APL's DIY axial-flux motor

[TorontoBuilder]

magnax.png
See how theres a jagged edge down the face without rounded edges, on traditional motors that would be hot spots on the steel.

Im aware the segments are built from laminations but they stack in a different oreantion theres a flat face of stator facing the magnets rather than edges face on.

These edges face the opposing coil im not expert but its 1 or a mixture of these 3 things help efficency, loeer high speed losses or smoother motor reduce torque ripple by sharing losses mutually coil to coil.

I see the jagged stepped stator lamination edges in the image you posted. No disagreement there, because I have a larger version of that artist's rendering that illustrates the stepped edge quite clearly.

I'm fairly sure that does not represents the current 'as built" motors, because I have the following photos of actual built stators that do not show any jaggedness to the stator edge.

another construction info pic

I'm fairly certain that the artists rendering is from a very early magnax prototype, photos of which I posted earlier in this thread. Here is a closer view showing the crude 6 step stators held between plastic end pieces.

While other images with stepped sided stators were posted from the one patent document, they are merely an artist's rendering and not how the stator is actually built. This is clearly stated within the long form text description in the patent filings, which I have excerpted here:

and let's end this post with a gratuitous animated gif...

 

[TorontoBuilder]

I might still fill in between the magnets with resin, to strengthen it further, and make it easier to clean off iron dust,
which seems to find it's way on to the magnets from everywhere.

Then a good coat of primer, to keep them from rusting,.. mount them, and I can finally move on to the stator wiring.

Hey APL, do you still have these old rotors, the ones where the magnets were glued to the back iron slightly misaligned?

IF so would you be so kind as to measure them for me? I am really interested in the bore and outside diameters, but the thickness and magnet protrusion from the inside face of the rotor would be helpful too.

I'd like to design an easy to construct 24 slot stator that would provide a .95 winding factor and not only let you reuse the rotors/magnets, but would be a very light design you could plug into another chopper project.

 

[Ianhill]

I thought it looked odd good to see im not going mad just yet then, id of thought as long as the step from lamination to lamination is less than its thickness the step up the side would look fairly uniformed to the eye but on the micro level would still reveal sharp edges.

Is it worth adding another machining step to each pole piece ? on its own i doubt it but if a pocket was made for the copper to sit in then the motor could be more compact as a result.

Really now its about gaining reliability through increasing strength of prototype 1 hopefully oneday theres enough free time to have another try, i like to sit on problems for five have a doc brown moment.

 

[APL]

TorontoBuilder, that's not shoddy workmanship,.. that's 'selective magnet skewing', an emerging technology.
(just kidding)

That motors in pieces at the moment, so yea, I can get some measurements. Still not all that bad of an idea for
DIY motors, but definitely needs some re-designs. I'll be interested in what you come up with. :thumb:

 

[APL]

I didn't want to hack HH's thread with this,.. but the discussion was splitting off into air core/iron core attributes,
and the possibility of a hybrid of sorts.

Got me thinking about a cad design I was playing with a while back and almost forgot about. I don't know what to call
it,.. linear radial? (be kind,.. no swear words :lol: ) I'm just calling it the fender motor.

Anyway, that's beside the point,.. the thing is that it solves the air cores gearing problem since the magnets are moving
at high speeds, and it solves the torque problem since it works on the wheel diameter, plus it uses much less power
relative to a hub motor, so helping solve heat issue as well.

It's a picture on the internet, so it should work, right? (I also have a magic wand if your interested) :?

Could be air or iron core, although air would have no freewheeling resistance and generally works with higher air gaps
anyway.

The magnets go inside the tire, and back irons overlap like shingles for flexibility. The tire would probably need to be
solid or a flat-less design of some sort, since it will attract every metal frag on the road.

Hey, it's just another crazy idea, and I've seen basically the same thing done in different ways. As Spock says,.. "curious".




Now beat me up Scotty...

 

[TorontoBuilder]

That motors in pieces at the moment, so yea, I can get some measurements. Still not all that bad of an idea for
DIY motors, but definitely needs some re-designs. I'll be interested in what you come up with. :thumb:

Great, I'll start working on some parametric elements and adjust once I get your measurements.

Whatever I design for you I will also build myself since wife and I spent an hour watching youtube videos of chopaderos outlaw bicycle club gatherings and other cycling parades with cruisers. She so wants one. They're not something we have in the suburbs of Toronto.

I'll be putting my motor on 1910 Harley belt drive inspired bike... because I've come to the realization I will never have the real deal

Edit: Make that a 1911 Merkel W, since I have always liked the v shaped of that frame better and I like the challenge of figuring out the spring frame and spring fork designs.

 

[APL]

TorontoBuilder, magnets are 10x40x3mm, and rotor is 122mm ID, 202mm OD. Look forward to seeing your thread.

 

[TorontoBuilder]

Thanks APL.

 

[timplett]

Kinda out there, but I could see crushing up the SMC and blending it into some PLA pellets. Getting the ratio of SMC to PLA high enough would probably be troublesome.

I read through this whole thread over the last couple days and made an account just so I could make a recommendation that MAY be useful.

Look into resin based 3d printers, there are people who have mixed large amounts of powder into resin and 3d printed parts from it. Look up Integza on YouTube, he has mixed powders with resin to make rocket grains, I believe successfully printed with up to 70% powder. From what has been said here, it may not make great quality cores as compression is recommended, but it may perform well enough for prototyping at least.

Hope this might be of use to someone.

 

[APL]

Don't know how I missed this post!

Thanks for checking out the thread timplett, and thanks for the input. Welcome to the Sphere! :thumb:

I checked out the Integza YouTube vid's,.. seems like most of them involved heat. 3D printing metals is interesting and still in it's infancy, I can only hope that it achieves usefulness for motor cores. It would make life soooo much easier for us! :wink:

 

[phowe11]

Hey APL, I sent you a PM with something you might be interested in, please check it out. Thanks!

 

[Legna199111]

I honestly loving this thread and I have one question which has been egging me what if you added v-shaped or even a hallbach array for radial flux on your existing rotors...if you put the radial magnets between the axial ones you'll should have plenty of torque while also having smoother power... I see this picture and don't get why these rotors are considered mutually exclusive

 

[fechter]

I've seen motors that have both axial and radial magnets. It can be done but adds more complexity. I can't remember what it was called.

 

[Ianhill]

Heres an interesting fact when a magnets poles are connected together with a magnetic material the newton value at the face of the magnet increases but is lower at a distance.

So a motor with the rotor magnets inside steel limits the rpm due to gathering more and more eddys with speed but at lower rpm there will be more torque avaliable for the same given input.

Thats how ipm motors have high levels of torque while having a much larger gap between the primary and secondary field of the motor plus theres a thermal path on the largest faces of the magnets helping keep short bursts of power under control more easily than having one side open to the air gap.

Theres some crazy steps forward in motor design been undertaken, i remember 10 years ago as ev started been told theres be more innovation in batterys than motors and controllers as their design has been refined more than the battery over the years but from what ive seen all aspects have moved on massively controllers commutation methods making a quiter drive and so on but that battery break through is desperately needed now we have packaged what we got as good as we can and a break through is needed.

 

[Jrbe]

I've seen motors that have both axial and radial magnets. It can be done but adds more complexity. I can't remember what it was called.

Koenigsegg has one they named "quark" and call the mixed design raxial. I have not seen pictures of the internals.
https://endless-sphere.com/forums/viewtopic.php?p=1706372

 

[huskberry]

Finally got to the last page of this thread and what a journey! I'm not saying I read it all; just perusing. Kudos!!



I'm interested in building my own axial flux motor.



TBH, I'm not sure if Emrax will come for me :lol: because it looks a lot like theirs. Whoever is interested in this tech, here is a Github link to the CAD files that I did in FreeCad while practicing to use it the program.


I'm a newbie to all of this so if you're interested in helping me make some enhancements to my design, feel free to reach out. Have an amazing day!

 

[BalorNG]

There is some discussion about coreless (ironless) PM motores here, but I didn't find answers to a few specific questions that I have.

So... ironless motors might not have 'iron', but still not exactly light and certainly not cheap, because lack of iron core must be compensated for by more far more powerful magnets and more copper and greater size.
They don't have powerful normal force between rotor and stator (due to magnet attraction to iron which is a plus, and work best with high KV because that means less copper losses and ironless PM motors do not have iron losses, only copper and friction losses... or do they?

I mean, unlike something like SynRM, an ironless motor is a pretty conventional tech and you can even 3d print it (just add copper and magnets).
Wouldn't it make it perfect application for a bicycle motor than can do assist AND regen AND zero losses (except some friction) when 'off'?

Also, should'nt their efficiency curve look very different from a typical PM motor, peaking very early and very high and than dropping exponentially due to I^2R copper losses?
(By this I mean that at high RPMs, but low very amps/torque it should have extremely high efficiency, unlike motor with an iron core where iron losses will absolutely dominate ohmic heating)

While not looks particularly tempting to 'screw fuel... er, battery juice economy' mindset crowd and certainly not partcularly good as a hub motor, it seems like a very interesting solution to intermittent assist/regen setup in a form of a middrive where you spend most of your time just pedalling, or using very low levels of assist, but still want regen.

Am I missing something? Are there more losses than just friction and ohmmic heating?

 

[Suresh N]

I don't expect much induction at that distance, but there are 52 magnets, and 300 to 600 Rpm, so I could be wrong
about all this. Recon were going to find out pretty soon!

How come its 52 Magnets, I thought it was a 21S/26P Motor, so it will have 26 magnets right

 

[Suresh N]

This would be also useful to find out saturation current.
If we short all three phase wires and measure the current between them while the motor get spun to higher and higher RPM.
Only problem is to find a gadget or drive train that is powerful and torquey enough

I'm currently testing my Axial flux motor just like you said by shorting the 3 Phases together and running it using a load, and measuring the phase current (Arms), This is my first time doing this, any tips of how to find the saturation current or if at all this graph makes sense.

I'm trying my best to learn from this thread, everything is very informative. Thanks to all for such a great package of information

 

[amberwolf]

How come its 52 Magnets, I thought it was a 21S/26P Motor, so it will have 26 magnets right

26 pole pairs is 52 magnets. Each pole pair is made of one north and one south magnet.

If the motor is actually just 26 poles, there would be only 26 magnets, but it's common to use pole pairs.