[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [FEMM] Loading external data + Discrete magnets



Steffen Jankuhn/IOM wrote:

Dear FEMM group,

first of all, I would like to thank Dr. David Meeker for
his excellent programme.

We are using FEMM in order to simulate the magnetic field
applied in our ion sources (see e.g.
http://www.iom-leipzig.de/muehlleithen/2004/jankuhn_04.pdf
). The cylindrical permanent magnets (AlNiCo, NdFeB) are
arranged around a cylindrical tube - a typical case for a
spherical, axisymmetrical approximation. An approximation
of the 0th order, because we have discrete magnets and not
a ring magnet.
How is B scaled in this case (with the number of magnetons
or, simple, with the volume or ...)?



The best way to go about it would probably be to think about magnetic circuit representations of magnets to derive an equivalent material. Eyebaling this quickly, it looks like that for NdFeB materials, you end up getting a material which has more or less the same permeability as air, but with an effective Hc which is the Hc of the material multiplied by the fill factor for the annulus that you are approximating as a continuous magnet. AlNiCo magnets are more of a pain to do in this way, because they are magnetically pretty nonlinear.

Now we have measured the real magnetic field and we would
like to compare the measured and the simulated field. Or
better, finding the factor which fits the simulated
ring magnet to the discrete magnets arrangement. For a
better presentation, I would like to load the measured
data into FEMM. Is this possible now?


The Lua language embedded in FEMM is pretty general with respect to allowing one to read and write data in arbitrary ways. However, it sounds like what you are really after is doing things like making graphs that have experimental data and results from FEMM on the same figure.

Or would it be a suggestion for the future?


I'm not planning on modifying functionality like line plots of field data from inside FEMM. Instead, I've spent a lot of time recently making detailed interfaces from FEMM 4.0 to both Mathematica and Matlab. The capabilites of these interfaces subsume what one can presently do with Lua. A great variety of graphing options are possible by using FEMM through one of these programs. Time-limited beta versions of these interfaces (MathFEMM and ActiveFEMM) are currently available from the FEMM website, although these beta versions are due to expire fairly soon.

Dave.