Oil cooling your hub- NOT snake oil!

[fechter]

It may have to do with more air moving past the outside casing at higher speeds or the oil is pumping around the insides better at full speed.

OK, don't use brake fluid. Thanks Bigmoose.

Here's some interesting info on silicone oil:
http://www.clearcoproducts.com/standard_pure_silicones.html

"Clearco Low Viscosity Pure Silicone fluids are clear, colorless, and odorless
% linear polydimethylsiloxanes (CAS# 63148-62-9) They range in viscosity from 5cSt to 20cSt (centistokes) and are characterized by their low pour points, low surface tensions, low viscosity change at temperature (VTC), excellent lubricity, high dielectric strength, high damping action, wide service temperature range and inertness to virtually all substates.

Clearco Low Viscosity Silicone Fluids also possess high resistance to shear, high water repellency, low vapor pressure, and low reactivity. They are excellent lubricants for plastics, foams and rubbers. Due to their low pour points and low V.T.C. (viscosity change at temperature), they are excellent choices for low temperature bath fluids

Properties:

•Low viscosity polydimethylsiloxane fluids..slightly thicker than water
•Low pour points (as low as -100°C)
•Excellent Low Temperature Stability
•More Thermally stable than the Super-low viscosities
•Low surface tension fluids (high spreadability)
•Low viscosity change at temperature (low VTC)
•Inert to virtually all o-rings, gaskets, seals and valves *
•High Dielectric Strength
•INCI Name: Dimethicone
•California VOC Exempt
* not recommended for silicone rubber due to possibility of swelling

Hmm... not sure what it would do to silicone glue?

Definitely need some kind of vent or air expanding with temperature will pressurize the housing and force a leak.
Oil will be good for those solid steel gears, or any gears for that matter.
Power steering fluid might be another possibility. Based on my garage testing, it seems to be a little thinner than ATF.

[Spicerack]

Many or even most ps fluids are ATF in a different bottle with a higher price tag.

ATF is a good choice for hubs as its cheap enough and easily available. Also easy to spot a leak being red. The silicone based oils are usually a more expensive and more difficult to find- or just use dot5 brake fluid as mentioned. I'll be sticking with ATF- doesn't attack silicone either.

[johnrobholmes]

I'm gonna have to give ATF a shot in one of my MAC hubs. Heck, i'll just drill one hole and see if it leaks out a stock motor.

[Spicerack]

I'm gonna have to give ATF a shot in one of my MAC hubs. Heck, i'll just drill one hole and see if it leaks out a stock motor.

Oh, I soooo wanna see before and after photos of that, especially your clothes...

After you've done that you might want to seal those side covers up....

[teklektik]

ATF is a good choice for hubs as its cheap enough and easily available. Also easy to spot a leak being red.

Yep - though I might be tempted to use one of the low viscosity brews from Red Line.

[johnrobholmes]

I'm gonna have to give ATF a shot in one of my MAC hubs. Heck, i'll just drill one hole and see if it leaks out a stock motor.

Oh, I soooo wanna see before and after photos of that, especially your clothes...

After you've done that you might want to seal those side covers up....

Bah, I think the wet paint during assembly is enough! I'll be sure to wear eye protection, I bet it will be like spin art.

[Spicerack]

Bit difficult to see, but here's a couple of shots of the oil holes. The first one is the black plastic bung I used to cover the air cooling holes. It has a hole drilled in the middle which is filled with a rubber bung. Oil goes in here.

The second photo is not very clear but you can just make out the vent hole through the gap in the disc- just to the right and up from the rotation arrow on the disc. The red drip of atf on the cover near the axle end is from overfilling it slightly this time- when the vent hole is down and the wheel's not turning, it just seeps out of the hole very slowly. Not enough to drip on the floor yet.

[zombiess]

I can't believe it, but you guys are actually making me want to try this myself now. I like the simplicity and no mess of just venting the covers which I have done on my 9C 2806 and plan to vent some covers on one of the Hubzilla motors I'm going to attempt to dump 20KWish into. I'm curious as to which will perform better. I know the holes I put in my 9C 2806 made a huge difference in cooling and allowed me to ride it much harder for longer before it would heat soak. Even with lots of abuse the windings still look new and the motor performs great and will most likely be used in a project for my wife which is how my race bike started in the first place LOL, sorry hun.

[Kingfish]

Mechanically-speaking, the method of oil splashing around inside the motor is very common for lawn mower engines: Of the two most common methods, the bottom of the piston has a simple blade protruding about an inch or two that dips into the oil basin at the bottom of the down-stroke, splashing the oil, atomizing it into vapor in the process, and ultimately lubricating the interior. The second method uses a lobe on the crankshaft that drives a spring-loaded plunger which pumps oil up through the piston via the journal.

Good lightweight motor oil should work fine, probably SAE 5W-20. Synthetic is optional. Most engine oil has a limitation of about 185°F/85°C wherein further rise will cause it to separate into degrees of paraffin.

Since the default cooling method is radiation (air) and conduction (axle), adding oil allows for convection which naturally boosts the cooling ability at the cost of reduction in spinning efficiency; there will be drag. Light oil though should greatly reduce the effect.

I am saddened to learn of this thread though AFTER I had my hubs drilled to vent water. However, a small vent hole near the axle is still possible, and I like the idea that oil is sloshing around to keep moisture away from my delicate unplated stators. At least they will have varnish.

Good thread, good ideas; thanks guys – this is very useful!
~KF

[Spicerack]

I am saddened to learn of this thread though AFTER I had my hubs drilled to vent water. However, a small vent hole near the axle is still possible, and I like the idea that oil is sloshing around to keep moisture away from my delicate unplated stators. At least they will have varnish.

Good thread, good ideas; thanks guys – this is very useful!
~KF

My 9C has seriously drilled covers for air cooling. I'll be filling the larger holes with perspex discs epoxied in. That way I'll be able to see the oil level easily! The smaller holes I'll just fill with epoxy like I did on this MAC motor.

[John in CR]

Smaller geared hubbies which are so thermally limited already, have a fast spinning rotor that needs cooling, and gears that like lube...sure...MWKeefer has been doing it for years. I seem to recall that he tested it on DD's too, and the efficiency hit made it not worth it. On direct drives I'll let you guys be the guinea pigs, since mine don't overheat and enough oil to reach the stator means enough to reach the fragile hall wires. Hopefully some of you will also run some comparative wh/mile numbers before and after. In the meantime, I'm focusing on more efficient motors, since having half the heat to reject is a better solution.

[flathill]

I seem to recall that he tested it on DD's too, and the efficiency hit made it not worth it.

Do you recall why? Could the increased drag be that much of a hit? Remember the cooler you keep the windings the less losses....Your 93% motor wont measure 93% hot. Are you the same John from diyaudio?

http://www.stereophile.com/reference/1106hot/index.html

Fig.1 Relative resistance vs temperature for copper (blue trace), aluminum (red), and silver (purple).

For drag racing analuminum wind motor might win (not likely, depends on winding to iron ratio and some other stuff) Aluminum windings can also be smashed tighter than copper without cutting the insulation so what you give up in conductivity you main gain in fill factor (assuming round wire) so instead of copper having a 20% advantage its more like 10 real world (not counting volume)

[fechter]

A somewhat simpler test would be to measure the no-load, full speed amps before and after adding oil. My guess is the oil will increase the no-load current quite a bit. There's no way it will make the motor more efficient, but will allow it to run at a higher power rating without failing. It would be nice to know exactly how much additional drag it causes. It may cause only a small reduction in range.

It should be much less significant on a direct drive motor as the velocity of the rotor is much slower.

[John in CR]

I seem to recall that he tested it on DD's too, and the efficiency hit made it not worth it.

Do you recall why? Could the increased drag be that much of a hit? Remember the cooler you keep the windings the less losses....Your 93% motor wont measure 93% hot. Are you the same John from diyaudio?

Speaker building was a hobby that seems a lifetime ago, though just 5 years or so ago, so it depends which John there.

Regarding the 93% efficient motors, there's exactly 0% chance oil is going in there as it is likely to mess up the series/parallel switching.

If people are so interested in cooling their motors, why don't they run smaller wheels and adjust voltage back to the same speed? On top of the greater efficiency resulting from the lower torque requirement, they'll also get much better cooling from the higher rpm. 90%+ who ventilate don't make an effort to optimize it, and a spinning pizza pan with some big round holes in it certainly doesn't make much of a fan blade.

The issue isn't whether a liquid cools better than air, since that's obvious. For me the question is whether it's more effective overall compared to a substantial fresh air flow, and in determining effectiveness you have to consider increased drag, risk of the more direct heat transfer through the magnets, the overall rate of heat rejection, etc. In comparing to ventilated you've got to consider that if you do it right and achieve flow that the air carries heat out with it, which totally bypasses the step of outer shell to the environment.

Here are my concerns about the oil in a DD approach:
1. Drag- The decrease in top speed should give some idea.
2. Halls- I haven't seen a response when I asked before.
3. Magnets- Is there more risk of heat damage to magnets with such a direct route for heat?
4. Optimum fill- This might be the hardest to determine, since at speed the oil is forced to the perimeter
by centrifugal force. I'd think just enough to just touch the lamination edges at speed would minimize drag
while maintaining a good thermal pathway. The best fill might even be even less so at low speed where efficiency
is low during acceleration that the oil falls onto the stator. Then at high speed and high efficiency where drag
if touching the stator would be great, centrifugal force pins the oil to the perimeter offering no drag at all and
almost no cooling effect, but it's not really needed anyway. OR since it's the windings that are the heat source
does the fill need to be much greater to fill to the copper even at speed despite the greatly increased drag.
5. Erosion- Obviously that moving oil is going to wear on parts not designed for a liquid. Other than the hall wires,
I'd be concerned most about the potential for the oil to get under the edges of the epoxy holding the magnets
and whether it could do harm getting between the laminations especially starting at the edges facing the magnets.

On top of these concerns I don't have a good visualization of the oil's behavior at speed with it inside the rotor, a short spinning cylinder, and having the stator, a very coarse stationary object, in such close proximate all around. If the liquid gets turned into a fine spray that would be ideal and greatly reduce drag.

Heat problems have causes that liquid cooling won't solve, so better cooling is just a bandaid. Addressing the cause, which is typically over-gearing or over-saturating the stator, is the real solution. How else could one of the biggest loads on the forum zip around at 50mph+, and climb mountains for over 3 years with a sealed motor that is very little different than those most DD hubs? I'm not even talking about recent addition to the stable of the 93% peak efficiency hubbie I'm still testing. Forget the western medicine approach and cure the problem instead of living with it and using bandaids that come with risks and compromises. Shoot, I don't even run a temperature sensor. Strap a 100lb backpack on most of these guys backs and let's see how their ebikes perform both on the highway and in the mountains. Liquid cooling won't be enough to keep up.

John

PS- Regarding drag, my son swears he feels a bit of drag with the motor I ventilated that he runs on his ebike. I can feel it too when I ride it, and if I had to guess I'd say it's about 50W of fan blade resistance. That's air, so just imagine the drag of a liquid at the speeds we run.

[John in CR]

It should be much less significant on a direct drive motor as the velocity of the rotor is much slower.

I think it's the opposite. MWKeefer did a lot of experimenting, and I think it's just enough fill in the geared hubbie that the smooth outside of the motor bell housing just touches the oil. Plus spinning the opposite direction is likely to create more of a spray effect. With a DD you have just the tiny air gap now filled with oil and the other odd shaped parts of the stator with lots more surface area.

Detailed results will be interesting, just not with my motors.

John

[flathill]

You bring uo some good points John from CR
The answer to all these thing is
it depends
You have guys in ev land swearing gearboxes are dead
But I dont see a future of big rig trucks carrying tons of cargo without gears or active cooling
I was just tring to bring up the point a cooler running motor is more efficient no matter the application
Passive cooling should have almost no downsides but you are right about oil seeping into the glue joints could be an issue depending on how they react

[John in CR]

You bring uo some good points John from CR
The answer to all these thing is
it depends
You have guys in ev land swearing gearboxes are dead
But I dont see a future of big rig trucks carrying tons of cargo without gears or active cooling
I was just tring to bring up the point a cooler running motor is more efficient no matter the application
Passive cooling should have almost no downsides but you are right about oil seeping into the glue joints could be an issue depending on how they react

Once battery tech get to where a big rig EV makes sense will be a happy day for me. Just imagine the ebike I can have. Of course I'm still waiting for my Jetson's type car I was essentially promised by the year 2000 back when I was a kid.

[Hillhater]

.....ATF is a good choice for hubs as its cheap enough and easily available..

Has anyone considered the oil they use for cooling in HV transformers. ??
Its formulated for cooling electrical devices, but i suspect it may be higher viscosity than ATF

[motomech]

A somewhat simpler test would be to measure the no-load, full speed amps before and after adding oil. My guess is the oil will increase the no-load current quite a bit. There's no way it will make the motor more efficient, but will allow it to run at a higher power rating without failing. It would be nice to know exactly how much additional drag it causes. It may cause only a small reduction in range.

It should be much less significant on a direct drive motor as the velocity of the rotor is much slower.

Somebody has. From this discussion,

Good! Someone finally oil cooled a hubbie!! HS3540

viewtopic.php?f=30&t=34369&hilit=cooled&start=45

[quoteAt 400ml fill, the oil level is just below the axle, where the filler hole is in the above photo, (you can just see the red coloured oil inside the hub).
With this much oil, the motor makes a satifying, just audible ,sort of sloshing noise when it is spun up.


No oil, no load current = 1.05 A
150 ml oil @ 18 degrees C = 2.0A
200 ml oil = 2.0A
250 ml oil = 2.1A
300 ml oil = 2.15A
400 ml oil = 2.15A

I will leave the oil level @ 400ml and install it on my bike to test the thermal effects.
The bike currently has an identical motor fitted, including a temp sensor (just no oil), so the comparison will be interesting.

Burtie
edit: corrected reference to Dexron II
Last edited by Burtie on Tue Jan 24, 2012 1:59 pm, edited 2 times in total.

Burtie
1 kW


Posts: 439
Joined: Fri Mar 27, 2009 10:45 am
Location: UK
][/quote]

This was a 9C motor

[johnrobholmes]

Regular mineral oil is used for cooling transformers.

[Spicerack]

I agreee with alot of the comments John in CR makes. I did it as a bandaid- but can also see an application for pushing cheaper motors well beyond their thermal limits- especailly geared hub motors like the mac or bmc where they are an insulated oven if you try and put too much power through them (remembering they are rated at 500-1000W, not the 3300W I was putting into mine).

We can't all afford to get a cro or don't want a 16inch scooter motor etc, which can certainly handle alot more power without causing dramas.

This type of mod is simple, very cheap to do and can elevate our lower powered hub motors into a much more fun level of power. In my case it makes my bike rideable again until I fix it. Even then, because I have time and money invested in what I have, I may as well oil cool the new stator and see what I can get out of it. As long as I monitor temp at the windings it will be fine. When I get to that point, I'll do some fairly decent dry and wet measurements just for people's (and my) curiosity.

[cell_man]

Interesting, I will have to try this myself. Now I have temp sensors fitted in some of the hubs it's easy enough to check and make comparisons.

BTW Andy, I have your new stator with the higher grade and thinner laminations and will get it out tomorrow. Would be interesting to see how it compares temperature wise compared to the stator you have are now using when run dry, but as this is a 10T and you are now running an 8T so not sure how that could sway the measurements.

Good stuff and food for thought, thanks

[Spicerack]

Interesting, I will have to try this myself. Now I have temp sensors fitted in some of the hubs it's easy enough to check and make comparisons.

BTW Andy, I have your new stator with the higher grade and thinner laminations and will get it out tomorrow. Would be interesting to see how it compares temperature wise compared to the stator you have are now using when run dry, but as this is a 10T and you are now running an 8T so not sure how that could sway the measurements.

Good stuff and food for thought, thanks

Ooh I'm excited! I was going to drop you a line and let you know about this thread- you beat me to it.

[John in CR]

I agreee with alot of the comments John in CR makes. I did it as a bandaid- but can also see an application for pushing cheaper motors well beyond their thermal limits- especailly geared hub motors like the mac or bmc where they are an insulated oven if you try and put too much power through them (remembering they are rated at 500-1000W, not the 3300W I was putting into mine).

We can't all afford to get a cro or don't want a 16inch scooter motor etc, which can certainly handle alot more power without causing dramas.

This type of mod is simple, very cheap to do and can elevate our lower powered hub motors into a much more fun level of power. In my case it makes my bike rideable again until I fix it. Even then, because I have time and money invested in what I have, I may as well oil cool the new stator and see what I can get out of it. As long as I monitor temp at the windings it will be fine. When I get to that point, I'll do some fairly decent dry and wet measurements just for people's (and my) curiosity.

I really like it in a geared hubbie too and want to try it in a Fusin geared hub I've had laced up for a few years. Geared hubs are so much more thermally limited since the entire motor is inside an insulating layer of air. The outside is just a shell to hold the ring gear for the planetaries and to mount the spokes.

Direct drives otoh cool themselves a lot better than people give credit. Despite having identical motors with great working ventilation, I've never bothered to ventilate the one on my daily rider. While called a scooter hubmotor, that's just because it has a longer stronger axle along with a big brake drum on one cover that ends up being an added heat source. Otherwise the motor is virtually the same as an H series Xlyte with a 40mm stator instead of 35mm, and has almost the same surface area outside to dissipate heat, same spoke flanges etc. It's not really any more efficient either, at 84% peak. When LFP was down 2 years ago we ran one of these motors at 10kw+, and that was in a 24" wheel. That motor is still good as new other than whatever bearing wear there is, and we hammered that motor for a week, so it's not just the smaller wheel that I always push people toward.

So what makes my motors survive all sealed up hauling almost a 400lb load on the highway at 55-60mph, climbing continuous grades approaching 10% at 30+ mph? It takes real power to do these things, and the motor has essentially the same surface area to cool and the same peak efficiency. I'll tell you how:
1. I know my motor's limits and will shut down before bogging down on too steep a hill.
2. The high speed winding- ie Because of the low turn count, 2, I've yet to own a controller that could put out the current to push their stators to saturation, because they can't handle the switching required by the current limiting demanded of such motors. The same current limiting also creates a much flatter efficiency curve, so my motors are more efficient during acceleration. They have to be because I zip around like a jackrabbit, definitely not taking it easy on the motors.
My big load more than makes up for the little bit of added potential power of the 1/7th longer stator than those H series motor that people have melted, so the difference has to be the turn count.

I'm that I participated in this discussion, because I never thought all the way through this analysis, and the low turn count s has to be the answer. Better cooling with oil will have limited benefit if you're pushing the motor into saturation, and I've proven that if you're not, then typically you don't need more cooling anyway.

Geared hubbies are completely different. The motor itself is capable of much more power than running inside an air insulated shell permits. With a DD the shell is part of the motor.



John

[Hillhater]

.... So does anyone know a car that has had reasonable production volume that has very small planetaries in the transmission that could be repurposed?

Audi A4 starter planetary set being repurposed with a RC motor..