Ultimate lightweight wheel-motor concept

[madin88]

Should be a major advantage in the cooling department having it open. Do not want to think about what could happen if a spoke breaks when it is like that and the motor eats bits of it: :x

if the motor is open it will not improve cooling (much). two sidediscs will better emit heat i think.
i more bother about if some dirt / little stones get between lams and magnets..

Maybe the joint won't be perfect and there will be a few hot spots but the stator will be fit into the Al ring frame and bonded in place. If you make the ring frame separate from the side panels/spokes the motor can be wound easy

yes i also think there must be a ring for lams which will be boltet to the other part of the case after wound. otherwise winding will be a torture..


edit:
hmm or maybe there is a way to leave a gap between the lams and the case sideways to keep winding easy?
a one piece case (at least on one side) would be better for centering the rotor (very consistent air gap). with a bolted case ring there could be some slackness..

 

[Punx0r]

Great stuff guys

 

[Miles]

Cassette versus Disc Mounting

Each version has things in its favour.

No reason why we can't develop designs for both. The active motor parts will be the same.......

I need to check some measurements for the disc side version:

-Distance from inside dropout face to disc mount face.

-Distance from disc mount face to to outside face of spoke flange.

-Spoke flange radius.

-Distance from hub axis to closest part of the frame on the plane of the disc mount surface.

 

[Miles]

edit:
hmm or maybe there is a way to leave a gap between the lams and the case sideways to keep winding easy?

You'd wind the core before fixing it into the case...

 

[Miles]

Have you been sizing/scaling it around a given available in N45SH magnet type my friend? I'm in China right now, and ironically just a few blocks away is a MASSIVE rare-earth magnet factory. I'm gonna try to get a tour if I can find time, and if nothing else hopefully get purchasing contacts and a list of what they offer off-the-shelf.

Thanks Luke. It would help if I had list of standard sizes. At the moment, I'm leaning towards using N40XX. I've scaled for 10mm magnet width. Length of 15mm seems about right, could be 16mm. N40 works at 4mm thick. Stronger would save a bit of weight but not much and have less height available for fixing.... Segmented circumferentially is another option. Could be 5mm x 4mm; 5mm x 5mm etc. Anything around these dimensions. We can even change the pole count to fit a stock item....

 

[speedmd]

Are we settling on a bonded magnet setup. Like the idea of stock, common size mags. Something that shares sizes with a bunch of the high performance grades for evaluating cost / performance trade offs for upgrades.

Xièxiè Luke, Bring us back some good news.

 

[Miles]

Are we settling on a bonded magnet setup.

Bonded magnets? No way! Unless you mean just a bonded fixing as opposed to a mechanical one? In which case, yes...

 

[flathill]

edit:
hmm or maybe there is a way to leave a gap between the lams and the case sideways to keep winding easy?

You'd wind the core before fixing it into the case...

so the stator core is removable if you need a rewind for whatever reason? to get good heat tranfer it has to be a super tight fit. it is like boiling a pot of water, press down hard and it boils faster

 

[Miles]

edit:
hmm or maybe there is a way to leave a gap between the lams and the case sideways to keep winding easy?

You'd wind the core before fixing it into the case...

so the stator core is removable if you need a rewind for whatever reason?

Heat the case and pull it out...? That's what you'd have to do with an Astro motor

 

[flathill]

ok cool

fine for me but if someone messes up their first wind only find out with power it is going to be a little more of a challenge with having to use a torch

I dont have hand on experience with hiperco so im not sure how brittle/fragile it really is.

it the lam stack going to be bonded or welded. just thinking about when trying to remove the core if you "catch a lip"

 

[Miles]

Here are some weights to ponder:

Copper: 695 grams
Core: 836 grams
Magnets: 306 grams
Backiron: 348 grams

Active wt. 2.2 kilograms

Total weight is going to be, at least, 2.5 kilograms. Too heavy?

What do other mid sized hub motors weigh?

 

[Miles]

fine for me but if someone messes up their first wind only find out with power it is going to be a little more of a challenge with having to use a torch
it the lam stack going to be bonded or welded. just thinking about when trying to remove the core if you "catch a lip"

The alternative is to machine the case side and rim separately. Which makes the one-sided case slightly less appealing and adds more weight.

 

[flathill]

BMC V4 rear 4.5kg

the cassette motor doesnt replace the hub like a normal hub motor so account for a few hundred grams

----
Long term the case needs to be one peice cast and machined

or two piece with case ring made from bending a strip of Al into a hoop, welded, and machined

I agree one piece machined is optimal weight wise and that is the goal so lets not worry about it for now. Just something to keep in mind if sales blowup

We really need to do a heat sim to see if the two sided case makes sense. Two structural sides.

 

[Miles]

Section-view showing crude outline of the case:

.

 

[speedmd]

Are we settling on a bonded magnet setup.

Bonded magnets? No way! Unless you mean just a bonded fixing as opposed to a mechanical one? In which case, yes...

Hi Miles

Yes sintered magnets glued down to the rotor core. Just wanted to establish thats what you were thinking and trying to get a feel on details in the core that could be a help in positioning or even possibly holding on to them better. Looks like you have drawn them in slight pocket which addresses my concerns.

2500 grams is not too heavy and a good bench mark. What do you think we may be able to loose in iron, copper and possibly even magnets before we start loosing (more than say 10%) power. When we get closer to finding out availability /dimensions of standard magnets that could work, we may have a better Idea what may be scaled back a bit at a minimum loss of power. Look forward to the simulations. Thinking this little pancake is going to have some amazing torque/power for its weight

 

[Miles]

Here are the first simulation results for the small motor. Not optimised for torque ripple etc. yet.
Simulated at 400rpm (226.7Hz) 15mm stack. Circa 2kg active motor weight. 0.55 slot fill factor.

[pre]Current Torque Eta Power out Heat out
14A 12.0Nm 94.6% 487W 28W
20A 17.0Nm 94.3% 701W 42W
30A 25.6Nm 93.0% 1058W 79W
40A 33.4Nm 91.4% 1382W 140W[/pre]

[pre].60A 49.1Nm 88.1% 2045W 275W
80A 64.2Nm 84.8% 2674W 480W
100A 78.4Nm 81.5% 3267W 744W[/pre]

 

[flathill]

Looks good. Just needs ribs to loose weight and fins to add weight.

Also think about mounting features to make the motor usable in other applications (mid drive, electric planes, robots, etc).

Make the torque arm wiring conduit a component (maybe someone will want to use a right angle adpater)

 

[Miles]

Here are the first simulation results for the small motor. Not optimised for torque ripple etc. yet.
Simulated at 400rpm (226.7Hz) 15mm stack. Circa 2kg active motor weight. 0.55 slot fill factor.

[pre]Current Torque Eta Power out Heat out
14A 12.0Nm 94.6% 487W 28W
20A 17.0Nm 94.3% 701W 42W
30A 25.6Nm 93.0% 1058W 79W
40A 33.4Nm 91.4% 1382W 140W[/pre]

[pre].60A 49.1Nm 88.1% 2045W 275W
80A 64.2Nm 84.8% 2674W 480W
100A 78.4Nm 81.5% 3267W 744W[/pre]

Kt is dropping away but not too sharply, even at 100A.... There might be some scope to increase the field strength a bit.....

 

[Miles]

I'm exploring the idea of having both sides open, with a dust/splash protection barrier at inner rim level. It just looks so much better!

Also, how cool it would be, on the Dualdrive. You'd see one "wheel" stationary and the two others going at different speeds...

 

[madin88]

You'd wind the core before fixing it into the case...

Heat the case and pull it out...? That's what you'd have to do with an Astro motor

wouldn't this small, brittle core break into pieces by doing so? i feel a bit sceptical this will work..

 

[madin88]

I'm exploring the idea of having both sides open, with a dust/splash protection barrier at inner rim level. It just looks so much better!

Also, how cool it would be, on the Dualdrive. You'd see one "wheel" stationary and the two others going at different speeds...

sounds like a good idea. by doing so we could make the side discs more lightweight (structual design like the rotor). what material you would use for the barrier?

ok, but heat dissipation will get worsening due to less area

 

[speedmd]

Having openings in the solid case will add much to cross ventilation and lighten it a bit also. Surface area losses would not be a bad thing if done well IMO. If I am not mistaken the added fresh air cross flow far outweighs the loss of conductive surface losses this area would bring in a sealed motor.

Pet duty mosquito screen glued into openings should keep the dangerous stuff out and still be able to see the action. More than 2KW of power and only 275 watts of heat is very good news. With the venting, this should easily handle much higher power levels than targeted. It looks to be worth the weight. Great work miles.

 

[speedmd]

Also, how cool it would be, on the Dualdrive. You'd see one "wheel" stationary and the two others going at different speeds...

If we mimic something like a standard mag road wheel in the stator housing and rotor support. Very cool indeed.

 

[Joe90]

First, kudo to those creatives ideas.

Maybe it has been discussed somewhere in this tread.
I like to be able to do regen. for around 10% range improvement
What do you think of mounting the motor on the brake side, no freewheel.
use a single speed freewheel on the crank side
and dish the wheel accordingly ?
This way you can make room for the motor, and have it direct drive for regen.

This motor is VERY interesting. I really like the choices made here.

 

[Miles]

Hi Joe,

This was the initial proposal by Jenny. It's a bit less straightforward but certainly doable.

I posted this but no one responded.......

Cassette versus Disc Mounting

Each version has things in its favour.

No reason why we can't develop designs for both. The active motor parts will be the same.......

I need to check some measurements for the disc side version:

-Distance from inside dropout face to disc mount face.

-Distance from disc mount face to to outside face of spoke flange.

-Spoke flange radius.

-Distance from hub axis to closest part of the frame on the plane of the disc mount surface.