Electric water pump and controller

[Yager200i]

http://eziautoparts.com.au/pdf/Davies_Craig_Electric_Water_Pumps.pdf

I think this would offer us quite a few advantages over the mechanical water pump currently used in our scooters:

1) We could use any coolant we wanted, even coolants containing silicon. Currently we can't use silicon-based coolants (which work really well with aluminum engines, as ours are) due to the problem of mechanical seal wear on the pump. The electric pump is magnetically-driven, so there are no seals to wear out.

2) Low current draw. It only takes about 15 watts maximum.

3) Decent flow. About 4 gal / min at maximum pump speed.
    {ASIDE: Does anyone know what the maximum flow rate is on the mechanical pump?}

4) Avoiding heat soak. After you shut down your hot engine, the coolant system can no longer do its job of removing excess heat. With an electric pump, the controller can be set to run for up to 2 minutes after shutdown to prevent heat soak.

5) More usable power at the rear wheel. Because the mechanical water pump has to work from idle RPM up to 8000 RPM, it's designed to provide sufficient water flow at idle, and more than sufficient at higher RPMs. So at max RPM, it's pushing more water than it needs to, which just gets recirculated in the engine if the thermostat isn't open fully. With the electric pump and controller, pump speed is set to what is needed to maintain temperatures, and nothing more. By removing the inefficient mechanical pump (with it's excess pumping power loss at high RPMs and the power loss of friction in the mechanical seals), and substituting an electric pump that takes at most about 15 watts, we have more usable power at the rear wheel.

6) Small size. 3.33" x 3.81" total. Hopefully, it'll fit right under that front cover below the front of the seat. If not, the metal tube coming from the radiator can be cut a bit, and it can be mounted on the inside of the kickstand support. The mechanical pump, pump housing and drive shaft would be taken out, and replaced by a similarly-shaped metal blank with a gasket to seal it.

7) The controller can be set up to give an indication (via LEDs mounted in the speedometer housing) of the pump speed (the brighter the LED, the faster the pump is running; or get a combination Red-Green-Blue LED, and set it up so that the LED shows blue at low temps, green at 'just right' temps, and red at maximum temp (and therefore maximum pump speed) triggered off the voltage the pump is running at), and whether the cooling fan is running (if the cooling fan LED is on, the fan is running).

I think I'll be ordering this and trying to fit it to the Yager when my wife and kids go back to Taiwan in March. A few advantages of doing it then:
1) I'll have more free time.
2) I can get dirty working on the bike without her nagging.
3) No kids picking up parts and making them disappear.
4) I can make a mess in the garage and she won't nag.
5) I'll have exclusive use of the car so the scooter can be torn down and worked on.

[Yager200i]

Big news forthcoming on this front. I'm having custom circuitry (a microcontroller) designed and built that'll allow two electric pumps and up to two cooling fans to be used (for redundancy purposes, and for extra cooling should one pump not be able to cool the engine sufficiently). It'll also have secondary temperature monitoring, on either the cylinder head, or on the tailpipe, to determine if the engine is overheating. Many other features that the Davies Craig system doesn't have that'll ensure motorcycle and scooter engines stay cool no matter what. All the temperature setpoints will be user-configurable by plugging into the microcontroller with a computer, via USB.

I've sourced most of the parts, and I'm currently trying to find a machine shop to create the metal blank that'll cover the hole left by the old pump, and which will hold the new pumps. I'm designing it such that the system will fit on pretty much any water-cooled bike (except for the blank for the old water pump, of course, since water pumps vary from bike to bike), and the parts are off-the-shelf (even the microprocessor is an off-the-shelf item, but the circuitry supporting it will be custom, as will be the program to run the microprocessor), so replacing failed components like pumps and temperature sensors is a simple drop-in matter.

The best part is, for our scooters, even with both pumps running at 100%, the whole cooling system should take somewhere around 20 watts (the 15 LPM mag-drive pumps don't take a lot of current), not including the cooling fan (which is a load that's already accounted for in sizing the generator on our bikes). Larger pumps can be used, of course, for larger bikes, up to 5 amps each at 12 volts. And up to 10 amps for each cooling fan.

So at full tilt with both 15 LPM pumps running, all of our coolant would be circulated roughly 30 times per minute, or once every two seconds... that's more than enough to keep our little engines cool. Of course, pump speed will be controlled so that they only run as fast as necessary to cool the engine.

I'll keep you all apprised of my progress. After the circuitry is built and has been tested for reliability and longevity for a year, I'll offer the plans for sale to recoup my design costs. It's costing me $2400 to have the circuitry designed and built.