A while ago, I posted a message describing a problem we were having
with steady but excessive liquid helium boiloff rates on a 400
superwidebore magnet. I received numerous helpful responses, most of which
pointed out the possibilities of either (a) helium gas in the vacuum can or
(b) a thermal short due to ice bridges.
Now, we are experiencing a much more serious and possibly related
problem with this magnet. Since replacing its helium transfer line, we have
experienced very poor liquid helium transfer efficiency during fills.
Specifically, the new line transfers only 20% of the liquid helium consumed
from the tank into the magnet dewar!
Here are some facts:
1. There are no visible or tactile cold spots on the transfer line,
either during or immediately after the fill.
2. The transfer line cools down to form a good plume within 1
minute of pressurization at 0.5 psi.
3. The old and new transfer lines have the same length and overall
diameter (0.5 inches).
4. When discharging either into the magnet or into open space, the
new transfer line requires about _twice_ the flow rate of helium gas to
maintain a given tank pressure as the old transfer line.
5. The new transfer line gave 20% efficiency for filling at either
0.5 psi (100 liters consumed in 70 minutes) or 1.5 psi (100 liters used in
39 minutes). The old line gave up to 80% efficiency and consumed 100 liters
in 60 minutes at 1 psi.
6. During most of the fill with the new line, the plume exiting the
helium manifold looks like the magnet is full (inner blue-white cone).
However, both the level sensor and thumper readings show the magnet is less
than half full. The plume does lengthen when the tank is empty.
7. The filling stack freezes up more quickly and dramatically than
the charging stack. This has been true with both the old and new transfer
lines.
8. At a given fill pressure, the liquid accumulation rate in the
magnet seems to drop off exponentially over the course of the transfer, as
measured by the level sensor during brief interruptions of the fill via
tank depressurization.
9. The transfer line can be inserted all the way into the funnel
and screwed in without excessive force.
10. A thumper rod pushed down the fill port reaches the bottom of
the helium dewar without hitting any tactile obstructions.
11. The liquid helium boil-off rate of the magnet has not increased
during the changeover from the old to the new transfer line.
Given the symptoms above, I would like to know:
A. Why does the new transfer line require so much more helium
pressurization gas for a given tank pressure?
B. Is a tank pressure of 0.5 psi too low for a 0.5" transfer line
(1m X 0.8m X 1.5m)? The manufacturer has suggested pressures as high as 3-4
psi, but this seems excessive.
C. Could the poor efficiency be due to an ice blockage somewhere
inside the helium dewar? If so, where could it be and could it be removed
without warming up the magnet?
D. Is it normal for the filling stack to get much colder than the
charging stack?
and, in summation:
E. Is there any way that the new transfer line could be defective,
given the lack of icing, or are we just using it incorrectly?
Since I suspect that the boil-off and transfer efficiency problems
might be related, I will combine the advice I receive for both questions in
a future posting to AMMRL.
Thanks in advance,
Ken Fishbein
Facility Manager, NMR Unit
NIH/NIA/GRC
Ph. (410) 558-8176, 8388, 8169
FAX (410) 558-8173
E-mail: fishbein@vax.grc.nia.nih.gov
Address: Ken Fishbein
NIH/NIA/GRC
4940 Eastern Avenue
Baltimore, MD 21224 USA