Machining/modifying motor laminations?

[DanGT86]

Trying to understand a little bit about rotor laminations and what if any effect modifications to a lamination stack would have on a motor.

If someone wanted to change the diameter of a hub motor axle could the entire stack be machined in a lathe as if it was one big chunk?

Is the shape of the lamination designed precisely enough where changes all the way in the center would have drastic effects on the field or magnetic properties?

Are the laminations varnish coated before they are assembled such that they are insulated from each other electrically or are they basically zero resistance between them?

I was thinking about this in regards to modifying a 3-5kw scooter hub into basically a giant RC outrunner. Ideally I would press the axle out and replace it with a hollow tube for bearings to run a shaft through. This would require boring out the stator lams.

Im thinking the advantages would be:

• cheap (not counting the machine work)
• sealed for water resistance and to keep in oil or statorade
• they are about 5" dia which is similar to lightning-rods and cyclone mid drive motors
• low enough KV to run single stage reduction. Tiny high RPM motors are cool but the weight of reduction pulleys and sprockets adds up to one large motor.
• Relatively easy for cheap controllers to drive.

 

[amberwolf]

Trying to understand a little bit about rotor laminations and what if any effect modifications to a lamination stack would have on a motor.

If someone wanted to change the diameter of a hub motor axle could the entire stack be machined in a lathe as if it was one big chunk?

Probably; depends on how "deep" the center section is, before it starts at the winding teeth. At the center, where it connects to the axle, it's unlikely to affect anything about the motor properties--it's really just structural at that point.

You can look at the FEMM images for various motors to see where the fields tend to be:

FEMM images for various motors - Google Search

Some of them have very little core around the axle or shaft, and some have a lot, and you can see where the fields tend to pass in those that have a lot, and leave a large volume of laminations "unused" magnetically.

Compare your motor to those images and you can guesstimate how much core you can remove.

Are the laminations varnish coated before they are assembled such that they are insulated from each other electrically or are they basically zero resistance between them?

Tehy are insulated, but in areas where there's no flux it shouldn't matter if you end up shorting them. They are likely already shorted at the axle mount from the key/slot or weld or splines or other axle/lam interface method.

I was thinking about this in regards to modifying a 3-5kw scooter hub into basically a giant RC outrunner. Ideally I would press the axle out and replace it with a hollow tube for bearings to run a shaft through. This would require boring out the stator lams.

The catch you may have with doing this, is that if you intend to run that outrunner at the speeds an outrunner would normally spin at, you may run into problems with induced eddy currents, since most of the hubmotors are meant to run at a few hundred RPM, while most of the outrunners seem intended for ten times that speed.

The hollow axle idea itself has been done before, including Farfle's double-wide MagicPie.

Not sure if any of APL's axial-flux motors have hollow axles, but I think it was explored in one of them as an idea.


Are you planning on running the wiring to the stator (that has to stay fixed in place relative to the frame/mount) thru the core of the axle? Or are you going to fit bearings inside the hollow axle for an output shaft, and run the wiring outside the hollow axle?

 

[DanGT86]

Good idea looking at the field images. I was thinking similar to what you said about the laminations being conductive to each other as soon as an axle is pressed into them but motors are a bit mysterious to me so I was not sure.

Looking at other scooter stators I think there probably isnt much extra material in the center. Most of them I found pics of online look like this.


I just picked up this motor really cheap which is what got me thinking about this in the first place.

As far as RPM and eddy currents go It is rated for 60v and I believe it is somewhere in the 30-35kv range so my 20s battery is not too far above its intended RPM.


Its 11 lbs right now. The bolt-on half of the rim is 1.5lbs. Im thinking cutting off the other rim half with spokes would probably subtract about 2lbs.

As for the wiring and axle I could leave the wires through the middle of the stationary axle but I would for sure want to cut as much length off that axle as I could. Probably have to turn new bearing journals on the axle and make new side covers. The drive side cover would need some kind of sprocket mount that could spin without touching the stationary axle.

Or an alternative if the stator allowed would be bearings in the stator and a shaft that attaches to the rear side cover. Then the wires could come out of the stationary drive side mounting surface. This would essentially be turning it into a large copy of an 80100 or similar. I guess the drawback there is that Id be adding some weird large bearings which would get heavy and expensive fast.


This is definitely some major surgery and a lot of work for a low quality motor. It just seems like these motors are about the size, power, and low kv I cant seem to find in the mid drive world.

A lightning-rods small block motor is 67kv and over 6lbs. for roughly the same general size. That would require a 2 stage reduction on my 26" wheel bike. I'm thinking the weight after pulleys, sprockets, jackshaft, bearings, and mounts would be about the same as this hub after I butcher it.

As usual I'm just thinking about ways to be super cheap while creating tons of extra work for myself on projects I'll never finish.

Also, I remember Farfle's mid drive hub. What an awesome guy and inspiration he was to so many. I think about him every time I look at the custom Genesis swing arm hanging in my shop.

 

[amberwolf]

Looking at other scooter stators I think there probably isnt much extra material in the center. Most of them I found pics of online look like this.

That one doesn't look like it has lams to the core, but rather an aluminum cast support frame that includes the axle mount, the three arms, and the ring under the (probably fr4 type stuff). The lams are probably clamped between the two halves of this frame; it's a farily comon build method.

If yours is made like this, and you need more space for the tube, beearings, and shaft than the axle support presently gives, you'd have to make a new core support frame, or modify the existing one perhaps with a block of alluminum stock machiend down to fit in there.

you'll have to check the inside of the motor you've got there t o see how it's built.

 

[scianiac]

If you wanted a cheap build I think using one of those motors will less modification could work but if you want a high performance build, high power to weight ratio I don't think this is a great play. An LR motor or similar are not that expensive and I don't think the power to weight ratio will be comparable. Statorade is great and all but it's still pretty far away from an innrunner where the coils are mounted to the case directly and part of their ability to make high power in a light package is that high RPM and the extended high torque range from it being an IPM motor further improves things. Gear reductions are always the trick though but I think maybe putting the same effort into that may be time better spent, two stages also means more gearing options and you aren't trying to run a tiny motor cog and such.

I've thought about options for cheap high power to weight ratio motors and I think something like an aggressively air cooled used A40 drone motor would be interesting. Although aggressive air cooling can dramatically increase the power output of any outrunner style motor. I picked up one of those awhile ago but haven't had a chance to play around with it yet.

 

[DanGT86]

Good points. Im kinda realizing as I am asking about it that my energy is better spent in building a quality reduction setup. If it was simply machining the rim mount off this hub and bolting a sprocket to the disk flange It may be worth some power testing. But my imagined rebuild adds just as many new bearings as a reduction.

I built this 80100 drive unit years ago. It was neat but would require a lot to be mud proof. Id like to revisit this design with a neumotor 8057 but enclose the belt reduction to keep it clean. I would also put the motor in a tube and mount a fan to the bell. Ideally it could intake and exhaust cooling air through ducts or directional vents to keep splashing water out. But again, reduction components, ducts, and enclosures add up to a bit of weight. The motor needs to be spectacular to justify that over a LR motor. Im just nostalgic for the ES of 10 years ago when the cool kids all had high strung RC motor drives.


A lightning rods motor would also fit on this bike but Its super close. I need to machine fins and some of the bolt bosses off the motor.



Another option bouncing around in my head is trying the CYC gen 4 motor. It ticks most of the boxes for me but I'm scared of it tossing magnets since gen 1-3 were kinda problematic.

All this is to say that everything is a compromise so doing massive redesign of a $100 scooter hub to net a 5"dia 9lb outrunner that any controller could drive didn't seem too crazy.

Perhaps maybe someone on the forum wants to join forces. I got the machining portion handled but dont think motor design or winding is something I'm ready for.

 

[bananu7]

This is definitely some major surgery and a lot of work for a low quality motor.

I think you hit the nail on the head with this one. If you can do machining, why not design an open hardware motor from scratch? Setting aside how many people would jump on a hub with a modern through-axle, it seems doable to beat the performance of cheap castings.

Virtually all "greats" (Alta, Stark, Electric Motion) use small diameter, carbon-sleeved motors that spin to 14k RPM with double stage gear reduction. There must be something in that.

 

[DanGT86]

For me the barriers to designing from scratch are lack of knowledge on the electrical side and cost of multiple iterations during development.

I'm lucky to have access to a machine shop for hobby stuff but absolutely cannot and ethically would not develop products to sell using resources and machines I dont own.

So I cant afford thousands of dollars in experimental laminations and expensive bearings. Since I wouldn't recover my costs.

This is why I always find myself trying to cheat by repurposing cheap crap then regretting it.

I agree that the successful formula seems to be high rpm small motors. I think packaging constraints and weight are a big factor for these companies. Also they can afford to experiment with gear reductions and materials that would cost way too much for me to try.

After years of looking around I'm convinced that the reason I cant find a 5-6lb 5kw @2000rpm waterproof inrunner is that physics doesnt allow it.

 

[scianiac]

Why not just mount the motor under the downtube with using the BB as the jackshaft like an LR drive? You have to consider where any jackshaft or motor if single reduction will be carefully to not have a crazy amount of chaingrowth. Unless you're completely reconfiguring a frame the BB is generally a pretty reasonable location to not suffer chain growth problems. Also of course using the BB means you aren't really adding much weight for the shaft, a belt first stage with the larger cog plastic and so on can keep weight down.

Honestly I think I was thinking exactly like you for a long time, I too love the high strung complex outrunner builds of old and I still think there is some advantges. But really I have yet to think of many ways to make a drive that is significatnly better than an LR kit, the constuction and reduction is pretty effiicent and combined with a powerful motor that spins fast enough to be very power dense but not fast enough to be a pain to slow down. And actually I think these motors can make more power with a bit more voltage. I'm trying to remember but when I tested my small block to see if I could gain efficiency from lowering the gear ratio I found that my already very very low gear ratio was better so that tells me the motor would spin faster without too much losses and I already run it at 72v so maybe 100v could squeeze even power power out?

 

[DanGT86]

You guys have saved me from my addiction to extra work. I'm returning that scooter hub.

As for motor mounting I really like keeping the complex bits in the triangle. I tend to bash the BB area pretty hard on obstacles. A rectangular battery box there is easier to protect/serve as a skid plate than a lumpy motor.

As for chain growth I'm really hoping to pull this off where use the rear swing arm pivot as the jackshaft and ideally only reduction before final drive. This provides all the weight distribution joy of a mid drive and lets me keep a tight chain with regen.

I am about 90% sold on a LR small block but my Nucular controller hates the old SPM version I have. It has me a bit skeptical about buying another one.

It is the most logical choice for simplicity and requires the least amount of trouble to get going. I could always change to any motor in any location in the future once the final drive is battle tested.