DIY electric water pump - on the cheap

[Daox]

Thanks Ben, and yep, printed out on the UP.

Nimblemotors, what were you going to use for a pump?

[nimblemotors]

Originally Posted by Daox

Thanks Ben, and yep, printed out on the UP.

Nimblemotors, what were you going to use for a pump?

I found it challenging to find a pump that can handle high temps.
I have a prius pump used for the inverter, low volume might still work,
but not high pressure.
Could just put electric motor on an existing car water pump

[Daox]

These are the same issues I ran into. Thus the decision to reuse the OEM pump and put my own motor on it.

[Daox]

Here is the model for the water pump pulley. It is currently has 14 teeth, so it is a 4:1 ratio with my 56 tooth motor pulley. This should allow me to ramp up the water pump speed to 5000 rpm which is what the water pump would be spinning at if the engine were at 5500 rpm on the stock setup.

The current plan is to use this pulley for testing. I want to measure amp draw vs water pump rpm. It is quite possible that all of this nonsense is for naught or very little. As I said a while back, this mod really isn't a first mod kind of thing. It works well with other mods like an alternator delete, block heater (to cycle the coolant with the engine off), and possibly even preheat the transmission. In and of itself, its probably not that great. I anticipate this test will show us that effect or not.






To address the heat issue with the plastic pulley, I plan on replacing the four bolts on the water pump with threaded rod. These will completely detach and space the new pulley away from the OEM pulley. This will transmit much less heat to the pulley. I'll also probably use some nylon washers to reduce that heat transmission further. As it spins, the threaded rod will act as a fan and cool itself down. This is a lot simpler than the previous idea.

[2000mc]

I wouldn't think that there would be an advantage to going more than a 3000ish engine rpm equivalent. Poking around trying to find evidence to support this theory I pulled out of thin air, it seems that most electric water pumps don't even flow half of what their mechanical counterpart would flow at redline. Also craigdavies site says it takes 8x more power to spin a mechanical pump 2x faster.

[2000mc]

On pulley cooling, if there is enough space for a reasonably sized hole through the center of the pulley, you could add vanes to the back side of your pulley, to try to make a tiny squirrel cage fan, to move air through the space between the pulleys

[Daox]

Originally Posted by 2000mc

I wouldn't think that there would be an advantage to going more than a 3000ish engine rpm equivalent. Poking around trying to find evidence to support this theory I pulled out of thin air, it seems that most electric water pumps don't even flow half of what their mechanical counterpart would flow at redline. Also craigdavies site says it takes 8x more power to spin a mechanical pump 2x faster.

I completely agree. I just want to test to show the power requirements at different rpms. That'll give us a much better understanding of what is actually going on.

Once we have that information, I'll probably print another pulley that gives me the max rpms I want at closer to max motor rpms.

Poking a hole in the middle of the pulley sounds like a great idea. I shall definitely do that.

[Daox]

This weekend I had some time to remove the water pump pulley from the Metro and measure it up. It looks like it'll be a little tighter than I had initially thought, but still plenty of room.

Here are some pics:


This is the clearance between the water pump and the frame.





This is the OEM water pump pulley removed. WOW the water pump is ridiculously easy to spin by hand. It spins and coasts. It'll be interesting to see how much power the pump actually draws once its spinning a few thousand rpm.





Here we have the bolt centers measured up so I can update my pulley design with the correct spacing. I'll also be incorporating 2000mc's idea of adding a hole in the center to aid in cooling.

[Daox]

With the OEM pulley measurements, I can finalize the design of the 3d printed test pulley for the water pump. Its been lightened up, the center hole has been added, and a lot has been tweaked here and there to make it fit nicer and be a bit thinner.

[California98Civic]

I lke the newer design. It'll cool itself better. Surprising the water pump turns so easily by hand. Suggests that your experiement might show relatively little to be gained? How much load could an easily turned pump and its pulley be putting on the engine? Maybe the OEM pulley and its etra belt line puts as much load on the engine as the pump itself?