Oil cooling a motor with transformer oil/dielectric fluid - the next frontier?

[Dui ni shuo de dui]

I think viscosity could be a factor in geared hub motors like the one I'm using in this small folding bike. The ratio that it's achieving is about 12 to 1. That runs at about 5000 rpm for the motor's rotor with the bicycle going 35 kilometers per hour.. At that RPM, the viscosity could be a significant drag. The direct drive is a different matter entirely. I'm gonna try to get silicon oil but if I can't get that I'll try paraffin oil.

The drag is quite easy to measure:
Test the no load current at a given rpm without any fluid, then do the same test again with oil.

This will directly give you how much power is lost to friction.
It's very easy to do, if you make some kind of draining hole in the cover you could be able to test many different fluids in a very short amount of time.

I think there will be a significant difference between no load dry and no load oiled, but in my opinion this will be completely unsignificant compared to the power levels needed to actually move the vehicle.

 

[Dui ni shuo de dui]

Is it possible to run a motor with oil in it yes definitely, can a working production example be mentioned ? I dont know of one and would love to see a detailed experiment on before and after and see if they keep it that way.

That's because there is no advantages of doing this from a production perspective.
But it doesn't mean there is no advantages for the people who want to extract the maximum amount of power to overcome the limitations of their existing motor. We cannot easily rewire our motors to limit heating from the source, so curing the symptoms is the next best thing.

In my opinion it has its place for people who ride their bikes hard. If you plan on doing this for a vehicle whose purpose is to carry you for long distances at relatively slow pace, or long straight lines in the countryside then it doesn't make any sense indeed. But if your goal is to do repeated drag races at red lights, or ride quickly in the city then it surely sounds like a good idea.

 

[AGS]

Oil cooling is a great idea until it leaks out all over your rear brake disc. I oil cooled a Bafang SWX02 a year ago so I could run it at 1500 watts continuously at under 90C and made sure I used good quality flange sealant on the motor shaft to seal it properly.

I am using ATF and it worked great for around 1000 miles until yesterday and the oil leaked out all over my rear disc.

Another problem that I have occasionally is oil leaching through the wiring and filling up the motor cable connector, so I have to periodically clean the connector with some switch cleaner when the motor stops due to oil contamination on the pins.

So be aware. Oil cooling isn’t fill and forget. It does have its downsides.

Now the motor has to come apart, cleaned out and fresh sealant applied as well as decontaminating my rear disc and pads.

 

[marka-ee]

I'm not trying to extract unusual amounts of power. It threw a magnet when I was simply going in a slight downhill with higher than normal speed. It was still very hot and it threw a magnet. I'm hoping that I can just cool things down a bit.

 

[Papa]

That's because there is no advantages of doing this from a production perspective.

Gear hub manufactures choose to avoid liquid lubrication because of the added cost and design complexities of sealing from leaks.

My garden tractor's 6spd Peerless transaxle uses bentionite because it has no axle seals. Works fine as long as I avoid heavy loading - which reveals it's limits via heat. And during cold weather, such as snow plowing, shifting gears is aggravating. It's a low speed vehicle so heat generation normally isn't a problem.

I have a new Gmac that'll get filled with liquid lube... Not grease, because it'll be expected to pull moderate loads up 6-10% grades regularly. Regen will hopefully replace the rear disc if leakage becomes a problem.

 

[neptronix]

Oil cooling is a great idea until it leaks out all over your rear brake disc. I oil cooled a Bafang SWX02 a year ago so I could run it at 1500 watts continuously at under 90C and made sure I used good quality flange sealant on the motor shaft to seal it properly.

I am using ATF and it worked great for around 1000 miles until yesterday and the oil leaked out all over my rear disc.

Another problem that I have occasionally is oil leaching through the wiring and filling up the motor cable connector, so I have to periodically clean the connector with some switch cleaner when the motor stops due to oil contamination on the pins.

So be aware. Oil cooling isn’t fill and forget. It does have its downsides.

Now the motor has to come apart, cleaned out and fresh sealant applied as well as decontaminating my rear disc and pads.

Sounds nightmarish!

Looks like this motor has is a small pocket where fluid can sit, and i would guess that the smaller the motor, the more discharge, lol.


I wonder if a motor like a MAC which is taller could retain the given cooling fluid much easier. Bullfrog seems to have good experiences with Motul mocool.

The MAC has a much longer distance from cooling fluid pocket to axle..



Does the oil heat up and leaks progressively more... or does it leak a little at the start then accelerate as it gets hotter?

Does leaking happen if the bike is tilted?

Justin (Grin Tech owner/operator) suggested an alcohol might be a good cooling fluid. After his suggestion, I tried distilled water with Motul MoCool as a corrosion inhibitor in my MAC and it worked quite well. Nice thing about using my mixture was if it leaked on my brake disc, it didn't appreciably affect the braking performance like ATF did.

Geared hub motors like a MAC have a different requirement than Direct Drive hub motors like the Leaf because you need about five times as much cooling liquid to thermally connect the windings to the shell for the Geared motors so viscosity can be a bigger problem for the Geared motor.

Alcohol and opticool are both on my list to try someday.
Maybe i should get a GMAC.

 

[Dui ni shuo de dui]

Oil cooling is a great idea until it leaks out all over your rear brake disc.

Interesting, finally some first hand experience!

Where did it leak from? Was it the cover flanges of the main axis seals?

Another problem that I have occasionally is oil leaching through the wiring and filling up the motor cable connector, so I have to periodically clean the connector with some switch cleaner when the motor stops due to oil contamination on the pins.

That's unexpected, very interesting.
You mean the oil crawled outwards by capillarity up until the controller connector? is this distance very short on your particular vehicle ?

So be aware. Oil cooling isn’t fill and forget. It does have its downsides.

Now the motor has to come apart, cleaned out and fresh sealant applied as well as decontaminating my rear disc and pads.

Do you have a venting hole or something so the pressure can equilibrate between inside and outside the motor?
When the motor heats up the air pressure inside will increase and it will tend to try any possible way to escape. If you didn't allow for a specific, controlled path for it then either pressure will increase until the weakest spot fails, or it might push everything through the main axis seals (and the wires as well, apparently).

What's your opinion about this now you'vve made about 1000 miles with it?
Does it worth the hassle?
Did you notice any significant drop in range?
Is there any drawback besides the leaks?
Does it change anything like noise, driving sensation, vibration?
During disassembly would you mind taking a few pictures of your process so we could see how the motor aged, how dirty the oil got, how much did you put, how you sealed the engine , etc? That would be awesome.

Thanks!!

 

[Ianhill]

Hub motors are difficult to cool at the best of times and im sure amberwolf could have linked a thread that had alot of test info detailed not sure if anyone else can.

When i joined the forum the main way for more power from a hub motor was been obtained from a rewind opening the case up for airflow and overvolt it use an electical varnish for a bit of protection.

Theres no more efficient way of controlling heat better for burst torque than adding more copper fill while performing an overvolt, if u go through the effort of adding stronger magnets to lower kv you end up with a motor that has very simular rpm per volt but the torque curve has jumped up the scale giving better climbing ability and burst torque and a slighty longer heat soak period but again theres no unlimited power strategy eddys will build on sharp edges of the core and heat will gather and move through the core till its saturated with heat, or to many amps and the end turns in the middle of a winding melts becuase they not in the stator core and normally suffer from excessive amps heating them first.

Could even go through the trouble of encapulating the windings in thermal potting for a extra few seconds at best and then the halls sensors cant be repaired easily so thats a mad mans game.

So i see the appeal of oil it seems easy and its a coolant no doubt but the practically of it sloshing about in a washing machine with little sealage waiting to flick out the axle on to your disc and tyre not so much.

 

[Dui ni shuo de dui]

So i see the appeal of oil it seems easy and its a coolant no doubt but the practically of it sloshing about in a washing machine with little sealage waiting to flick out the axle on to your disc and tyre not so much.

Yet this is what happens everyday in every petrol engine ever made. I don't think this is a problem that cannot be solved at all.

For example, it is easy enough to add anti splashing covers in the inside of the motor to avoid having too much oil on the bearings or the shaft. A kind of 3D printed sleeve that would go around the axis and prevent most of the oil from coming in contact with it.
As for sealing the flanges, In my opinion this should be the easy part, either with caulk/silicone, or a custom made paper gasket.

The difficult part might be venting, but in my opinion a breather hole could be fitted somewhere near the axis, where it shouldn't be too exposed to oil projections. It is also possible to add some kind of labyrinth path before it so it would make oil escaping almost impossible.

Theres no more efficient way of controlling heat better for burst torque than adding more copper fill while performing an overvolt, if u go through the effort of adding stronger magnets to lower kv you end up with a motor that has very simular rpm per volt but the torque curve has jumped up the scale giving better climbing ability and burst torque and a slighty longer heat soak period but again theres no unlimited power strategy eddys will build on sharp edges of the core and heat will gather and move through the core till its saturated with heat, or to many amps and the end turns in the middle of a winding melts becuase they not in the stator core and normally suffer from excessive amps heating them first.

Yeah, well this is very easy to say but also very difficult to do.
Changing magnets is pretty much out of reach beause they are very difficult to come by, plus they are surprizingly expensive (it's almost half the price of a new motor).
Rewinding a motor isn't impossible but it is very difficult to say the least (I tried). And also expensive.

As you said yourself, heat will gather until the core is saturated. So the idea here is to try to at least delay the saturation time, in order to either push more power during the same time, or the same power levels but for a longer time.

 

[Chalo]

There's no replacement for displacement. Use a motor within its rating, speed, voltage, and load. Then no prollems! For more of the above, use a bigger motor. Leave the filthy drippy BS to the ICE dingdongs. Their garbage is only like 25% efficient anyway.

Industrial users (get stuff done sector) wouldn't tolerate any of that nonsense if it wasn't absolutely necessary. They just use as much motor as the application requires.

 

[Dui ni shuo de dui]

There's no replacement for displacement. Use a motor within its rating, speed, voltage, and load. Then no prollems! For more of the above, use a bigger motor. Leave the filthy drippy BS to the ICE dingdongs.

Well yeah, obviously.
But then it can be said for any kind of mod.
Why doing overclocking when you could just spend thousands of dollars on a better CPU?
Why increasing the turbo pressure on your camry when you could spend hundreds of thousands of dollars and purchase a Ferrari ?

I don't understand what is the point in such comments, really. The whole point of this thread is to try to increase the output from a given motor, we are all aware of the fact that better, more expensive motors are available. Unless maybe you were offering to pay for the swap? That would be nice.

If this doesn't interest you then why not just let those who find this interesting experiment the way they like?

 

[Chalo]

The whole point of this thread is to try to increase the output from a given motor, we are all aware of the fact that better, more expensive motors are available.

Bigger ≠ more expensive. I don't fool with "more expensive"; I'm a bike mechanic for a living. But annoying yourself and everybody who has to share a floor with you so you can get 1500W from a 350W motor, when you could use a 1500W motor instead, is kinda goofy.

To be fair, I dump as much as 48V x 60A, 2880W nominal, into 1500W (Leaf) and 1200W (QS212) rated motors. But if they got more than slightly warm, I'd turn down the power.

Surely there's some point where you want tens of kilowatts and the motors to do that get rather expensive. But I'd rather let up on my notions of how much power I require, than live with a leaky mess that might as well be an English sportscar or a Harley Davidson. No thanks.

 

[Bullfrog]

Using any fluid to cool a MAC/GMAC is problematic...when I used ATF, the fluid actually leaked when capillary action pulled the ATF inside the insulation on the wires. I was successful using distilled water and a little Motul MoCool as a corrosion inhibitor. Even the water needed "topping off" occasionally because it would evaporate over time. With a MAC/GMAC you need to add between 5 and 6 ounces of cooling fluid for it to reach the stator and do any additional cooling. With a Direct Drive motor, you can get away with a lot less fluid.

All electric motors "breathe" to some extent...when they get warm the pressure inside increases and as they cool the pressure inside drops. My point is completely sealing a motor to be air tight and "fluid" tight might not be the best idea.

I'd first size my motor and pick my Kv so it should not get too hot...you can use the Grin Tech Motor Simulator to get an idea about your temperatures. Then if you need more cooling and it is a Direct Drive motor, I'd add Statorade. Hub Sinks will add a little more cooling capacity but not nearly as much as the Statorade.

Although the water/MoCool mixture worked in my MAC, lacing the motor in a smaller diameter wheel and staying below about 28 mph will help a LOT...the smaller diameter wheel and lower speeds will keep any geared hub motor cooler. The diameter of the wheel is what determines the final gearing.

 

[Chalo]

Furthermore, it's not like we don't have fluids already in our e-cycles... Ever heard of brake fluid? Are you also recommending we stop using it because it might leak?

Hydraulic brakes on bicycles are silly moto-fetishism, and yes they do leak and ruin the brakes of fools who fall for the hype. I see it all the time: donated bikes that come in with mechanical brakes usually still have brakes that work, most of the donated bikes with hydraulic brakes do not.

If what you really have is a motorcycle, then use motorcycle brakes-- hydraulic or whatever. Bicycle hydraulic brakes are toy nonsense, and they demonstrate it again and again.

Motors are rated for continuous power. Their rating is completely irrelevant to the power bursts you can feed into them.
My old QS205 motor was rated for 2KW and it took between 16KW to 22KW peaks on a daily basis for years.

"My Civic makes over 500 hp with the stock crank and it's just fine!"

Wildly exceeding rated limits is okay, until it isn't. Do you think the design engineers behind these things are trying to deprive you of the potential of the products they want to sell?

If you want to use some problematic liquid cooling, why not try liquid nitrogen? Who knows, maybe that would let you rewind your stator with a superconducting alloy wire. Then you wouldn't have to pay extra for a bigger motor.

 

[marka-ee]

The amount of oil I should add to this gear reduction type hub motor is in question here for me. I just came to the revelation that yes, my rotor threw a magnet because it became too hot, but the rotor with the magnet is not the reason or the source of the heat in the first place. It's the field windings of the surrounding stator . So I don't have to add so much oil to reach up to the rotor. It's just really adding enough oil to enable some thermal path for the poor little windings to shed their heat to the outside aluminum housing. This is the design limitation or problem in this type of motor. There's this wonderful 5mm air gap, which is a nice insulator between the actual motor and the outside housing. So if I just add enough ATF to provide a good thermal sink, then that should be enough. maybe 120 cc's. Am I misguided in my thinking? Now, because the two items are spinning against each other at a rate of about 270 RPM, then there will be significant swirling and sloshing of this oil. So I should add enough to make sure that even during the sloshing and whatever, there's still enough thermal conductivity between the two. Just to clear up a point of my motivation on this, I'm not trying to extract huge amounts of power from this little tiny 250 watt rated unit. I just don't want to throw it away. I want to use what I have.

Edit: I ended up putting in 50 cc's of synthetic ATF and a teaspoon of molybium disulfide powder for good measure. 50 cc's when I saw it in the syringe looked like it was just the right amount. I did a 40 kilometer ride to test it and there's a little bit of seepage from the axle shaft but very minor. Molybium disulfide makes it easy to spot, makes it rather black.

 

[AGS]

I cleaned all of the grease from the gears and put around 75ml of ATF in my SWX02 using a syringe. It was enough to keep my temperature down to under 90C at a sustained power of 1000 watts.

I carefully drilled through two of the brake screw holes making sure not to damage the threads. One hole for filling and the other for venting during the fill.

I also added a LM 35 temperature sensor inside the motor at the same time using the white wire from the speed sensor. I am using a KT controller so I don’t need the speed sensor.

Temperature monitoring with an LM 35 is ok but it’s only accurate when the motor isn’t on due to emf interference. So I stop peddling and let off the throttle when I want to check the temperature.

An NTC 10k temperature sensor is much better if you have a couple of wires available, because that won’t suffer from electrical interference so much.

 

[Ianhill]

I like the ideas been brainstormed for overcoming issues like 3d print around the axle and the covers are the easy bit, as for venting pressure through a liquid but not the liquid itself i think that would need a stainless pore insert that has a desired micron mesh.

As for how well it works thats the experiment, dont be to hopeful and u may be suprised but maybe not to lol.

I predict alot of cavitation in the cooling fluid that reduces its performance as a coolant drastically, the motor would need to turn very slow for any hope of measurable results i believe.

I wish u well on your experiments none the less.

 

[marka-ee]

So far my experiment with the small 250 watt geared hub motor is working out okay. 50 cc of synthetic ATF with a small amount of powdered moly ultrafine lubricant powder. There's a small amount of seepage from the axle bearing shaft, but I just wipe it off with a paper towel occasionally. It's not much. I think the temperatures are way better controlled now since there's more than just an air gap insulator between the motor and the outside housing, the biggest design weakness in these geared hub motors, in my opinion.

 

[scianiac]

Does that motor have all metal gears? Generally moly is recommended to not be used with polymer surfaces as it is abrasive to them, PTFE is the preferred EP additive for polymer components. IDK how much it actually increases wear in this case.

 

[marka-ee]

It's actually a very strange 'geared' motor because not only does it not have plastic inside, but it does not have gears inside. It has metal rollers with needle bearings. So far it's not slipping , so it's working out okay.