Definitive Tests on the Heating and Cooling of Hub Motors

Hummina Shadeeba said:

Maybe people aren't thinking of their frames as heatsinks. With a connection between the motor shaft and the bike being more designed for heat transfer I bet it'd be better than air. And using air to cool almost always means wind drag

Click to expand...

@HS Steel is a poor heat conductor, so transporting large amount of heat from your hub motor via your steel axle into your steel frame will not be as effective as you might think. I would for sure put my work other places to better cool a hub. If you find yourself unsatisfied by FF and air cooling perhaps water cooling or even oil cooling is more for you?


@CD Does the varnish fill too much space in the hub, so that when added FF things are a little tight? Or you think it is the FF itself grinding as the motor turns, no matter if there is varnish or not?

Cowardlyduck - It'd be interesting to take one side cover off and see directly how the FF reacts being attracted to the magnets. For example, does it cover in an even spread or pool directly where it has been insterted, creating a fixed or movable 'puddle', in effect creating a 'wet' slopping sound and the more heat the 'sloppier' it becomes in the tight spaces. Kinda like when your in the bath and you cup and squeeze your hands to 'propel' water out, if you catch my dift .

Dude... Guys. Justin stated MEASURE YOUR CURRENT DRAW WHILE ADDING INCREMENTALLY. When you start to get an increase in current draw while running max unloaded, you've added enough FF to your motor. You want barely any contact. Adding too much doesn't improve cooling ability, and instead just adds friction to the motor. The point is to give just enough fluid to form "fingers" between the magnets that barely make contact with the stator. Typically for the most common hubmotor sizes, that's 5ml. If you added 5ml to a small outrunner? You've flooded the thing. Further, Justin said there's very little advantage to FF in small outrunners. The whole point is you use the large-diameter rotor ring from a hubmotor to act as a large heat-sink, and some of that heat also bleeds off into the side covers. Typically guys will re-seal the side covers with thermally conductive heatsink grease as well to maximize thermal transfer. Either way, it all depends on wind-speed across the motor, and if you aren't getting a lot of cooling while running at slow speed, add an external fan to blow on the outer diameter of the motor. You'll see a difference.

The reason you hear noise is that the fluid is making contact. Further, if you pot a motor to extremely high tolerances unloaded, (Justin tried this,) you'll end up fighting the flex you get from the materials the motors are made of when you put the weight of bike and rider onto it, and moving parts will start colliding and scraping. People don't realize how much "give" even the steel shafts have. I'm sure if you build a motor heavy enough with enough bearing contact, you might be able to reduce tolerances to something tiny, since there will be less flex and give in all parts, but in the "usual" motors we run, it's a very diminishing rate of returns.

The increase in noise when heat increases also makes sense; the motor heats up, metals expand, gap between rotor/stator likely decreases, creating more FF contact.

macribs said:

Hummina Shadeeba said:

Maybe people aren't thinking of their frames as heatsinks. With a connection between the motor shaft and the bike being more designed for heat transfer I bet it'd be better than air. And using air to cool almost always means wind drag

Click to expand...

@HS Steel is a poor heat conductor, so transporting large amount of heat from your hub motor via your steel axle into your steel frame will not be as effective as you might think. I would for sure put my work other places to better cool a hub. If you find yourself unsatisfied by FF and air cooling perhaps water cooling or even oil cooling is more for you?


@CD Does the varnish fill too much space in the hub, so that when added FF things are a little tight? Or you think it is the FF itself grinding as the motor turns, no matter if there is varnish or not?

Click to expand...

heat transfer of steel compared to air is a huge difference:
http://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html steel is 17- 43
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html air at sea level .024
http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html granted an airFLOW will help but looking at this it seems not much

carbon is very low at 1.7 but maybe a powder swirling around of something else would be less friction than FF and maybe more conductive. transferring the heat directly to the magnets seems not ideal as if its too hot could damage the magnets and also the heat then has to transfer through the magnets and then the glue and then the housing.

i bet some good thermal paste alone would add allow a lot of heat to get to the fork

Hummina Shadeeba said:

macribs said:

Hummina Shadeeba said:

Maybe people aren't thinking of their frames as heatsinks. With a connection between the motor shaft and the bike being more designed for heat transfer I bet it'd be better than air. And using air to cool almost always means wind drag

Click to expand...

@HS Steel is a poor heat conductor, so transporting large amount of heat from your hub motor via your steel axle into your steel frame will not be as effective as you might think. I would for sure put my work other places to better cool a hub. If you find yourself unsatisfied by FF and air cooling perhaps water cooling or even oil cooling is more for you?


@CD Does the varnish fill too much space in the hub, so that when added FF things are a little tight? Or you think it is the FF itself grinding as the motor turns, no matter if there is varnish or not?

Click to expand...

heat transfer of steel compared to air is a huge difference:
http://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html steel is 17- 43
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html air at sea level .024
http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html granted an airFLOW will help but looking at this it seems not much

carbon is very low at 1.7 but maybe a powder swirling around of something else would be less friction than FF and maybe more conductive. transferring the heat directly to the magnets seems not ideal as if its too hot could damage the magnets and also the heat then has to transfer through the magnets and then the glue and then the housing.

i bet some good thermal paste alone would add allow a lot of heat to get to the fork

Click to expand...

Measure the cross section of the frame of forks. Measure its length.

Sadly by a few inches from your motor, a steel frame will be nearly back to ambient even while the motor is smoking hot.

macribs said:

@CD Does the varnish fill too much space in the hub, so that when added FF things are a little tight?

Click to expand...

No I don't think so...it's less than 0.3mm thick.

macribs said:

Or you think it is the FF itself grinding as the motor turns, no matter if there is varnish or not?

Click to expand...

Yes, but it's not 'grinding'...just sloshing.

ccmdr said:

Cowardlyduck - It'd be interesting to take one side cover off and see directly how the FF reacts being attracted to the magnets. For example, does it cover in an even spread or pool directly where it has been insterted, creating a fixed or movable 'puddle', in effect creating a 'wet' slopping sound and the more heat the 'sloppier' it becomes in the tight spaces. Kinda like when your in the bath and you cup and squeeze your hands to 'propel' water out, if you catch my dift .

Click to expand...

It would be interesting to see it, but once a side cover is removed you loose the centering action of the side cover and the stator just magnetically stick to the magnet ring making this kind of impossible without a clear side cover as I previously suggested...I can't do it though.
I didn't inject all the FF in one place once the motor was assembled, I evenly distributed it, so as to avoid any pooling effect, so I don't think that's the cause of the sound.

Kodin, I'm not sure which of your comments were directed at me, but what you say about the lower tolerances causing the increased noise when hot does make some sense, but I do think there might be more to it than that.

As mentioned, I did add thermal paste, so the side covers should still help a bit, but as LFP points out, steel doesn't transfer much heat...so the thermal path is;
Windings-->Stator-->FF-->Magnets-->Steel Ring-->Aluminium side covers.
Granted depending on how much FF is added there is a chance the FF is in direct contact with the windings, and direct contact with the side covers. If that's the case then thermal conductivity would be far greater.

Cheers
Cheers

Over the last year I have been using a Nine Continents M3006RC direct-drive hub motor as sold by Grin Technologies on one of my bikes that I frequently ride in the hills with climbs and descents as high as 1000m vertical. Since I use a CA3 with a temperature rollback between 120 and 140C, I would never see more than 140C when driving the motor. But, since the CA3 does not regulate regen behavior, I would often see the temperature exceed 140C during regeneration, as high as 156C on one occasion.

Upon opening the motor my nose was met with the acrid smell of burnt something. But, after I pried the rotor off the stator I was pleased to see relatively shiny varnish on the wires and no evidence of severe overheating: darkened or dry, cracked varnish. In a few places the varnish was a slightly darker color, but I could not be sure it not was this way when the motor was new. The adhesive (rated up to 150C) I had used to install a 10k thermistor was intact although it had darkened slightly.

Nothing else in the motor appeared to be damaged by heat except for the non-drive side axle bearing (6904-2RS, 20x37x9 mm) that had burnt grease residue on its face exposed to the interior of the motor. The bearing turned roughly, so I pressed it out and will re-assemble everything pending the arrival of a replacement.

The larger drive-side bearing (6812-2RS, 60x78x10 mm) did not appear to be damaged at all, which is fortunate as replacing it would require the additional inconvenience of removing the freehub.

I learned that these rubber seal bearings are designed for temperatures up to 120C. While the axle bearing may not get quite as hot as the coils on the stator, the non-drive side bearing apparently got hot enough to burn off most of the lubricant and to suffer damage. It appears that (without Statoraid) the non-drive side bearing is the weakest link when this motor overheats. I will probably lower my rollback temperature range to something like 110-130C or 100-120C (at the coils) to avoid premature bearing failure.

After I get the motor reassembled I plan to run some tests to assess the qualities of running Statoraid in the motor, having already run my control trials without Statoraid.

But steel does transfer heat well compared to air. Even asbestos is better than air.
Look at this list:
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

surely there's something on here you could fill every void in the stator and this will more easily get the heat through the axle and maybe even better it could, with a very thin air gap everywhere, transfer directly to the rotor.

Here's a plastic that's light and does 10 w/mk:
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html


A tiny air gap everywhere in the motor and the heat could transfer.

The more of anything jammed in the motor would be better than a bunch of air rolling around inside isolating so much with big spaces that stop radiant heat.

Air as measured stagnant for insulating value is kinda irrelevant if you're circulating it as a working thermal transfer medium rather than dealing with it stationary.

Think of what's happening inside the motor perhaps as air contacting hot surfaces, like the winding perhaps, then rather than conducting through that air, it gets turbulently flung into contact with cooler than the windings surfaces, and they calm that thermally excited air down to a slower vibration so it can touch the windings again and get accelerated to a higher vibration again and repeat.

Even if the air magically had no thermal conductivity to itself somehow, if the particles still bounce off the windings and heat up and bounce off the aluminum sides and slow down then its working.

Beef, lean (78.9 % moisture) 0.43 - 0.48
Apple (85.6% moisture) 0.39
Butter (15% moisture content) 0.20


How hard is this:

https://en.m.wikipedia.org/wiki/Thermoelectric_cooling

Too hard. forget it. Cool though.


I haven't figured this link out either but this, which I also already posted, should give an answer to how well air dissipates heat :
http://www.engineeringtoolbox.com/convective-heat-transfer-d_430.html
But there's math here that seems to show air flow is pretty ineffective

Hi, I've been catching up on the FF. I was going to send some of my motors off to get vented this winter but now I want to know if I seal the covers on a vented hub cromotor, will the FF leak? I was looking forward to the weight savings of vented covers...

Longshot said:

Hi, I've been catching up on the FF. I was going to send some of my motors off to get vented this winter but now I want to know if I seal the covers on a vented hub cromotor, will the FF leak? I was looking forward to the weight savings of vented covers...

Click to expand...

Not so far on any of these vented motors:




I did still seal the side covers to the magnet ring to prevent capillary action taking effect.
The FF seems to stay put pretty well even with the holes right next to the windings.
I think the bigger question is what happens over time with the FF being exposed to air/ dirt/ dust, etc.

Cheers

Longshot said:

I was looking forward to the weight savings of vented covers...

Click to expand...

FWIW, unless your side covers are thick steel, venting isnt' going to save significant weight. The entire side covers on most motors don't weigh all that much, especially compared to the motor itself, and it's magnets and the steel backing those are on. I'd guess that just your rim, or your tire, either one, weighs more than the entire side cover or even pair of them, for most motors.

amberwolf said:

FWIW, unless your side covers are thick steel, venting isnt' going to save significant weight. The entire side covers on most motors don't weigh all that much, especially compared to the motor itself, and it's magnets and the steel backing those are on. I'd guess that just your rim, or your tire, either one, weighs more than the entire side cover or even pair of them, for most motors.

Click to expand...

I agree it's not much in weight savings, but does it make more of a difference because it's rotational mass though?

Cheers

Yes removing rotational mass is good, but you likely will be able to save more by going tubeless.
If weight removal is essential why not shave down the space between the flanges? Possible to remove more then a kilo of material. Solid steel.
Side covers can be made from aluminum and be light weight, meaning when you start drilling those covers weight saving will be close to nothing.

Then again, drilling side covers do have effect on the cooling, so if you would like more cooling do drill em.

tell me if im wrong but wouldnt there be good transfer of heat through radiation if the gap between the stator and rotor were small everywhere and not just the "air gap"? If every stationary part were comprised of a highly thermally conductive material it would absorb the widings' and stator's heat directly through conduction, and then if the rotor also were comprised of highly thermally conductive material and a "black body" and it were as close as possible to the stationary part of the motor it would be able to absorb heat through radiation?

http://www.auburn.edu/academic/classes/matl0501/coursepack/radiation/text.htm
Another airgap and "flux" again


Basic:
http://edinformatics.bmobilized.com/?ref=https%3A%2F%2Fwww.google.com%2F&url=http%3A%2F%2Fwww.edinformatics.com%2Fmath_science%2Fhow_is_heat_transferred.htm

A black body material is wanted

Am I on a wrong track because this seems like a great idea to me. Conduction and radiation seem much more effective than convection with air as its such an insulator

I added some Ferro Fluid to a small '250W' geared hub motor the other day.
I fully realise that this would on it's own, essentially do nothing which is why I also added oil to the hub motor shell.
Details on the bike build here:
https://endless-sphere.com/forums/viewtopic.php?f=3&t=53322&start=25#p1153542

I added 3ml of FF, and 10ml of oil. I didn't want to spend any money so just used the chain saw lube oil I had lying around. It's a bit thick, but at only 10ml and once warmed, it seems to be doing a good job.
The idea being that the FF gets heat to the magnet ring inside the hub, then the oil bath gets the heat further out to the hub shell. The FF really just means I can get away with running far less oil than would otherwise be necessary.
After about 5min thrashing it around at about 1000W the shell got to 45C in ambient temps of around 20C.


I think I should be able to reliably run around 500-750W now.

Cheers

Can someone explain the differences about "statorade" and "normal Ferrofluid?
read the last hours this topic, some wrote commercial stuff maybe burns at 90° and stuff like that.

problem: justin is out of stock. ebay is full of FF....but how can i know if its the right FF?!

i want to feed my Cromotor with that stuff

I have done a bit of research on the topic. The ferrofluid most available on ebay is Ferrotec EFH1. The carrier for this fluid is a light hydrocarbon with a flash point of 90 degrees C. Given overdriven hubmotors (which is what we are all interested in this fluid for) regularly record temperatures well above 90o I consider it risky to use EFH1.

Ferrotec produce several grades of ferrofluid for speaker damping that appear much more suitable for use in hubmotors. I suspect Justin is using these higher spec fluids for his product. I have contacted the Ferrotec distributor in Australia to see if the other grades are available in quantities of more than one or two mls. They aren't getting back to me at this stage.

Ok,
i found some Infos about that stuff.

i had a talk with someone from ferrotec (Europe)

He could tell me some things.

Like most, it depends on Temperature....

- on Room Temperature theres no aging in 20 Years
- Its "Build" for 10 Years of use.
- Its in use since years in China Elektro Motors (Yes ebikes..but its in this rikscha/Rickshaw with 3 or 4 wheels Motors)
- Engineers over the World are testing it in Windmills to cool and...
- Corrosion Protection for Magnets
- Speakers that are used on Boat/Ships have an filled coil with Ferrofluid.
- oxygen is not good and helps aging the fluid. (not as much as rain or heat will do)
- condensation contact is ok, but "water" or "rain" is not...accelerates aging.
- viscosity from APG 1110 is 100mP. Aging means more Viscosity. How much? Depends all the time in using.

High Temperatures, Rain, Oxygen helps accelerating Ferrofluid.

From what i have talked with him, i could only take a look in my glas "sphere" (haha)

since it is build for 10 Years of use in Speakers. I think 1 or 2 Years wouldnt make the Viscosity to bad.

Everyone who use it have to decide how long he will ride his Bike in different conditions and how long he will wait to take a look inside his motor whats going on there.

But he said it is no Problem to use it in elektro Motors.
If you are every day above 120°, maybe take a look inside after a few months and change it every year, if you go with the same motor longer then a year

so it sounds like a closed motor is the best use for FF.

My biggest concern with venting and FF is mixing a a bunch of water in with it. Does anyone have any info or data yet after riding in some heavy rain or slush? Anyone tried spraying their motor magnet ring with water to see how the FF reacts or holds up?

hydrocarbon is hydrophob....it doesnt react with water.

thats why you can make funny experiments with ferrofluid IN water.

dont compare it with the L5N ferrofluid from ebay. That has 10 times less viscosity. Its like Kerosin. If you drop something on your table it takes just a few hours to vaporize.
this stuff is for our "needs" a bit danger, because the "steam" of it can burn easy with a spark and after 90°c it starts "cooking"....at this state you get a lot of "steam" and if something make a spark, maybe we had our first sxploding hubbie

Merlin said:

hydrocarbon is hydrophob....it doesnt react with water.

Click to expand...

Agitate it though and it forms an emulsion. That may or may not be a problem.

Punx0r said:

Merlin said:

hydrocarbon is hydrophob....it doesnt react with water.

Click to expand...

Agitate it though and it forms an emulsion. That may or may not be a problem.

Click to expand...

Yes, that is my point. Oil will emulsify much easier given the heat and high level of agitation spinning around in the motor. I could see it washing out or just temporarily losing its properties until the water evaporated. Curious to see who if anyone has tested this yet with there vented motors. I have a vented motor I may try it on the spring, but have been reluctant to get the bike it is in wet as it is my shiniest and prettiest one! LOL

I don't have the time to read the whole thread so I'm sorry if this has been answered. You can kindly point me to where the answer is as it has not come up in searches I did. Am I correct in thinking that Statorade will not help a scooter hub motor like mine because the rotor is actually the rim and heat transferred to the ''rotor/rim'' would simply heat up the air inside the tire? And that heat would have nowhere to escape?

Thanks
Ray