The best info I could find with a comparison showed roller bearings having just over double the friction of ball bearings, but handle higher loads. Based on the info Toolman2 has shared just the lubrication of the bearings can make a lot more difference than that.
I love the idea for offering a means to both beef up and axle and/or bring larger wiring out of the motor. I also like the concept of liquid cooling, but without the motor being made specifically for liquid cooling, there are good reasons no one has pulled it off successfully. The reasons I see are:
-The much lower temperature differentials involved compared to ICE's.
-The difficulty in getting a short thermal pathway from the sources of heat to the cooling fluid.
-The far lower temperature limits of the components inside our electric motors.
It's nothing like cooling an ICE where cooling liquid passes close to a multi thousand degree heat source with only metal between the source and the cooling liquid to relatively quickly transfer heat away at ambient. It's also nothing like liquid cooling a PC with fractions of 1kw of heat to reject and a relatively large amount of liquid, because to even consider it you're looking at quite a few multi-kw of heat to reject at similar temp limits.
The exception would be having the cooling liquid freely flowing over the stator and piping it out to a radiator. Get the halls or other timing moved outside the motor and figure out how to prevent leaks and count me in. One is beyond my ability and the other nobody has solved despite quite a few attempts with no pipes or radiator.
Everyone seems to forget that no matter what method you use you're still dissipating the heat to the outside air. That last step of heat transfer is based on the surface area, the temperature differential, and the coefficient of convective heat transfer (a quite complex number based on many factors but a good turbulent flow inside our motors can maximize it, though only reasonable guesstimates could reasonably made without lots of testing).
Our motors have quite a bit of surface area at or near the heat source, and it's relatively easy to flow significant ambient air through the motor. With a mid-drive arrangement it's also relatively easy to both protect the motor from ingesting problematic gunk, and increase flow at the same time with a weight and space penalty far less than a much less effective liquid cooling system. Have you seen the numbers Toolman2 has posted WRT the leaf blower fan he's used?
Figure out a way to flow the cooling liquid directly over the stator without problems and I say go for it. Otherwise unless the motor is designed specifically for liquid cooling, so you just have to do some piping and pumping to the radiator, then forced air cooling is sure to prove both easier and more effective. All the heat goes to the environment anyway, so just short cut it all and bring the environment into the motor to dissipate the heat directly at the maximum temperature differential possible.
John
