Building a triple stator axial flux motor

Lebowski

10 MW
Joined
Jun 27, 2011
Messages
3,412
Location
beautiful Zurich, Switzerland
This winter I'll try to build a triple stator axial flux motor for my 'wheelchair' (what my GF calls the recumbent :D )
It's basically v2 of the motor shown here:
http://endless-sphere.com/forums/viewtopic.php?f=30&t=30061&p=434496#p434496
This motor is on my daily commuter, has done 2000 km and, through the ice, snow
and rain, has not let me down. Power is about 1 kW, delivered straight to the rear wheel.

The plan for the wheelchair is to use the motor to drive the cranks and to have full
benefit of the bikes gears. Power will be around 1.5 to 2 kW. The motor will cool
itself when it spins, this works really well in v1 and will be improved a little bit for v2.
Hopefully it'll be able to deliver 1.5kW continuously...

It's not really difficult to build a motor like this and you don't need any special tools
or anything like that. What you need is access to the basic materials and to be able
to drill and saw accurately (like what they tought you in kindergarten).

Main tools I use for this build (just to show that you too can do this :D) :
View attachment 5
DSC01086.jpg

The raw materials:
DSC01078.jpg
DSC01077.jpg
DSC01083.jpg
DSC01082.jpg
 
The motor will have 9 stator windings and 10 magnetic poles. This is dictated by the outter dimensions
of the motor. The wheelchair will fit a 192x192mm motor, keeping this in mind 9 coils of 4 cm diameter
fit perfectly.

Magnets are 25 mm diameter, 7 mm thick: http://www.supermagnete.ch/eng/S-25-07-N

Professionally done design CAD drawing:
DSC01076.jpg

I started with the nastiest bit, the pieces that connect the rotor to the axle. First I centered
all the plates for the rotor with a 2.5 mm drill. To scratch circles in the aluminium I use:
DSC01080.jpg
A good rule to work by is to keep the 2.5mm center hole until the very end, and to cut the
outside of whatever you're making first. Then you'll always be able to use the center until
the very end. In v1 i didn't do this so, the rotor is not exactly in the middle but luckely it'll do 1000rpm
quite smoothly.
The axle with half-finished connector pieces:
DSC01079.jpg
Note how the 16 3mm holes in the connector pieces are offset over 11.25 degrees. This so the 4
6mm holes line up when the pieces are attached to the axle (small screws through the spoke holes).
The center holes were sawed with the jigsaw. Use a sharp metal blade and it'll go through 6mm
aluminium like butter.
 
Lebowski said:
It's not really difficult to build a motor like this and you don't need any special tools
or anything like that. What you need is access to the basic materials and to be able
to drill and saw accurately (like what they tought you in kindergarten).

Looking good. I often wish for better tools but find that ingenuity is the best tool of all
 
COOL.
 
Awesome!
I'll be watching this closely. I hope to build me one like that too. I even went so far as to buy some parts... a bunch of magnets (when it looked like rare earth magnets were poised to skyrocket in price). For stator, I bought some light weight (honeycomb core) fiberglass panel, about 3mm thick if I remember correctly. I got carbon fiber panel for the rotor.

I'm not sure if these composite parts are any good, I have no experience with this sort of stuff...
 
Man this is pretty awesome.. Looked back thru your other post also , but I cant find a picture of your motor attached to your bike.
You have one of those handy ? What a sweet project. Look forward to watching it evolve.
 
ohzee said:
Man this is pretty awesome.. Looked back thru your other post also , but I cant find a picture of your motor attached to your bike.
You have one of those handy ? What a sweet project. Look forward to watching it evolve.
http://endless-sphere.com/forums/viewtopic.php?f=6&t=39338&p=574726#p574726
 
Lebowski,

I have been considering in emulating your motor design and was wondering if you had considered casting your stator coils in acrylic resin?
 
I admired the lovely cnc jobs that pop up but, when I see
The products produced by members with only hand tools
I'm REALLY impressed, best of luck with the motor build
Buddy, soooooooo subscribed!!!!

KiM
 
salty9 said:
Lebowski,

I have been considering in emulating your motor design and was wondering if you had considered casting your stator coils in acrylic resin?

I didn't really simulate the motor design. Back when I started with v1 the main purpose was to have something
to develop a controller with. I did some calculations at the time which showed me it would be a good bicycle
motor (based on power and efficiency). Calculations and practise was quite a good fit so, I'm very happy with
the way v1 turned out. I've used the same equations on v2...

The very first coil plate I made, I used a special acrylic glue. This did not work at all.... Acrylic glue is solvent
based, it dissolves a bit of the acrylic and this is kind of how it works. When glueing the coils however, everything
is clamped tight to make sure the coils are in the plane of the acrylic glass. This however prevents the acrylic
glue from curing as the solvent cannot 'vent'... it needs a glue that cures based on a chemical reaction, not on
a solvent evaporating. This is how I ended up with using epoxy.
 
Lebowski, I'm going to subscribe to this thread as well. I have all the tools needed to do a motor like this. Thanks for taking the time to record the process as you make the motor.

I have 1 question. Do you think this will also make a low power generator ? I need something for my Micro Hydro system, that can be in a hostile environment, splashing water, even though I will try me best to keep water out. I can only get 100W max from the system, at low flow rates, when we are out of the rainy season, otherwise, I can get up to the 1KW with ease.

I need to do some figuring for charging 12V battery systems, so, if I can push the AC from this motor, up to the house, this will be perfect, and, I can build a spare, in case one goes out for some reason.

This is going to be a GREAT thread. Thanks again.
 
I mean, any motor can be used as a generator so yes, that's possible. What's nice about this
motor (or, v1) is that unloaded it runs as smooth and easy as a bicycle wheel. I managed
to make v1 rotate with only 0.03 W, so it should be good as a generator even if very
low power is available. There's virtually no 'binding'

About a hostile environment, I think that just depends on how good you waterproof it. v1 has
no waterproofing whatsoever (except this black shrink stuff in the wiring) and does fine
even though some of the untreared iron starts rusting now (with all of the salt on the roads
etc... remember salt on the roads from back in the day when you were not living in a sunny
tropical paradise ? :D ). I think if you spray the magnet plates with paint this should be no
problem though...
 
You have confirmed exactly what I was thinking.

Proceed with the build. 8) 8)

Thanks Lebowski 8)
 
found an interesting site for calculating the magnetic fields and such in a motor like this. With the magnetic field
the back-emf of the motor can be calculated. If you then calculate the wiring resistance a good picture of the motors
efficiency at different rpm's and power outputs can be made.

http://www.kjmagnetics.com/gap.calculator.asp

I use 25mm magnets, 7mm thick with probably a 7 mm gap. The configuration to use ín the calculator is
the 'two disk magnets with yoke.

The interesting thing is that the gap is small enough that adding iron to the motor doesn't bring that much (factor 2 only).
Two magnets with a 7 mm gap: about 8000 gauss
Same but with a 0.5 mm gap: about 12700 gauss

Since you've to integrate over the whole area and the 7mm gap has a more bowl shaped curve, lets say the difference
integrated over the magnet area is about 2.
 
Jon Stewart tells me there is a new book out - "The Dude and the Zen Master". Don't know if it has any insights into motor building but it sounds as if it might be interesting.
 
DSC01104.jpg

With the two middle magnet plates finished it's time for the two end plates. These plates carry magnets and have a thick iron backplate.
Typically said these backplates are to close the magnetic field, which is a load of BULL, magnetic field lines know how to close
themselves and don't need any help from a bit of iron. No, the iron backplates in effect doubles the internal magnetic field strength in the
motor by routing the returning field lines through a neighbour inductor.

DSC01106.jpg
the above picture shows the situation without backplates The magnetic field from the center two magnets returns in whatever way it
sees fit. The returning field lines have no usefull contribution.

DSC01107.jpg
The iron backplates (in green) act as a conductor for the magnetic field. The (effectively) shortest route for the field lines to
return is now via the backplates THROUGH THE NEIGHBOURING COILS. So in effect the returning field lines contribute
to the field in the neighbouring coils. Similarly, the field lines from both neighbour magnet pairs returns through the central coil,
doubling the magnetic field.

IM000124.jpg
In v1 the backplates were made using a 25mm by 5mm iron bar cut into wedge shaped pieces. This is however a lot of work (especially
since I had to cut the bar by hand) so for v2 I decided to go with metal sheets instead.

DSC01105.jpg
For each endplate I use 4 1.5mm sheets of galvanised steel.
 
At least some (not sure if it's your case) axial flux motors are closed only by the magnet's force. So, if we could slightly adjust the separation of rotor and stator "on the fly" by some screws or whatever, that would give us a kind of transmission box (continuous or discreet), having a torque-oriented final reduction and increasing separation to achieve higher speeds.
 
Njay said:
At least some (not sure if it's your case) axial flux motors are closed only by the magnet's force. So, if we could slightly adjust the separation of rotor and stator "on the fly" by some screws or whatever, that would give us a kind of transmission box (continuous or discreet), having a torque-oriented final reduction and increasing separation to achieve higher speeds.

Ehm, this is torque control by decreasing efficiency, not the way I would want to go :?
 
Now to make the endplates of the rotor. each plate consists of an aluminium holder with holes for the magnets, and
4 1.5mm thick metal rings as the magnetic backplate.
DSC01108.jpg
DSC01109.jpg
DSC01110.jpg
lots of sawing later....
DSC01113.jpg
DSC01114.jpg
now for the stator plates...
 
I'm excitedly following this thread Lebowski. :) I'm excited to see how it turns out, it's a clever design.
 
Back
Top