Where i'm at 1+1=3

[Smudge]

It turns out "pushing" the Fe plate back in every 60 degrees is actually a strong pull in, you gain energy.  Pulling the plate out at the appropriate points requires little energy.  Summing the energy gained at pull in with that gained from rotation tells me this machine is capable of 10 watts output at 6000 RPM.  I need to go through all this again working with a finer mesh and checking everything.  One thing I got wrong, I actually did the runs using the opposite rotations yielding negative torque, but simply changed the signs for reversed torque.  But that screwed up my pull out at 40 degrees, it is actually at 20 degrees with the correct rotation.  But the pull out force is negligible compared with the pull in force.

Smudge

[Centraflow]

Here is my work so far.  The first image is my FEMM arrangement with the Fe plate at its innermost position.  The next image shows the plate at its outermost position, that is a movement of 4mm.  Next is the torque v. angle for the two positions.  This is just one torque, it has to be doubled for both rotations geared together.  The average torque in each case is small, it would show as zero on these charts.  Next I show the results for the Fe plate being suddenly pulled out 4mm at the 40 degree angle (and of course pushed back in at the 60 degree angle).  The average torque is now non zero, so the machine would keep running with this Fe pumping action.  My next step is to see whether the energy consumed for a single in-out pump is less than the energy gained from the 60 degree rotation.

Smudge

You have drawn it better than me, but that is what I was trying to explain.

The angle changes as the wheels rotate.

[Smudge]

Well I have to report that on checking everything the system is not OU.  It takes energy to do the in-out move and that consumes more energy than that obtained from the rotation.  Still looking into other possibilities but not hopeful.

Smudge

[Verpies]

As i happen to have the same programs as mentioned in the attached PDF some posts of my ago, i tried to analyze / simulate the forces / flux density etc. of 2 opposing 10x20mm N52 magnets Graham is using.

EM Works ?

[Itsu]

Yes.

[PhysicsProf]

Well I have to report that on checking everything the system is not OU.  It takes energy to do the in-out move and that consumes more energy than that obtained from the rotation.  Still looking into other possibilities but not hopeful.

Smudge

Thanks for this!