FEMM simulation showing COP 3 and 7

[Omnibus]

The forum may not be @broli's own forum but the truth is it is infested with arrogant incompetents such as @exnihiloest. So what that femm uses the laws of physics to modelyze the system. As I have shown, the laws of standard physics taught in colleges and universities contains inherently violation of CoE, as I have already shown. This fact has been missed so far in theoretical physics but it desn't mean it isn't there. Therefore, there is nothing unusual for femm to uncover situations whereby OU shows itself. One such situation is the one studied by @broli and he deserves admirations and encouragement for that finding of his and not attacks by incompetents such as @exnihiloest.

[lumen]

At this point I will say that the energy gain exists when modeled in Maxwell 3D also. It's like the fields pushing against each other within the iron, exerts no force back into the magnets. The force needed to separate the iron when the magnets are close is greater than what is gained by moving the iron back to the magnets, but the energy gained from the separating magnets is still greater and in the end the process does show a large gain.
I also examined another modification (shown below) to this process by dividing the iron into two pieces and moving the iron apart in the same direction as the magnets would separate. Surprisingly, the iron already wants to separate when the magnets move into place and as the iron moves apart the magnets move apart with greater force which can be used to move the iron apart, which increases the force to move the magnets apart!
Something like a self starting separation that increases in force. After the magnets are separated, the iron will slide back to the opposing iron with no force and again the magnets can be moved back to center with no force to restart the event.

[fletcher]

Only a build will confirm the if the two different sim software results are accurate or not - two products showing the same or similar gain means either the sims have it right, which can be replicated in real world - OR - both sims are not sophisticated enough to model real world.

[Omnibus]

Like I said, the sims are correct in their OU conclusion because it coincides with the results from the experiment, as already seen.

[broli]

Thanks for your constructive contribution lumen.

This is a rather simple problem for such simulators which are used to accurately predict much more complex systems. It would defeat the purpose of simulating anything if these simulators couldn't predict such a simple system. To add insult to injury even vizimag shows the same behavior.

Again, we are talking about tens of joules in gained energy per complete cycle. on a palm sized setup, this is too enormous to just brush off as a "bug". Especially when experiments show similar behavior.