Definitive Tests on the Heating and Cooling of Hub Motors

[Kodin]

As I haven't seen this question addressed yet, is there any disadvantage to trying a higher viscosity fluid since the contact area is very small and we're only adding until we reach a slight increase in load? Would a thicker fluid hold up better long-term?

 

[macribs]

For my fill hole I drilled a hole about 2" from the perimeter and then plugged it with a tiny sintered metal "breather" vent from McMaster (threaded into hole) and still had no FF leakage out the vent. The sintered metal vent helps equalize pressure and prevents water ingress near the axle which is the root cause of corrosion

¨


Hm my dictionary falls short here, what is "sinter metal breather"? Do you have pics of your motor with this breather?
If this really equalizes pressure that should help maybe even for those who run ATF oil, as well as for [strike]ferrofluid[/strike] Ferrochill?
(maybe ATF will not be be forced out of axle/bearing due to high pressure?)

 

[fechter]

There are all kinds of those. Frequently used on fuel systems. Generally, they look like this:

 

[Offroader]

I just wanted to let people know of the cooling method I built for my cromotor. The FF looks promising and I look forward to more testing. The only thing I want to see is how well it performs with low speed offroad riding, basically any really hard trail riding where you are riding very slowly and need to climb steep hills.

I have put together fan based air cooling in my cromotor using an air duct to circulate fresh air in and hot air out. I believe the idea was madin88 and I built the idea into my cromotor. It sucks air through the center holes and pushes it through the windings on one side, then through the magnet/stator gap, then it pushes it through the windings again on the other side then out of the motor only through the outer most holes. Holes have to be drilled only on one side for this to work. The fan used to push the air is a 30mm edf a very high powerful fan that is much powerful than muffin fans, but runs at like 60-80 watts at high speed.

It wasn't the easiest thing to get it built, the hard part was getting the airduct to seal against the side cover of the motor so that the air doesn't recirculate in the motor. If any air recirculated inside the motor the results were very bad as it wouldn't cool effectively. Even if the air duct that sealed the motor inside to stop air recirculating had like a small 2mm gap that was enough to recirculate almost all the air and the cooling wasn't good.

It took me to mill a custom air duct that needed to be tapped / screwed down to the stator. To seal the air duct against the side cover I used silicone on top of the air duct and pressed on the motor covers to flatten the silicone and molding a tight seal. Remember the air duct has to seal against the spinning side covers of the motor. I used a piece of tape to make a small space (around .2 mm) between the silicone and side cover to give a little bit of space if there were any tolerance issues in the case or else the silicone would rip off.

For the fan I used a high speed 30mm EDF fan that spins over 50K RPM and blows a lot of air. These are like tiny jet engines, holding it to your face feels like you stuck your head out the window at over 50MPH.

How well does it work? It works amazingly well to cool the motor. The air coming out of the exhaust holes when the motor reaches over 70c is so hot that holding my hand close to the output vent holes will start burning my hand in about 5 seconds. It is throwing out a tremendous amount of heat. If I put my face to the side of the motor, but far away (50 cm), I can feel the full breeze of heat coming out of the motor.

I went from not being able to use my bike in the hot summer days because of overheating problems to being able to drive as hard as I want when running the EDF at full speed. When climbing steep hills where the motor is almost stalled at 7000 watts, the windings will still jump up quickly to 110c or higher but will cool back down to 65c in just maybe 30 seconds to a minute on a fully heated motor. This would take me like 30 minutes to cool down with a sealed motor in the summer and that is why I didn't want to ride in the summer, because once that motor became fully saturated with heat (after about 800 watt hours of power through it) it wouldn't cool off after that and any acceleration would overheat it above 100c. Considering I have about 2400 watt hours of battery power my ride was pretty much over at only 1/3 my battery power because the motor was fully saturated with heat and wouldn't cool down I had to limit my driving to keep it from overheating.

How well does it hold up? After numerous trial and error and having to open up the motor like 6 times in the first week, I got it built right and it has held up perfectly with lots of hard riding and has not had any issue since for many months. The EDF fan has also held up without any issues other than having to glue on the propeller because it would slip off the axle, this is very common as the propeller is only press fit on.

There are also lots of ways to improve on the design, first the hot air when blown out of the motor may have some of it getting sucked back into the motor because they vent on the same side, this can be fixed by directing the hot air away from the inside holes with a small air duct on the outside. I can also add heat sinks to the inside of the stator to exhaust hot air quicker out of the stator. I also am running the EDF at lower volts and can easily raise the speed to get an even higher airflow or add another EDF as the weight is like 20 grams or so. The other thing I could do is direct the air more directly thorugh the windings on both sides by closing up the space above the windings.

I really believe this is one of the best designs for slow speed riding because you can cool when stopped and driving really slowly. It may also cool better than water cooling if I add heatsinks to the interior of the stator because it will push heat out of the interior of the stator like water cooling blocks through the heat sinks, but it also has the benefit of running fresh air through the windings on both sides of the motor and the magnet/stator gap.

Most of the time I don't run the fans and only use them for hard riding or those 90F+/32c+ days. Having drilled the side cover for air cooling seems to have made a huge difference by itself. The drawback is that it is somewhat loud, like a noisy computer at full speed. You won't hear it driving 20MPH+ but when driving really slowly you can hear it. I can easily control the fan on the handlebar with a servo tester.

If anyone is wondering why I drilled the holes so small the reason was first I only wanted air cooling when using the fan so I wanted them small to protect from anything large getting inside the motor as I do a lot of woods riding. I did some tests to get enough holes drilled to make sure the air flowed out with good force.

[youtube]iU9ti000LuU[/youtube]

 

[justin_le]

As I haven't seen this question addressed yet, is there any disadvantage to trying a higher viscosity fluid since the contact area is very small and we're only adding until we reach a slight increase in load?

I suppose a better question is would there be any specific advantage? I can do tests, but it takes bucketloads of time (and hence $$), so there has to be some kind of compelling reason why this would confer benefits before running that course.

For my fill hole I drilled a hole about 2" from the perimeter and then plugged it with a tiny sintered metal "breather" vent from McMaster (threaded into hole) and still had no FF leakage out the vent. The sintered metal vent helps equalize pressure and prevents water ingress near the axle which is the root cause of corrosion

I didn't know sintered metal actually blocks water ingress, I thought it was just for particulates no? The industry standard for allowing breathers for pressure equalization while blocking the flow of both liquid water and water vapour are expanded PTFE "Gore Vents", and our pseudo-plan is to have a snap-in gore vent that can be used to plug the hole used for adding the FF. Likely one from these series:
https://www.gore.com/MungoBlobs/236/167/GORE_PTV_SnapIn_en,1.pdf

Considering the popularity of the MXUS motors, how hard would it be to put together a group-buy for a few hundred finned-with-magnets-installed rings made? Why engineer all of these complicated retrofits when the manufacturer already has the facilities to build these quickly and in large volume?

I understand the enthusiasm on the one hand but you're also really jumping the gun quite a bit! Wouldn't a next step be to measure and confirm what kind of actual measured benefit is had by the addition of circumferential fins before talking about tooling up for mass manufacturing? The best way to know what to tool up is to first engineer and experiment with retrofits, and then determine what is the best design and whether the gains are actually worth it.

 

[justin_le]

I just wanted to let people know of the cooling method I built for my cromotor. The FF looks promising and I look forward to more testing. The only thing I want to see is how well it performs with low speed offroad riding, basically any really hard trail riding where you are riding very slowly and need to climb steep hills.

Well I've been running most of my more recent tests in the wind tunnel starting at 10kph, so I have all that data. I'm presuming that by very slowly you are still talking above 10kph / 6mph?

One thing that the FF would allow at low speeds is the ability to keep the motor sealed up without drilling all the intake/outtake holes and instead have an external fan that blows air over the motor housing. The FF is now thermally linking the copper windings to the motor shell so you don't really need internal airflow inside the hub. I mean if there's room inside the motor for the fan and you're not concerned about the holes then inside is cleaner and cooler, but there may be cases where an being able to blow air outside the shell could be preferred and just as effective.

If you can use an instrument to measure and record the temperature profile of your core while it is cooling down so we have a temp vs. time plot, then since we know the heat capacity of the MXUS hub we'd be able to cross reference that to an effective conductivity value and see how it stacks up with my current tests.

BTW the internal ducted fan build there is pretty awesome. I've played around with those things quite some time ago on UAV projects and they sure can move a serious amount of air.

 

[hillzofvalp]

This stuff won't like stick to magnets, right? It just reacts to them? So I must have a sealed motor.

Also, how long (hrs) do you estimate it will last in a sealed environment at 75C continuous 100C peak?

I didn't seal mine. Doesn't seem to be dripping out. FF does seem to cling to the magnet ring, even harder than centripetal forces try to eject it.

Say if you turned the motor so that the axis is coming out of the earth, would it drip out the vent? or does it stay on the magnets? If you shake, does it then come loose?

 

[flathill]

For my fill hole I drilled a hole about 2" from the perimeter and then plugged it with a tiny sintered metal "breather" vent from McMaster (threaded into hole) and still had no FF leakage out the vent. The sintered metal vent helps equalize pressure and prevents water ingress near the axle which is the root cause of corrosion

¨


Hm my dictionary falls short here, what is "sinter metal breather"? Do you have pics of your motor with this breather?
If this really equalizes pressure that should help maybe even for those who run ATF oil, as well as for [strike]ferrofluid[/strike] Ferrochill?
(maybe ATF will be be forced out of axle/bearing due to high pressure?)

here is what I used. Obviously sintered metal wont withstand submersion but rain ingress has not been an issue. i see less corosion with a breather than without

instead of loctite to hold it in (which may not be FF compatible) I just put a small ring of silicone to prevent it from backing out with vibration

 

[justin_le]

Say if you turned the motor so that the axis is coming out of the earth, would it drip out the vent? or does it stay on the magnets? If you shake, does it then come loose?

It's impossible to get it out of the vent hole even if you tried as hard as you might. Once the stuff gets to the magnets it stays at the magnets. If you overfill the gap then it can potentially leak from the side cover plate if you haven't sealed the side plate closed and run the motor at high speed.
But a) most side plates are already sealed, and b) once the excess FF has leaked out this way assuming you didn't seal the side covers, then the remaining FF in the gap is all you needed in the first place and it then no longer leaks. So it's a bit self regulating.

This post is relevant
https://endless-sphere.com/forums/viewtopic.php?p=1106508#p1106508

At the moment I'm running a set of FF mL vs conductivity tests on a Crysatlyte H3525 hub motor, and in this one I just drilled a hole in the motor plate without removing the side cover and breaking the original seal. As a technique, assuming that you don't allow chips to fall into the side cover hole when you drill it (ie drill upside down so that gravity causes all the chips to fall out) then it means you don't even have to deal with sealing the side plates, you can leave the original seals in tact.

 

[hillzofvalp]

I'm really curious how it will perform with a vented motor near the surface of the road... where a lot of dust and crap can get in. I wonder how easily the grime will dissolve in the FF and how that will impact thermal conductivity

Send me some please?

 

[flathill]

if you dont seal the side covers and some has already leaked out it is not self regulating

(but remember some are already sealed from the factory)

over time more will wick out or wick in between the plates, usually only the carrier fluid wicks, leaving the ferro particles behind which is worse than if the ferrofluid (carrier fluid plus ferro particles) was leaking

use of thicker fluid is more dependent on saturation than viscosity. in this application all we care about is heat transfer not air gap reduction (less resistance to flux)

 

[hillzofvalp]

I think Justin means it is self-regulating to a point.. it can only retain so much. Once you've reached the max amount, you shouldn't lose any unless the suspended particle distribution vastly changes, and there is nothing to draw the fluid inward (higher concentration towards the magnets, less around the edges, idk)? At that point you're only losing solution and not particles? No?

 

[flathill]

No once an "exit" path has been established it will continue to "wick" into the cover plates and/or out of the cover plates. When it wicks between plates it may not be even visible from the outside

It is capillary action on steroids. I have seen fluid leak out between two tightly clamped/bolted steel plates (Finely machined precision totally flat plates) traveling a few inches until all the carrier fluid was gone and all the particles were left in the motor. Other times with the same motor its was the ferrofluid that leaked leaving almost nothing left in the gap.

Once again... when the carrier fluid only leaks (aka ferrofluid separation of the carrier fluid and ferrous particles) it is worse than if the ferrofluid had leaked because now ur motor is filled with grit

 

[Punx0r]

I'm really curious how it will perform with a vented motor near the surface of the road... where a lot of dust and crap can get in. I wonder how easily the grime will dissolve in the FF and how that will impact thermal conductivity

Send me some please?

Why would you put ferrofluid in a vented motor?

 

[spinningmagnets]

No, once an "exit" path has been established it will continue to "wick" into the cover plates and/or out of the cover plates

I'm willing to risk it.

Why would you put ferrofluid in a vented motor?

Why does anyone around here do anything? Its a fun experiment, even if it doesn't make sense...

 

[Punx0r]

True. After that person put a pancake on a bunny's head humanity didn't really have anywhere to go...

 

[fechter]

The surfactant on the FF particles will make a certain amount of oil 'cling' to them and won't be drawn out by capillary action between the cover plates. They test the stuff by spinning it in a centrifuge. After centrifuging, the particles remain suspended. I think if all the oil somehow was sucked out of the FF, the volume of the particles would shrink so much that it would no longer bridge the gap between magnets and stator and not cause a real problem. You could also just add more oil at that point.

A higher viscosity of FF would likely increase the drag losses. Runny should be better.

There has been lots of discussion on various vent arrangements in the oil cooling threads. The vent only has to equalize pressure changes caused by heating/cooling of the air inside the motor. A very tiny hole would be adequate. Some way to keep rain and hose spray from entering would be good. The Goretex stuff should work to keep water out but it may not work if any oil gets on it. This could be tested.

Something like a drip-emitter for irrigation systems might work. Those have a tiny labrinth passage that would allow air to pass but be pretty resistant to hose spray.

 

[Kodin]

I would think a thicker fluid might be more stable long-term; since we are targeting only minimal contact, the drag a thicker fluid induces may not matter. Further, the thicker fluids may have a higher particulate count, which may enhance thermal as well as flux radiation. ...That said, I really have no idea if it matters. It's more just a possible-worthwhile-data-point to compare against. If it's a matter of money, I'm happy to chip in $50-100 to help with purchasing a thicker fluid, and I understand that your time is the main resource in short supply. I'm patient, so no rush on my request. Just figure it'd be nice to not make assumptions and actually test the alternative. Also, as soon as the FF kits are ready for shipping, let me know; I'm buying one as soon as you want to start boxing them up.

 

[liveforphysics]

I would think a thicker fluid might be more stable long-term; since we are targeting only minimal contact, the drag a thicker fluid induces may not matter. Further, the thicker fluids may have a higher particulate count, which may enhance thermal as well as flux radiation. ...That said, I really have no idea if it matters. It's more just a possible-worthwhile-data-point to compare against. If it's a matter of money, I'm happy to chip in $50-100 to help with purchasing a thicker fluid, and I understand that your time is the main resource in short supply. I'm patient, so no rush on my request. Just figure it'd be nice to not make assumptions and actually test the alternative. Also, as soon as the FF kits are ready for shipping, let me know; I'm buying one as soon as you want to start boxing them up.

Thicker fluid offers the potential for more loss and worse cooling only. A thicker fluid is not less prone to wicking or weaping out of the motor, that is a complex surface energy interaction. A thicker fluid has worse thermal transfer from longer chain molecules moving slower, it doesn't have a higher solids content nor the capacity to hold higher solids.

Likewise, these solids are effectively homogeneous in the fluid, they are nano particulates with insubstantial particle mass to particle surface tension interactions keeping them suspended. .

Modeling FF like some fine sand tossed in oil is not accurate with respect to the behavior, I would look at it more like a uniform oil.

That said, a wicking the oil up may be one of few ways possible to seperate the oil from the particles. I would personally not use anything like a porous breather or gortex breather for that reason (and because everytime Ive ever used a gortex breather in projects it just hopelessly clogged almost immediately.)

Perhaps use a very fine tipped needle, to shoot it into a very fine drilled hole, then call it done.

 

[justin_le]

I'm really curious how it will perform with a vented motor near the surface of the road... where a lot of dust and crap can get in. I wonder how easily the grime will dissolve in the FF and how that will impact thermal conductivity

Send me some please?

Why would you put ferrofluid in a vented motor?

Well, for kicks, at the end of all this I'm going to do one run in the wind tunnel of a motor that is potted / vented / air ducted / ferrofluid filled / and externally finned just to see what the overall compounding effect is like. Even in a vented hub, the FF should still provide a meaningful alternate conduction path for the heat to the motor shell (alternate to the passing air vent), and the more parallel paths for heat flow the less overall thermal resistance. Whether we'd want to have FF exposed to the elements in practice I find unlikely, but for those one-off race events where you want to use every trick up your sleeve it might prove worthwhile.

 

[liveforphysics]

I'm really curious how it will perform with a vented motor near the surface of the road... where a lot of dust and crap can get in. I wonder how easily the grime will dissolve in the FF and how that will impact thermal conductivity

Send me some please?

Why would you put ferrofluid in a vented motor?

Well, for kicks, at the end of all this I'm going to do one run in the wind tunnel of a motor that is potted / vented / air ducted / ferrofluid filled / and externally finned just to see what the overall compounding effect is like. Even in a vented hub, the FF should still provide a meaningful alternate conduction path for the heat to the motor shell (alternate to the passing air vent), and the more parallel paths for heat flow the less overall thermal resistance. Whether we'd want to have FF exposed to the elements in practice I find unlikely, but for those one-off race events where you want to use every trick up your sleeve it might prove worthwhile.

I bet it helps substantially even in a vented motor. Excited for that test, I bet that motor will have 2-3x higher continous power handling than stock.

 

[hillzofvalp]

Why do I want to try it with a vented motor? Because I already have ton of them. If it is the type of deal where you add more every 100 miles, that I could live with. If it gums up and destroys the motor.. not so much.

Haha I might have to restore life into my electric MB5.. with the 30kW hubmonster system. It is vented.. boy could it use some FF

 

[flathill]

The surfactant on the FF particles will make a certain amount of oil 'cling' to them and won't be drawn out by capillary action between the cover plates. They test the stuff by spinning it in a centrifuge. After centrifuging, the particles remain suspended. I think if all the oil somehow was sucked out of the FF, the volume of the particles would shrink so much that it would no longer bridge the gap between magnets and stator and not cause a real problem. You could also just add more oil at that point.

is that what u read?

my small collection from a few different vendors

 

[flathill]

Why do I want to try it with a vented motor? Because I already have ton of them. If it is the type of deal where you add more every 100 miles, that I could live with. If it gums up and destroys the motor.. not so much.

Haha I might have to restore life into my electric MB5.. with the 30kW hubmonster system. It is vented.. boy could it use some FF

FF works fine with vented motors for hot rodders but you have to considering the flux weakening effect of heating the magnets and the the possibility of overtemping the magnets. Just put a fine mesh over the holes. Forced air venting with FF is not recommended unless you filter the air and don't blow thru the gap.

When I make recommendations about leakage/wicking and lifetime, that is more directed to Justin who has to put his money and his customer satisfaction on the line, when and if he decides to offer an off the shelf motor with FF. If your motor leaks it is not big deal, just fill it up and buy some new clothes. If a customer buys a motor with FF and it leaks and/or gums up and begins to be hard to pedal and make strange sounds it is a different story. Justin has to design his motor to last maybe 8 years with no service.

FF also works with liquid cooled sealed outrunners to further improve dissipation but you have to considering the flux weakening effect of heating the magnets and the the possibility of overtemping the magnets (that that you can't OT the magnets without FF, but it changes how and when it happens). With FF your weakest link (temp wise) becomes the FF itself (long term, short term it may not be). Without FF the weak link is the magnet grade. With FF the magnet OT margin is effectively reduced as the air in a standard sealed motor acts as an insulator to shield the magnets from the heat source (as long as the motor is moving and you don't consider heat soak).

Remember weaker magnets can make a motor more efficient (at certain loads/speeds) and run at higher speed but it will have less torque upon the second acceleration if the magnets are hot.

 

[jmz]

Yes, I think some input from magnetics and motor experts would be valuable here.

The higher the operating temperature, the lower the field strength that will cause permanent demagnetisation.

Some discussion is here: https://www.kjmagnetics.com/blog.asp?p=temperature-and-neodymium-magnets , the required elements are

a) the magnetic work cycle of the rotor magnets under various operating conditions of D and Q-axis current.
b) the actual grade of the magnets used. May be hard to get from the manufacturers.


Hence I would be extremely careful when using field weakening (negative D axis current through the higher-permeability magnetic circuit of the D-axis) in combination with elevated magnet temperatures. A quoted figure for critical temperature of NdFeB magnets is 150 C, however this cannot be truly specified without the exact magnetic work cycle being employed.

The net effect is that I believe extremely good thermal bonding of the magnets and rotor backing to the shell will be required if field weakening is to be employed while increasing heat transfer through the rotor, hence potting of the rotor up to and perhaps over the magnet surfaces will be required when using FF.