Re: sources at cryogenic temp

[benesch@cebaf.gov (Jay Benesch)]

       This is mostly a reply to Rob Gunter's comments; I replied 
privately to the original post and didn't keep a copy.  My expertise is 
in applied superconductivity and related areas of ME.  Guy White, 
Experimental Techniques in Low Temperature Physics, is one basic text.  
What follows is based on a decade's experience.   

1.  4.2K and 77K are essentially the same for stress purposes, because 
90% of the thermal contraction takes place by 77K.  
        This is not true for niobium, which is subject to hydride 
formation ("hydrogen embrittlement") between 70K - 150K.  One has to 
cool through this region at greater than 200K/hr (0.05 K/s) if one wants 
to avoid problems with Nb.  There may be a few other metals that have 
this problem below room temperature, but I don't know of any.  There is 
lots of data on hydrogen embrittlement at 300K and above; if it happens 
in that range to a metal it doesn't happen below 300K.  

2. Find relative thermal contractions of the various materials in the 
source.  Then look at room temperature clearances.  If nothing 
interferes cold, you're fine because tensile strengths go up as you get 
colder.  If there is interference cold, one looks at the stresses and 
compares with literature values for allowable, or tensile results from 
coupons.  

3.  Cooldown rate is not significant except for objects large enough 
that thermal gradients occur during cooldown.  If the object is compact, 
cooldown is essentially static until one gets to 1000 K/s or higher, 
because the sound speed in the metal governs.  I suspect your source is 
small enough that this isn't a concern.  "Radial" temperature gradients 
could be a concern if the thermal conductivity from the center of the 
source to the outside is poor, for then the outside shrinks ~0.3% while 
the inside is still at room temperature, and the stress is maximum.  
Then one needs to limit cooldown, again to limit thermal gradient. 
        I used to screen composites and plastics for cryogenic use with 
a open mouthed dewar of  LN2, a beaker of boiling water, and a mallet.  
I would alternate plundes into LN2 and boiling water.  If the item 
hadn't cracked after ten cycles, I started hitting it with the mallet 
immediately after the LN2 bath.  You'd be surpised how many things can 
take this regimen.  

4.  Fatigue is a concern.  If the "interference" stresses are moderate, 
and the number of cycles is 1000 or so, there shouldn't be a problem.  
If both of these aren't satisfied, go to the literature for cold fatigue 
testing, or pay for it to be done.  There is an outfit near Boulder that 
does this, but I can't remember the name.  Talk to the cryo group at 
NIST there.  If you are at 10E6 cycles, find some real data.  The best 
literature source for such stuff is the 40 volume series "Advances in 
Cryogenic Engineering", published by Plenum.  The journal Cryogenics is 
another source.  

5.  Martensic stainless steels tend to be brittle at or below 77K.  
Austenitic steels, if fully stabilized, are not - but the common alloys 
partially convert to martensite on cold work, possibly leading to 
brittle behavior.  As purchased, most austenitic stainless steels have a 
small martensite component.  316L is better than 304.  There are 
Nitronic (TM) alloys which are quite good.  

Jay Benesch
benesch@cebaf.gov