Magnetic Shielding of Detectors

[Cmertin]

My professor has given me a project, figure out how to magnetically shield some detectors which surround a solenoid. Now, we have mumetal tubs which fit the detectors, but he wants numbers. The solenoid outputs a maximum of 6 Tesla, and it's in a dewer, so there's no way for me to perform actual data measurements. He told me to perform a simulation, given the solenoid's dimensions. The problem is, I'm not sure where to start. He told me, at the beginning of this project, that he's "ignorant in this area" and that I have to find it out myself, including the resources. I have yet to take Intermediate E&M yet, so I'm sort of at a loss at the magnetostatics of this all. I have some resources on magnetiostatics, but I was wondering if any of you had any good resources on magnetic shielding, possibly even on modeling it in something like MATLAB or something? I can't really find anything. I got a book, Applied Electromagnetics using Quickfield and MATLAB, by Claycomb, though it doesn't really talk about how to model it, or even how this magnetic shielding occurs.

Am I trying to find an answer to something that is above my level, and that I probably won't understand until grad school?

Thanks in advanced.

Edit: My professor is in Nuclear Physics. I have asked people in the High Energy Department, hoping that they would have some information, as they would deal with stuff like that at places like CERN and the like, but they couldn't help either and didn't have any references for it. All I was able to ascertain from a professor is that "common practice" is to put the PMT 1 mumetal tube diameter inside the tube.

 

[berkeman]

Holy cow! 6 Tesla is a big field. Yikes.

I have used products and consulting from Magnetic Shield Corporation in the past, and have been happy with them:

http://www.magnetic-shield.com/faqs-all-about-shielding.html

The applications literature at their website may be of help in starting to get your head wrapped around the project. You might also give them a call to discuss your application -- I have found them knowledgeable in the past.

Could you post a sketch or pictures of the setup, and say more about what needs to get shielded from what? Where is the 6T number referenced too, how close to the solenoid source? And how far away will the shelding be, and how far away are the things that you are trying to shield from this field? What level of B-field is allowed at the objects that are to be shielded from the strong field? What is the $ budget for the shielding?

 

[Cmertin]

The good news is that the solenoid is only approximately .2 meters in diameter, and is encased in a dewar that is approximately .5 meters in diameter. Leaving approximately .3 meters at a minimum of playing, but I'm thinking that this is way too small. My professor didn't give me the design specifications of where the detectors are to go yet, as he's also having me model the solenoid to get the off-axis elements of it. The 6 Tesla is the on-axis element at z = 0.0cm. I only have the data for the on-axis elements, so if my model has these correct, we're going to assume that the off-axis are also correct.

I'm thinking that the budget is low, as he told me he wants to do this the cheapest way possible. The mumetal came from the high energy department when they were getting rid of them for some reason or another. Apparently the type that he got is "pretty thick stuff." The idea came up that we might have to wrap each detector in it's own solenoid to counteract the magnetic field inside the mumetal tube itself.

The level of the magnetic field that the PMT's can take is sort of unknown. I'm not sure 100% the model of them. I know that the current PMT's are good at not having a magnetic field effect them, though are poor with energy resolution. I think that they are designed to withstand 50 Gauss or so. However, he wants to "upgrade" the PMT's into ones that are more accurate, though those can only deal with magnetic fields on the order of a tenth of a Gauss I think? Again, coming from the high energy department.

I will try to draw up a diagram and ask him some more about it tomorrow. In the mean time, I will read up on the link that you gave me. It seems like it's a really good start. Thank you!

 

[berkeman]

Are these dewars cryogenic? If so, I don't think you should think about putting the shielding inside of the cryogenic environment, if that's the initial proposal. Magnetic materials do not respond well to extremes of temperature.

You could shield the outside of the dewar, and/or you could shield the PMTs themselves. Also, if you are wanting to re-use mu-metal material, keep in mind that there is an annealing process that needs to be used on the mu-metal after it is re-shaped, in order for it to keep its high-mu properties. Also keep in mind that when you have a very high field like 6T, mu-metal saturates quickly and becomes ineffective. It takes layers of different materials to deal with shielding of high B-fields (like iron, netic and co-netic materials).

 

[Steve Barlow]

Magnetic Shielding DANGER!

The magnet you describe ~200 mm bore and 6 T is NOT A TOY! Attempting to place mu metal shields anywhere near it can result in hazardous conditions that can kill or maim and even worse damage or destroy equipment. I am not kidding. The reason is simple; mu metal or any other ferromagnetic material wil be strongly attracted to the magnet, if it is not held properly it can fly across the room to the magnet destroying whatever is in the way--equipment, professors, students, etc. A number of years ago, the NMR/MRI folks produced safety guidelines for using large magnets, I do not have the reference, but it should be easy to find on the internet or from your local safety folks--get a copy and read it.

With that said, shielding detectors in the vicinity of such a magnet is normally not a simple task. These large bore superconducting solenoids normally have very large fringe fields in both magnitude and extent--this is what produces the hazardous conditions referenced above. (There exists a class of modern large bore magnets with "active shielding." These solenoids are much safer, but still need to be respected.) Most mu metal shields on detectors are made of one or more layers of thin material 1-2 mm and are designed to shield out the Earth's field or other relatively weak fields. In the large fields near a superconducting solenoid, the shields will simply saturate and provide very little effect.

So as a first cut in tackling your problem, you need to know what the fringe fields are. The fields can be calculated by analytical means for an unshielded magnet--a little messy on account of elliptic integrals. The stray field maps for actively shielded magnets are usually available from the manufacturer.

Second, you need a model of your shield--shape, size, wall thickness, etc. You should then be able to determine whether the shield will saturate at the proposed location(s). With a little help from the shield vendors mentioned in a earlier post, you can also determine what kind of shield design would actually work. For instance you may need a large, thick outer shield made of soft iron instead of mu-metal. The shield design will also depend on how much shielding you really need for your detectors to work properly. Again, the experts are at the shield manufacturer's company.

Third, the presence of all this magnetic material around the solenoid will distort the solenoid's field. Depending on what you are doing, the distortion can render the system useless.