Dear All,
Thanks to all who responded to my question, 22 people all in all. The
full text of responses with names removed are below.
The general consensus was that a commercial probe operating at RT cannot
cause a magnet to quench.
A number of scenarios (many more than I expected) were suggested where a
probe may cause a quench:. These were:
* Freezing the O-rings during low-T work.
* Melting the O-rings or overheating the bore during high-T work or due
to uncontrolled heating (VT controller failure).
* Overheating the probe due to arcing or excessive RF/PFG power levels.
* High powered gradients yanking on the field of the main coil.
* Forceful and rapid removal or insertion of the probe into the bore.
This could cause a quench either by excessive physical movement of the
dewar or via eddy currents.
* "High energy" samples exploding inside the probe.
Unfortunately, none of these applied to our situation. The probe in
question uses low power levels relative to 5 mm probes, and was running
a non-gradient 2-D experiment when the quench occurred.
As if to exonerate the "specialty" probe, the magnet quenched again,
four weeks after being brought back to field... I'll send a follow-up to
the group about the quenches shortly.
Thanks again for the help
Ivan
Original message:
Dear All,
Our 600 magnet quenched a while ago without any warning. The quench
occured during the night when a "specialty' probe was installed. The
probe in question is commercially available and is used excusively by a
single research group. The probe was acquiring a 2-D spectrum at 25 C
when the quench occured. Since the quench had no other apparent
explanation, the specialty probe has been suggested as a contributing
factor.
Are any of you aware of any instances and/or can you envision any
scenarios where a probe at RT could cause a magnet to quench?
I will summarize responses if there's interest.
Responses:
Can you describe the experiment in more detail? For example, was the
probe operated with temperature regulation (to do a kinetic study at 25
degrees)? Was the probe set up for optical irradiation? Is this a very
large diameter probe (say 20 mm sample diameter)? I ask since I have
seen quench problems with all three probe variations mentioned above.
Of these, the temperature controlled experiment is the situation that
has caused many difficulties over the years.
It is possible that the problem was with the physical probe switch
itself. It does not take much of a mechanical shock to quench a
high-field magnet. If the experimenter did his/her own probe change,
perhaps he/she was careless and caused a mechanical transient. Pulling
the probe very rapidly from the magnet bore could do this via a coupled
eddy current mechanism.
***********
It's a long shot, but possibly the probe was arcing (possibly from a
mis-setting of the power levels?)and a hot spot transferred through.
***********
In my opinion, a commercially-available probe operating at 25 C can not
be the cause of your quench. Sometimes magnets spontaneously quench.
***********
I can think of absolutely no reason why a probe at RT could quench a
magnet.
***********
No, the only "probe related" quench I have seen was VT related and
occurred when the room temperature bore tube got so cold it froze the
"O" ring at the top. The two cause of quenches "without warning" that I
have seen are:
1) Ice blockage of the helium vents leading to pressurization of the
helium dewar. This may give warning, if you notice it, in the form of
an excessive drift rate before the quench.
2) Running out of helium
Was there frost or condensation on the room temperature bore tube after
the quench? This would indicate a loss of vacuum.
***********
Sounds extremely improbable to me. I've heard of many "spontaneous"
quenches, but none caused by a probe at RT. The only quench I've heard
of that could have been regarded as caused by a probe was when a system
was run at very low temperature for a very long period and the RT bore O
rings froze.
***********
This sounds like someone is looking for a scapegoat. It is undoubtedly a
coincidence that the "specialty" probe was being used at the time.
***********
There aren't many possibilities where a commercial probe could cause a
quench. Low cryogen levels will increase the risk in each case.
One scenario is if the probe was inserted or withdrawn with excess
force. The force of insertion/withdrawal of a probe is exerted
mechanically on the magnet and it's supporting structure inside the
dewar, and the eddy current is inductively coupled to the magnet
current. If excess force is used, the field, magnet, or its dewar could
shift, causing a quench.
Another remote possibility that comes to mind is if a gradient coil in
the probe was pulsed with so much current that it yanked on the magnet
or field. An off-center probe would contribute to this risk.
I would also check the past operation to see if the drift or shim
history indicated a problem. Spontaneous quenches are relatively common
in young magnets.
***********
I don't see how unless it over heated which might cause a problem. How
old is the magnet and what make?
***********
Not a Chance! If they're working on high energy materials (explosives),
you could have exploded a sample in the probe, but your bore and the
probe would both be severely damaged.
Did you notice a decrease in boil off of He or N2? Did you have a
tornado (extremely low pressure front) move through? Have your TA's
really been refilling the N2? Ice blocks of N2 or He are much more
common a problem. How new is the magnet? If less than ~6 months, it
could be a training quench.
***********
I'd be more interested in what the drifts rates had been approaching the
quench, changes in He flow rates and etc. I've never heard of a probe
causing a quench.
***********
I guess that _at_25 C implies that the heater was on, so it could have
malfunctioned badly enough to melt o-rings and quench magnets. But that
would also probably melt the probe. I suppose that an unbalanced
spinning sample could shake the system, but its unlikely to shake hard
enough to quench the magnet. The probe change could have put enough
torque on the shim stack to cause a vacuum malfunction.
I'd probably call it an "act of god" and hope that it doesn't happen again.
***********
A deep temperature probe may cause a quench, but in the case of room
temperature
I can see no reason. A very rare reason for a quench is a sudden switch
off of the
shim currents.
***********
I haven't had direct experience, but from my NMR experience over the
years, I'd say it was highly unlikely the speciality probe caused the
quench.
If it was operating at extreme high or low temps, there would be a
direct correlation. Extremely hot or cold feeder gas or exhaust could
damage o-rings on the magnet and crack the magnet vacuum.
The only other ways I can invisage a particular probe damaging the
magnet would be:
i) if the probe damaged the magnet while being inserted - however this
would require such force the probe would probably be inoperable
immediately ii) if the probe employed extremely high pulsed field
gradients, the gradient coils were not cooled correctly, they
overheated, and caused damage to the magnet. Again, the gradients would
probably burn out themselves and cease functioning before this very
unlikely scenario.
I'll be interested to hear what others say - but I think it was just a
case of bad luck, and something else about the magnet caused the quench.
***********
In 40 years of dealing with many magnets I have never heard of any such
thing.... only the low/hi temperature, cryogen on the flange O rings,
and stuck/impaled metal pose such threats... I can't imagine any
rf-magnetic field coupling doing it unless there is something odd in the
probe you refer to as 'specialty' let us all know if anybody does come
up with a possible scenario!
It seems a curious case...though I am sure any manufacturer will tell
you that there is always that possibility of an imperfectly epoxied turn
moving...
***********
I've never heard of such a thing, and after thinking about how a probe
could quench a magnet, not sure it's even possible. A really big
magnetic field gradient might do this, but it would take a very powerful
gradient driver, beyond likely what you have. VT could quench a magnet,
but you were at 25C. Very likely the use of the specialty probe was a
coincidence the quench event. Be interested to hear what happened,
if/when you find out.
***********
If this this was a capLc probe no way. If it was a high power grad probe
and it overheated maybe.
***********
First, is not logic that a probe could cause a quench.
Therefore your questions to that group could be:
Who install it? How was the probe installed? Which tools were used to
install it? Something wrong done during the installation?
I'm assuming that you were not present during that process.
Do you have more info about it (probe brand, etc, etc.)
***********
I assume that some probe would be in the magnet at all times. Theirs
happened to be it. Even if the probe got very warm it because of BVT or
air flow failure, it would not cause a quench. If it was not burned up
with scorched metal obvious, it is not guilt.
***********
I would love to hear possible explanations on how a probe at ambient
could cause a quench - I cannot think of a mechanism that a probe could
directly cause a quench of a magnet. The probe would have to somehow
interact with the magnet coil or the dewar system.
Shaking the magnet, insufficient cryogen, soft vacuum, "warm" He bubble
hitting the wrong place at the wrong time, stressing the magnet (like
taking a T tank a bit too close)...
It might be useful to speak with someone from the magnet dept of the
vendor of your NMR...
--
Ivan Keresztes, Ph.D.
Director, NMR Facility
Department of Chemistry and Chemical Biology, Cornell University
Baker Laboratory, Ithaca, NY 14853.
phone: (607) 255 0709, fax: (607) 255 4137, email: ik54_at_cornell.edu
Received on Mon Oct 03 2005 - 19:27:33 MST