My new water injection plans
 

I have ran water injection on previous cars and the big issue was always keeping the water flow metered correctly. I was browsing around looking at unrelated stuff and came across this pump. It should be here today so I will try hooking something up this weekend.

Peristaltic Liquid Pump

It is 12V and 300mA max current. Flow rate: up to 100 mL/min.

It has a low enough flow rate that I am hoping I can just run it in parallel with the tbi injector in the car and have it pretty close to the right gas/water ratio. My plans are to hook this pump up and plumb it into the throttle body in an unused vacuum port. I will put a gallon of water in the car with 1/10 gallon divisions so I can adjust it by putting a resistor in series with the motor until I get the ratio of 10 gallons of gas to 1 gallon of water.

If it works out the injector PWM signal should give me proportional water flow and after getting the speed right I can just change over to a larger water tank and have a pretty reliable system that will finally meter the water properly. It would be a pretty easy system to duplicate and get working on just about any car.

Just curious, are you attempting to introduce a second fuel to your engine similar the how propane is used in diesel engines?

It is an old idea that is used sometimes on engines to suppress knock. It has been generally shown to improve mileage a small amount. With water injection you can increase the compression ratio without knocking and advance the timing a little bit. I am of opinion that using excess heat to turn the water into steam thereby increasing the cylinder pressure is where the gain in mileage comes from.

I ran it before on my first Metro and I gained like 3-4mpg when it was working properly. The problem with the setups I tried and the commonly used fish tank gang valve setup is that it never stays at the proper ratio unless you fiddle with it before every drive. That is a huge hassle and not worth it. I always wanted to use a fuel injector setup sized at 10% the flow of the stock injector and just let the computer manage it. But I don't think a fuel injector would live long running straight tap water.

The nice thing about that pump listed in the first post is it can pump any fluid without having to interact with the fluid so it won't care about pumping water for a long time. I just have to get the speed of it right and I should have a more or less exact ratio all the time. The pump came in today but my variable power resistor didn't show up so I won't get to hook it up and test it till Monday most likely.

Verry intersting. I will follow this thread :)

I have a test setup in the car. I used some clear hose going from a water bottle in the cup holder to the pump then to a vacuum port on the engine. I wired it to the injector with no resistor on the wire just to get a feel for the max flow of the pump.

I started the car and the motor just jumped back and forth. The inductive kick from the injector was high enough to keep it from doing anything. I found a random diode I had laying around and put it inline with the motor to get rid of the reverse current. I didn't have a meter or a wiring diagram with me so I have no idea yet if I have it wired in the right direction but at idle it was pumping a very tiny amount of water so it was good enough for a quick test.

I will get a meter and make sure the power is going the right way to the motor and the diode is facing the right way. Then I can dial in the system but it looks like it should be no problem to get it working very well.

So far the total parts list is that pump listed on the first post and a 25W 200 Ohm Round Ceramic Wirewound Potentiometer and a diode. The diode needs at least a half amp current capacity or so. I will leave it hooked up and drive to work tomorrow and see how much water it uses. Then I will put the variable resistor in and try and dial it in.

I would take a picture but right now it is just a ball of duct tape on my battery handle and a bit of hose so it doesn't look like much.

Interesting... have not heard of it improving economy for gasoline engines (diesel engines, however, there's lots of testing supporting it)... I suppose the gains came after you advanced the timing?

That kit is made for turbo setups that are trying to get more power. That isn't really applicable to the setup I want to try. I want to meter tap water, not washer fluid with Methanol in it. Plain tap water will plug up a lot of pumps and injector setups over time. I won't advance the timing or do anything to the motor once I get this setup dialed in and just do a flat comparison of water vs no water before trying to tune anything. I still think the primary reason for the gain in mileage is the increased cylinder pressure so not changing anything else will be the easiest way to test that theory out.

With my test setup the motor does not turn on the pump when driven directly from the injector signal. The pulse just isn't wide enough to get it spinning with the pump on the end of it. I will experiment with a few things I can do to get it working a bit better before I resort to using a microcontroller to regulate the flow of water. I might have to find a smaller pump anyway. If I wire this pump to power and run it as slow as possible with the variable resistor it can empty a 16oz bottle in probably 10 min. That is going to be way faster than I can deal with.

I will have some time to experiment with it this weekend, I will hook it to a pwm controller and see how slow it actually can run. If it won't go slow enough I will get a much smaller pump and try again. The good thing about these pumps is they are pretty cheap so experimenting won't cost much.

what is the nozzle look like? wonder how much pressure it can pump.
i tried water injection on my 81 VW 1.6 diesel. it was crude. the engine kept running, no matter how much water i pumped in it. i got scared i would destroy it so i took it off when water came out the tailpipe.