I have found a way to easily and succesfully cool hub motors

[dogman dan]

Still only cooling the outside of the motor, while the hot part is the winding. The windings ONLY metal to metal contact is at the axle.

But if you can get the cover to cool faster, it is in contact with the magnets. Cooler magnets near the stator could help. But if you are really pushing the motor, it's not going to save your halls. I like the idea of some heat sink right on top of the magnet ring, vs the side covers. But it's too bad the sides don't have some small ribs anyway. Every bit helps some theory.

This kind of cooling is not going to make your motor handle 10x its rated wattage, but nothing wrong with the motor running 50Fcooler or so, if you live in a climate or haul a load that's enough to make it reach temp equilibrium above 250f. Much better to keep your motor below 200, even though you can get away with 300 to some degree. If you never reach temp equilibrium on the ride before the modification then you won't save the motor with these fins. That's just pushing the motor too hard. Starting point is finding out what temp your motor is running out. If you don't know, you are still just guessing whether you even need it. No need for it if you typically reach equilibrium at 220f.

Re the water. If you live in a wet place, every single time you park your bike and the motor cools, it sucks in humid air. It's not sealed, it sucks down the wire sheath. This can collect in the motor, till it's full of water enough to freeze on a cold night locking the motor.

What I meant was,,,

Drill some holes. Cover with high temp metal tape. Then periodically, let the water out of your motor by running it with the holes uncovered. Two holes will ALSO let you cool the motor faster if you do find yourself in the situation that your stator is pushing 300F. Shit! my motor is frying! Pull the tape and let some hot air out fast. That might save your halls.

 

[MadRhino]

I don't bother anymore, not until they make 40 HP liquid cooled hub motors anyway. Doing proper cooling yourself is time consuming, and motors are cheap. Have a replacement on the shelves, feed your motor with crazy power until it fries, replace in half an hour and ride again.

 

[dogman dan]

I definitely practiced that for awhile too. But it took me longer than that to fix the motors. No big deal if you really had to ride hard to melt the motor. It's the ones that toast a motor in less than 30 min that get tiresome.

Because of the time it takes to transfer the heat from stator, to air, to cover, I kind of doubt cooling fins on the cover will help much if the motor is being run that much past it's limits.

When you run a thermometer right on your stator, it really does educate you the difference between running a smaller motor like a 9c, at 2000w or 3000w. At 3000w, it's like OMG, my stator heats up to 200F in 5 min! Still only takes about 20- 30 min at 1000w. But it takes much longer to see the heat on the cover. So cooling the cover is always going to be at least 15 min behind your stator temp.

 

[made_in_the_alps_legacy]

Still only cooling the outside of the motor, .

well... you can't really see it this way (when takling heat transfer concept, first determine which is the "heat source" and which is the "heat sink", then you can built up ideas...)
and I was just recycling "ebikedelight" idea to "pipe" the heat from Inside (heat source) to outside of the motor through a preferred known way, then you can release it (heat sink)

anyway, heat transfer seems far to complicated for most people, while some "common sens" rules for cooling, should apply :


1- by convection : increase heat transfer area, increase flow velocities, increase fluid density
2- by conduction : reduce the conductive resistance by reducing the thickness and/or get a material of better conductivity
3- by radiation : paint it black to increase emissivity

NB :
- this is without considering any potential control on the "fluids" temperatures
- ebikedelight does 1 + 2
- 1 is what you do when you drill your flanges, and/or add an impeler and/or add fans : you increase the internal velocities.
In addition : filing up a hub with liquid, is not really "cooling". it's increasing its mass to store more heat. (unless this liquid flows too a dedicated heat exchanger)

Wonder why no one talked about "peltier" cells yet

 

[macribs]

Peltier as in peltier elements?
Where would you place those without getting the wires tangled? You could place them on the stator to cool the stator but remember that as the one side of peltier gets cold, the other side heats up. So that might very well lead to even higher inside temperature.

 

[MadRhino]

Hub motor development is still primitive, frying them cheap Chinese motors is our contribution to the urge to improve :wink:

Proper cooling is by means of liquid circulation.
Ceramic stator with integrated cooling circuit, integrated centrifugal liquid pressure booster...
Make it a doughnut, composite rim ring...
Our kids will see the 100 Kw 100 Lbs TRON bike

 

[gogo]

…
In addition : filing up a hub with liquid, is not really "cooling". it's increasing its mass to store more heat. (unless this liquid flows too a dedicated heat exchanger)
…

Re: Definitive Tests on the Heating and Cooling of Hub Motor
… the purpose of transmission fluid is to get almost perfect heat transfer from the stator to the shell, not to provide more random mass to absorb heat, though as you calculated it does contribute a bit to that effect. When you do have a motor with fluid in it, then the core and the shell effectively become one single thermal mass, and since the results from our tests showed that the motor shells typically have more heat capacity than the motor core, adding the small amount of fluid will more than double the immediate heat capacity around the windings. So in a roundabout way it does increase the effective motor heat capacity much more than it's own small contribution would suggest. …

 

[made_in_the_alps_legacy]

not sure why justin le focusses his studies on "heat capacity"

I think what he is trying to say is that the "heat" is better transferred through a moving liquid than a gas :
nothing against what I said, "heat" is still "contained" ("capacity"!!) rather than "released" in those kind of experiments

(I hope he is not saying : "I put a 1000x heavier fluid in my hub than original and guess what ? it takes more time to heat it up!") <= kidding

 

[gogo]

not sure why justin le focusses his studies on "heat capacity"

I think what he is trying to say is that the "heat" is better transferred through a moving liquid than a gas :
nothing against what I said, "heat" is still "contained" ("capacity"!!) rather than "released" in those kind of experiments

(I hope he is not saying : "I put a 1000x heavier fluid in my hub than original and guess what ? it takes more time to heat it up!") <= kidding

I took it to mean that liquid makes the outer shell conduct heat from the core more effectively than gas, making the shell into more of a dedicated heat exchanger.

 

[MadRhino]

Yep, liquid puts the heat in contact with the cover, thus a 'radiator' motor cover. Yet, to optimize liquid cooling, the fluid has to be circulating in the core of the coil, and to a radiator that has proper capacity. 5 Oz of properly managed liquid coolant is better than a liter just bathing inside a hub.

 

[ebikedelight]

When I 1st got my 1000 watt hub motor, 4 years ago, I immediately took some silicone sealant and forced it into the inside of the axle where the phase/sensor wires enter. This evidently did a good job of keeping the inside of the hub motor a sealed system because when I opened my motor for the 1st time, to do this heat sink mod, there wasnt any moisture inside the motor or evidence that moisture had been in there { surface rusting , corrosion, etc} .

I realize that cooling holes strategically placed and left open, would be great in allowing air to circulate inside the hub and keep it cooler, but I just never liked the idea of a scenario where you easily invite debris and moisture inside the hub motor itself. I just dont think these motors are designed for that situation , and thats why im spending time and money on the heat sink ideas.

When I run my 1000 watt motor to the 2500 watt specs...I dont do it for 5 minutes at a time , nor will I ever be doing that. I just needed a simple, effective cooling method to deal with heat issues I experience when I run at those high wattages in short bursts...for example, we have a steep gravel road in my area and at full throttle , im pushing 2500 watts , for about 20 seconds....then im back down to the 500- 1000 watt ranges.

I believe most people that push their hub motors above factory specs, are in a same situation as me....they dont push their motors above spec for 5 minutes at a time.....my heat sink mod is a system that is trying to simply deal with heat issues for people in my situation. ...where they periodically push their motors way above specs, in short bursts , and want a simple, clean way to deal with the heat buildup , as they throttle back down in their spec ranges.

I have read about some people doing the open holes method and using urethane to coat the inside of the hubs components to deal with rust issues....some people claim it works, some people claim it didnt , and that the urethane eventually flakes off the components.....I also would imagine that coating the coils/ magnets, etc...would result in some type of effecieny loses..since you are putting a layer of something over components that werent meant to have coverings on them.

I do not believe that my heat sink mods are the best means to provide the absoloute best means of cooling, but I do believe it will prove to be a good way of cooling, that is simple, cheap and effective for most users. Theres no messy oils/ fluids involved, or having to wire up and place fans in the hub , or having wide open holes to allow debris and water inside the hub itself.

For anyone that is gonna be racing their ebikes or going up mountains and pushing their hub motors way above specs for several minutes at a time, then I doubt my mod will be adequate for them ...they will probably need something more complex to deal with their heat issues.

 

[Kyle201]

I agree with most of what you said there.

So what did you use to adhere the heatsinks to the motor casing? I'm still looking for something that will hold strong, but still not be quite 'permanent'... in case I do want to remove them later... and it also needs to have a high thermal conductivity... generally something with a high silver content I assume. Most options seem to be designed for computer units, mostly greases, or 'permanent' epoxies.

 

[dogman dan]

That was what I meant, when I oversimplified calling it just cooling the cover. My point was if you push a motor that hard, the stator will heat up very fast. I have put my hand on the still cool side cover, of a melted motor.

So convection by the air really does slow the effect of cooling the winding. But if they do, like winter weather does, give you a lower motor equilibrium temp that's good for the motor, then fine. Better to run a motor that reaches temp equilibrium at 250f, than one that reaches equilibrium at 300f. But if your temp equilibrium will be 450 or more, the only solution is to stop in time. Perhaps this fin idea gives you more time, perhaps it does not. It won't, if you are pushing the motor too hard.

WTF do I mean by pushing too hard? Not a watt number per se. It's a matter of whether you are stalling the motor, what you are asking from the motor, while it's running at full power. What I mean is, take a typical 28 mm motor. Give it say, 4000w. But if you only draw 4000w for 1% of the ride, and the rest is cruising 35 mph using 1500w, the motor won't overheat. Take the same setup to a wicked bad dirt trail, and draw 3000w for 50% of the ride, and you can melt that motor in 20-30 min in hot weather. Up those 25% grade hills, you might be running that motor as low as 50% efficiency. Takes a lot of cooling to get rid of 1500w of heating.

But a different motor, with more copper and wider magnets could do the same ride with a lot less stalling, a lot less heating.

My main point, cooling it does not make a 28mm wide motor into a powerhouse. They just aren't that strong a motor. But, it's sure fun to run them hard as hell, and see what happens.

 

[made_in_the_alps_legacy]

yes, 've been thinking about it... this is maybe where the "roads meet":

1- for rides where "power excursions" are rare, then you can rely on "heat capacity" has a "sort of cooling" : the "excessive heat" due to those "power excursion" stores itself in the hub but will (eventually) have time to release to the environement during the "less demanding" part of the ride

2- for rides where "power excursions" are not rare, then you have to "release heat" with cooling "add-on" because the "excessive heat" due to those "power excursion" would be too high to be stored and won't anyway have time to release to the environement during the "as demanding" resting part of the ride.

All in all, both directions are somehow complementary (they can be implemented together) while being somehow paradoxal (storing the heat is not "cooling")

NB: I would personnaly always prefer to release the "heat" than storing it ...not blaiming my motor but riding alpine mountains...

 

[ebikedelight]

NB: I would personnaly always prefer to release the "heat" than storing it ...not blaiming my motor but riding alpine mountains...

Using the heat sinks, is not " storing heat" ..it is a alternative method of releasing heat. Yes, having several large holes drilled in the hub side covers, would result in better releasing of heat, but as Ive explained, I did not want to use that method due to the negatives it presents.

Lets look a lil deeper into this heat sink idea.

Not only is the heat sink designed to release more heat, then just the flat side covers of the hub motor, but adding the heat sinks, adds more surface area to allow the release of heat.

example :

if we use 4 heat sinks , each 3 inchs long ,on 1 side of the hub , and place them directly above the coils, where much of the heat is generated ..and each heat sink has 3 rows of fins, that are 3/4" tall.

then each heat sink adds a additional surface area of 6.75 inchs....multiply that times 4 heat sinks, and you have added 27 " more surface area in which to release heat . If you do this mod for both sides, you have added 54 " additional inchs of surface area to cool heat { if my rough math is correct} ....and since the heat sinks are aluminum, the weight they add is negligible.

Also, since the fins are vertically elevated off the hub, the effects of " wind chill" will provide even better cooling effects then just a flat hub motor side cover......and then finally factor in that I have removed portions of the side cover, by drilling holes directly lateral of the coils....so the coil heat does not have to travel thru side cover and then thru the heat sink , to the fins....the heat has a much easier access straight to the thin base plate of the heat sinks, then to the cooling fins.


On another note...on the left brake disc side of my motor, I have enough room to have added heat sinks that have 2" of vertical rise...which would have greatly increased surface area for cooling....but the heat sinks I bought, had 3/4" or less of vertical rise on the fins .

 

[ebikedelight]

I agree with most of what you said there.

So what did you use to adhere the heatsinks to the motor casing? I'm still looking for something that will hold strong, but still not be quite 'permanent'... in case I do want to remove them later... and it also needs to have a high thermal conductivity... generally something with a high silver content I assume. Most options seem to be designed for computer units, mostly greases, or 'permanent' epoxies.

I used silver thermal heat conductive adhesive...bought it off ebay for $5 ..its a 2 part epoxy

 

[macribs]

NB: I would personnaly always prefer to release the "heat" than storing it ...not blaiming my motor but riding alpine mountains...

Using the heat sinks, is not " storing heat" ..it is a alternative method of releasing heat. Yes, having several large holes drilled in the hub side covers, would result in better releasing of heat, but as Ive explained, I did not want to use that method due to the negatives it presents.

Lets look a lil deeper into this heat sink idea.

Not only is the heat sink designed to release more heat, then just the flat side covers of the hub motor, but adding the heat sinks, adds more surface area to allow the release of heat.

example :

if we use 4 heat sinks , each 3 inchs long ,on 1 side of the hub , and place them directly above the coils, where much of the heat is generated ..and each heat sink has 3 rows of fins, that are 3/4" tall.

then each heat sink adds a additional surface area of 6.75 inchs....multiply that times 4 heat sinks, and you have added 27 " more surface area in which to release heat . If you do this mod for both sides, you have added 54 " additional inchs of surface area to cool heat { if my rough math is correct} ....and since the heat sinks are aluminum, the weight they add is negligible.

Also, since the fins are vertically elevated off the hub, the effects of " wind chill" will provide even better cooling effects then just a flat hub motor side cover......and then finally factor in that I have removed portions of the side cover, by drilling holes directly lateral of the coils....so the coil heat does not have to travel thru side cover and then thru the heat sink , to the fins....the heat has a much easier access straight to the thin base plate of the heat sinks, then to the cooling fins.


On another note...on the left brake disc side of my motor, I have enough room to have added heat sinks that have 2" of vertical rise...which would have greatly increased surface area for cooling....but the heat sinks I bought, had 3/4" or less of vertical rise on the fins .

I think your solution is an improvement. And you have also removed some of the covers for easier heat dissipate directly through the aluminum fins. As aluminum is a better "heat remover" then the steel cover that should work in your favor.

Don't know what motor you have but have you given any thoughts to putting fins on the inside of the cover as well? Probably would not work at the out most part where the coiling is but should work closer to center. If the inner fins are connected to the out fins I would think the cooling effect would be even higher. And it would still mean a full protected motor. All your holes in the covers would be plugged by heat sinks, some of them even on both sides of the cover.

Have you given any thoughts to combine your fins with the 20$ impeller? I am all for finding a good way of cooling - but it seems most methods comes with side effects, some even severe. And your way seems to have no side effects - so how can we increase the thermal removal even more so that your passive cooling would suit even high powered overvolted motors?

 

[ebikedelight]

NB: I would personnaly always prefer to release the "heat" than storing it ...not blaiming my motor but riding alpine mountains...

Using the heat sinks, is not " storing heat" ..it is a alternative method of releasing heat. Yes, having several large holes drilled in the hub side covers, would result in better releasing of heat, but as Ive explained, I did not want to use that method due to the negatives it presents.

Lets look a lil deeper into this heat sink idea.

Not only is the heat sink designed to release more heat, then just the flat side covers of the hub motor, but adding the heat sinks, adds more surface area to allow the release of heat.

example :

if we use 4 heat sinks , each 3 inchs long ,on 1 side of the hub , and place them directly above the coils, where much of the heat is generated ..and each heat sink has 3 rows of fins, that are 3/4" tall.

then each heat sink adds a additional surface area of 6.75 inchs....multiply that times 4 heat sinks, and you have added 27 " more surface area in which to release heat . If you do this mod for both sides, you have added 54 " additional inchs of surface area to cool heat { if my rough math is correct} ....and since the heat sinks are aluminum, the weight they add is negligible.

Also, since the fins are vertically elevated off the hub, the effects of " wind chill" will provide even better cooling effects then just a flat hub motor side cover......and then finally factor in that I have removed portions of the side cover, by drilling holes directly lateral of the coils....so the coil heat does not have to travel thru side cover and then thru the heat sink , to the fins....the heat has a much easier access straight to the thin base plate of the heat sinks, then to the cooling fins.


On another note...on the left brake disc side of my motor, I have enough room to have added heat sinks that have 2" of vertical rise...which would have greatly increased surface area for cooling....but the heat sinks I bought, had 3/4" or less of vertical rise on the fins .

I think your solution is an improvement. And you have also removed some of the covers for easier heat dissipate directly through the aluminum fins. As aluminum is a better "heat remover" then the steel cover that should work in your favor.

Don't know what motor you have but have you given any thoughts to putting fins on the inside of the cover as well? Probably would not work at the out most part where the coiling is but should work closer to center. If the inner fins are connected to the out fins I would think the cooling effect would be even higher. And it would still mean a full protected motor. All your holes in the covers would be plugged by heat sinks, some of them even on both sides of the cover.

Have you given any thoughts to combine your fins with the 20$ impeller? I am all for finding a good way of cooling - but it seems most methods comes with side effects, some even severe. And your way seems to have no side effects - so how can we increase the thermal removal even more so that your passive cooling would suit even high powered overvolted motors?

Yes...when I 1st came up with this idea a few weeks ago...I was thinking about doing the heat sink mod on the freewheel side of the hub motor...and then doing a impeller mod on the brake disc side....I have more room on the brake disc side, so I could go with a wider impeller unit with wider blower fins.....but if I did this, I would not want to use a plastic impeller, because it does not conduct heat away from the hub cover and since the impeller unit would need to be attached to the hub side cover , it would only make sense to me to use a aluminum impeller to get the benefits of both conductive and wind chill cooling. Using plastic impeller, doesnt make sense to me.

But once again....if just the heat sink mod works in the hottest part of the summer, then I wont even have to go any further with these mods....im not out to get perfect cooling...im just out to make sure that my motor never gets VERY HOT , under the stress/ circumstances I use it for.

I have no idea if placing fins, on the inside of the side covers, would work or be beneficial. Since my mod does not have open holes to allow for heat dissipation, I doubt inside fins would do much..other then circulate the hot air thats already trapped in the hub motor.

 

[macribs]

Yeah that might be true, could be circulating the hot air like one of them modern kitchen ovens
So I guess what is need is rapidly removing vast amount of heat before inside of hub turns into a hot air oven.

However more thermal mass should also help in temp rise situation. And as long as those inside fins are connected directly to even large outside fins - with no cover that heat need to travel through I would think that the speed of the wheel, wind and the impeller would cool the outside fins so good that the heat from inside would be "sucked out" via the fins.

If the fins have embedded heat pipes filled with cooling liquid that will circulate when heated that will remove heat in the fastest way. I mean like a closed loop we have seen on computer passive cooling kits. That require no circulation pumps and no radiator as the heat sink and fins act as the radiator.

The heat pipe would transfer hot liquid from inside the hub to the outside end of the heatpipe where pipe is connected to large surface heat sink. As liquid cools it will return inside by capillary action. Those pipes would much higher thermal conductivity the the solid heat sinks them self. Even if you made your fins from pure silver.
Cool air being funneled over them outside heat sinks/fins/heat pipes should cool instantly and that would create circulation inside the pipes and cool down both fins, cover, inside temperature and stator/magnets/copper.

Such cooling solution would also includes insides fins, heat pipes, outside fins and an impeller. It would use priciples from thermal convection, conduction, radiation and a "wind blown" outside heat exchanger that forces liquid to rapidly travel through the closed loop bringing inside heat outside.

This system would be relatively light weight using light weight high thermal capacity material. No extra complexity regarding electric pump, large radiators.

 

[ebikedelight]

On another note, I notice many people claim that painting the inside of the hub side covers, black, will result in noticeably more heat dissipation. This is another subject that im not convinced is true.

I have read some opinions about this subject , from thermal dynamic specialists, who claimed there is no real noticeable benefit to doing this. ..which makes sense to me.

the inside of the hub motor assembly, is completely dark anyways and it doesnt make sense to me that painting the inside hub covers, in a setting where its already dark, will have added benefits. Also, it can be very difficult to get spray paint to stick to aluminum surfaces, especially when those surfaces get very hot and are in constant movement. IMHO..spray painting the inside of a hub motor black, is not useful, and could even result in the paint flaking off/ chipping away , under stress , and having negative repercussions in regards to the coils, magnets, etc.

 

[pucksterpete]

Do you have any images of what you did ?

My wifes away visiting relatives and she has the digital camera. When she returns, Ill try to get some photos uploaded.

This mod really worked better then I anticipated. Its simple and effective..and takes away the hassles of messy oil cooling , installing fans and having large open holes for moisture/ dirt to enter the inside of the hub and cause problems.

Its so simple, it seems no one else has done it before .

On another note, I ran the 1000 watt hub motor I have, up to 2500 watts on several occasions, on these dirt trails, and the motor still did not overheat.

Thats how successful this mod was.

When does your wife get back?

 

[ebikedelight]

Do you have any images of what you did ?

My wifes away visiting relatives and she has the digital camera. When she returns, Ill try to get some photos uploaded.

This mod really worked better then I anticipated. Its simple and effective..and takes away the hassles of messy oil cooling , installing fans and having large open holes for moisture/ dirt to enter the inside of the hub and cause problems.

Its so simple, it seems no one else has done it before .

On another note, I ran the 1000 watt hub motor I have, up to 2500 watts on several occasions, on these dirt trails, and the motor still did not overheat.

Thats how successful this mod was.

When does your wife get back?

Hopefully sunday....

shes already got the photos on her digital camera , but she forgot to upload them to my computer before she left.

 

[dogman dan]

Be interesting to finally see the pictures. Cooling the covers (in hot weather anyway) has to be better than nothing. Always good when the motor can reach an equilibrium temperature less than 200F.

 

[ebikedelight]

I just ordered different sized heat sinks...to experiment more...

I have question .

Does anyone know for sure, if a heat sink that has a base plate that is 4.5 m.m thick will conduct heat better/ quicker to the fins of the heat sink then a heat sink that has a base plate that is 2m.m thick ?

In other words, does a thinner or thicker aluminum base plate result in quicker/ better removal of heat to the heat sink fins

 

[liveforphysics]

If both heatsinks are being bonded to something with uniform heating, and the thermal resistance of the bond is the same between the two, the thinner backed heatsink would be the highest performance, as it has less Rth induced delta-T through the aluminum itself.

Get both and try both if you care enough.