Rotating halbach array idea.

[Low-Q]

Hello. Long time no see. I just want you to take a brief look into this design. It is a magnetmotor with two stator magnets, and the rotor consist of 5 magnets.
The 5 rotor magnets themself spins in opposite direction of the rotor, and they turn one round clockwise while the rotor turn half round counterclockwise.


At all possible positions of the rotor, the rotor magnets are align in such a way that they make up a halbach array.


I have tried to simulate this in FEMM 4.2. FEMM tells me that the over all torque is high, and in one direction at every position I simulate.
I have in mind that the rotor magnets themself, also rotates around each separate axis. See animation for this. These 5 rotating magnets is driven by the rotor, so that can be a problem for sure.
However, what FEMM tells me, when I simulate the torque that is around each of the rotor magnets axises they are both positive and negative, and when I sum up these torques, and any random position of the rotor, the sum is zero. So, what FEMM suggests, is that keeping these 5 magnets rotating, does not enhance or hold back rotation of the rotor.


With these results in mind, I started to 3D-print parts for this model.


One large and fixed gear in the middle. On the 5 spoke rotor I place 5 gears in connection to the large gear, and a second set of 5 gears connected to these. This will make a counter rotation of the outer most gears.
Inside those outer gears, I have made room for a stack of 6 round neodymium 10x2mm magnets.


As stator magnet, I stack some ferrite magnets as an arch. Notice that the same pole is pointing in the same direction for both stator maglets.


The print is almost done, so assambly will take place today. I do this becaus I cannot understand how FEMM can EVER suggest that a permanent magnet motor could work. Fingers crossed.


PS! How in the world do I share an animated GIF? It seems like this GIF is much smaller in filesize, and is not animated....


Br. Vidar

[ayeaye]

I do this becaus I cannot understand how FEMM can EVER suggest that a permanent magnet motor could work.

Yes evidently FEMM has shown overunity in some setups. Though as much as it's known, none of these were never physically replicated. Why FEMM can show that, as my experiment shows, an asymmetry of the field (non-Coulomb irregularity) can provide overunity. Such asymmetry requires difference from the Coulomb model for magnets. But the Gauss equation (Maxwell equations) provides such difference, and FEMM does modeling based on Maxwell equations.

I don't know about that design, maybe some asymmetry of the field may provide overunity there. But consider that when you physically build something modeled by FEMM, you may not get gain of energy even if FEMM shows it. Because FEMM doesn't consider friction, and the biggest problem used to be that the overunity is not greater than friction.

[Low-Q]

Yes evidently FEMM has shown overunity in some setups. Though as much as it's known, none of these were never physically replicated. Why FEMM can show that, as my experiment shows, an asymmetry of the field (non-Coulomb irregularity) can provide overunity. Such asymmetry requires difference from the Coulomb model for magnets. But the Gauss equation (Maxwell equations) provides such difference, and FEMM does modeling based on Maxwell equations.

I don't know about that design, maybe some asymmetry of the field may provide overunity there. But consider that when you physically build something modeled by FEMM, you may not get gain of energy even if FEMM shows it. Because FEMM doesn't consider friction, and the biggest problem used to be that the overunity is not greater than friction.

To be more accurate about friction and over unity. If the machine isn't overunity, and you remove friction completely, it will not start to work. Overunity is when the energy out is more than energy in, and that extra energy would fight friction, and still have more to make a rotation. However, as we do know (If we reeeeally look deep into our sane mind), overunity isn't something that can happen.


I know that FEMM sucks. It is a 2-dimentional simulator, and it is very time consuming to find torque for objects that is not centered at 0, 0. Then I must move the model up and down, side to side, to align the test object in center to do the measurements.


I have absolutely no faith in my design. I just want to see in practice how the thing works by playing with it. There is for sure something I have overlooked in FEMM also.


Vidar

[Low-Q]

Aaahhh. After 14 hours of 3D printing, I discovered that the outer small gear only takes a 180° counter turn instead of 360°, when the rotor turns 180°. Ofcourse it does, because the gear is going 360° counter clockwise when the rotor is going 180° cloclwise. Didn't think of that!


So I am printing a new, and smaller one, but with a large outer stationary gear instead. This large gear have 52 teeth, while the small gears still have 13 teeth. That is a 4 to 1 gear ratio instead og 2 to 1 as the previous one. That should be correct. Still 8 hours left before the printer is done....



Vidar

[onepower]

Low-Q
You Rock, my friend.

Even if this never pans out you have made the effort and that is what counts. I fail 99% of the time because reality is seldom what I think it is, I am often mistaken. However success has little to do with being wrong but recognizing when it happens and changing what we do and how we think. If one cannot change then one cannot move forward and learn... the only metric or measure that counts is learning.