Can someone confirm this in FEMM?

[Low-Q]

I have been using this simulation software called vizimag to establish the field strength between magnets and the attraction they cause on a moving magnet. I'm simulating a certain setup now and every simulation I have performed shows COP > 1 for this setup. The highest I got so far is 4. This setup is as basic as it gets, I stumbled on it while looking at something unrelated. I now want to confirm it's not the software giving me wrong results by having someone else confirm the data in FEMM.

I think the graph is wrong. This shows that the approaching magnet is pulled harder and harder until it hits the blue line, and then, in an instant, it is counterforced instead of finding its sticky spot . This is just not what is hapening with a magnet setup like that - trust me, I have tried in practice, and simulated in FEMM too.

The graph should show a green area that is increasing, but also decreasing untill it hits the blue line, where the force is gradually going from pull to push. The red area should therfor be on the opposite side of the x axis of the graph, and the graph should also look much more like the graph that is displayed later in the thread.

When the moving magnet is between the stator magnets, they are so close, that the counterforce in that area should be very much greater than the force you have in any point outside it. However, the time and distance that any point of force is representing will be the same in both force and counterforce.

Vidar

[lumen]

@Robbie, I wonder how accurate the FEMM simulator is? I have a configuration of three magnets and a movement method that shows a gain in femm 4.2.

If you have some time and want to graph something that shows gain, let me know and I'll post the setup.
Plus, I need to be sure it's not just something I'm doing that's causing a problem in the simulator.

[robbie47]

@Robbie, I wonder how accurate the FEMM simulator is? I have a configuration of three magnets and a movement method that shows a gain in femm 4.2.

If you have some time and want to graph something that shows gain, let me know and I'll post the setup.
Plus, I need to be sure it's not just something I'm doing that's causing a problem in the simulator.

Hi Lumen,
Accuracy of FEMM depends on a few factors:
- mesh density. This defines the number of points that are calculated. The higher the mesh density, the higher the accuracy, but of course you need a PC that has a powerful CPU to allow fast calculations.
- material definition correctness
For the simulation in this thread I used default Neo magnet setting that comes default with Femm. To caluculate COP the absolute correctness of the magnets  is not very relevant. If you want absolute forces to be calculated it is of course.
I used a quite high mesh density that I defined manually instead of the default ' auto mesh' setting.
On top of that it is important to know that FEMM simulates in 2D not in 3D.
The depth of a setup can be somewhat realistic by defining the thickness of the magnets.

In general my experience is that for COP calculations of simple setups FEMM can be quite accurate.
I verified this by simulating a few different setups that have clearly COP=0 and checked how close FEMM calculates it to be 0. With the calibration setups I used, COP has indeed values very close to 0, e.g. 0.000000000001. But again, choosing a correct mesh density is key here.
FEMM, but also other simulators, does not take into account Eddy Currents that e.g. occur within Neo magnets when you move them very close to each other. Something to take into account when real builds are compared with these simulations. Ceramic magnets suffer much less from internal eddy currents.

I could have a look at your setup and simulate it in FEMM if it can be defined in 2D.
Just let me know your setup by posting it here or send me a PM.

[lumen]

The setup is based on magnets I have and I plan to do some actual testing on this setup with my digital scale to confirm the results also. (I also posted the FEMM setup.)

This is the setup:
The small magnets are .125 x.75 x 1
Large magnet is .125 x 1 x 2

The lower small magnet moves into and out of the page and is shown already inserted at the starting position.
(there is no way to check this in FEMM and this needs physical testing)

It is inserted .015" from the center line of the upper small magnet and shows a positive Y force at this location so it should pull itself in.
The edge of the large magnet is .350" from the edge of the upper small magnet.
At this point, the large magnet will move in the positive X direction and needs to be stopped at .350" from the left edge of the upper small magnet. At this point the lower small magnet shows a minus Y value and should be easy to remove.

The process is reversed by again inserting the lower small magnet with reversed polarity, at .015" from the center line of the upper small magnet. Now the large magnet will move back to the starting position.



A graph showing the force of the large magnet along the .550" X direction travel along with the force of the lower small magnet's force in the Y direction should confirm any gain.

[jonifer]

Never use for calculate 3D problem program designed 2D only. Try program Maxwell 3D from Ansoft (for example) - and you find no overunity in this system.