Hi all,
I've heard back from several of you on the potential for a "domino" or "sympathy" quench - a quench induced in one magnet because one nearby undergoes a quench; in our case, we were worried about an old unshielded 600 being adversely affected by a potential quench of a new 400 during installation. What a pleasure to receive such detailed, definitive information on this topic!
The gist of the feedback is that we're almost certainly safe as long as we keep the 5G lines apart - over 3 feet from one another in this case. Many facilities have magnets whose 5G lines overlap - in some cases the 5G line of one intersects the magnet bores of others! - and no one reported problems associated with one of them quenching, aside from the need to reshim.
There was some additional discussion about the problems of a field "bloom", in which the shielding of a shielded magnet fails before the main magnet, leading to a temporary increase in the instrument's stray field. It seems that modern magnets are equipped to minimize such effects.
Many thanks to all those who responded.
- Josh
COMPENDIUM:
****
It is my understanding that the distance between the two magnets would be over 1.875' + 3' + 13' = 17' 10.5".
The stray fields produced by the new 400 at the location of the 600 magnet are extremely small, thus the change in the magnetic field that the 600 would see in case the 400 were to quench should not pose problems.
Based on the description provided below the 5 Gauss lines of the two magnets do not cross.
We did gather some experience from sites where new shielded magnets had to be placed in the same room with older non-shielded magnets.
One particular site comes in mind where an older Oxford non-shielded 800 magnet is located exactly in the center of a 40' x 40' square room, with the 5 Gauss touching all walls in the middle.
* We were able to install new 600, 700, and 850 shielded magnets (UltraShield Plus and Ascend) in the same room, one magnet in each of three corners that were available (the forth corner was taken by other electronic equipment).
* We ended up having the 5 Gauss line produced by the older Oxf 800 non-shielded magnet almost reach the center of each of the new magnets. So in this case the 5 Gauss lines were crossing.
* The distance between the magnets was ~ 20'. We were not so much worried about the negligible effects produced by the new magnets on the older 800 non-shielded magnet, but the other way around with possible effects of the 800 given that its 5 Gauss was reaching the center of the new magnets.
* Everything turned out fine and everything is running well.
* One of the new magnets did experience a quench during the installation but it not affect the non-shielded 800.
****
I can't vouch 100% for Oxford Instruments magnets, but this sounds like it's not quite as close together as the magnets are in our test lab. While we do not have so many training quenches nowadays and certainly have far fewer than Oxford Instruments used to, there are enough that if there were an issue, we would have picked it up. And to my knowledge we have *never* seen such a "domino" effect.
...I do not believe that any complex (such as NMR) shielded magnets will be 100% "protected" from the bloom field, since the only way to guarantee it totally is to not subdivide the magnet but protect it in one loop. However, there are several design measures that we incorporate to minimize it. Our quench modelling shows that at the 5 gauss line, the "bloom field" or "fringe field burst" is limited (without the design measures, the field could grow to >>50 gauss); however, measurements on various magnet prototypes have shown that it is even further limited beyond this and typically only increases marginally beyond 5 gauss.
This is all pretty much public domain, but I cannot go into any further details than this without a CDA.
I know that later Oxford Instruments included their own (different) measures to limit bloom field as well. I am not so sure about the very first ones. Some earlier Magnex shielded magnets did not include these measures, but most did at some level.
****
I have worked over 100 magnets, many in crowded labs with overlapping 5G lines, and I have never seen a sympathetic, or induced quench through the radial stray field. Our 400WB and 600NB overlap 5G lines and the 400 quenched 3 times before reaching field and did not bother the 600 at all. In the Magnex factory in Yarnton, Oxfordshire the magnets undergo training quenches in closer proximity than anywhere else, and we didn't experience cascading quenches. Also, the longer a magnet stays at field, the more difficult it becomes to induce a quench, due to equalisation of internal stresses over time, so your 600 is likely very, very stable. BTW, considering stray fields by their diameter can be misleading, since only one radius is ever between the magnet and any neighboring object.
One thing that I have seen in many crowded locations is to reverse the field of an adjoining magnet, which creates a Null Field between them. Otherwise the stray fields between would add up. This provides a safe avenue between magnets for people and equipment. You might need to special order the 400 magnet in order to do that, due to the possibility of polarised protection circuitry.
****
I have never seen a domino quench unless the magnets were really close to each other. If the 5 G lines are not even overlapping there is only the most minute of chances. I think a person with a big bunch of keys is a bigger disturbance.
The one thing you should make sure is that the magnetic field of the 400 does not "bloom" during the quench. This means the shield goes first and then the main field follows. Our magnets are protected against this, but I can not speak for other brands.
The thing that might happen is a sympathy or compassion quench. Magnets just don't like it when their brethren are not feeling well.
...I was just kidding with the "sympathy" quenches, magnets have very little feelings for each other.
It is very unlikely that the 600 will be affected by a quench of the 400, if as you say the 5G lines are 3 feet or more apart. For our 400 this would put the 0.5G line at the 5G line of the 600.
As for the bloom, if one would assume that the shielding coil quenches first and the main field would expand to the size of an unshielded 400 then the 5G line would be at ca 7 - 8 feet. this would mean the 5G line of the 400 would be at the 10 - 20 G line of the 600. I think, but one never knows for sure, that this is still safe.
Just have your magnet maunfacutrer give you a specification what the maximum extent of the 5G line will be in the event of a quench.
****
That's a first for me... I've never experienced a domino quench nor have I ever heard of one.
I've attempted to run down a wide bore Oxford 400 with diodes, after opening the main switch it quenched.
There was a Bruker 600 very close by but the wide bore quench had no effect.
I'd say, run up the 400! The 5 gauss line is practically contained within the can!
You'll have no problems.
****
Can't help with the specifics of your question (would a 400 quench trigger a 600 quench), but a suggested answer would be to bring the 400 up to field another 5 ft away or so, then carefully move it into place after it is at field. Seems this shouldn't be too hard to do. Might need to finish with a cryoshim on the 400 in place, but if it's been at field for a day or two, the risk for quench would be minimized.
****
if the 5G lines are 3 ft apart, I doubt if you will have any problems.
****
I have heard of this happening but only between two magnets of the same field. I will forward your note to my service engineer and see what he has to say.
...The service engineer says that he has never heard of a domino quench. He also says that with the separation of 5 gauss lines you have described, he would anticipate no problems.
****
I find it hard to believe. We brought an 800 to field which was alongside a 600 whose 5G line was about 2 ft away. The 800 quenched just as it came close to field and the 600 never even lost it's lock!
...The 600 in question was a first generation shielded magnet - probably about 8 years old when the 800 was installed. Not exactly sure because we bought it used from Bruker.
We also have a 19 year old unshielded 400 which is our oldest magnet. We actually moved it cold from the basement of one building to the third floor of an adjacent building (up an elevator). Bruker brought it to field first, and then delivered and installed a 500 wide bore on one side and a 600 on the other. The 5 Gauss lines just meet on all three magnets. The 400 was not damaged when it's neighbors came to field, just a bit of shimming. However, fortunately, there was no quench.
Just an other anecdote. When the 800 quenched, the emergency exhaust system took too long to deploy and the chilled water lines in two ac units ruptured (these were located at the extreme opposite end of the lab). In the ensuing flood, the 600 was left standing in about an inch of water. So when I say that the 600 did not even lose it's lock - totally remarkable.
****
think you're quite a bit more likely to get struck by lightning.
Is this a widebore 400 or something? I thought new ultrashielded 400
magnets had a 5 gauss line just barely outside the magnet can (that's the
way ours is).
****
At XXXX the center of the 800 is at the seven (7) Gauss line of the 900. Our first 900 quenched spontaneously three times before it was decided to send it back to Oxford Instruments. All we had to do for the 800 was to reset the lock frequency to its original value and do a little shimming by hand. Need I say more?
Good luck with your installation of the shielded 400. At the time there were only unshielded 900s! The perturbation at your 600 from a quench of a shielded 400 should be negligible as compared to what our 800 went through.
****
We have a number of old Oxford 500's and one 600 of the same vintage as yours their 5 gauss lines all overlap , In the two quench we have had over the last 15 years neither induced a quench in the othe 500s or 600 we did have a slow quench on one of the 500s that made the 500 nearest to it unshimable until the quenching magnet was completely down.
Josh Kurutz, Ph.D.
Instructor and Senior Scientist for NMR
IMSERC, Chemistry Department
Northwestern University
2145 Sheridan Rd.
Evanston, IL 60208-3113
847-467-1681
imserc.chem.northwestern.edu
NMR Blog: www.imserc-nmr.org
Other: www.joshkurutz.com
Received on Fri Aug 12 2011 - 13:19:22 MST