Discussion Forum

Hi,

I posed this problem before but I received no response so I try it again: I would like to simulate the dynamic process of demagnetisation of a liquid sample and study the magnetic field around it. In essence the real process is as follows:
(1) The sample is placed inside a complex magnet setup (similar to Halbach arrays)
(2) The sample becomes magnetised, hence an initial static sample magnetisation M is established.
(3) The magnets are removed, hence the sample demagnetises. This demagnetisation process is dynamic (Larmor precession) and can be described by Bloch's equation.
(4) During the demagnetisation process the magnetic field around the sample varies with time.

Now I have previously achieved numerical results covering (2-4) but only with the assumption that the initial sample magnetisation (point 2) can be described analytically. I would like to extend this by assuming arbitrary initial distribution of M, which cannot be described.

My solution was be to separate this problem into two:
(A) A static AC/DC model with arbitrary magnet setups, which yields a spatially varying magnetic field distribution B0 within the field of view
(B) The second identical model is created (with magnets but assumed zero remanent magnetisation), hence number and distribution of meshes should be the same. The initial field distribution B0 from (A) is imported which determines the initial sample magnetisation M as described above in (2), since they are related by Curie's law, B(r)~M(r).
(C) From there I can use my previous approach
I am now wondering how to link both A and B, or with other words, how to import B0 from one model to the next.

Kind regards
Michael