Power in Flux
Likes Likes:  4
Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 24

Thread: Water cooling a 2013 Zero Motor

              
   
   
  1. #11
    Moderator Nuts & Volts's Avatar
    Join Date
    Aug 2010
    Location
    Columbus, OH
    Posts
    1,793
    Post Thanks / Like
    U can get some very low viscous oil that won't add much more resistance than the air passing the windings. Any drag increase is evened out with the lower copper losses you have with a lower temperature. I'm sure adding oil cooling with a radiator loop could increase power by 30% at least.

    This is all just perspective from my enegineering mind. Pretty much speculation haha

    Sent from my SCH-I605 using Tapatalk
    Whats under my tank may shock you!!! R6 Build, Blog/, [/URL] OSU Current webpage

  2. Likes gasfreeearth liked this post
  3. #12
    Member Jay64's Avatar
    Join Date
    Jul 2013
    Location
    St. Petersburg, FL.
    Posts
    46
    Post Thanks / Like
    From my gas race testing, viscous drag robs a LOT of power. I'm sure you could get more power out than drag introduced so overall it would be a better scenario, but I would have to disagree that it wouldn't add too much. But then again, there is only one way to find out for sure.

  4. #13
    Member
    Join Date
    Feb 2013
    Location
    Sweden
    Posts
    48
    Post Thanks / Like
    I guess it depends on how smooth the surface of the rotor is. If it's just like a turned solid cylinder, it may be ok with some fluid in there.

    What kind of oil has only slightly higher viscosity than air?

  5. #14
    Moderator Nuts & Volts's Avatar
    Join Date
    Aug 2010
    Location
    Columbus, OH
    Posts
    1,793
    Post Thanks / Like
    Unfortunately I have no numbers or examples of fluids. Just haven't had the time to investigate this much.

    Two things I know for sure. The Remy motors used by a ton of teams/vehicles has a design in which the entire rotor (7-8" dia) is flooded in oil for cooling. This motor is 94-95% eff peak and spins to 10000 RPM no problem. If viscous effects were that bad it would be escalated at that extreme speed. The second thing is that frictional losses make up a small percentage of overall motor losses. So even doubling viscous drag won't add that much. Refraining from doing calculations without researched numbers.

    Sent from my SCH-I605 using Tapatalk
    Whats under my tank may shock you!!! R6 Build, Blog/, [/URL] OSU Current webpage

  6. #15
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    ATF is what you want to fill electric motors with. It will add some drag to any motor of course, some designs are mechanically more accepting of it than others. Fortunately, it's not a like the rotor has a bunch of spikey teeth hanging off it twirling in much heavier gearbox oils... so it's not nearly as much drag as some other devices gas bikes turn if you've got that as your comparison metric for expected windage loss.

    The Rth on electric motors is often very high between the teeth to the case shedding the heat. I know of many hubmotors I've roached that were only a bit above ambient on the outside case while the inside 260degC wire insulation is reeking and boiling inside as an extreme example. In a setup like that, filling the little hubmotor with a bit of oil can easily give it >2x continuous power at the cost of maybe 1amp additional no-load current. If the oil filling would let a smaller motor (with less core loss) be used, than it would easily be possible to reduce motor size, oil fill, and come out ahead on efficiency over simply running a physically larger motor.

    IMHO, absolutely worth trying on any motor reasonably easy to get sealed up well enough to give it a shot. If I remember correctly, the MissonR race bike used circulating oil flowing through the rotor and out to a cooler, in addition to a water jacketed stator outside case. That doesn't seem too hard to do, and done right should be pretty hard to beat cooling for any motor.

  7. #16
    Senior Member
    Join Date
    Oct 2010
    Location
    North America
    Posts
    758
    Post Thanks / Like
    I worked on a project years ago where we looked into this idea of running the rotor submerged in liquid coolant. I found some notes I still have in my file. We tested using an inverter driven induction motor. The rotor was machined smooth on all surfaces. It was approximately 5 inch diameter by 3 inch long. The whole motor was tested in air and then submerged in a tank of ATF. It was an open frame motor so the oil easily flooded the motor. The additional power due to the oil was:

    0.5 kW @ 1800 RPM cold,
    2.4 kW @ 4000 RPM cold, 1.5 kW with hot oil,
    3.3 kW @ 5000 RPM cold, 2.2 kW with hot oil,
    4.6 kW @ 6000 RPM cold, 3.3 kW with hot oil,
    7.0 kW @ 7000 RPM cold,
    9.8 kW @ 8000 RPM cold.

    Cold means room temperature and I don't recall the hot temperature but was likely close to boiling water. Clearly there is a significant power required just to rotate the smooth cylinder in oil. Better cooling methods exist. For instance, the Remy motor is not flooded. The oil is passed through the shaft and coats/splashes the rotor and then the inside of the stator.

    BTW, if I remember correctly, the test motor was rated at 7.5 hp, 1800 RPM in free air.

  8. #17
    Senior Member
    Join Date
    Oct 2010
    Location
    North America
    Posts
    758
    Post Thanks / Like
    I also was introduced to Boundary Layer Turbo Fluid Dynamics. That is a blast from the past. But it was interesting although I didn't pursue it. But the basis is that the fluid element is acted upon by the boundary layer adhesion which accelerates the fluid to the same surface speed as the rotating disk at a given radii, while the centrifugal force moves the fluid particle radially. I should give credit to Mr. John Kuczaj for that sentence. There are actually turbines and pumps made with smooth discs. See Turbonique.

  9. #18
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    Quote Originally Posted by lugnut View Post
    I worked on a project years ago where we looked into this idea of running the rotor submerged in liquid coolant. I found some notes I still have in my file. We tested using an inverter driven induction motor. The rotor was machined smooth on all surfaces. It was approximately 5 inch diameter by 3 inch long. The whole motor was tested in air and then submerged in a tank of ATF. It was an open frame motor so the oil easily flooded the motor. The additional power due to the oil was:

    0.5 kW @ 1800 RPM cold,
    2.4 kW @ 4000 RPM cold, 1.5 kW with hot oil,
    3.3 kW @ 5000 RPM cold, 2.2 kW with hot oil,
    4.6 kW @ 6000 RPM cold, 3.3 kW with hot oil,
    7.0 kW @ 7000 RPM cold,
    9.8 kW @ 8000 RPM cold.

    Cold means room temperature and I don't recall the hot temperature but was likely close to boiling water. Clearly there is a significant power required just to rotate the smooth cylinder in oil. Better cooling methods exist. For instance, the Remy motor is not flooded. The oil is passed through the shaft and coats/splashes the rotor and then the inside of the stator.

    BTW, if I remember correctly, the test motor was rated at 7.5 hp, 1800 RPM in free air.

    Your values are around an order of magnitude higher windage loss than my own experience with a similar sized rotor at similar speeds in ATF. One substantial difference however is my tests didn't involve the rotor being more than about 1/3rd submerged in the oil, and your test was full submersion if I understand correctly?

    What's odd about your result, is that you've got more windage drag on a tiny smooth rotor than the entire engine windage loss of my honda insight gas internal combustion engine with a ~18" long crankshaft counterweights and rod ends dipping into the oil pan slinging it everywhere, and that's motor oil with drastically higher drag than ATF.

    In the many documented bicycle hubmotor tests, the no-load current increase of the oil filled motors was quite low.

  10. Likes Skeezmour liked this post
  11. #19
    Senior Member
    Join Date
    Oct 2010
    Location
    North America
    Posts
    758
    Post Thanks / Like
    Quote Originally Posted by liveforphysics View Post
    ... and your test was full submersion if I understand correctly?
    Yes, the test was done completely submerged. That is what I understood "flooded" to infer. Partially submerged or splashed makes a big difference in drag. However can be nearly as effective in cooling. As for the bike hub motor; the RPM is substantially lower. And was it completely flooded? I wouldn't see the need to.

  12. #20
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    Quote Originally Posted by lugnut View Post
    Yes, the test was done completely submerged. That is what I understood "flooded" to infer. Partially submerged or splashed makes a big difference in drag. However can be nearly as effective in cooling. As for the bike hub motor; the RPM is substantially lower. And was it completely flooded? I wouldn't see the need to.
    In the hubmotor applications, the oil level gets filled to about 1-2" below the axle bearing. This eliminates the need for a special bearing seal (just RTV sealing the side plates). It works wonders, easily 2x continuous power in exchange for ~1-2amp higher no-load.

  13. Likes Athlon liked this post
Page 2 of 3 FirstFirst 123 LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •