Simple axial flux motor

[thepronghorn]

I am thinking about building a smallish, low power axial flux motor as an experiment and test to see if it would be feasible and reasonable to build a higher power version for usage on my bike. I want to be able to mount a sprocket to it, so it needs to have some rotating shaft or side disc where a sprocket could mount. I have some questions about the design so some help would be appreciated.

Planned details:
Coreless to keep it simple, cheap
5" diameter
2 rotors, 1 stator
Thin magnet wire to reduce eddy currents in the copper
Stator coils glued together with some sort of epoxy (What sort of glue/epoxy would be good for this application?)
1/8" thick magnets, 3/8" thick stator (What would be a good stator to magnet thickness ratio?)
1/8" thick back iron for rotors (Is this enough?)
circular magnets (Using fewer magnets means the magnets can be larger in diameter. What are the drawbacks to using 8 large magnets (1.5" diameter) as opposed to something like 16 smaller magnets?)

I hope to be able to build this at home using just a drill press and some hole saws to make all the round parts. Obviously the motor will not be perfectly balanced but by close attention to detail I hope to get pretty close. Any other comments or suggestions?

 

[Miles]

1/8" thick back iron for rotors (Is this enough?)...................................................... What are the drawbacks to using 8 large magnets (1.5" diameter) as opposed to something like 16 smaller magnets?

There is a relationship between these two factors. The greater the number of poles, the lesser the amount of back-iron needed for the same flux density. Unless you mean more than one magnet per pole?

 

[Lebowski]

http://endless-sphere.com/forums/viewtopic.php?f=30&t=30061&p=434496#p434496

I built the exact motor you're talking about, see the above thread. If you have any questions...

The motor from the above thread works very well, I use it at a street legal 500W but the
built-in cooling that you get with axial flux means it'll probably take 1kW or more.

 

[thepronghorn]

For coreless axial flux motors:

I've been looking at litz wire, and it's all twisted up. Does this affect the performance of the motor because each strand is longer and takes up more space that could be used to add more turns or more strands of wire?

Would lots of little strands that are not twisted up be better, sort of like the way most RC motors are wound?

And what do you guys think about putting a 1/32" or 1/16" sheet of garolite/fiberglass material in the middle of the windings where the flux density is weakest to help support the windings and add stiffness to the stator?

 

[thepronghorn]

1/8" thick back iron for rotors (Is this enough?)...................................................... What are the drawbacks to using 8 large magnets (1.5" diameter) as opposed to something like 16 smaller magnets?

There is a relationship between these two factors. The greater the number of poles, the lesser the amount of back-iron needed for the same flux density. Unless you mean more than one magnet per pole?

Ok so I have decided on 10 magnets for each rotor spaced around a 4.5" circle - 1" diameter, 1/8" thick, grade N42
How thick should the back iron be? Is 1/8" thick enough, or would 1/4" be more appropriate?

 

[Miles]

Somewhere around 3/16", I'd guess. If you draw it up, we can analyse it in femm.

 

[thepronghorn]

For coreless axial flux motors:

I've been looking at litz wire, and it's all twisted up. Does this affect the performance of the motor because each strand is longer and takes up more space that could be used to add more turns or more strands of wire?

Would lots of little strands that are not twisted up be better, sort of like the way most RC motors are wound?

And what do you guys think about putting a 1/32" or 1/16" sheet of garolite/fiberglass material in the middle of the windings where the flux density is weakest to help support the windings and add stiffness to the stator?

Looking at the kj magnetics simulator here here, the flux density is actually strongest in the middle, so my above idea is less than ideal. Anyone have any thoughts on stator stiffness/strength?

I'm still not sure about the first two questions above. Any ideas or help would be much appreciated. I am looking to start ordering parts within a week, so I need to commit to a design.

 

[Lebowski]

have you seen this thread ? I've started the build of a tripple stator axial flux motor

http://endless-sphere.com/forums/viewtopic.php?f=30&t=46476

 

[thepronghorn]

have you seen this thread ? I've started the build of a tripple stator axial flux motor

http://endless-sphere.com/forums/viewtopic.php?f=30&t=46476

Yes I have seen your thread. Your motor looks very nice, and it helps show me what I can do with about the same tools as you - drill press, jigsaw, hard work... I was hoping to make my motor a little smaller but spin it faster for around the same power, hence my interest in reducing eddy currents in the copper. I'm probably going to just order some smaller gauge wire (~27 gauge) and parallel several (~10) strands. If I need to go even thinner once the motor is working, then I'll probably buy some litz wire from here.

 

[bearing]

I feel that you need to put all strands in series to remove the eddy currents. If you put them i parallel, each strand will generate a different voltage, and cause circulating currents. Maybe it's not that bad... but my reasoning is this: since the flux can vary like 20% in the air gap, then the EMF between two strands will differ 20%, which will put 20% of the EMF across the small resistance of two strands in series. To me that sounds like a huge loss.

 

[thepronghorn]

Well I've been doing a lot of reading about coreless axial motors and eddy currents on the lrk-torquemax forum, especially the posts by username herbertkabi. He seems to have a done a fair bit of research/work on motors and has posted a little data set of his measurements for power loss from eddy currents in coreless motors. His data encompasses four winding configurations - thick copper strips, thick copper wire, paralleled thin copper strips, and litz wire - power loss decreases as the list progresses. I believe he does mention that the paralleled thin wires must be twisted in a configuration similar to litz, so for now, I will use litz wire.

One thing I've been trying to keep in mind is inductance which will play a large role in the controllability of my motor unlike the motors of the lrk forum which are built to spin propellers.

Current design - 1" x 1/8" disc magnets, 1/8" thick back iron, 5/16" airgap, 5" diameter motor

 

[Kingfish]

>>Controllers
That’s part of the challenge. Smaller high-spin motors are likely better suited for R/C type controllers.

~KF

 

[thepronghorn]

>>Controllers
That’s part of the challenge. Smaller high-spin motors are likely better suited for R/C type controllers.

~KF

RC controllers are definitely not the answer for low inductance motors. Still hoping that this motor won't have problems with inductance, but I'm just going to build it and see. I have spring break in a couple weeks, so I'm hoping to get some orders for materials out and start building. Here's a little emachineshop + paint drawing. Copper color is the wires, grey is the magnets. Outside diameter will be 5.25 inches. Any comments or suggestions? More copper?

 

[Teh Stork]

Are you going to run it direct drive or through an reduction gear?

I think the magnets should be spaced just a little bit, but I'm not entirely sure.

 

[Miles]

More copper?

Absolutely.....

 

[thepronghorn]

Are you going to run it direct drive or through an reduction gear?

I think the magnets should be spaced just a little bit, but I'm not entirely sure.

It's not a hub motor, so yeah I will be running it through a chain reduction to the rear wheel. No gearbox though, if that's what you're asking. I have no idea how fast this thing is going to spin, so I'll figure out gearing after I build it.

More copper?

Absolutely.....

How's this look? A lot better obviously, but maybe I shouldn't go quite so thick...

 

[Lebowski]

with the magnets so close to each other (practically touching), don't you think the magnetic
field line will jump sideways to the neighbour magnet instead of jumping through the air gap
to the opposite plate ?

 

[Thud]

definatly space your magnets about 1/2 of their thickness...

 

[Miles]

For round magnets, having them touching might be the best tactic, for a given motor diameter? If the alternative is reducing the diameter of the magnets......