3D printing a structure for an experiment with magnets

[Low-Q]

Here is my two models. As you can see, the small one is not good...

[Low-Q]

Here is a design that is based on the magnetic interaction of thin magnets to the back side of opposing magnets.
It seem that when a thin magnet overhangs another thin magnet the back faces cause a huge change in the normal result.
Even opposing faces will attract with enough overhang.

Check out the simulation!

I have watched the animation. What I see is that the rotor magnets are twisting opposingly to the ring magnets magnetic field. That happens in both sides. It appears to me that this rotor would rather go clockwise than counter clockwise. Hard to analyze without the correct data.


Vidar

[lumen]

I have watched the animation. What I see is that the rotor magnets are twisting opposingly to the ring magnets magnetic field. That happens in both sides. It appears to me that this rotor would rather go clockwise than counter clockwise. Hard to analyze without the correct data.


Vidar

The twist is only to apply a different reaction between the stator ring of magnets and the two rotor magnets.

At the point where the rotor magnet is vertical (or forming a "+") to the outer ring, then there is more connection to the outside face of the stator ring.

This changes the reaction to the (sticky spot) that would rob any gained energy from the rest of the rotation where energy is gained.
The gain is from the expanding distance while repelling and then the contracting distance during the attraction. By then the magnet has rotated and increases the connection to the back side making the connection almost neutral while crossing the sticky spot.

I am running a 3D simulation to test the process but sometimes in a complicated setup it becomes easier and faster just to build the thing.
Here is the clip to show the backside connection with real magnets. 

https://youtu.be/ZoDg6hVHttU

I don't know if it's something one would try to build with a 3D printer but if I get good simulation results I might build it.
From your projects it appears that printing is a reasonably fast method for prototyping.

[Low-Q]

.......
I think about what I have missed,, or might of overlooked or not even considered.

What is the reaction to the pin that will keep your ring magnets in orbit,, is all of the "torque" from those countered by the long rod? or can the long rod reflect that torque in a way you have not considered??  see,, stuff like that is where I tend to go.

I imagined the outer magnets was connected to a spring that forced them to flip over. If I force them back in right position by "winding up " the springs, they will use the tension in the spring to rotate the outer rotor. This seems logic.
So I made a time consuming simulation and calculated torque for every outer magnets center in 12 steps (10° steps for outer rotor, and 15° steps for inner rotor (120° outer rotation, 180° inner rotation)).
In my model in FEMM I have in average -9.5Nm torque in all three outer magnets combined.
The outer rotor has an average torque of -17.5Nm.
The inner rotor have an average torque of +19.5Nm.


Total torque of outer rotor is therfor -27Nm.


Outer rotor have 1 revolution.
Inner rotor have 1.5 revolution.


So:
Outer energy pr. revolution is -169.6 Joule
Inner energy pr. 1.5 revolution is +179 Joule


I need to consider the resolution and accuracy of the simulations. As I have left 9.4 Joules per 1 complete cycle, it is a possible error reading of 5.5%.


Or, if I was right about my first assumptions:
Outer energy pr. cycle is 110 Joule
Inner energy pr. cycle is 179 Joule
Resulting in excess energy of 69 Joule pr. cycle...


Even if I'm a sceptic, I will finish the project, but I need some heavy duty ball bearings that fits 10mm bolts. Friction between PLA plastic and a 10mm bolt thread is so high when the magnets are attracting or repelling eachother, it is impossible to make a good practical experiment. And the printed structures make the outer magnets too wobbly. However, I locked the outer magnets by tightning the bolts that is holding them, and put the inner rotor in place. The inner rotor REALLY wants to rotate (Hard to hold it back), and drags the outer rotor with it, but that was when the outer magnets was fixed in one position on the outer rotor. That proves only one thing so far: The inner and outer rotor have similar torque, but the gear ratio will force the inner rotor to rotate - when the outer magnets are locked/fixed.


Vidar

[Low-Q]

The force that is interacting on both is the same quantity,, the radius is what creates what we call torque,,  If the force is the same but the distance to center is not then the longer one has more torque and the shorter one less and then it is radians of change ( degrees of rotation) so the shorter arm needs to move more radians than the longer one to have an equal amount of energy.

Now the gears are then converting that "force of rotation" and they are interacting as a lever with each other,, so that is a sliding face contact and might also have a high friction,, they are passing the same amount of energy as your magnets interaction but on a smaller size.

Thanks for the tip regarding ball bearings and magnetism. The magnet will stand upright with nort/south horizontally, and the bearings also horizontally, but with some 5-6mm space between magnet and bearing.


I have to finish this project. I am comfused, and the only way to confirm its operation is to build it 


Vidar