast, I don't think helium has such bonds, so its atomic mass would only be 4. Being so much lighter than N2 would add to its volatility, I would think.

 

[Q_Goest]

Hi Dave. It sounds like you could make something work but the next question regards cost. Liquid helium isn't cheap. How much can you spend on this test and roughly where are you located?

 

[Davephaelon]

I was hoping no more than 200 dollars, maybe 300 at most, and I'm located in southern New Hampshire, so within easy driving distance of Boston, New York City.

 

[Davephaelon]

Oops, I meant to say southwest New Hampshire. I used to do part time technician work at Columbia University's Lamont-Doherty facility in Palisades, NY, right across the Hudson from New York City. I remember it was 150 miles to get there, so not too bad a drive. But I would drive further if necessary.

 

[Q_Goest]

Sorry, I was getting sidetracked with that last question. I see you already mentioned you have access to a facility with LHe. So I'll assume that.

Your thermos may work. If it still has a bit of LIN in it if left overnight, the heat flux is on the order of 10 Btu/hr, round numbers. If you filled it with liquid helium, it would boil off in roughly 10 minutes. If you can control venting say through a small hole, that could work. You would want to flow LHe continuously into the container at a slow rate. I'd think you would want a temperature probe (silicone diode) to tell you what temperature you were at. Transfer would have to be done with a vacuum jacketed flexible hose. Generally, you pressurize your LHe source to 2 to 5 psi to start flow. The thermos will vent helium continuously. This presents a few hazards, asphyxiation and overpressurization of your thermos, depending on your test set up.

 

[Davephaelon]

Q Goest:

Thanks for all the suggestions. I'm not absolutely sure that LHe is used at Lamont-Doherty. I did have the acquaintance of an individual there, who was a chemist. He visited our department at the Woods Hole Oceanographic Institute (WHOI), from which I retired 4 years ago. However, I haven't done any contract work at Lamont-Doherty since the Fall of 2005, so am not sure that individual still works there. I do believe that the Chemistry department at WHOI uses LHe, and I could check with them.

Assuming I'm not able to run the experiment at WHOI, or possibly Lamont-Doherty, would your facility be within driving distance of Southwest New Hampshire?