Re: FEMM simulations errors ??

["David Meeker" <dcm3c@xxxxxxx>]

--- In femm@xxxx, harti@xxxx wrote:
> To explain it a bit more:
> 
> Look at the posted GIF pic.
> 
> I drew a clockwise red bouding box
> at the additional 4 points in the air
> which I placed additionally in the fem
> model near the 2 runner magnets (called "runner stack")
> 
> When I integrate it this way and use the "Force from Stress Tensor"
> I get around +640 N/m into x-Direction,
> but as you can see, the runner stack must move to the left side,
> due to the reduced magnetic field at the left side, so the 
calculated
> force is into the wrong direction !
> So is this a FEMM error or do I still have to make the mesh still 
> finer ?
> The dimensions are:
> 2 neodym stator magnets 10 cm x 1 cm and 2 runner magnets
> 3 cm x 1cm, 2 cm thick steel core as back flux core.

The way that you have your arrangement configured, the magnets should 
be pushed to the right--femm is giving a reasonable answer.

The way to see this is to think of each magnet as being composed of 2 
point poles of strength Q. For a N pole, the force on the pole is 
B*Q, where B is the flux density created by the stator permanent 
magnets. For a S pole, the force is -B*Q. When you look at it this 
way, you can see that the vertical force should be zero, and there 
should be a net horizontal force to the right. One reference for this 
trick is M. Lindeburg's "Engineer in Training Review Manual, 6th ed.," 
1982.

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