MULE1.2 Axial flux test motor/bicycle specific

We are deep into the area where I have the least to offer.
Tis a mystical & confusing land of compromise & what if senarios........
That said, I am starting to model the stator configurations. (in a triptophan induced stuper..I love thanks giving)
The visuals will help clarify some design questions & hopefully expose any glaring flaws in the thought process.
that will get us zero'd in on the the first test stator. I hope the scrap yard is open tomorrow. The junk men may be shopping on BLACK FRIDAY :shock:

It is not very scientific but I plan to wind as mutch copper into the coils as possible & see what the day brings. Time to get my CAD on.

rhitee, you post as I make spelling checks. Looks like we are on the same page.
 
Thud,

I'll look forward to some drawings. As you say, I think this is a good time to get something more concrete to discuss.
 
rhitee05 said:
You'd definitely want to space the magnets out. Those magnets are 1/8 arcs each, so if you made the rotor the same OD as the magnets you'd probably want to use 4 or 6 to leave some open space. 4 might be a bit sparse, so 6 seems like a good place to start.

Assuming the design goal is still relatively low Kv, you'd want a fairly high coil/magnet ratio, maybe 2, 3, or even 4x as many coils as magnets. So, 18 as LFP suggests might be good. At this point you've pretty much defined what the size of the coil is, so I'd put as much copper in there as you can fit. That'll give you coils that are pretty tall and narrow, which should be fairly optimal. You can connect the coils in any of several arrangements to get the desired phase resistance.
Click to expand...

Now you lost me. If we have so many times as many coils as magnets, wouldn't coils get energized when they are outside of the flux of a magnet, and be wasteful?
 
oh im completely lost here as well . BUT i am starting to think this is where the majic that the controller works comes into play since it is reading back emf (like i know what that is) . my extremely basic understanding of it is that the controller is creating a rotating magnetic field by running juice through the coils . more coils per phase means the longer it takes for the field to make a rotation.
 
John in CR said:
rhitee05 said:
You'd definitely want to space the magnets out. Those magnets are 1/8 arcs each, so if you made the rotor the same OD as the magnets you'd probably want to use 4 or 6 to leave some open space. 4 might be a bit sparse, so 6 seems like a good place to start.

Assuming the design goal is still relatively low Kv, you'd want a fairly high coil/magnet ratio, maybe 2, 3, or even 4x as many coils as magnets. So, 18 as LFP suggests might be good. At this point you've pretty much defined what the size of the coil is, so I'd put as much copper in there as you can fit. That'll give you coils that are pretty tall and narrow, which should be fairly optimal. You can connect the coils in any of several arrangements to get the desired phase resistance.
Click to expand...

Now you lost me. If we have so many times as many coils as magnets, wouldn't coils get energized when they are outside of the flux of a magnet, and be wasteful?
Click to expand...

Perfect coil over magnet alignment during attraction works like brake to hold the rotor, and perfect coil over magnet alignment during repulsion can't create any torque until it gets some offset between magnet and coil position. It's kinda like when an engine is at TDC, all the cylinder pressure in the world can't create a bit of useful torque until you get some moment arm off-set in there, and then you can generate some torque.

To generate torque, you gotta be comming towards the rotor, or leaving the rotor. At least that's how I've always looked at it, but I'm no expert. :oops:
 
liveforphysics said:
Perfect coil over magnet alignment during attraction works like brake to hold the rotor, and perfect coil over magnet alignment during repulsion can't create any torque until it gets some offset between magnet and coil position. It's kinda like when an engine is at TDC, all the cylinder pressure in the world can't create a bit of useful torque until you get some moment arm off-set in there, and then you can generate some torque.

To generate torque, you gotta be comming towards the rotor, or leaving the rotor. At least that's how I've always looked at it, but I'm no expert. :oops:
Click to expand...

LFP,
That much I do understand, and is why I want to understand exactly how the firing sequence works with our brushless motors, but I keep coming up with goose eggs on my searches. Is it correct that 2 phases fire on each pulse, in opposite polarity, or is it just one at a time? I'll look at a couple of my motors' rotor/stator combos, and figure this issue out. My knee jerk reaction is that a higher than typical coil count would increase the benefits of going to 6 phase using 2 identical controllers, but maybe I'm over complicating in my head a commutation that is actually quite simple in the motor.
John
 
John, here is a link to an automation showing the coil charging pattern & the effect in an outrunner senario.
I think you will see that brushless motors are effectivly "step" motors.
http://www.gobrushless.com/kb/index.php?title=Brushless_101_-_Chapter_3

I need some input on the potential core material & some option i have in hand.
coreoptions.jpg


On the left is a simple core made from shim stock/insulated & bonded then ground to shape. it could be a slip in fit or the coil could be wound directly on it.

on the right is a core made up from thin gage wire. again insulated & bonded to shape. same applications.

Eric,anyone, do you see any advantage of one over the other in terms of inductive performance?

there is much less continuous area in the wire senario. what I am not certain of is that to our advantage pertaining to inductance. seems a larger plate/surface would allow graeter area exposure to the flux feild & raise inductance potential.
but tiny wires if inducted wil heat that much faster?
is there a simulation to study the effect?
I know wich one will be easyer to make. :p
still need to inqure on the SMC material previous posts.
 
here is a drawing using The wedge magnets.
lukesmagnets.jpg

lmassy.jpg

you can see the motor is getting seriously large in diameter. Still reasonable in width in my opinion.
I spaced the magnet array by the thickness of the mags. 14 poles
far right is an 18tooth stator array. If my math is right I can get 77 turns of my 14g wire on each tooth. roughly 36 feet.
someone care to do the math & see what the therorical resistanc is on one coil & then a 6 coil serries.(my brain hurts)
If we go to circular coils & only use the upper 3rd the magnet we can get more turns with less wire(no end loss on the copper). Any ideas if its better to make efficent circles.edit: strike the last sentence. it doesn't work out witht the wedge mags.
but that may be in a red zone regarding inductance.

time for every one to chime in........
 
John in CR said:
Now you lost me. If we have so many times as many coils as magnets, wouldn't coils get energized when they are outside of the flux of a magnet, and be wasteful?
Click to expand...

John,

LFP is correct in his explanation, and the link posted by Thud is also informative. One other way to think of it, the rotor turns by one coil arc at a time. So, if there are more coils with smaller arcs, the motor will turn more slowly. The firing sequence would be (positive to negative) AB, CB, CA, BA, BC, AC.

Thud said:
On the left is a simple core made from shim stock/insulated & bonded then ground to shape. it could be a slip in fit or the coil could be wound directly on it.
on the right is a core made up from thin gage wire. again insulated & bonded to shape. same applications.
Click to expand...

It would be more work, but I think the 2nd version on the right would be a better choice. The laminations in the 1st version are oriented in the wrong direction and might cause some eddy losses. It might not be too bad for a lower speed motor, but the eddy losses get worse as the speed goes up. The wire construction on right would definitely have lower eddy losses, its just hard to guess the magnitude of the difference.

I take it the circular magnets you're testing with are 2" diameter?
 
Using the plate construction, you could incorporate teeth into the cores by having a step in the laminations. It wouldn't matter that they didn't continue around the ends....

This would still be possible if the lams were turned through 90 degrees.....
 
For circumferentially oriented plates:

Bond the lams with insulation and then machine to wedge shape with step to create teeth?

The ends could have 3 facets rather than being round?
 
Good idea, TD.

You could mill radial slots for the teeth and then cut down the middle of each slot to separate....

Polyimide tape layer in the coil, for insulation.....? If one could get a good enough bond.............
 
I like where we are going with all this. :)

I like the 18pole 14 magnet rotor layout, and I like the 8" rotor diameter. As long as we don't cripple the efficiency somewhere, this motor will be enough to get me to 100mph on a bicycle. :) It will be very light and easy to mount in a bicycle as well. :)

I'm liking all the creative core designs. It there something that makes laminations perform better than bundles of small iron wires?

I know there are many types of material for laminations. For our application, would we have anything noticeable to gain from using premium magnet core materials over regular lamination material?

If there is substantial gains to be had from one material over another, I'm willing to put up some cash for anything that will improve efficiency at high power levels.
 
Hello,
This is a very interesting discussion. I am planning to use sintered powdered metal parts for my windings. My interest is specifically multi 3phase designs. Hence, the increased number of windings shown in the JPGs. This is not any kind of comment on anything being disucssed here. We need to try/experiment with many different designs/concepts. I admire anyone who is taking the time to understand these motor designs. The motor drive component is even more interesting in its current implementation. I am trying to get some reasonable quotes on the sintered powered metal components and their finishing. Not too many companies are willing to spend time doing small prototype work at reasonable cost. Somewhere in someone's mind is a simple design that can be built easily with readily available or machineable parts. This discussion is the best I have seen on any forum like this. Keep on with this good work.
kenkad
 

Attachments

  • afm coilflatfoila.jpg
    afm coilflatfoila.jpg
    23.1 KB · Views: 846
Kenkad,
I had at one time a large box of salvage sintered powderd iron I intented to do the same thing with. Unfortuantly I couln't find the box anywhere at the parents house. A lot of stuff is being thrown out since dad passed a year ago.
I do think that the sintered metals may be the ticket.
I am more interested in keeping the iron to a minimum & only placing it in the interior of the coils as long as we are using cores.
A corless stator will be a cake walk once the stator molds are complete. I really need a test motor running. I am anxious to see exactly how flat wound coils are going to compair.

Rhitee, that core comfiguration is for the wedge shaped magnet rotor.
 
Thud said:
I am more interested in keeping the iron to a minimum & only placing it in the interior of the coils as long as we are using cores.
Click to expand...
Increasing the area facing the gap will make a significant difference, I think... Maybe we could model it?
 
Another possible core fabrication method. Obviously, you could do it this way with, or without, teeth.... Back-to-back teeth could be milled with a T-slot cutter, or with a normal cutter after separation..
 
Looking good Miles,
We are assembling a lot of options there.
 
How about having a bonded core lamination stack, water jet cut into individual cores? You could easily shape the ends that way... The teeth could then be cut as a second water jet operation, maybe?
 
Miles said:
How about having a bonded core lamination block, water jet cut into individual cores? You could easily shape the ends that way... The teeth could then be cut as a second water jet operation, maybe?
Click to expand...
Would the lams blow apart?
 
You must log in or register to reply here.

Similar threads

E
    • Like
Eldridge Axial Flux Motor/Generator
Replies
19
Views
4,270
Logic11
L
I
Axial flux motor not rotating but vibrating.
Replies
4
Views
1,683
imraj
I
jasonbay13
Burnt my old motor, looking for something a bit more capable, intending on building my own.
Replies
4
Views
746
jasonbay13
jasonbay13
A
The NEFASTUS Axial motor (Toy Version)
Replies
5
Views
1,170
ardillo lambo
A
E
    • Like
Homemade Axial flux PM motor
2
Replies
27
Views
12,129
Erik952
E
Back
Top