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

Ianhill said:
Ive seen a raspi cooled submerged in a transparent liquid in a sandwich box i imagine that was opticool ?

That was one of opticool's tests. I believe they ran the computer submerged for 6 months, inspected it for any pitting or degradation, and saw none.

https://dsiventures.com/wp-content/...Reliability-in-Liquid-Cooled-Applications.pdf

Others have done the same experiment:

https://rc.library.uta.edu/uta-ir/bitstream/handle/10106/25911/THAKKAR-THESIS-2016.pdf?sequence=1

Ianhill said:
Thing is arm board is easy enough to cool with a decent heatpipe pc cooler and nothing else.

You may not even need a heatsink with a CPU that small. But this solution scales to large servers as well.
 
Bought two motors in a week. Today an Mxus XF39 30H. I believe Grin calls it a 3006. Price was too good. I don't know what I'll use it for, but I'll definitely coat it for use with mineral oil... or something.
 
Ianhill said:
Heres where i show you a trick of an ice engines coolant systems.
IMG_20210419_005300096.jpg

Even i missed a trick by there becuase looking at the graph the temperature range is intune with a nuclear reactor not a car engine and they operate at very high pressure no taking the cap of that expansion tank.
 
When investigating high-performance coolants (that is, substantially superior to common 0W oil or ATF, and assuming filled to axle), if choosing between 40C and 100C to compare, I'd actually suggest to compare at 40C. I suspect that common ebike motors would magnetically saturate (or fuse) long before coolant temp could hit 100C. 40C coolant is probably better representative of bursty loads.

Water is the first choice for cooling challenges (from ICEs to ICs):
kinematic viscosity
@40C: <0.7cSt, 6x lower
@100C: <0.3cSt, 5x lower

thermal conductivity
@40C: 0.629W/m*K, 5x higher
@100C: 0.677W/m*K, 5x higher

specific heat
4.2J/g*K, 2x higher

Distilled water has dielectric strength of 65-70MV/m, notably stronger than air, and indeed is used as dielectric insulator.
Distilled water is cheap and widely available.

Not surprisingly, it has been used as an ebike motor coolant. For this purpose, I would hesitate to add a corrosion inhibitor, which is likely to decrease dielectric strength / increase conductivity.

Water isn't perfect. Being polar, water readily dissolves salts to form ions, which do conduct electricity. So it wouldn't hurt to flush the motor a few times with distilled water first.

Still, these transformer oils and exotic dielectric fluids are spec'd for much different environments and conditions than ebikes. They're used where much higher voltage requires more robust electrical insulating properties, and where operating temperatures extend below freezing and above boiling.

I would try distilled water first.
Another cheap and widely available alternative would be a very light suspension oil: Redline Likewater, viscosity @40C: 4.0cSt and @100C: 2.34cSt.

But ultimately, for the DIYer, the added mass and drag are never going to compare favorably to simply more copper that doesn't lose energy to heat in the first place. I get there are packaging and cost constraints that make coolants attractive, but don't lose sight of the big picture.
 
I would never consider putting a conductive fluid inside a motor. The second you have a miniscule breach in the sealing of your stator and hall board, you have a short. The motor is also inoperable in freezing temperatures.. so you need additives to remedy that.

Water would also not work in the geared motors or mid drive motors, which are the most common motors today.

isoparaffin based cooling fluids are very competitive with water's thermal dissipation properties.
 
Castrol is now in the market with a product that looks an awful lot like it's isoparaffin based, and is using it for battery cooling. This one graph looks much like the viscosity-temperature curve that opticool has, IE a relatively narrow band of viscosity changes from 40C to 100C.

6a00d8341c4fbe53ef0263e9a0d16d200b-800wi.png

https://www.greencarcongress.com/2021/04/20210427-xing.html

They also sell a 'grease e-fluid' for motors, but details are even more scant on what that means:
https://www.castrol.com/en_gb/unite...c-vehicle-fluids/castrol-grease-e-fluids.html

If the above is some sort of paraffin derivative oil that's optimized for lubrication and cooling in one, then we're talking about something that's potentially better than opticool H.

Petroleum companies increasingly getting into making advanced cooling/lubricating fluids for electric vehicles is a nice trend to see. :)
 
I've always been interested in liquid cooling (I built evaporative and chilled-water computers), so I genuinely appreciate the interest and research.

neptronix said:
I would never consider putting a conductive fluid inside a motor. The second you have a miniscule breach in the sealing of your stator and hall board, you have a short. The motor is also inoperable in freezing temperatures.. so you need additives to remedy that.
Water? Distilled water isn't a conductive fluid -- as above, it's insulating. You'd deliberately overcurrent your motor to the point of burning up without liquid cooling, but wouldn't consider water?
Definitely wouldn't work in freezing temperatures, but that doesn't seem to be common riding weather, due to the impact on lithium ion batteries -- and a motor ridden in freezing temperatures would have least need of auxiliary cooling.

neptronix said:
Water would also not work in the geared motors or mid drive motors, which are the most common motors today.
Why not? Use marine grease on gears.
Why not on mid-drives?

neptronix said:
isoparaffin based cooling fluids are very competitive with water's thermal dissipation properties.
Well, they're very competitive because they have almost opposing properties (robustly insulating, extended temperatures, low volatility) -- as above, they're certainly not competitive on viscosity or conductivity...

It's great research and they're great products, but they're also unobtainable or hundreds of dollars per liter.
And that's their biggest limitation: since any liquid cooling adds mass and drag without improving efficiency, it needs to be cheap/free with minimal mass and viscosity to make sense as useful hack -- if you have to pay for it, it makes much more sense to simply put that money and mass into more copper instead, which doesn't suffer from parasitic drag or resistive electrical losses.

Again, I'm not saying water is perfect or even desirable, but within the narrow window where liquid cooling makes any sense (primarily, running such high current that you'd burn the motor up otherwise, but without money to buy a bigger motor), I'd at least consider it..
 
I don't think money is the barrier to getting a bigger motor

it's those looking for "fun" use cases, cycling for sport,

looking for high power at light weight.

500# cargo bikes are unlikely to get involved in this topic until reliable solutions have been proven by the zoom-zoom crowd.
 
fatty said:
Water? Distilled water isn't a conductive fluid -- as above, it's insulating.

My bad on this. But even distilled water does not fit the bill for an environment where you need thermal dissipation and lubrication all in one.

fatty said:
Why not? Use marine grease on gears.
Why not on mid-drives?

Marine grease would be displaced by the cooling medium in the motor, wouldn't it?

fatty said:
It's great research and they're great products, but they're also unobtainable or hundreds of dollars per liter.

Opticool H is $100 a gallon and the amount you need for a single motor is around $5 worth of product. Not necessarily out of reach. Some people on this forum spend that kind of money for a big tube of exotic lubricant they'll only use an ounce or two of.

fatty said:
And that's their biggest limitation: since any liquid cooling adds mass and drag without improving efficiency, it needs to be cheap/free with minimal mass and viscosity to make sense as useful hack -- if you have to pay for it, it makes much more sense to simply put that money and mass into more copper instead, which doesn't suffer from parasitic drag or resistive electrical losses.

In overdriven applications, liquid cooling induces some drag but allows more power output but ends up as a net benefit by increasing power density when a more powerful option is not available... which is universally the case with 2 stage geared motors and to a lesser degree, mid drives.

The current limit of a geared motor is it's ability to shed heat. The same dimensioned stator which is rated for 1000W in the MAC is rated for 2000W when sold by MAC as a lawn mower motor where the stator has a dramatically better path for stator heat to exit. Since you cannot buy a 2000W rated geared motor, you only option is to alter the heat dispersion characteristics of an existing one.

You could instead use a 2000w rated direct drive motor that weighs 2x more, but the additional unspring weight will hurt your rear suspension's ability to do it's job, and also, changing a tire on a 20lb wheel versus a 10lb one is a pain in the ass.

There are situations where a tradeoff between power density and efficiency makes sense.
 
neptronix said:
My bad on this. But even distilled water does not fit the bill for an environment where you need thermal dissipation and lubrication all in one.
Marine grease would be displaced by the cooling medium in the motor, wouldn't it?
Similar, so addressing together:
I'm actually not sure that distilled water would be worse for lubrication. Bearings and gears must have an appropriate lubrication film thickness -- in our applications, supplied by grease, which includes appropriate anti-wear additives. Tacky (especially marine) grease is quite resilient to and protective against water -- the nonpolar grease and polar water don't mix, so I wouldn't say there is a risk of water displacing grease. Instead, you'd want to carefully pack the bearings full of marine grease to displace the water.
On the other hand, exposing nonpolar grease to another nonpolar oil, like a mineral transformer oil, would wash out the bearing grease. Of course, you'd then have oil in the bearing, which is better than running the bearing dry, but the lighter oil wouldn't provide a sufficient lubrication film thickness.

It's kind of a known compatibility/risk (and thus bearing maintenance) with water, versus an unknown compatibility/risk with proprietary dielectric fluids. I'd want to see a lubricant compatibility list from the company.

So I'm definitely not saying water is better, but I don't have the data to say it's worse, either.

neptronix said:
Opticool H is $100 a gallon and the amount you need for a single motor is around $5 worth of product. Not necessarily out of reach. Some people on this forum spend that kind of money for a big tube of exotic lubricant they'll only use an ounce or two of.
Ah. I checked Amazon and eBay, but didn't find any. $100 shipped direct from the company?
$100 isn't bad, but 6oz sounds more like Statorade quantity, not filling to the axle. The published figure was 650mL for a 9C = 22oz = $17?

neptronix said:
In overdriven applications, liquid cooling induces some drag but allows more power output but ends up as a net benefit by increasing power density when a more powerful option is not available... which is universally the case with 2 stage geared motors and to a lesser degree, mid drives.

The current limit of a geared motor is it's ability to shed heat. The same dimensioned stator which is rated for 1000W in the MAC is rated for 2000W when sold by MAC as a lawn mower motor where the stator has a dramatically better path for stator heat to exit. Since you cannot buy a 2000W rated geared motor, you only option is to alter the heat dispersion characteristics of an existing one.

You could instead use a 2000w rated direct drive motor that weighs 2x more, but the additional unspring weight will hurt your rear suspension's ability to do it's job, and also, changing a tire on a 20lb wheel versus a 10lb one is a pain in the ass.

There are situations where a tradeoff between power density and efficiency makes sense.
Yeah, these packaging constraints (geared hubmotor and bottom bracket drive) make sense, and I don't mean to discount them.

Liquid cooling definitely has a place when pushing these packaging envelopes, and I'll be interested to see future research, products, and testing. I just wouldn't dismiss water out of hand, at least until there's actual testing data to support doing so.
 
john61ct said:
looking for high power at light weight.
Liquid coolant improves power-to-weight when compared to the same motor uncooled, but conventional wisdom is that putting the same added mass instead towards more copper gives higher power with fewer compromises.

Which isn't to say that liquid coolant doesn't have a place, but rather it's used when packaging-constrained -- just as OEMs use it.
 
I've seen people try to sell the remaining grease in ziplock bag packets on ebay like I've seen with the aviation Mobil 28 grease that people talked about packing their geared motors with. It only came in boxes of tubes, gallon and barrel quantities. If you go to the auto/boat parts store, you can buy single tubes of other varieties for dirt cheap compared to legit Mobil 28 sellers.

Nep
Opticool H is $100 a gallon and the amount you need for a single motor is around $5 worth of product. Not necessarily out of reach. Some people on this forum spend that kind of money for a big tube of exotic lubricant they'll only use an ounce or two of.
 
john61ct said:
The current limit of a geared motor is it's ability to shed heat
Unless the torque limit of the weakest gear is sooner right?[/quote]

Yes, that too. But very often gears can fail for thermal reasons and this would help them. Any motor with white nylon gears uses a nylon formulation with a lower melting point than say, the blue/gray gears that came in MAC and BMC motors.

In a dual or single reduction geared motor, you could greatly extend your 'hill climbing until melt' range and see higher efficiency during hill climbs due to the copper being at a lower temp. So even without boosting power, we get a benefit from an ideal cooling medium.
 
markz said:
I've seen people try to sell the remaining grease in ziplock bag packets on ebay like I've seen with the aviation Mobil 28 grease that people talked about packing their geared motors with. It only came in boxes of tubes, gallon and barrel quantities. If you go to the auto/boat parts store, you can buy single tubes of other varieties for dirt cheap compared to legit Mobil 28 sellers.

I am about to move my garage into a properly equipped warehouse and will be experimenting with a lot of new things i wasn't able to before. I'd be happy to be the first taker on a gallon of opticool H and resell it in smaller containers to members here interested in playing with it.
 
As expected, OptiCool is not compatible with mineral or synthetic oil or grease:
Since OptiCools Fluids have some solvency characteristics, I would expect that the OC Fluid would dissolve the grease over time. OC Fluids are miscible with just about all mineral and hydrocarbon-based oils, which are the base materials for these greases. So, the grease shouldn’t hold up in OC Fluid service.
Best,
David Childs
VP Technology and Technical Services
Cell: (903)216-5132
Tyler Office: (903)526-7577
www.DSIVentures.com

I'd suggest using a hydrocarbon oil like Redline Likewater, or immiscible distilled water.
 
Yes, anything you put in the motor is going to dilute or destroy the grease and turn into a mess.
The fluid you put in must be lubricating itself or able to disperse a lubricant.

Do you having any testing for materials compatibility on the Redline Likewater?
Materials compatibility is a big problem in geared motor and is why ATF does not work long term.
 
neptronix said:
Yes, anything you put in the motor is going to dilute or destroy the grease and turn into a mess.
The fluid you put in must be lubricating itself or able to disperse a lubricant.
Well, not anything (see silicone oil below), but certainly any nonpolar hydrocarbon.

neptronix said:
Do you having any testing for materials compatibility on the Redline Likewater?
Materials compatibility is a big problem in geared motor and is why ATF does not work long term.
Yep -- as above, hydrocarbon Redline Likewater would still eventually dissolve grease, but provide better residual lubrication.
Besides grease, the other materials compatibility I'd be worried about would be adhesives and phase wire insulation, which is trivial to change for oil-resistant wire.

Silicone oil (being used a transformer oil) was actually my next choice after water, with good and cheap availability, good materials and lubricant compatibility, high dielectric strength/resistivity, and cSt down to 0.65. But silicone oil flash point scales down with viscosity, so 0.65cSt oil has a flash point below ambient. I even thought about purging oxygen with an inert gas, but this wouldn't help in an accident.
 
fatty said:
Silicone oil (being used a transformer oil) was actually my next choice after water, with good and cheap availability, good materials and lubricant compatibility, high dielectric strength/resistivity, and cSt down to 0.65. But silicone oil flash point scales down with viscosity, so 0.65cSt oil has a flash point below ambient. I even thought about purging oxygen with an inert gas, but this wouldn't help in an accident.

Why don't you go for higher cSt Silicone oil then?
Had 20 cSt in a g310 (flash point avove 200°C)... Drag wasn't measurable. Leaking from the sealings was a problem though. Could imagine this would get worse with cSt <1.

I'm now testing 100 cSt. So far no leaking, have not measured drag yet
 
fatty said:
Yep -- as above, hydrocarbon Redline Likewater would still eventually dissolve grease, but provide better residual lubrication.
Besides grease, the other materials compatibility I'd be worried about would be adhesives and phase wire insulation, which is trivial to change for oil-resistant wire.

Silicone oil (being used a transformer oil) was actually my next choice after water, with good and cheap availability, good materials and lubricant compatibility, high dielectric strength/resistivity, and cSt down to 0.65. But silicone oil flash point scales down with viscosity, so 0.65cSt oil has a flash point below ambient. I even thought about purging oxygen with an inert gas, but this wouldn't help in an accident.

I can see how redline likewater could be an interesting candidate with some changes to materials in the motor.
If it is a concern with adhesives then the enamel around the windings may also be a concern.

I invite you to detail any experiments you engage in with alternative coolants in this thread. May the best one win. :)

My new warehouse space is still in disarray but i plan to do the opticool experiment in my eZee geared hub very soon.
 
Yamarlo said:
Why don't you go for higher cSt Silicone oil then?
Had 20 cSt in a g310 (flash point avove 200°C)... Drag wasn't measurable. Leaking from the sealings was a problem though. Could imagine this would get worse with cSt <1.

I'm now testing 100 cSt. So far no leaking, have not measured drag yet
Cool -- glad somebody else also thought of silicone oil.

I don't know how to model drag by viscosity. It may well be insignificant, but conceptually, lower viscosity = lower drag.

Please keep us posted.
 
neptronix said:
I can see how redline likewater could be an interesting candidate with some changes to materials in the motor.
If it is a concern with adhesives then the enamel around the windings may also be a concern.
Maybe. Part of the problem is impossibility of getting any materials list for a Chinese motor.
But really, for the low cost of these motors, the low temperature at which a cooled motor would operate, and the short duration which they are operated, I think just about any of these coolants would work. If coolant gets you another year out of the motor until you can save up for a proper motor that doesn't have such high losses, well, that's good enough in my book.

neptronix said:
I invite you to detail any experiments you engage in with alternative coolants in this thread. May the best one win. :)
I'm curious to learn, but I'm well past thermal limits and deep into stability/control limits. If I do have to revisit thermal limits, it would probably be in a sanctioned race environment, which may well exclude oil as a hazard if leaked onto the track (not to mention exceedingly dangerous for me as well). At least when I raced ICE, only distilled water coolant was permitted for this reason.

neptronix said:
My new warehouse space is still in disarray but i plan to do the opticool experiment in my eZee geared hub very soon.
:thumb: Keep us posted

Here's the guy using distilled water and Motul MoCool.
 
neptronix said:
Materials compatibility is a big problem in geared motor and is why ATF does not work long term.

To go along with my above assertion that anything would be good enough for a year, can you link to these ATF-induced failures? I'd be curious to read what failed.
 
Low-viscosity dimethicone.

Chemically non-reactive, non-toxic, easily accesible, good flow and heat characteristics...
I run my own MXUS 3K filled (to 1/3) with this thing.


"Silicone oils are primarily used as lubricants, thermic fluid oils or hydraulic fluids. They are excellent electrical insulators[2] and, unlike their carbon analogues, are non-flammable. Their temperature stability and good heat-transfer characteristics make them widely used in laboratories for heating baths... "
"Aerospace use includes the external coolant loop and radiators of the International Space Station Zvezda module, which rejects heat in the vacuum of space.[3]"
-- en.wikipedia.org/wiki/Silicone_oil


"The polymer is manufactured in multiple viscosities, ranging from a thin pourable liquid (when n is very low), to a thick rubbery semi-solid (when n is very high)."
-- en.wikipedia.org/wiki/Polydimethylsiloxane

Sure you get hit with a few more watts of motor drag, but is it even worth a consideration in a multi-kilowatt system? And with rising heat, the drag gets lower...

EDIT: i see it has showed up in the discussion already. My experience is that the "rated power" increases like twofold when coupled with two sets of hubsinks. I now actually have problems getting the motor past 60-80 deg.C.
As in, I am trying intentionally to overheat the thing... Full throttle up a long steep hill, where before i had to stop half-way up and let the motor rest even if going carefully... Now I rarely get upwards of 70C.
I will let you know if it has some catastrophic effect on my motor, but knowing how stupidly inert this thing is I can hardly see a bad influence.
 
fatty said:
To go along with my above assertion that anything would be good enough for a year, can you link to these ATF-induced failures? I'd be curious to read what failed.

Unfortunately nobody has documented them other than in passing, so the posts are hard to find. What i have read is that ATF eventually eats wire housings at a bare minimum over the course of a year. Nobody has posted pictures but i've heard this twice.

xfrankie said:
My experience is that the "rated power" increases like twofold when coupled with two sets of hubsinks. I now actually have problems getting the motor past 60-80 deg.C.
As in, I am trying intentionally to overheat the thing... Full throttle up a long steep hill, where before i had to stop half-way up and let the motor rest even if going carefully... Now I rarely get upwards of 70C.
I will let you know if it has some catastrophic effect on my motor, but knowing how stupidly inert this thing is I can hardly see a bad influence.

That's a very impressive gain from dimethicone and a testament to how much additional power density one can yield from a motor.
The question of course is how does it do from a materials compatibility standpoint over the long term.
Due to it's very long list of applications, i think it would have good materials compatibility, but this listing shows it has major issues with oils and solvents:
https://tym.co.uk/pages/silicone-pdms-chemical-compatibility

No mention of materials inside a geared motor or mid drive, which is a bummer.
I looked at some dimethicone oil spec sheets and never found a materials compatibility listing.
This listing on nylon states 'excellent' compatibility with nylon: https://www.calpaclab.com/nylon-chemical-compatibility-chart/

I'd certainly like to hear your long term report and the results of a visual inspection around 6-12 months to see if any materials end up being pitted.
 
Back
Top