How Can I Minimize Liquid Helium Boil-Off in Dewar Flasks?

  • Context: Graduate 
  • Thread starter Thread starter swooshfactory
  • Start date Start date
  • Tags Tags
    Boil Helium Liquid
Click For Summary
SUMMARY

The discussion focuses on minimizing liquid helium (LHe) boil-off in dewar flasks and cryostats, emphasizing the importance of design features such as neck diameter and material choice. Stainless steel is recommended for its low thermal conductivity and durability compared to glass. Key factors influencing boil-off rates include radiation, conduction, and convection, particularly in wide-neck cryostats. For optimal performance, users should consider the type of cryostat, whether it operates continuously or in a single-shot mode, and the specific temperature requirements of their experiments.

PREREQUISITES
  • Understanding of cryogenic principles and heat transfer mechanisms
  • Familiarity with stainless steel and glass dewar flask materials
  • Knowledge of Joule-Thompson cryostat operation
  • Experience with temperature regulation techniques in cryogenic environments
NEXT STEPS
  • Research the design specifications for stainless steel cryostats
  • Explore the operational principles of Joule-Thompson cryostats
  • Investigate methods to calculate boil-off rates based on neck diameter and thermal insulation
  • Learn about closed-circuit cryocoolers and their advantages over traditional liquid helium supply
USEFUL FOR

Researchers, physicists, and engineers involved in cryogenics, particularly those working with liquid helium and low-temperature experiments, will benefit from this discussion.

swooshfactory
Messages
61
Reaction score
0
I've been looking at various papers which describe the mechanisms and provide calculations describing heat flow into helium. I'm trying to minimize helium loss and am in the market for a new dewar flask/cryostat. I plan on asking the supply companies directly what the proper dimensions are to reduce helium loss while still maintaining low temperatures for our needed time period (~10 min), but I would like an idea of the physics behind these dimensions.

Currently, we have a glass dewar flask containing LHe sitting in a glass dewar flask containing LN2. Purchasing a metallic cryostat is an option, but I don't know much about them.

Here is one of the papers I found:
http://jap.aip.org/resource/1/japiau/v22/i12/p1463_s1#tabs_1_113_1274104113_tab3 (Evaporation Rate of Liquid Helium. I). It provides good qualitative description, but the dewar flask I was looking at doesn't have a reduced-diameter neck as the one they describe does.

Any input would be valuable!
Thanks!
 
Last edited by a moderator:
Physics news on Phys.org


These days most dewars and low-temperature cryostats tend to be all-metal designs. Stainless steel has a rather low thermal conductivity and can be made much thinner than glass without risk of breaking.

There are several sources of heating: Radiation, conduction, and convection, especially if you have a wide-neck cryostat. Estimating these quantitatively is not so easy.

10 min holding time at 4.2K should be trivial to achieve. If you want to go lower, you need to pump. Then there are several options of how to build the cryostat, essentially with a Joule-Thompson stage (continuous operation) )or a secondary, closed bath (single shot operation).
JT-type cryostats you can get as flow-type without a bath. They are fed directly from the storage dewar and can be very efficient.

Other things to think about:
Will your sample be immersed in liquid, or in exchange gas or in vacuum?
Do you want a single temperature, or do you need variable temperature, to be regulated with a heater. What is the base temperature you want?
Do you want a cryostat that relies on (external) liquid He supply or a closed-circuit cryocooler that only needs electricity?
 


A good transport dewar that can hold say 60-120 liters will have a boil-off rate of about 1-2 liters/day.

However, in many experiments one needs a fairly wide neck-diameter in order to be able to fit a probe (50mm or even wider), in which case the boil-off rate goes up quite a bit (say 5 liters/day); and the boil-off rate will increase while the probe is inserted (and will of course go up even further if you are runnning a 1K pot).

I wouldn't bother trying to quantify this using formulas etc. , there are so many different factors that come into play that I don't think you can ever get a realistic estimate.