Definitive Tests on the Heating and Cooling of Hub Motors

[kk64]

That area is covered in magnets! On my typical DD motors.

Read my post again. I'm writing about geared motor, not DD. Geared motor don't have magnets on shell in this place. By the way - DD motors cooling down systems are discused many times in this thread with working/great solutions. Now please let's stick to geared and let's discuss my scenario.

(for newcomers - proposed solution details on previous site nr85)

 

[Bullfrog]

That area is covered in magnets! On my typical DD motors.

Read my post again. I'm writing about geared motor, not DD. Geared motor don't have magnets on shell in this place. By the way - DD motors cooling down systems are discused many times in this thread with working/great solutions. Now please let's stick to geared and let's discuss my scenario.

(for newcomers - proposed solution details on previous site nr85)

My MAC (geared hub motor) is covered with magnets in that area as well.

All Outrunner motors are going to have magnets where you want to drill holes but Inrunners will not.

The problem you are going to run into is the area around the motor is a dead zone as far as airflow...the tire pushes the air out of the way and it doesn't circulate very well around the motor. Liquid cooling is going to work a lot better because you need to get the heat from the windings to the shell before you can reject it to the atmosphere. I used both low viscosity ATF and water (at different times) to cool my MAC. I did mix some corrosion inhibitor with the water before adding about 5 ounces total to my MAC and it dropped the temperature of the windings about 20 degrees C.

 

[kk64]

I did mix some corrosion inhibitor with the water before adding about 5 ounces total to my MAC and it dropped the temperature of the windings about 20 degrees C.

I like distilled water with an anti-corrosion additive the most. but in this thread someone wrote that raised to a very high temperature can damage hall sensors. I don't quite understand what he meant. rather, he wrote about the temperature at which water turns into a gaseous state and steam enters the sensor.

Second problem mentioned in this thread was that distilled water will mix with grease and it will become big mess.
But the grease is not able to mix with water. maybe someone meant that the water will wash the grease from where the grease should be and it will be moved to places where it shouldn't be?

Isn't that a reason not to use distilled water? or did I misunderstand?

 

[Bullfrog]

I did mix some corrosion inhibitor with the water before adding about 5 ounces total to my MAC and it dropped the temperature of the windings about 20 degrees C.

I like distilled water with an anti-corrosion additive the most. but in this thread someone wrote that raised to a very high temperature can damage hall sensors. I don't quite understand what he meant. rather, he wrote about the temperature at which water turns into a gaseous state and steam enters the sensor.

Second problem mentioned in this thread was that distilled water will mix with grease and it will become big mess.
But the grease is not able to mix with water. maybe someone meant that the water will wash the grease from where the grease should be and it will be moved to places where it shouldn't be?

Isn't that a reason not to use distilled water? or did I misunderstand?

I can't speak for the person that posted the comment but I never had a problem and my MAC had a hall sensor. My normal running temp was about 120C and water boils at 100C. I ran it off road and constantly stopped/started...that is why my temps were so high. I used Motul MoCool for my corrosion inhibitor. And by the way, I did use distilled water.

I went to the distilled water/MoCool mixture after some ATF leaked on my rear brake disc...which made it feel like my bike sped up when I hit the brakes :lol: .

 

[ZeroEm]

You want a lithium-based grease to repel water.

 

[amberwolf]

Geared hubmotors don't have magnets between the flanges, but they do have the large ring gear (though this is usually confined to the side where the planetary gearing and clutch is located, so is parallel to just one flange).

Now, there *are* magnets on the rotor bell (for outrunners) on the motor itself, but that is not in contact with the shell where kk64 wants to drill the holes. (if it were, the gearing of the geared motor could not work as the rotor would be locked to the shell, and it would then be direct drive motor ).


To kk64: To find out if it will work, you need to build it and test it, first without any modifications, and then with your suggested modifications, with temperature monitoring instrumentation inside the motor's stator and if possible mounting temperature monitoring instrumentation on a "boom" from the axle up to as close as you can get to the outside of the rotor bell, to monitor the temperature of that as well, and another sensor on the boom to monitor the air temperature inside the shell.

Then you can compare the before and after to see what effect it really has.

Testing is what this thread is all about.

Myself, just as a first-look thought: I don't think it is going to help, and may actually make motor heating worse, because with the wheel cover addition, you'll be adding yet another layer (there are already two) of airgap and thermal transfer between the outside air passing by the wheell and the motor itself inside the hub shell and rotor bell. While the air venting of the hub shell between the flanges may allow some airflow out of the hub shell into the wheel cover area, it's still going to take the same amount of time to get the heat from the motor thru the air gap to the rotor bell *** and clutch / planetary gearset, then out of those into the air gap to the hub shell.


*** There is a solution to assist this part of the heat transfer, and that is to use statorade (discussed elsewhere in this thread) between the motor stator and the rotor bell magnets. Since this is not an encloseable area like a DD motor is, I don't know how long the statorade will stay in place, so it might have to be topped off more often than in a DD motor. But it has the potential to help with this specific part of the heat transfer.



Regarding the effects of water from the environment on the motor:

If you have the holes in the outside of the hub shell, and no wheel covers, any water that gets in can immediately drain out. While it's getting in there it will significantly help cool the motor, if that matters. As long as water does not stay in a motor, and isn't drowning the motor, it may not cause you any problems.

If you have salted roads in the winter, or other conditions that could get corrosive materials into the motor, then open holes it it are probably a bad idea. Similarly, if you have roads with magnetically-attracted debris that coudl get flung up into the holes, that's also potentially a problem, as it may get into the rotor bell and stick to the magnets causing grinding between rotor and magnets.

Other debris (grit, dust, etc) that can get into the holes and get stuck on the gear grease could cause wear problems with the gearing.


If you don't have any of those on your riding paths / roads, you could just leave the holes open, with no wheel covers, and that probably would help a little bit with cooling the motor.

I have used a couple of DD hubmotors with side covers drilled open (one of them with bigger-than-quarter-sized holes), and neither one has had a problem with water, dirt, debris, or grit despite the road conditions here--there has been dust in them, but no debris and no water corrosion. But I have not tested this on a geared hubmotor, where the duts would probably stick to the grease and cause extra wear.

Myself, if I were to need to cool a geared hubmotor, I would use one of the partially-fluid-filled methods, probably one of the oils, if one is available that is compatible with the plastic of the gearing. I started to try ATF in a small Fusin geared hubmotor, but the sealing method I tried wasn't compatible with the ATF and it rapidly leaked out, so I never really got to test it; other stuff came up so I never did get back to solving that and trying it out. Since then I've used DD hubmotors, though I'm probably about to try a GMAC (unmodified).

 

[amberwolf]

I like distilled water with an anti-corrosion additive the most. but in this thread someone wrote that raised to a very high temperature can damage hall sensors. I don't quite understand what he meant. rather, he wrote about the temperature at which water turns into a gaseous state and steam enters the sensor.

If you heated the water enough to make steam, that in itself is a high enough temperature to damage hall sensors--the typical sensors probably aren't rated for more than 70C (100C is boiling, to make steam).

So the steam itself doesn't have to damage anything--the heat generated by the stator (where the sensors are embedded) would be enough to do the job by itself.

 

[Bullfrog]

Geared hubmotors don't have magnets between the flanges, but they do have the large ring gear (though this is usually confined to the side where the planetary gearing and clutch is located, so is parallel to just one flange).

Now, there *are* magnets on the rotor bell (for outrunners) on the motor itself, but that is not in contact with the shell where kk64 wants to drill the holes. (if it were, the gearing of the geared motor could not work as the rotor would be locked to the shell, and it would then be direct drive motor ).


To kk64: To find out if it will work, you need to build it and test it, first without any modifications, and then with your suggested modifications, with temperature monitoring instrumentation inside the motor's stator and if possible mounting temperature monitoring instrumentation on a "boom" from the axle up to as close as you can get to the outside of the rotor bell, to monitor the temperature of that as well, and another sensor on the boom to monitor the air temperature inside the shell.

Then you can compare the before and after to see what effect it really has.

Testing is what this thread is all about.

Myself, just as a first-look thought: I don't think it is going to help, and may actually make motor heating worse, because with the wheel cover addition, you'll be adding yet another layer (there are already two) of airgap and thermal transfer between the outside air passing by the wheell and the motor itself inside the hub shell and rotor bell. While the air venting of the hub shell between the flanges may allow some airflow out of the hub shell into the wheel cover area, it's still going to take the same amount of time to get the heat from the motor thru the air gap to the rotor bell *** and clutch / planetary gearset, then out of those into the air gap to the hub shell.


*** There is a solution to assist this part of the heat transfer, and that is to use statorade (discussed elsewhere in this thread) between the motor stator and the rotor bell magnets. Since this is not an encloseable area like a DD motor is, I don't know how long the statorade will stay in place, so it might have to be topped off more often than in a DD motor. But it has the potential to help with this specific part of the heat transfer.



Regarding the effects of water from the environment on the motor:

If you have the holes in the outside of the hub shell, and no wheel covers, any water that gets in can immediately drain out. While it's getting in there it will significantly help cool the motor, if that matters. As long as water does not stay in a motor, and isn't drowning the motor, it may not cause you any problems.

If you have salted roads in the winter, or other conditions that could get corrosive materials into the motor, then open holes it it are probably a bad idea. Similarly, if you have roads with magnetically-attracted debris that coudl get flung up into the holes, that's also potentially a problem, as it may get into the rotor bell and stick to the magnets causing grinding between rotor and magnets.

Other debris (grit, dust, etc) that can get into the holes and get stuck on the gear grease could cause wear problems with the gearing.


If you don't have any of those on your riding paths / roads, you could just leave the holes open, with no wheel covers, and that probably would help a little bit with cooling the motor.

I have used a couple of DD hubmotors with side covers drilled open (one of them with bigger-than-quarter-sized holes), and neither one has had a problem with water, dirt, debris, or grit despite the road conditions here--there has been dust in them, but no debris and no water corrosion. But I have not tested this on a geared hubmotor, where the duts would probably stick to the grease and cause extra wear.

Myself, if I were to need to cool a geared hubmotor, I would use one of the partially-fluid-filled methods, probably one of the oils, if one is available that is compatible with the plastic of the gearing. I started to try ATF in a small Fusin geared hubmotor, but the sealing method I tried wasn't compatible with the ATF and it rapidly leaked out, so I never really got to test it; other stuff came up so I never did get back to solving that and trying it out. Since then I've used DD hubmotors, though I'm probably about to try a GMAC (unmodified).

Your explanation/description of the magnets in a geared hub motor is correct and better than mine. I have not tested it but I don't believe drilling holes in the case of a geared hub motor would help much. First problem is the bell with the magnets is really close to the shell so air would have difficulty getting in or out.

Hall sensors....Justin and I both have run our MAC/GMAC at temperatures of 150C without any problems with any part of the motors.

I don't believe statorade would work very well in a geared hub motor because it would be attracted to the magnets in the bell and would not tend to bridge the air gap like other fluids that are not magnetic. It might work but it would take more than the 4.5 ounces of water/ATF I used to bridge the air gap with my MAC.

 

[kk64]

Let's try to sum up the 86 pages of this thread, which is already 9 years old.

DD motors - statorade - proven efficiency and zero problems with electronics, leakage, resistance and temperature. The perfect solution - no need to look for another. Justin was pasting the temperature graph.

Gear motors - distilled water with anti-corrosion agent and relubrication with lithium grease that is extremely resistant to leaching by water. The solution is 100% safe for brake disc and tire leaks (no accident risk), and water is cheap and easy to refill. Hall sensors will withstand steam and will not be damaged by temperature and corrosion, and if anyone has concerns, it is worth flooding the sensors and paths with epoxy resin.

I omit engine cooling solutions with oils and ATF due to problems with sealing bearings and possible leaks on the brake disc and tire.

Am I right about the summary of solutions for gear motors and DD? (I am most interested in summarizing the solution to the problem in gear motors ).

I must have read all 85 pages of this thread. Maybe new users don't need to do this anymore, just see the summary

 

[amberwolf]

First problem is the bell with the magnets is really close to the shell so air would have difficulty getting in or out.

That's true.

For an example to show those that have not opened their motors, here's a pic of the Fusin "1000w" hub internals, where it is easy to see the side cover's threads are barely outside the outside diameter of the bell, and those threads are even with the inside diameter of the hub shell.
https://endless-sphere.com/forums/download/file.php?id=223424
https://endless-sphere.com/forums/download/file.php?id=223421

Hall sensors....Justin and I both have run our MAC/GMAC at temperatures of 150C without any problems with any part of the motors.

Interesting. I just looked up hall sensor temperatures, and it looks like there are variations capable of -40C to 150C
https://www.allegromicro.com/en/products/sense/switches-and-latches/three-wire-hall-effect-switches/aps11205
though I don't know what version the sensors in any of these motors actually are. (I just expected all the sensors would be the cheapest ones, which typically implies the lowest capability specifications).

I don't know what temperature my Fusin was at (a different one than the above pictured one), when I had hall problems with it. In my case the sensors were not "killed" as they resumed operation after it cooled down enough, but they could have been damaged (I didn't use the motor much after that, as I was working on other solutions and had other motor cores I swapped out, IIRC; long time ago so I don't recall the specific details).

I don't believe statorade would work very well in a geared hub motor because it would be attracted to the magnets in the bell and would not tend to bridge the air gap like other fluids that are not magnetic.

The bolded part is exactly the way it works in any motor. It is magnetically attracted to the magnets, and then whenever the fields are created within the stator it may be attracted to that too and bridge the gap between magnets on the rotor and laminations on the stator to conduct the heat better than the air.

The only things that are very different from how it would physically operate within a DD hubmotor vs a geared hubmotor is that the rotor in the geared hubmotor rotates several times faster than that of a DD hub for the same wheel speed, and the geared hub rotor is not "sealed" or lipped in a way that allows keeping rotational forces from flinging the statorade away from the open edge of the bell. The DD hub is (or typically can be) sealed to keep that from happening.

 

[Bullfrog]

First problem is the bell with the magnets is really close to the shell so air would have difficulty getting in or out.

That's true.

For an example to show those that have not opened their motors, here's a pic of the Fusin "1000w" hub internals, where it is easy to see the side cover's threads are barely outside the outside diameter of the bell, and those threads are even with the inside diameter of the hub shell.
https://endless-sphere.com/forums/download/file.php?id=223424
https://endless-sphere.com/forums/download/file.php?id=223421
dsc06867[1].jpg
dsc06863.jpg

Hall sensors....Justin and I both have run our MAC/GMAC at temperatures of 150C without any problems with any part of the motors.

Interesting. I just looked up hall sensor temperatures, and it looks like there are variations capable of -40C to 150C
https://www.allegromicro.com/en/products/sense/switches-and-latches/three-wire-hall-effect-switches/aps11205
though I don't know what version the sensors in any of these motors actually are. (I just expected all the sensors would be the cheapest ones, which typically implies the lowest capability specifications).

I don't know what temperature my Fusin was at (a different one than the above pictured one), when I had hall problems with it. In my case the sensors were not "killed" as they resumed operation after it cooled down enough, but they could have been damaged (I didn't use the motor much after that, as I was working on other solutions and had other motor cores I swapped out, IIRC; long time ago so I don't recall the specific details).

I don't believe statorade would work very well in a geared hub motor because it would be attracted to the magnets in the bell and would not tend to bridge the air gap like other fluids that are not magnetic.

The bolded part is exactly the way it works in any motor. It is magnetically attracted to the magnets, and then whenever the fields are created within the stator it may be attracted to that too and bridge the gap between magnets on the rotor and laminations on the stator to conduct the heat better than the air.

The only things that are very different from how it would physically operate within a DD hubmotor vs a geared hubmotor is that the rotor in the geared hubmotor rotates several times faster than that of a DD hub for the same wheel speed, and the geared hub rotor is not "sealed" or lipped in a way that allows keeping rotational forces from flinging the statorade away from the open edge of the bell. The DD hub is (or typically can be) sealed to keep that from happening.

I think it would be best to keep the temps for a geared hub like the MAC/GMAC to 130C or lower...if your bike is set up correctly and you hit over 130C, thermal runaway is a real thing and your motor can go from 130C to 150C very quickly. My point is your comments on temperatures and Hall sensors are legitimate.

 

[Bullfrog]

Let's try to sum up the 86 pages of this thread, which is already 9 years old.

DD motors - statorade - proven efficiency and zero problems with electronics, leakage, resistance and temperature. The perfect solution - no need to look for another. Justin was pasting the temperature graph.

Gear motors - distilled water with anti-corrosion agent and relubrication with lithium grease that is extremely resistant to leaching by water. The solution is 100% safe for brake disc and tire leaks (no accident risk), and water is cheap and easy to refill. Hall sensors will withstand steam and will not be damaged by temperature and corrosion, and if anyone has concerns, it is worth flooding the sensors and paths with epoxy resin.

I omit engine cooling solutions with oils and ATF due to problems with sealing bearings and possible leaks on the brake disc and tire.

Am I right about the summary of solutions for gear motors and DD? (I am most interested in summarizing the solution to the problem in gear motors ).

I must have read all 85 pages of this thread. Maybe new users don't need to do this anymore, just see the summary

I did not run any lubrication for the gears in my MAC when I ran the distilled water/Motul MoCool mixture and my gears looked great after running it about 6 months. The ATF obviously is a good lubricant as well as a good way to move heat from the windings to the shell if you can seal your motor against leaks. If you are not running a cooling fluid, I would cover the ring gear teeth with Mobil SHC 100 grease. It was recommended by the Engineer I spoke with at length several years ago when I first started running a MAC.

Just my opinion...if I was running a geared hub motor, I would run a small enough wheel and limit my hill climbing so that my motor did not get over about 130C on a regular basis. All of the cooling methods are just a band aid for a more serious problem. And I am probably one of the biggest geared hub motor (just another way to say MAC/GMAC) fans anywhere.

Just my opinion again...if I was running a DD hub motor, I'd either use Statorade or a low viscosity ATF to bridge the air gap if my motor was running too hot.

No matter what type of motor you are running, the diameter of the wheel that the motor is in determines the final gear reduction ratio i.e. a 29" wheel is geared pretty tall/high where a 20" wheel is geared much lower. The diameter of the wheel has a lot to do with how hot your motor runs...if I was having temperature/heat issues, the first thing I'd investigate it if my brakes were dragging and the second thing I'd investigate is using a smaller diameter wheel.

 

[kk64]

The only things that are very different from how it would physically operate within a DD hubmotor vs a geared hubmotor is that the rotor in the geared hubmotor rotates several times faster than that of a DD hub

Yes and no. You forget that geared hubmotor has two gaps to fill. Ferrofluid may help you to transfer heat between first gap on the photo, but we still have problem because there is second gap.

Transfering heat over first gap not solving problem, because you have still heat inside motor and the second gap is bigger and makes shell cold.

So maybe distilled water is better, because water will cover two gaps and transfer heat to shell - and transfering to shell is our point.

 

[Cowardlyduck]

Well I know it's not ideal, but I had some good success a few years back with some super thick chainsaw oil in a small geared hub.

Yes it leaked a bit, but nothing too bad. What ended up killing the motor ironically was a rubbing drum brake I didn't notice due to it's extra power. The brake was rubbing so bad it overheated the motor shell way too much and cooked the windings.

Anyway, simple is usually better and I still think oil is the way to go in geared hubbies.

Cheers

 

[amberwolf]

The only things that are very different from how it would physically operate within a DD hubmotor vs a geared hubmotor is that the rotor in the geared hubmotor rotates several times faster than that of a DD hub

Yes and no. You forget that geared hubmotor has two gaps to fill. Ferrofluid may help you to transfer heat between first gap on the photo, but we still have problem because there is second gap.

No, I didn't forget that--I posted about that specifically in my first reply to this portion of the thread above, covering the general problems of heat transfer within the motor.
https://endless-sphere.com/forums/viewtopic.php?f=2&t=48753&p=1742888#p1742781

In my reply you've quoted, I'm only discussing the physical differences in the way ferrofluid operates differently between a DD's and geared motor's rotor/stator--it doesn't cover anything beyond those since FF isn't designed to help anywhere else.

 

[kk64]

anobody found solution for gear motor cooling? something safe for leaks on disk brakes and corrosion inside motor? summer hit already and my motor today was around 96 celcius for one hour during crazy ride. 96 sounds a lot(or maybe 96 is ok?).

 

[amberwolf]

anobody found solution for gear motor cooling? something safe for leaks on disk brakes and corrosion inside motor? summer hit already and my motor today was around 96 celcius for one hour during crazy ride. 96 sounds a lot(or maybe 96 is ok?).

Before we start, which solutions have you tried so far and what were their results?

 

[kk64]

Before we start, which solutions have you tried so far and what were their results?

None of proposed here looks safe. Oil can leak on disc brake. Distilled water can mix with grease and make real mess. So I did not try anything yet. Even ferro is not 100% solution for gear motors(but it is for DC). Maybe somebody found something new here?

 

[neptronix]

Grin makes a number of geared motors with regenerative braking ability these days, so you could theoretically ditch the rear mechanical brake completely and ditch the oil on disc brake problem.

 

[Hummina Shadeeba]

None of proposed here looks safe. Oil can leak on disc brake. Distilled water can mix with grease and make real mess. So I did not try anything yet. Even ferro is not 100% solution for gear motors(but it is for DC). Maybe somebody found something new here?

Maybe the only reliable solution for getting more continuous power out from a geared hub is planning for more heat with wire and magnets that can take it and metal gears. Far from an easy solution involving rewinding the stator and building a custom rotor with likely custom made magnets.


Or maybe periodically blast compressed air in somehow?

Or use skewed rotor magnets that pump coolant in one side and out the other with rotation..and circulating back around to the inlet via a tube on the outside of the wheel.

 

[neptronix]

..now you understand why mid drives are popular..

You can bolt metal to them to increase heat shedding capacity or pump coolant in/out because the motor case isn't spinning all the time.

 

[offGridDownUnder]

anobody found solution for gear motor cooling?

I don't put my 'gears' inside the motor.

For most traditional bike designs, and with this mounting, the motor is coupled to the rim - 'hub motor'.

I use a hub motor, but mine is decoupled from the rim which permits me to introduce my own 'gears' - a reduction. I have seen other schemes to do similar. It's the third in my linked builds in my signature.

The summary is that if the rim is attached to the motor you've forced a 1:1 'gearing', and this leads to an attempt to put the gears inside the motor.

Ok with me if someone devises an arrangement that permits all inside the motor and still cools, but I don't see an obvious solution to that yet.

The constraint is the historic form of the 'safety bicycle', which does have so many other advantages.

 

[AnthonyC]

How about spiral grooves or ridges on the rotor bell? Skin effect against the hub shell drags a fluid from one side of the motor to the other where it re-circulates mainly through the windings. Axial grooves in the hub shell might increase the effect.

 

[Bullfrog]

To cool my MAC Geared Hub Motor, I ran about 5.5 ounces of distilled water with Motul MoCool added as a corrosion preventative. It worked great and did not affect my brakes if it got on my rotor. I did have to occasionally add additional water/MoCool since the water evaporates.

I sealed the side plate/motor case interface using Permatex Ultra Gray.

I added the mixture through one of the disc brake rotor bolt holes...they go all the way through to the inside of the motor on a MAC.

 

[kk64]

I'm thinking about cutting out a space in the side cover for a copper heat exchanger (copper conducts heat better than aluminum by 50%). I first need to measure how much space I will have - because the cassette is also mounted on this side. If there is not enough space, perhaps I will add a flat copper plate. or perhaps three flat plates in the shape of a "pizza slice" - only three narrow plates to maintain the stability of the cover. installation, most likely with screws + sealed with silicone resistant to high temperatures. I'm analyzing this idea - unless someone thinks it's an idiotic idea, I'm listening
edit:
or maybe placing there only flat plate of copper will be great. I mean wery thin to get very good heat transfer. on this side most of you have motor windings so place for put there thin copper exchange is good.
Better than a copper plate, it could be to add a membrane that allows air to pass through but does not allow water or dirt to pass through, but I have no idea if they exist.