[femm] Re: Axis-symmetric modelling

[Dcm3c@xxxxxxx]

In a message dated 11/1/99 10:24:59 PM Eastern Standard Time, 
peter.ellis@xxxxxxxxxxxxxxx writes:

> First I would like to congratulate Dave and his co-workers on a first-class 
> piece of work. From the time I was introduced to the FEMM package by one 
of 
> my colleagues to obtaining validated results from three planar permanent 
> magnet models was just one day.

Coworkers? Anyhow, I'm glad that the program has proven to be useful to you.

> I have a few (constructive) suggestions and a question:
> 
> It is sometimes useful to be able to divide the air spaces around the 
magnet 
> and airgap into regions, so that the mesh density can be increased in 
> critical areas. This is easy using Segments or Arc Segments, but I would 
> like to be able to hide these segment lines on the Femm View display and 
> printout, since they are not part of the problem. A provision for 
> suppressing the appearance of selected lines would be a handy user feature, 
> at least at the Femm View stage.
> 
> It would also be useful to be able to vary the range of colours or 
> equivalent grey tones for the density plot and its legend. This can be a 
> help when preparing output for eg; publication.

The line removal and greyscale shading are a good suggestion, as well as 
being not too hard to incorporate. I'll see about putting these into the 
next release.

> As a raw beginner, I had some difficulties with lack of explanation in the 
> manual for specific dialog box requests. In the "Properties for selected 
> block" box, I couldn't understand what the "side length" referred to, until 
I 
> connected it with the fact that I had generated over 32000 triangles in one 
> small bounded area!

The "sidelength" parameter has been almost universally confusing. I probably 
need to explain this one better for the next release, perhaps by putting a 
note as to what it is on the dialog itself.

> My question concerns the axis-symmetric mode of operation. It seems to me 
> that it should be possible to model a wedge of a cylindrically symmetric 
> system, which is presumably the case for the "axi1.fem" example given. I 
> would like to generate a model based on a cylindrically symmetric permanent 
> magnet, in which the left vertical boundary is the apex of the wedge and 
> coincides with the magnet axis. The magnet field direction is parallel to 
the 
> cylindrical axis. My efforts so far seem to produce nonsense, probably 
> because I do not yet understand how to set the correct boundary conditions. 

> If someone can provide a simple example I would be grateful.
> 
> Thanking you in anticipation. Peter Ellis

For an example of an axisymmetric permanent magnet, you could easily modify 
axi1.fem. This example is supposed to represent a cylindrical, air-cored 
coil. 
An example of an axisymmetric permanent magnet would be to take the axi1.fem 
example and change the the square coil region's block label from "coil" to 
"magnet". The problem would then represent a washer-shaped permanent magnet 
with the magnetization directed along the z-axis. The direction of 
magnetization defined in the material property for "magnet" is 90 degrees, 
corresponding to the z-axis. An orientation of 0 degrees would denote that 
the magnetization was oriented radially. The H_c parameter defines the 
strength of the magnet. For rare earth magnets, a good approximation is to 
pick the magnet material to have a relative permeability of 1, and choose H_c 
according to 

(H_c in A/m) = sqrt(Energy Product in MGOe) /( 2*Pi*10^(-6))

For example, a NdFeB magnet with an energy product of 40 MGOe would have a 
H_c of 1006584 A/m, which is pretty much how the "magnet" material in the 
example is defined.

Anyhow, thanks for the feedback.

Dave.
--
http://members.aol.com/dcm3c