Magnetic Gears For Power Transmission

[person123]

This thread is an attempt to better understand the pseudo direct drive, or PDD, gear. What I do know is it's made up of a set of magnets which can serve the function of a reducer- a basic diagram of it as well as a video can be seen below. In this mechanism, the inner rotor is the input while the pole pieces is the output. The outer rotor is said to usually remain stationary. Without any physical contact, the power is transmitted from the inner rotor to the pole pieces.
Would it, however, be possible for both rotors to become the inputs simultaneously? If, for example, the two rotors are going in different directions and with different speeds, would the pole pieces rotate based on both of the rotors, or would it instead fail to function? I couldn't find the answer to this question in any videos, and I would be interested in the opinion of this forum.

 

[Jonathan Scott]

Rotating the outside ring is equivalent to rotating the whole apparatus. The same relative rotary motion would be observed, but whether that is practically useful depends on how it is used externally. Note that connecting a third external rotation would be more complicated; for just two one can simply have an axle coming out either side connected to the relevant part.

The effect is fairly similar to that of having a pair or ring of smaller gears between a gear ring corresponding to the outer rotor and a gear on the inner axle corresponding to the inner rotor.

The way it works is that the number of pole pieces is less than the number of pairs of magnetic poles around the outside, so they match up with different points in each pair, effectively connecting a wave of different points in the magnetic field cycle to the inside. The number of cycles of that wave around the inner rotor is equal to the difference between the number of pole pieces and the number of pairs on the outside rotor. This wave then matches up with the pole pairs on the inside rotor, pulling it into alignment. As the ring of pole pieces moves past a single pair on the outside rotor, the wave pattern completes a cycle, which means that the inside pattern also moves around by one cycle, but the cycle at the inside rotor is a larger angle, effectively gearing up the rotation speed.

 

[Jonathan Scott]

I see that on their website they show a diagram of the equivalent set of planetary gears as I mentioned above, in the section "Magnetic Gears Explained" on this page:

http://www.magnomatics.com/pages/technology/low-ratio-magnetic-gears.htm

The Wikipedia page on Magnetic Gear discusses a similar mechanism under "Second-order Device" but but where the ring of pole pieces is on a fixed "stator" and the outer ring of permanent magnets is the other rotating component, but this is very similar. The Wikipedia page also contains a link to some patents, including one from 2005 which appears to match the Magnomatics design.

It appears to be a useful and ingenious design, avoiding friction and wear on gear wheels, and automatically allowing slippage as necessary in the case of overload. It is probably far more practical with modern rare Earth magnets than it would have been with older types of magnet. For higher speeds, the rapidly changing magnetic fields might perhaps cause some heating of the materials, but apart from that it seems very neat.

 

[person123]

Thanks for the insight. I was wondering about one more possible function for the magnetics gears which may be useful for a design I'm currently working on.
If a torque were to be applied on the middle unit as well as the inner and outer rotors, would the mechanism continue to function properly? In the case of planetary gears, I believe the parts would simply lock up. However, I would imagine that for magnets, all three units would change their angular velocities in relation to one another, slipping when necessary. If this were the case, do you know of any equations I could use for this scenario?
I admit that I am new to the concept of magnetic gears, so I might not be showing a full understanding of how the mechanism works.

 

[Jonathan Scott]

Why do you think that planetary gears would simply lock up? When one ring is fixed, torque is still being applied (probably via the mounting of the gear system as a whole) to keep it from rotating. From a gearing point of view, I think that the magnetic gear system is very similar to the planetary gear system, except that it has more elasticity and allows smooth slippage when the relative torque overcomes the strength of the magnetic forces (although a similar effect could be achieved in theory by replacing the planetary gears with rubber wheels).

 

[person123]

Why do you think that planetary gears would simply lock up? When one ring is fixed, torque is still being applied (probably via the mounting of the gear system as a whole) to keep it from rotating.

I admit that my previous comment was poorly written.
Let's say the planet gear was held stationary, while torque was being applied to the ring gear and planet carrier in opposite directions. I would imagine that with an ordinary planetary gear system, they would either be stuck in one position, or they would be forced to move in the same direction.
What I would like to achieve is allowing both parts to move in opposite direction with torque still being applied between the two- I assume, but am not sure, that that's what it means by magnetic gear systems being allowed to "slip".

(although a similar effect could be achieved in theory by replacing the planetary gears with rubber wheels).

I actually have thought of using wheels to transmit power, but I was afraid that too much energy would be lost due to friction.

 

[Jonathan Scott]

I admit that my previous comment was poorly written.
Let's say the planet gear was held stationary, while torque was being applied to the ring gear and planet carrier in opposite directions. I would imagine that with an ordinary planetary gear system, they would either be stuck in one position, or they would be forced to move in the same direction.

No, I think you've missed how this works. If the planet gear ring was held stationary, and the outer ring is rotated, then the inside ring will rotate the other way, faster. The same applies to the magnetic version. "Slipping" is when the torque on a ring is too much for the magnetic field (or friction wheel or gear) and it slips relative to the normal position.

 

[person123]

I think there's a bit of a misunderstanding. Based on the diagram below, I'm calling the yellow gear the sun gear, the red gear the planet gears, and the blue gear the ring gear. The planet gears would be connected to a single planet carrier. In this situation, I feel confident that turning the ring gear would cause the planet carrier, not planet gears, to rotate in the same direction, not opposite direction.

I realize that I said the planet gears would be held stationary, when I meant the sun gear. I think this may have been the source of the confusion.

 

[Jonathan Scott]

The original system had the outer ring fixed, in which case rotating the planetary ring carrier would cause the sun to rotate much faster in the same direction.

If the planetary carrier ring is fixed instead, then rotating the outer ring will obviously cause the sun to rotate in the opposite direction to the rotation, but not quite as fast as in the original system.

This is equivalent to taking the original system while it is in operation and rotating the whole system backwards at the original rate of the ring carrier, bringing it to a halt. This would reduce the rotation rate of the sun by the same amount, and would mean that the outer ring is then rotating in the opposite direction at the rate at which the carrier was originally rotating.

The magnetic system works the same way (assuming the original spacing scheme for the magnets and pole pieces).

 

[person123]

If the planetary carrier ring is fixed instead, then rotating the outer ring will obviously cause the sun to rotate in the opposite direction to the rotation, but not quite as fast as in the original system.

As stated previously, when I wrote that the planet gear was held stationary, that was a mistake on my part. I meant the sun gear.

 

[Jonathan Scott]

Provided only one of the three things is held still, turning one of the others should result in the other one turning. It's still all relative. If the sun is fixed, then turning either of the remaining rings will cause the other ring to rotate in the same direction but at a different speed.

 

[person123]

If the sun is fixed, then turning either of the remaining rings will cause the other ring to rotate in the same direction but at a different speed

I would imagine that with an ordinary planetary gear system, they would either be stuck in one position, or they would be forced to move in the same direction.

Would my comment before, after correcting my mistake, now be correct?

 

[Jonathan Scott]

Would my comment before now be correct?

Probably. You can work out the relative angular velocities very easily. Just choose a convenient case, for example where the carrier ring is fixed, and work out the relative angular velocities for some example (for gears you can count the teeth). You can then add or subtract an angular velocity for all three rings to make a different one fixed, and that will tell you the relative angular velocities of the other rings.

 

[person123]

What I would like to achieve is allowing both parts to move in opposite direction with torque still being applied between the two- I assume, but am not sure, that that's what it means by magnetic gear systems being allowed to "slip".

So therefore, would this be correct as well?

 

[Jonathan Scott]

Maybe, but you'd get very uneven torque - a bit like gears slipping past one another, with alternating back force then forward as it slips into the new position. Seems an odd thing to want to do.

 

[person123]

I'll give a brief explanation of the design I'm working on to explain why I want to use this.

The mechanism would be used in a bicycle to automatically store and release energy in the form of a spring. The design includes two rotating friction plates pressed against one another to transmit power. The amount of force holding the plates together increases as the rate of the bicycle wheel increases. One of these friction plates is driven by the spring. The other friction plate rotates is driven by the bicycle wheel in the opposite direction.

The friction plate connected to the spring my cause the other plate to rotate in the same direction. If it was directly connected to the bicycle wheel, the wheel would turn backwards. Therefore, I would like to include a magnetic gear set between the friction plate and the wheel. Power could be transmitted between the two, but they could rotate in the opposite direction when necessary.

This is only part of my design, but hopefully it would be possible to know if a magnetic gear set would be applicable.

 

[Jonathan Scott]

I'm not an engineer and I don't have the time to try to understand the details, but have you even looked at the figures for the amount of energy that can be stored in a spring of the size that would fit conveniently on a bicycle? I suspect it wouldn't be very useful, and if any possibility of slippage is present it would be more likely to generate heat than useful help. There are electric assist mechanisms already available, some of which can charge up while riding normally or downhill and can then provide a boost for uphill sections.

I suggest you venture into the Engineering area of PF if you want more help, but I think you probably need to do more homework and pin down more details first.