Energy of magnetic field created by magnetic dipoles in a shphere.

[sotnet]

Dear everyone,

I am researching genetic algorithms and at the moment I am trying to solve a problem how would magnetic particles (dipoles) orient themselves in a thin hollow sphere.

Suppose that I have N magnetic dipoles placed in the sphere. There is no external magnetic field. The dipoles create their own magnetic field, thus the system has some magnetic energy. I am writing a genetic algorithm to minimize this energy.
I did the problem in 2 dimensions (on a plane) already, and it worked perfectly!

The problem is that I am not sure how to calculate the magnetic field energy created by dipoles in 3 dimensions.

The formula for calculating magnetic field created by one dipole with magnetic moment \vec{m} at point \vec{r} in SI system is:
\vec{B}(\vec{r}) = \frac{\mu_{0}}{4\pi}(\frac{3(\vec{m}\hat{r}) - \vec{m}}{r^{3}})

To find the energy, I just sum over all dipole moments \vec{m} and multiply by \vec{B} created by others (with minus sign).

In 2D I set the problem so that all the dipoles had \vec{m} and \vec{r} perpendicular, that is \vec{m}\vec{r} = 0 (magnetic dipole moment \vec{m} was perpendicular to the 2d plane the dipoles were on).

In 3D I have the \vec{m}\vec{r} term, which I am not sure how to calculate.

Any advice how to calculate this term to find energy in 3 dimensions?



Stan

 

[Lojzek]

You can define orientation of each magnetic dipole moment with two angles in spherical coordinates (theta,fi). Then you transform this vector into cartesian coordinates:

mx=m*sin(theta)*cos(fi)
my=m*sin(theta)*sin(fi)
mz=m*cos(theta)

And the product of r and m is:

mx*x+my*y+mz*z

where (x,y,z) is a vector from the source dipole to the second dipole.