Good! Someone finally oil cooled a hubbie!! HS3540

[bigmoose]

In this thread http://endless-sphere.com/forums/viewtopic.php?f=3&t=27328&p=396680& I stated:

...also I wish someone would seal the side covers and fill one of these hubs 1/8 to 1/6 full of Dow Corning Heat transfer fluid, SYLTHERM XLT or SYLTHERM HF. Then we can add convective fins to the HubOD between the spokes to convect the heat out. Again, not a lot of oil, but a moving mist inside the housing.

Well dave b on EV Geek Forum NZ DID IT and reported his results here:

http://www.ev.geek.nz/viewtopic.php?f=32&t=162&start=30

With gaskets already fitted to the motor side covers to prevent water ingress, I was already in a good position to add 170 mls of INOX MX3 lubricant http://www.inox-mx3.com/inox.htm. This is in order to provide a heat transfer medium to dissipate heat from the windings to the outside of the motor. INOX MX3 has a viscosity of 10.04 centistokes @ 40oC and 2.54 centistokes @ 100oC. In comparison, water has a viscosity of 1 centistoke @ 20oC. The additon of this oil has added no detectable drag when pedalling alone (no load current @ full hub motor speed has increased by 200 to 300 milliamps), but it has significantly solved my heating problem. I can now get all the way up my hill and the windings only reach 65oC. The outside of the motor now gets hot which is good because airflow over it can now absorb the heat generated by the windings. I found this to be so when cruising along the flat into a head wind. Normally the motor windings would reach 70oC when doing 35kph into a 15kph headwind (55 to 60oC at 35kph with no head wind). But with the oil, more wind is of course resulting in more load, but it now reduces the temperature due to the increased air flow. For example, cruising on the flat around 35kph with the oil added now results in a temp of 38 to 42oC (instead of 55 to 60oC). When I went into a 15kph headwind, this temp actually dropped 2oC rather than increased. The pic below shows my setup. The small tube is a tiny breather to allow volatiles to escape keeping the motor from pressurising. The gaskets have been more than adequate for preventing oil leakage. The oil also seems to be fairly stable regarding drying out and going gooey. Time will tell as to how well INOX continues to perform.

Whooppee I say! Hat's off to the NZ men. Now let's get this cooling method going on ES... it's about time! It should have been an innovation out of our board, but I digress.

Note that he added a 0.5 mm gasket to each side cover is all. (Edit from below: and a breather tube) This is foundational! It will also end the rusting problems within the motor. It will take AJ's Frocks to a higher level! MarkCycle should consider these results also.

[Alan B]

Plus he added a little breather tube.

What is a good source of this oil? Rough cost?

Any hazards associated?

Any downsides?

[AussieJester]

I thought HAl already watercooled a frock long ago BiGMooSE? I remember seeing one on his bike
with hoses coming out of it..?

Mineral oil would do the trick Alan B , tiz what the PC geeks use with submersed pc builds... Can only see plusses for it, now summer is here in Oz i been thinking more about enclosing my controller and fitting a pump and rad to the bike...

BiGMooSE....oil cooled or not though, frocks still a frock and no more impressive ...to me

KiM

[TylerDurden]

Actually, it was done a while back by an ES member, but mostly for lubrication... no temperature data was reported.

viewtopic.php?f=6&t=7454&p=148040&l#p148040

[bigmoose]

AJ I did see Hal's water cooled manifold. A work of art.

Tyler I did not see that thread by Reed. He was way ahead of his time and 95% there! He even specified what I would recommend for a DIY fluid, synthetic ATF, the thinnest that you can find by shaking the can or using a Zahn cup. He had the right level also, 1/4 to 1/3 full. Kudos to him!

I hope we can get a little buzz going and some more folks interested in this concept. It will advance hub development I think.

[John in CR]

I'm pretty sure MWKeefer experimented with oil bath cooling several years ago.

Did that guy have one of the sensorless motors, because that's my problem with the approach. ie I don't want the turbulent flow of a liquid coming near my hall sensor wires, since it would seem to guarantee a failure.

[Harold in CR]

There is an Engineering outfit, in Pa. that sells an 8 KW hubmotor that is oil cooled. Zehrbach or something like that.

You have to be special schooled to be able to buy them

[markcycle]

In this thread http://endless-sphere.com/forums/viewtopic.php?f=3&t=27328&p=396680& I stated:

...also I wish someone would seal the side covers and fill one of these hubs 1/8 to 1/6 full of Dow Corning Heat It will also end the rusting problems within the motor. It will take AJ's Frocks to a higher level! MarkCycle should consider these results also.

Yes we have had our own flooded liquid cooling program testing begins testing next week. All feed through are solid copper in epoxy that transition back to stranded wire before exiting the motor. the seal lip ride on a outer aluminum sleeve not epoxy as the picture appears.

since this picture was taken the motor has been assembled, we are waiting on the pump and radiator which should be in this week.

I'll talk more about it in the Motorcycle hub motor thread http://endless-sphere.com/forums/viewtopic.php?f=10&t=20931&start=75 once we get some data.

[amberwolf]

I'm tempted to try it on the "36V" Fusin geared motor on DayGlo Avenger, whcih seriously gets overheated at 48V (even at low throttle amounts) pulling a trailer full of dog, but I don't currently have a way of monitoring the temperatures and logging them, to collect data on before and after.

No real reason to try it on the CrazyBike2 2807, as it doesnt' really seem to have any heating problems (with or without ventilated covers) in normal use. It probably would on the racetrack, though, if I could stop crashing before I get more than a lap or two around.

[hydro-one]

if your going though all the trouble to liquid nitrogen cool, may as well use a high temperature superconducting coil too

im rooting to discover one that works at room temp. solve all our problems.

mike

[Arlo1]

There is some good reading there.

[Green Machine]

Really great stuff..enjoyed watching the video that AJ posted with the oil submerged PC case.

[bigmoose]

if your going though all the trouble to liquid nitrogen cool, may as well use a high temperature superconducting coil too ...mike

mike our lab, in the development of a superconducting motor to run in LH2, used ultra pure copper in LN2 as a step along the way. The UP-Cu at LN2 temperatures was amazingly successful as the Cu resistance at LN2 temperature is about 14% of room temperature, IIRC. So you gain to the first order 86% of superconducting performance by cooling the UP-Cu to LN2 temps.

[Alan B]

We use superconducting magnets, and the leads need to come out to external room temperature power supplies. For the 4K to 80K portion of the path we use high temperature superconducting leads. These are tubes full of powder. The high temperature superconductors, at least a few years ago, were not really suitable to make a coil. Unless perhaps a turn packed into a tube would work. Vibration is a problem, and there are still limits to the current before the field intensity prevents superconductivity.

[The Mighty Volt]

What effect would carbon dioxide have if it was introduced to the hub-internals under pressure, say from one of those little cartridges used to pump up wheels??

[Alan B]

CO2 cooling has been suggested and discussed. Luke even tried it at the race in So Cal, by spraying a CO2 fire extinguisher on the exterior of his motor.

It can work but will take a lot more than a 12 gram cartridge. Think 20 pound cylinders. With the right injection into the motor, and a good path out for the expanded gas, it could do significant cooling in the pits between races. Carrying the stuff for cooling is not so easy - a cylinder carrying 20 pounds of CO2 weighs about three times that total. So it is pretty heavy to carry.

[gestalt]

No, those small cartridges would do nothing. And spraying the co2 on the exterior is a waste. But I do think that a paintball tank would have a good deal of cooling power. Think of it like this, almost in to the end of the race and your motor is running too hot and you're still behind and can't dump more current in or you'll overheat. Flip the switch and liquid co2 gets piped right into the motor casing and washes over the coils due to centrifugal force and how you have the vents set up. Now you have how ever long that tank can cool to gun it and pass that motherfvcker. Ahh the sweet taste of victory.

[hydro-one]

Bigmoose.... Ive heard of ultrapure copper.....should be alot easier to form into coils. But isnt "14% of the resistance at STP of copper", is still alot of resistance compared to a real superconductor?? Ive seen that YCBO ribbons are for sale for use in windings. here http://www.amsc.com/solutions-products/hts_wire.html

So would superconducting coils cooled with liguid nitrogen benifit say an astro? I cant see it helping controllers much although a superconducting controller is probably a good idea by that time......

edit : Thanks LPF Bigmoose is right!! I see now we are comparing copper to superconductor (zero resistance). so we are getting 86% of the way there. What im saying is thats still alot of resistance compared to nothing. eg heat will be created and ruin the effect???? but state of the art i guess would dictate,,,, ultrapure copper is the way to go , still has excellent resistance at room temp, pour on the liquid n2 and your ready for the track. how can i make ultrapure copper??????

ULTRAPURE COPPER WIndings with N2 injection.

[liveforphysics]

Bigmoose.... Ive heard of ultrapure copper.....should be alot easier to form into coils. But isnt "14% of the resistance at STP of copper", is still alot of resistance compared to a real superconductor?? Ive seen that YCBO ribbons are for sale for use in windings. here http://www.amsc.com/solutions-products/hts_wire.html

So would superconducting coils cooled with liguid nitrogen benifit say an astro? I cant see it helping controllers much although a superconducting controller is probably a good idea by that time......

He said pretty clearly, that ultra pure copper at liquid nitrogen temperature has 14% the resistance of pure copper at room temp, getting you 86% of the advantage superconductor windings.

If I'm not mistaken, true superconductor windings would not even be conductive at all until they are in liquid helium or nitrogen, so the motor would be useless if you ran out.

[John in CR]

Regarding the original topic, oil cooling what is essentially an outrunner, something that's been nagging at me is the thought that an oil bath on a sealed hubbie doesn't accomplish anything more than create a more direct heat path from the stator to the magnets. We're still left with the same exterior surface area to transfer heat to the environment, so how can there be much gain from a continuous operation standpoint?

[miuan]

Regarding the original topic, oil cooling what is essentially an outrunner, something that's been nagging at me is the thought that an oil bath on a sealed hubbie doesn't accomplish anything more than create a more direct heat path from the stator to the magnets. We're still left with the same exterior surface area to transfer heat to the environment, so how can there be much gain from a continuous operation standpoint?

Since the stator and case have much better mutual thermal conduction, there will be less temp difference between them, resulting in hotter case and cooler windings than in air cooled setup. As a result, the case will dissipate more heat and stator will get even cooler. But, there is more heat generated due to liquid drag.
We can argue that a drilled cover will do better to cool the windings and is lighter for two reasons, but a sealed hub is better.

[Alan B]

The oil reduces the thermal impedance to the heatsink (case). The resulting thermal equilibrium will result in lower winding temperatures.

[John in CR]

The oil reduces the thermal impedance to the heatsink (case). The resulting thermal equilibrium will result in lower winding temperatures.

At what expense?...roasting magnets, which have a far lower temperature limit. Passing higher temps more directly through our magnets seems like a fail to me. Windings can handle plenty of heat, and too hot in the copper means you're geared wrong anyway. Motors we see on the forum with blackened copper are typically due to cheap magnet wire or just not paying attention or being the least bit familiar with your motor. I've heated my motors up to where performance is severely downgraded for a short time (I back off when I notice it of course), and the copper still looks like it did when new. Ask LFP how far he pushed the one in a 24" with a controller pushing 120A battery side when he was down. He had the stator hot enough to expand and make rubbing noises.

Oil baths have a place, just not involving magnets and hall sensors. There's a lot less thermal impedance inside hubmotors due to the turbulent air and small spacing than people give credit anyway. People just run their motors improperly, and don't take the time to understand their system's limits. It's not luck or some special motor that enables 250lb me + 125lb bike to zip around like I do climbing mountains and racing cars at 7000W peak controller settings for 3 years and never cooking a sealed motor and the one I ventilated barely gets warm. My bikes pull one of the heaviest loads on the forum in daily almost total car replacement use, yet I have no temperature sensor other than the pads on my fingers that I used frequently early on to learn my system and pretty rarely now. Of course I manage to ride my bikes with no in-use battery monitoring without killing a single cell, and I ride aggressively in traffic with no close calls or feeling like the target of driver mistakes.

John

[Alan B]

Your opinion, John. Need a finite element analysis to actually know for sure if the magnet temperature is increased by lowering thermal impedance. It is not clear at all. One thing that lowering thermal impedance does is to allow windings to operate at a lower temperature which lowers their resistance which lowers heat production in the first place.

Very many folks have burned coils, very few have demagnetized magnets, according to the reports we see on the sphere.

[zombiess]

The oil reduces the thermal impedance to the heatsink (case). The resulting thermal equilibrium will result in lower winding temperatures.

At what expense?...roasting magnets, which have a far lower temperature limit. Passing higher temps more directly through our magnets seems like a fail to me. Windings can handle plenty of heat, and too hot in the copper means you're geared wrong anyway. Motors we see on the forum with blackened copper are typically due to cheap magnet wire or just not paying attention or being the least bit familiar with your motor. I've heated my motors up to where performance is severely downgraded for a short time (I back off when I notice it of course), and the copper still looks like it did when new. Ask LFP how far he pushed the one in a 24" with a controller pushing 120A battery side when he was down. He had the stator hot enough to expand and make rubbing noises.

Oil baths have a place, just not involving magnets and hall sensors. There's a lot less thermal impedance inside hubmotors due to the turbulent air and small spacing than people give credit anyway. People just run their motors improperly, and don't take the time to understand their system's limits. It's not luck or some special motor that enables 250lb me + 125lb bike to zip around like I do climbing mountains and racing cars at 7000W peak controller settings for 3 years and never cooking a sealed motor and the one I ventilated barely gets warm. My bikes pull one of the heaviest loads on the forum in daily almost total car replacement use, yet I have no temperature sensor other than the pads on my fingers that I used frequently early on to learn my system and pretty rarely now. Of course I manage to ride my bikes with no in-use battery monitoring without killing a single cell, and I ride aggressively in traffic with no close calls or feeling like the target of driver mistakes.

John

You make some great points John, makes me wonder what the difference is and when I first read this I didn't consider that higher temps could now be transferred to the magnets in a really hot scenario since everything will try to reach equilibrium with the oil temp.

Just like you, I have been putting way more power into my little 9C than it was ever supposed to see (~7KW) but with cooling holes drilled, small diameter wheel and monitoring, the coils look brand new. I have noticed that anything over about 110C winding temp is when I noticed the performance really start to go down and I can start smelling the motor at 120C. Hottest I've ever gotten it though was 135C after letting a 220lb friend abuse it for a few mins after my 40mph 4 mile commute to his place and he was playing with regen a lot at 105% throttle LOL.

I'd definitely like to see some more studying done on this with comparisons to air cooling as as a reference point.