Definitive Tests on the Heating and Cooling of Hub Motors

Get all your technical information about electric bikes here.
User avatar
spinningmagnets   100 GW

100 GW
Posts: 11752
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by spinningmagnets » Jan 14 2020 12:08pm

From "bullfrog" posted with his permission:
Just a little data on my latest experiment...I tried to cool a MAC using a mixture of distilled water and Motul brand MoCool. The MoCool is a corrosion inhibitor and used in mainly motorcycle cooling systems to help prevent corrosion when ethylene glycol is not allowed...i.e. when race tracks will not allow the ethylene glycol to be used because it is slippery and a spill on the track would be extremely dangerous as well as take a long time to clean up.

Total volume of fluid used in my MAC was about five ounces...just enough to thermally connect the motor to the housing.

Well after running it for about four months, I thought my clutch was acting was not freewheeling and would go back and forth between normal operation and not freewheeling. Disassembled the motor yesterday and it turns out the failure mode was the bearings in the planet gears. The gears themselves were fine. Just FYI, the planet gears and their bearings are the highest stressed portion of a planetary gear train...learned that a long time ago while doing accident investigations on helicopters.
As far as better...the water/MoCool had about the same core temps as the ATF so from a cooling perspective they work about the same.

The ATF lubricates the planetary gears and doesn't affect the the planet gear bearings BUT it can be extremely dangerous if you get a tiny amount on your brake rotor. I was riding along after running the ATF for a few months and all of a sudden I had no rear brakes.

The water/MoCool does not cause a problem if it gets on your rotor but it washes the lubricant out of the planet gear bearings so they eventually fail.

Turns out the bearings in the planet gears are "608" bearings and the same ones used in skateboard wheels. I have been researching options this morning for ceramic bearings produced for skateboard use in an attempt to find some good ceramic bearings at a reasonable cost.

As far as the water mixture affecting other parts of the motor, the only thing I noticed was that the copper windings had a dull appearance compared to when they were new and shiny. I never had any problems with any part of the motor, hall sensors, temp sensor, or anything not working due to corrosion or issues from the water mixture. I must qualify that with I only ran it for approximately four months so I can't really comment on the "long term" effects.

I did absolutely nothing to prepare the motor before using the water mixture...Justin at Grin Tech suggested coating the electronics to prevent problems but all I did was orient the axle so that the Hall sensors were at the highest elevation possible so that when I used about five ounces of cooling fluid, they did not sit in the fluid when the bike was stationary.

IMO the best thing to do for most people is to run either a 10T or a 12T MAC depending on how much you weigh and how many hills you have as well as use a wheel/tire no larger in diameter than a 26" Maxxis Hookworm and don't more than a 14s battery. That way overheating is never an issue.

I really wanted to try my 12T MAC with a 72v battery but it will definitely require some auxiliary cooling. I know the best bet for hot rodding like I am doing is a DD motor but I really like the small size and lightweight of the MAC.

Thanks for the recommendation on the ceramic bearings.

John in CR   100 GW

100 GW
Posts: 14629
Joined: May 20 2008 12:58am
Location: Paradise

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by John in CR » Jan 18 2020 2:22pm

I'm running dual 12F Nucular controllers on a stock sealed HubMonster at 20kw peak input (pushing a 110kg bike geared to a max speed of 130kph), and a single 12F Nuc at 10kw peak on my son's MUXUS 3k speedier wind in a 23" OD 18" moto tire (geared to a 95kph top speed). Both bikes have 21s packs, and both are running markedly cooler with significantly lower wh/km than they did at lower power and lower performance than they did with common controllers. Note that like with all controllers I've used, speed settings are always set at 100% with 0 field weakening, because increases in either significantly impacts energy consumption, therefore heat. If your bike isn't fast enough, then you're running the wrong voltage or use the wrong motor.

IMHO the best and first thing anyone should do in terms of motor heat is to upgrade to a modern controller with FOC, hall sensor position correction and variable regen braking such as a Nucular controller. I assume that other FOC controllers have similar benefits, but the Nuc controllers are the only ones that I've seen other users comment about the remarkable efficiency and range increase.

Hardergamer   10 W

10 W
Posts: 96
Joined: Jan 07 2017 3:51pm

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Hardergamer » Sep 09 2020 7:38pm

Update. I have been running ATF for over 2 years in an MXUS 3K 4T turbo without any leaks. The motor has internal-external cooling fins and a 0.7mm air vent hole on the sprocket side with bigger windings and phase wires, etc. I'm running 108mm of ATF Dexron 2 at 96v 220a 500pa, 24s 27p 648x Samsung 30Q cells. Sabvoton 72200, with motor temperature rollback set at 105c. Max temperature I get to is around 92c, and after a 30-mile ride at 40-50 mph, the motor temperature is only 60c. Mods I did to stop leakages is when adding new thicker phase wires is to add silicon around them in the Axl. I added new crank seals with sealed bearings and soldered the phase wires solid in the ends inside and outside of the motor, I have the phases going into xt150's 130mm outside the motor.

Bullfrog   1 kW

1 kW
Posts: 331
Joined: Apr 08 2017 10:37am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Sep 10 2020 10:47am

I would NOT change cooling fluids if what you are running is working but if you have your motor open switching to a Low Viscosity Full Synthetic ATF like Dexron VI or the "Super Tech" brand of Low Viscosity Full Synthetic that Walmart sells will do three things for you...
A. It will go to a higher temperature without degrading...which you probably don't need :lol: .
B. The low viscosity will have less drag...which you probably won't notice in the big picture :D .
C. It will lubricate better...which doesn't matter in a hub motor since there isn't anything to lubricate :wink: .

Most electric motors can run 150C without any damage...not sure I'd want to push it that hot unless you can verify it with the manufacturer but going to 120C should be no problem whatsoever. Only reason I bring this up is you could run a little less cooling fluid which would give you less drag as well as a lower chance of leakage but just like the comments above, if what you are doing is working there is no need to change anything :thumb: .

Controllers...I agree FOC is the best. All the commutation strategies have advantages and disadvantages and the FOC is more complicated and challenging for software/hardware as far as getting it to work properly. Although trapezoidal is not very good at low speeds it is about as good at high speeds as FOC. Check out this article and you can jump to "Table 2" in the article for a summary that helped me: ... 1c74f6.pdf

Post Reply