Keyne's DH Comp Build - now with liquid cooled HX35! :)

[binlagin]

I'd say we are somewhere between the fighter and the bomber!

[crusoe]

Hey Keyne, how goes the

...roll-able scissor-lift stand for the bike to make it easier to work on

I need something similar!

[keyne]

The scissor lift works ok, but I might redesign it some day, it's just not wide enough to be completely stable.

I've been gearing up to do some efficiency tests to see how much more range I get because of the cooled motor - it could be quite a bit!

[Stielz]

Really nice build Keyne, keep up the good work!

How does your controller hold up to a bit of thrashing?

Just an idea, have you thought about liquid cooling your controller also? You've got all the hardware there, all you would need to do is make a channel through the FET heat sink and run the cooling liquid through in series with the motor. Then you could up the current limit and possibly push a few more kW through your system

[keyne]

The controller temperature has never gone over about 55C, but yes, liquid cooling my controller is on my list, mainly so it doesn't have to be an external component on the bike.

[megacycle]

I was thinking to put a copper jacket over the fins on the side of the controller were the fets are bolted, the fins then making the coolant channels ( no pulling the controller apart )or burst fire freeze spray with a hydrocarbon propellant and piezo igniter for afterburn and faster heat draw, for buyouts .

[liveforphysics]

Nice job! I've seen a couple water cooling hubmotor attempts before this one that looked like they would be less effective than simply drilling the covers or adding oil, but this one is exceptional in how you bonded the outer layer of water jacketing in a location that can absorb some stator heat effectively.

My favorite way to know if a cooling system is working is to simply watch the delta between water into the radiator and water out of the radiator. If you know the flow rate, and you know the delta, you know it's ~4.2J/gram of water raised 1 deg C. A Joule is a watt-second of energy.

So, just as an example calculation if you measure your flow rate at say 100L/hr through the cooling loop, and you observe a peak delta-T you can hold across the radiator continuously of say 10degC difference between inlet and outlet temps, then you know the rate your system can shed heat.

Crude calculation example would be something like 4.2KJ/Kg*100L*10degC = 4200KJ/hr of heat transfer, which is ~1.1Kwh of thermal energy transfer per hour capability. If you were operating the motor at a point where it was say 80% efficient, this would enable it to handle 5-6kw of input power continuously. That would make it better than any air or oil cooling alone setups to the best of my knowledge, but of course these are just make-up guess numbers just to show how to do the math. For the numbers to mean something you will need to take real data from your setup of course.

Great job!

[keyne]

Thanks Luke,

It's been a while since I've posted here, also a while since I've ridden the bike. (Damn PhD getting in the way )

The best cooling numbers I saw were about 700W (8deg drop, 1.2L flow). Done by measuring temp difference between entry and exit of the motor (to include heat loss from the tubing). This was however before adding the themal epoxy to the windings and the motor definitely cools even better now (I was seeing up to 25degC delta T between coolant and windings, this has dropped to 15degC max), I just haven't formally measured it since the change - I should probably do that. I wouldn't be surprised if I'm getting 1kW of cooling now.

I doubt my bike ever gets ridden for long at the 80% efficiency point, but then on my trails there's no way to keep 6kW on for long either!

Oh yeah, just to make it go faster I've added blue flames to the side panels