Faraday's cage, but for Magnetism

Question

Consider a (slightly) thick, i.e. meaty, conducting shell. This can be considered to act as a Faraday's cage. Any charge put inside it can affect the electric field out of the shell(even though the field is symmetrical regardless of where you place the charge in the conductor). However, any charge outside the shell can't affect the field inside the shell(a consequence of Gauss' law).

Is there any material or a particular distribution that does the same thing but for magnetic fields only? In other words, is it possible to have a region of space in which anything magnetic inside that region can affect the magnetic field out of it, but anything that produces a magnetic field kept out of the region cannot affect space in the region, or vice versa.

Answer 1

Yes, for static fields

Magnetic shielding can be achieved by placing high magnetic permeability around the volume of interest. This doesn't block the external magnetic field, but it does draw it into the shielding material, creating an area of low field strength.

Since magnetic shielding relies on magnetic permeability, and electrostatic shielding relies on conductivity, to achieve magnetic but not electrostatic shielding you need a material which is magnetically permeable but not electrically conductive. Essentially, such a material can be made from any magnetically permeable material by powdering it, and separating the grains of the powder by (very thin) electrically insulating gaps.

There are also materials which have this property in bulk, for example ferrites (iron oxide containing ferromagnetic ceramics). You can think of a ferrite as an extreme version of powdered iron with non-conductive gaps between the grains: the individual iron atoms are separated by oxygen atoms, the iron atoms are still close enough to interact magnetically.

As a practical construction, you can either use a soft ferrite ceramic (typically a manganese-zinc type), as used in transformer cores; or powder mu-metal and coat it or mix it into a nonconductive carrier (eg epoxy, silicone). The powder doesn't have to be very fine, anything 0.1mm to 1mm could work okay, just as long as the gaps break up the single continuous conducting path; the trick would be figuring out how much powder and how much carrier to use. Ferrites can reach permeability around 10,000-15,000, mu-metal 80,000–100,000, but powdered mu-metal would be a lot less than that, probably pretty comparable to a ferrite.

Note that this type of shielding would become significantly less effective for variable magnetic fields, because those are always coupled with variable electric fields. But for static or slowly variable fields, yes, it is possible to achieve the effect you're describing.