FEMM simulation showing COP 3 and 7

[Low-Q]

This depends on your boundary conditions. I've used the mixed BC just like in that famous FEMM tutorial, 1/uo*R*mm where R is the radius of the boundary in mm. The calculation of force is quite far from this boundary especially for the magnets. You can actually see the paths for the force calculation.

And again calculating force or "work" on a stationary object is meaningless. I have no clue why this is even mentioned.

The simulation files have been posted, if you think there's something wrong with it I'd like to see your "corrected" results. It's easy to criticize, but contributing is something else.

Thanks for the feedback @broli! I will try this experiment myself - under your conditions, and mine.
Btw. It wasn't meant to be negative critizism in my post. I just shared my experience about FEMM, which might disturb the results greately. I think such feedback are important. Who knows how much experience you got with FEMM? I didn't. Therefor my little advice. I hope you understand

Edit:

And again calculating force or "work" on a stationary object is meaningless. I have no clue why this is even mentioned.

You're right, provided that the "stationary" object never changes position.


Vidar

[Low-Q]

I did a test in FEMM. It looks like the net energy required to move the steel bars out and in are pretty much regardless of position of the magnets. The next test is to see the energy required to move the magnets back and forth in one complete cycle.

Vidar

[Low-Q]

I get COP 12... but that is calculated in only 4 positions:

1 - Initial alignment,
2 - Magnets close to eachother
3 - magnets close + iron bars separated
4 - magnets in initial alignment + iron bars separated.

All forces are calculated as average between those positions, and the distances varies with 20 units on the bars, and 104 units on the magnets.

When I got more time I will take more samples to get a better average.

Vidar

[broli]

I get COP 12... but that is calculated in only 4 positions:

1 - Initial alignment,
2 - Magnets close to eachother
3 - magnets close + iron bars separated
4 - magnets in initial alignment + iron bars separated.

All forces are calculated as average between those positions, and the distances varies with 20 units on the bars, and 104 units on the magnets.

When I got more time I will take more samples to get a better average.

Vidar

I suggest you use a lua script. I'm pretty sure that 12 will converge down to a lower value with more steps.

[Omnibus]

@broli,

You mentioned earlier you have an idea for a more efficient construction. Probably it would be a good idea to have @lumen try it with his Maxwell3D.