Why not direct drive?

[AndyID]

Gears multiply torque but they do that at the expense of power lost to friction and the greater the gear ratio the greater the friction. Why not dispense with gears entirely and drive locomotive axles with motors on the axles? In the past there were no motors capable of producing enough torque to do that but it might be possible now with neodymium magnets.

 

Here's a sketch of a motor. It's basically a sort of stepper motor but it's not all that different from the 3 and 5 pole motors you might be quite familiar with. Rather than having two magnetic poles this one has twenty-two. The other major difference is that the ring of magnets rotate while the windings are fixed.

 

There is no commutator as such. Speed is controlled by the frequency of the pulses supplied to the coils. That does require a little bit of electronics but nothing too exotic.

 

[Michael Hodgson]

May not have a commutator but you presumably need slip rings to get power to the coils. 

I assume it's got to be DC too

[Grovenor]

Direct drive motors have been used on locos in the past, look up early electrics in USA and Europe.

More recently there are designs similar to your suggestion used in wheel hubs in vehicles.

 

 

[billbedford]

These are called Brushless DC Motors, which you can look up. The problem with small versions of them is that they only produce useful torque at high speed. Really good for drones, etc, but not for small railway locos. 

[Skinnylinny]

2 hours ago, Michael Hodgson said:

May not have a commutator but you presumably need slip rings to get power to the coils. 

I assume it's got to be DC too

The post mentioned that the coils are stationary while the permanent magnets rotate, so no need for commutators.

[Michael Hodgson]

I've seen such motors and can see how that works with drones or computer cooling fans which are driving large diameter propellers from one end.

 

But I don't see how you feed power to stationary coils if they are completely surrounded by axle.  And in any case unless the size is absolutely tiny, I don't see how that design is going to fit onto a model railway axle - because  if it's too big to clear the track you're back to using gears.  Wouldn't it be to easier to make it fit an axle if you locate stationary coils outside the axles, with the rotating magnets inside, and you also don't have the problem of feeding the coils through the axle?

[JeremyC]

7 hours ago, AndyID said:

Gears multiply torque but they do that at the expense of power lost to friction and the greater the gear ratio the greater the friction. Why not dispense with gears entirely and drive locomotive axles with motors on the axles? In the past there were no motors capable of producing enough torque to do that but it might be possible now with neodymium magnets.

 

Here's a sketch of a motor. It's basically a sort of stepper motor but it's not all that different from the 3 and 5 pole motors you might be quite familiar with. Rather than having two magnetic poles this one has twenty-two. The other major difference is that the ring of magnets rotate while the windings are fixed.

 

There is no commutator as such. Speed is controlled by the frequency of the pulses supplied to the coils. That does require a little bit of electronics but nothing too exotic.

Is this not similar to the Maerklin C-Sinus motor.

https://second.wiki/wiki/c-sinus-motor

 

[AndyID]

7 hours ago, billbedford said:

 The problem with small versions of them is that they only produce useful torque at high speed.

 

Actually they produce maximum torque at low speed. The large number of poles on this version should (I hope ) act a bit like a reduction gear. In some respects it's more like a stepper motor.

 

Both coils will be energized simultaneously to maximize starting torque but if it can produce enough torque it might be far too noisy to be useful. The outer diameter of that version is under 15mm. It should be possible to use them on more than one driven axle.

 

But it's also quite possible it won't work for toffee

 

[AndyID]

5 hours ago, Michael Hodgson said:

I've seen such motors and can see how that works with drones or computer cooling fans which are driving large diameter propellers from one end.

 

But I don't see how you feed power to stationary coils if they are completely surrounded by axle.  And in any case unless the size is absolutely tiny, I don't see how that design is going to fit onto a model railway axle - because  if it's too big to clear the track you're back to using gears.  Wouldn't it be to easier to make it fit an axle if you locate stationary coils outside the axles, with the rotating magnets inside, and you also don't have the problem of feeding the coils through the axle?

 

The axle would not completely surround the coils. The magnets are on a drum which is attached to the axle at one end. The other end is open to accept the coils. Not a brilliant mechanical configuration but it could work. There are ways to support both ends of the drum but it gets a bit more complicated.

 

That version is less than 15mm diameter but smaller versions are possible.

 

It might be possible to flip it inside-out and make the coils external. I happen to like the idea of containing everything inside and hanging it all on the axle(s) to keep it compact but it might not work. I think the only way to find out is to try to make one and test it.

 

 

[billbedford]

Have you got one of these to work?  I'd love to see it if you have. 

[jcredfer]

Brushless motors are of two types, one has the magnets moving round the stationary windings {Can motor - Outrunner}, the other has the magnets stationary and the windings rotate {Inrunner, much faster x many times}.  BillBedford mentioned that these Brushless motors rely on speed to produce their power - and they are massively effective at doing that.  One of my smaller Cans was 6,000 revs/volt and went like a bullet.  I have an aerobat model, with an Inrunner, which is much faster / volt, which would accelerate vertical.  At slow speeds they show the steps from one set of magnets to the next, in a very jerky motion.  They also need a controller PCB, but it isn't very large, in size for what is suggested.

 

They are great, lightweight motors for high revving model aircraft, but direct geared to slow moving train wheel-sets wouldn't really be their forte.

[AndyID]

1 hour ago, billbedford said:

Have you got one of these to work?  I'd love to see it if you have. 

 

So would I

 

No, it's purely theoretical so far but I will take a shot at it. Nothing ventured.........

[AndyID]

1 hour ago, jcredfer said:

Brushless motors are of two types, one has the magnets moving round the stationary windings {Can motor - Outrunner}, the other has the magnets stationary and the windings rotate {Inrunner, much faster x many times}.  BillBedford mentioned that these Brushless motors rely on speed to produce their power - and they are massively effective at doing that.  One of my smaller Cans was 6,000 revs/volt and went like a bullet.  I have an aerobat model, with an Inrunner, which is much faster / volt, which would accelerate vertical.  At slow speeds they show the steps from one set of magnets to the next, in a very jerky motion.  They also need a controller PCB, but it isn't very large, in size for what is suggested.

 

They are great, lightweight motors for high revving model aircraft, but direct geared to slow moving train wheel-sets wouldn't really be their forte.

 

Yes, flight motors are optimized for power. I think they usually have a much smaller number of poles. This proposed configuration has 22 poles and each commutation advances by half a pole (8.2 degrees) although that might still be too much for smooth operation. It might be possible to reduce some of the jerkiness with a compliant torque connection between the stator and the loco's frames but that's assuming the thing is even capable of producing enough torque to turn the wheels, and that remains to be seen.

 

The magnets are 1/16" diameter and it's likely the number of poles could be increased with different geometry magnets. I think I better build a test rig to get an idea of how much torque to expect.  It should be possible to get a sense of that with one coil and a few magnets but that might take a while. We are stuck under a high-pressure "dome" which is making my workshop uncomfortably warm.

[jcredfer]

16 minutes ago, AndyID said:

 

Yes, flight motors are optimized for power. I think they usually have a much smaller number of poles. This proposed configuration has 22 poles and each commutation advances by half a pole (8.2 degrees) although that might still be too much for smooth operation. It might be possible to reduce some of the jerkiness with a compliant torque connection between the stator and the loco's frames but that's assuming the thing is even capable of producing enough torque to turn the wheels, and that remains to be seen.

 

The magnets are 1/16" diameter and it's likely the number of poles could be increased with different geometry magnets. I think I better build a test rig to get an idea of how much torque to expect.  It should be possible to get a sense of that with one coil and a few magnets but that might take a while. We are stuck under a high-pressure "dome" which is making my workshop uncomfortably warm.

 

With 22 x 1/16" poles that's over an inch in magnet circumference {with can diameter on top of that}, so there's not likely to be a problem with torque, given even short length magnets.  You might find that the diameter is a bit much for direct connection to wheels on diesel size wheels, whilst the greater diameter on Steam loco wheels might allow for an even greater diameter motor and more/smaller magnets for smoother slow running.  It rather depends on the balance between more smaller width magnets, with the lesser torque, against larger width ones with more torque.  A trip to a model aeroplane shop might prove useful, to look at what's available for the very small indoor model planes.  At least that would give you a head start on what might be practical -  fun too.

 

[GWR-fan]

As stated the design is a brushless outrunner.  Due the rotating mass these have more torque than an inrunner of similar size.  Inexpensive brushless controllers are available on eBay.  The idea has been suggested before but fails to gain traction.  In my experience with brushless motors in fast electric boats,  the ESC does not like to run at partial throttle and fails prematurely.  Perhaps they could find acceptance in a HST.

[AndyID]

1 hour ago, GWR-fan said:

Due the rotating mass these have more torque than an inrunner of similar size. 

 

Torque is independent of mass but an outrunner will probably have more inertia (flywheel effect) than an inrunner of similar size.

[Harlequin]

Hi Andy,

 

This is a very interesting idea.

 

Have you investigated non-cogging pancake motors? There are various different forms (no commercial items small enough for use in models as far as I know) but they have some possibly useful features you might be able to adapt. The lack of cogging would make direct drive movement smoother and the simple disc-in-a-can form might be both easier to miniaturise (if the physics works!) and a useful form factor to fit in a loco.

 

Edited July 6, 2021 by Harlequin

[RAF96]

Probably simpler to go for piezo-leg motors which come in linear and rotary forms. Again they need a controller board.