Shield low-frequency magnetic fields, but not high-frequency ones

[kvt]

I am wondering if it is possible to shield strong (1 Tesla) DC magnetic fields with high attenuation factor (10^4 or better) WITHOUT also shielding AC magnetic fields above frequencies of ~10kHz.

I have looked at mu-metal, which as far as I can tell shields DC but also AC fields to some extent, and active magnetic shielding, for which I could not find anything with a high enough shielding factor.

The to-be-shielded region would be approximately 0.1m x 0.5m x 0.5m in size.

 

[marcusl]

To shield your volume from a 1T ambient field, you need a material with the highest coercivity (or saturation field level) possible. Mu-metal is inappropriate because it sacrifices coercivity in order to get high permeability. You'll be using iron, and lots of it. If you could settle for a small shielding factor, you might use disconnected slabs at a large distance--the windows between them would let high frequencies penetrate. It won't be easy to achieve a 10^4 shielding factor this way, however.

Another thought is to build a magnetic shield with open ends that functions as a waveguide beyond cutoff below 10 kHz. I expect it will be quite large compared to the dimensions you list.

A wider, more "out there" thought--maybe you could install something akin to a cell-phone "repeater" inside your closed shield, that is, put a big antenna inside that is connected to a big antenna outside.

 

[rigetFrog]

There are two parts here.

1) Sans AC shielding. This means you're going to either need either a thin metal (E&M textbooks calculate the frequency dependent penetration depth using Maxwell equations), or an insulator. The only ferromagnetic insulator that comes to mind is Yttrium Iron Garnet (YIG) and its 4pi Ms isn't very large.

2) Magnetic shielding. The geometry and dynamics of your field source is important here. Without specific details, I can only give a general approach:

Any ferromagnetic material you do use, can be thought of as a collection of of little dipoles (i.e. North South magnets). Your task is to arrange them such that they cancel the field in the desired location. If you use high coercivity material, you can think of your little N-S magnets having a fixed orientation in space. A highly permeable material means your little N-S are free to re-orient themselves. They will try to align themselves with the local field. The local field will be a combination of the neighboring N-S magnets, the applied field, exchange field, and anisotropy fields...

Ok, this is getting a little more involved than I expected. If you have time, download a copy of OOMMF or NMAG and just numerically simulate the system. They're both open source (i.e. free). Expect about a week of learning. OOMMF get's kind of tricky, but I don't know if NMAG is any better. Post here if you get stuck.