Miles' 90mm inrunner build thread

[Miles]

Had an idea that may work to potentially greatly improve cooling by air crossflow. Pole count would allow you to stack the laminations and offset them relatively small amounts, one pole per lam (or in groups of two or three laminations) so the spokes of the stack would form a helix and act as a passive fan.

It's a cunning idea but I think the radial fans taking the heat directly away from the endcoils is the first thing to try

 

[speedmd]

It would be a totally enclosed passive approach. Active fans would certainly work also if added on. Certainly you would want air returns to run by the coils as on some industrial (TEFC) enclosed fan cooled units.

 

[Miles]

It would be a totally enclosed passive approach.

To circulate the air in a closed system. That makes sense. Thanks for the idea.

 

[madin88]

It would be a totally enclosed passive approach.

To circulate the air in a closed system. That makes sense. Thanks for the idea.

yes, and it keeps dust and dirt out of the motor.

 

[Miles]

I'll be experimenting with vented and enclosed versions. I haven't give that much thought to the enclosed version, yet. This is the initial plan for the vented version.

 

[dotrick]

A message to have some updates of this post... I would like to do a mid-drive motor, an outrunner.

 

[Miles]

I would like to do a mid-drive motor, an outrunner.

I'd suggest you start playing with Emetor It now has a template for using rectangular magnets in outrunner configuration.

 

[madin88]

i think dotrick more wants to buy a motor and not build it completely by his own

very nice motor with very good specs. can't wait to see this thing with gear in real installed on an ebike.

 

[dotrick]

No no... I want to build my motor and my frame... I have a CNC, I want to use it!

Emetor, I know it, i already use it... only the free version. I don't know if it's a good tool. I learnt a lot with Emetor. I want to validate my design with Femm before to start.

I want a strong motor... I didn't find a mid-drive as audi ebike... I want to do mine!

 

[Miles]

Emetor, I know it, i already use it... only the free version. I don't know if it's a good tool. I learnt a lot with Emetor. I want to validate my design with Femm before to start.

Hi dotrick,

Why not do a few non-linear simulations using Emetor?

Why not start a thread on your motor, so we can make suggestions?

 

[Lebowski]

Oh man I never knew building a motor was so difficult

 

[Miles]

 

[Miles]

Comparison with higher speed operation and reduction of lamination thickness

[pre]30Amps Power out Eta Copper Iron Heat out

3Krpm-Sura18 1112W 95.35% 38.40W 15.79W 54W
6Krpm-Sura18 2214W 96.55% 38.40W 40.68W 79W
6Krpm-Sura12 2219W 96.75% 38.49W 36.22W 74W[/pre]

 

[Miles]

That's it folks. No more simulations for this build...........

 

[speedmd]

Nice work Miles. 96.75% is amazing. Did not think it be possible. Will not even get warm on days like today. No need for fans if you run at reasonable power levels.

 

[Miles]

Thanks speedmd. Let's see what's possible for the real motor....... 0.6 fill factor is quite a challenge.

 

[speedmd]

Good point Miles. If I remember correctly, the theoretical maximum fill for round shapes is in the low 70% range. Now if you go with rectangular, square or hex shape wire..... 8)

 

[emetor]

I did a short investigation about the maximum possible slot fill factor using round conductors a time ago. It can be found here:
https://www.emetor.com/blog/post/how-big-maximum-possible-slot-fill-factor-using-round-conductors/

 

[speedmd]

Hi emetor

Chart looks reasonably /correct for thinner wire up to around 1mm or so. Looks to be going off the tracks a bit with the fatter wire sizes. Not seeing that above 80 is even possible with my math/assumptions. What are the two types of insulation and its average thickness. Also, if twists (of square or rectangular wire) are done outside the slot, they should not reduce the fill factor. Interested in how much the wire shape / fill can help here. May be significant.

 

[Arlo1]

Guys. I know I have to read a few pages but has anyone though about wrapping the magnets with Kevlar and resin on the other axis? I ask this because there would be less tensile strength required on the Kevlar being as it would need to keep fewer magnets on at a time.

 

[Miles]

It might be a bit tedious to do? It wouldn't be so effective if the magnets were segmented axially?

I'm rather hoping that I might be able to get away without even bonding the magnets....

 

[flathill]

I think the next step now is to get a quote on the slotted magnets. Remember you dont pay strictly by volume, more based on the fall out from cracking. Certain geometries are very prone to cracking. You pay for the scrap one way or another

 

[Miles]

I think the next step now is to get a quote on the slotted magnets.

Yes, that's definitely the top priority.

 

[emetor]

Chart looks reasonably /correct for thinner wire up to around 1mm or so. Looks to be going off the tracks a bit with the fatter wire sizes. Not seeing that above 80 is even possible with my math/assumptions. What are the two types of insulation and its average thickness.

The chart assumes that the conductors are placed in perfect order, that the slot is infinitely large, and that there is no slot insulation. All rather difficult to achieve in reality :lol:
According to the manufacturer, the insulation is a THEIC-modified polyesterimid with an outer layer of polyamidimid. The average thickness depends on the conductor diameter, eg. for 1 mm nominal conductor diameter, the Grade 1 insulation thickness is 0.034 mm, whereas it is 0.063 mm for Grade 2. The chart is a simple visualitation of the number of conductors per square cm provided by the manufactuerer for different conductor diameters.

 

[Miles]

Useful paper on "modular" type designs (like mine).