I did some testing of the "new black" 150PSI ShurFlo pump that everyone seems to be selling for a water/alky pump. The test setup was one water pump, one brass check valve (22 PSI open), one brass TEE, and one or dual nozzle setup. All hose was poly .17ID, 1/4" OD plastic.
Power was supplied by a car battery through a GM coolant fan relay.


All testing was done at 12 volts (measured at pump).

1) The pump doesn't pump constantly (it surges)
Seems like it pumps until it hits 150PSI and then backs of a bit and ramps up again.
2) Tested a 2 GPH @ 100PSI (M2) nozzle and it flowed 2.8 GPH at 150PSI
3) Installed a dual setup with one 2 GPH (M2) and one 3 GPH (M3) at 100PSI nozzles. They flowed an actual 5.3 GPH on the test setup.

Notice that the GPH dropped a lot with the dual setup. It should have flowed approx. (2.8 + 3.9 = 6.7 GPH), but actually only flowed 5.3 GPH. I ran the test three times and each time they flowed 5.3 GPH. The interesting part was that the 2 GPH nozzled flowed 230 mL and the 3 GPH flowed 250 mL in the test. I checked everything for blockages. There was nothing in any of the lines or nozzles. The 3 GPH should have flowed way more. In the end, the 2 GPH and 3 GPH dual setup only flowed about 330mL/min.
It seems like SHurFlo rate the PSI at VERY LOW flow rates. The PSI must be dropping way off with the dual nozzle setup.

I also did a code change to the AUJP 730 ECM to toggle the relay on and off at a 50% duty cycle. It lowered the flow by 33%. The pump pressure seemed MUCH smoother with the pulsing of the power.

Do the 100PSI ShurFlo pumps sort of pulse and not run smooth/constant?
Anyone else see the strange on/off pump motor thing with the 150PSI pump. I read one post about someone saying it does this. Mine does.

 

It looks like the DevilsOwn person found out the same pressure vs. flow stuff I did. Also, posted info about the fluctuating PSI.
Info is located:

HERE WE GO AGAIN (bogus info) - SRTforums.com

End result, buy nozzles bigger than you think you need for your pumps "rated" PSI. I am ordering two 5 GPH for my setup.

 

what style of nozzle are you planning on using, and what flow rate are you shooting for?
How are you figuring how much water/alky flow you need?

I was toying with the idea of rigging up something like this in the future, basically running a pump full out, with a variable regulator in line basically, so it'll just recirculate when not "on", then when "on", raise the PSI flowing into the motor linearly with boost sorta thing. That kinda your plan as well?
I figured mcmaster carr misting nozzles, perhaps 2, would work. Raise pressure to increase volume of liquid going through, but I had no idea how much GPH would be "right" for the application.

 

I got some parts from features. The pump is the ShurFlo 150 PSI black unit ($80 ebay), Poly .17ID hose ($6 ebay), 2GPH @ 100 with injector holder ($15 ebay), 3GPH @ 100 with injector holder ($15 ebay), brass check valve (opens at 22PSI, spring loaded) ($25 ebay). The brass TEE and all NPT to .17" ID compression fittings from Home Depot ($15). Using a one gallon jug of water for testing in the garage along with two 2 quart measuring cups. Power source is a car battery at 12 volts.

I was targeting a total amount of 450 - 500 mL/min. I wanted a two injector setup so I could put a nozzle in each charge pipe (two turbos). Theoretically, this amount of 50/50 water/alky is good for 450FWHP. See Snow PerformanceHome installation guide for settings. The Not2Fast: Where Science Hits The Road site has a great calculator for water/alky injection and models IAT drop from water injection.

The nozzle I have are the mcmaster-carr standard ones. One is a 2GPH and one is a 3 GPH. That is at 100PSI of pump pressure. At the 150PSI the volume should be 5*sqrt(150)/sqrt(100) = 6.1 GPH.

Doing the real test with constant power to the pump yielded only 5.3 GPH. That means that the pump was only putting out 110PSI with both nozzles installed (5*sqrt(110)/sqrt(100)) = 5.3 GPH. That is what was measured in the measuring cups. That is real world data.

Then I did some testing with PWM the power to the water pump. I did this using a GM coolant fan relay. The power/switched pins were to 12V car battery and the water pump. The relay control (coil) was connected to the car battery and the 730 ECM pin F3 (A/C clutch control in the stock AUJP code). I changed the AUJP code so that it toggled the F3 pin ON for 100ms and OFF for 100ms. This reduced the flow by 66%. The dual nozzles flowed 3.5 GPH. The relay was only used for the test setup. I am going to use a real power transistor for the actual car install. It is more reliable.

Since I have complete control of doing PWM (pulse width modulation) to the water pump, I am going to install two 5GPH @ 100PSI nozzles from mcmaster-carr. I can adjust flow by PWM on the power. I wrote some code that increases the voltage to the pump as boost increases. The plan is to turn it on at 4PSI and increase the spray from there.

What all the testing showed is that ShurFlo's rating of 150 PSI, 1.5 GPH is not a simultaneous measurement. If you install nozzles greater than 3 GPH @ 100PSI, then the pressure starts to drop quickly and you need to increase nozzle size for that. I am guessing that the dual 5 GPH @ 100PSI will have a pressure of 90PSI and flow about 9.5 GPH.

The pressure drop and reduced volume seems to be the main reason all the water/alky sellers say you need to experiment. If the pump was better, it would be more cut and dry.

Also, I am not all that excited about the poly hose everyone sells. All that I found is that most is good for 150PSI at 70* F. The PSI drops as temp. increases. I will probably use braided alky/meth hose in the car. The poly seems to be too close to the max. rating. I might try it for a week or so just to see how it holds up.

 

This information is digital gold right here! Excellent thread!
Won't the PWM control on the shurflo pump wear it out much faster due to the quick on off cycles? My knowledge of electrical motors/pumps is shaky, but I that's a thought that entered my mind.

-forgive my grave ignorance, but can you explain the compression fittings? Kinda like the Poly line, on a hose barb, and a hose clamp, but somehow using a compression fitting? I've seen compression fittings, just never understood how they worked.

-poly tubing, at upwards of 100psi scares the crap out of me... I guess the worst that can happen is a wet engine bay, perhaps some knock counts if you hold it at WOT.... braided alky meth hose, that's specific braided hose meant for this purpose? What's the base material of that tubing? Is this like the normal "braided line" that most people use with AN fittings? (i'm extremely ignorant of some parts of automotive plumbing, WRT AN in particular).
-what about using hardline? like aluminum?

Not sure how easy this would be, but what about using a rotary ****, like with a potentiometer, to vary the rate at which injected alky/water is increased? ie, the **** will go from 5psi water pressure per 1psi of boost, to 10psi water pressure per 1psi boost, etc?

hmm, your idea about using software to vary the voltage at the pump, in order to increase liquid flow sounds better than my idea of using a return style regulator. I'm just thinking about electric motors - drop voltage, and they try to suck up more current to compensate, and tend to overheat. This true?

Again, thanks for the post, that's just loaded with goodness. Keep us in the loop!

 

Pulsing the motor is hardest on the relay. The motor spins when voltage is applied and produces voltage when it is spun. So, by applying power it will spin and then create a voltage when the relay is switched off. At the critical moment of the relay switching off, the motor tries to create a large voltage due to the inductance of the motor windings. The voltage V=L*dI/dT is created. It is a way of saying that the faster the relay turns off, the larger the voltage is. Large voltages want to arc the relay contacts and arc at the motor brushes. Yeah, PWM is harder on everything. Brushes last far longer than relays usually do. I am not worried about the motor, but the relay is not a good way to do PWM. It is fine for testing purposes. You could use one in a car, but I would put another one in series that is controlled by a toggle switch on the dash. It would turn off power in case the PWM relay contacts weld shut. PWM is basically creating a mini welder out of the relay contacts. The pump flows at 10AMPs.

In the car I am going to use a power transistor to do the PWM. It can handle the large voltage better (with a zener diode). Solid state devices are much more reliable than mechanical devices (relays).

Compression fittings. Think of it like a wooden wedge kicked under a door to keep it closed. The harder you kick the wedge, the tighter the door is closed. In terms of the compression fitting, the outer brass piece that you tighten is the floor, the poly tube is the door, and the ferrule on the tube is the wedge. As you tighten the fitting it is like kicking the wedge (ferrule gets jammed in more clamping the hose). Inside the tube is a brass insert that holds the poly from crushing.

The S/S line is just the regular line everyone sells. It is rated for alky/meth and much larger temps. and pressures. Aluminum doesn't like methanol. It needs to be anodized.

Using a potentiometer would be fine if you could find one to handle the power. Suppose you want to cut down the motor voltage to 6 volts. That means the pot needs to drop more than 6 volts of car battery. The motor is drawing about 4 amps at 6 volts. So the pot sees 6*4 = 24 Watts. That is a lot of power for a pot.

A regulator would work great. I didn't want to take the time to find one that handles water/ethonal/methonal at such high pressures. If you look at the "progressive" water/alky controllers out there, you will see the packaging is pretty small. That means that they are doing PWM and not a linear voltage control (linear transistor, potentiomenter kind of thing). Doing linear control generates the power/heat that I was talking about with the pot setup. The packaging/box the progressive controllers use is too small to dissipate that kind of heat. They cost $125 and up. I do it for free in the 730 ECM and don't have extra hardware around that can break and **** me off if it breaks and the only solution is buy another one.

The motor will generate a little more heat. I am not worried about that. It is a large enough heat sink and the amount of time in PWM would be on the order of seconds to minutes. A tank of water/alky only lasts so long. Just the re-fill time would be enough for a lot of motor off time/cooling. The big problem is the relay taking a beating during PWM and trying to weld the contacts shut. It is the same problem with a NOS solenoid. That is why they sell progressive controllers. They are just a box with a power transistor inside of them and a few digital chips to read boost, TPS, RPM, etc.

I played with the 730 AUJP code some more. I changed the ON and OFF time so that I can control it better. I used the 6.25ms loop so that I can turn it ON and OFF in 6.25ms increments. It gives precise control of the motor and allows a greater range of % spray. It allows you to do all kinds of stuff.....like pulse the motor ON for 100ms and OFF for 1.4 seconds. It gives it a tiny shot every 1.4 seconds. Not that this is useful, it just shows what kind of control you can get out of a factory ECM. I need to order the 5 GPH nozzles tonight and test some more next week when they arrive.

 

This is interesting thread, I though about using the ecm for pw control but was not sure how to do it.
Tomorrow I should be able to pick up a parcel in the mail with a Fjo pump (shur flow with fjo stickers) controller and 2 400cc nozzels. Fjo has a optional pressure tranducer that reads line pressure and adjusts duty cycle based on it. Fjo looks like a good system. About $900 can. for the above parts.
I have down loaded the soft ware it is alot like a v/e file from a bin. Map vs Rpm. There nozzels are attached with the solenoid for minum delay and hopefully acurate duty cycle.

Some where I read that teflone lined braided s.s. hose is the way to go. We just happen to have lots at my place of employment. Our hose ends are steel crimp type with a gold (colour) coating. They are just plain hydraulic hose ends. Are they safe with meth/water? For the price of 4 an aluminum fittings maybe I should go with them.

 

I played around with the duty cycles tonight (one 2 GPH and one 3 GPH nozzles rated at 100PSI). All test were performed for 1 minute of flow time. I had it set up for a max. pulse width of 100ms. I did one test full ON and the flow was about 290 mL of water.
I then tried a 50ms ON, 50ms OFF. It flowed 290 mL (same as full ON). I then tried less on time and they still flowed 290mL. The flow should have been less. This shows that the pulse rate is too high and the motor doesn't wind down enough. I was using this "fast" pulse rate to try and keep the spray consistent.

I then used a max. pulse with time of 200ms. Pulsing it ON for 100ms and OFF for 100ms gave a flow of 250ms. I am going to try less ON and OFF times tomorrow night using the 200ms period.

*** WARNING ***
If using a brass check valve that is spring loaded, you can air bound the pump and it will not prime. For instance, you are pumping fluid and the tank goes dry. You then fill the tank and then think all is fine for your next drive. IT IS NOT. The pressure due to the check valve and air in the line on that side will not allow the pump to prime and spray water. YOU MUST CLEAR THE AIR FROM THE CHECK VALVE LINE in order to get it to prime fluid again. This was consistent and happened every time I was pumping and ran out of water. Without the check valve (spring loaded) it would prime fine.
This will also happen with solenoids, they need to be open to prime the pump.

The SPRAY vs. MAP vs. RPM is how the 730 ECM will do the control. Exactly like a VE 3-D lookup table. I get the SPRAY value every 100ms and apply it at a 6.25ms rate. So the update is really every 100ms. They setup you are getting will be like what I have in the stock ECM.

I am guessing that the hose ends are stainless steel. All the heavy equipment hoses are made with S/S ends. I was going to have a hydraulic shop make me lines after I finally put the setup in the car and know the line lengths that I need.

I have about 20 minutes of run time on a used GM relay and it is still working fine. I will take it apart when I am done testing and see how the contacts held up. The motor runs cold to the touch after consecutive 1 minute runs. The water flowing through it provides a huge heat sink.

 

I like your experment, it will alow for a cheap and very tunable system.

Here is another Idea if your relay will not hold up, run a secound 730 ecm, parrel the required inputs, tach signal, map, then use the solid state injector drivers in the ecm to run your pump.

Just a idea have not fully thought it threw.

Are you using a turbo or charger, or am I the only one trying a high compresion normally aspirated engine?

 

I might grab a "junk" 730 ECM and take the fuel pump output driver from it and put it in my ECM. It would provide a driver for the pump at ECM Pin A2. The circuit is the same as the fuel pump driver, but GM never install this chip. It will cost about $30 each to build three driver boards. A used ECM can be as low as $25. Yeah, I am a cheap and like playing with junkyard parts. If I go with the custom driver boards I will put about 5 drivers on them. One for water/alky, one for N2O solenoid control, two for wastegate control, one spare. In the end it would probably be better just to build the custom boards and get it over with.

Low boost turbo vehicle. More of an experimental rig than anything. Water/Alky would be good for a high comp. N/A setup. Some IAT cooling and higher octane is a good thing.

 

I love it, if I had less things on the go I would be right in there with the junkyard experments. I have lots of oportunity to work over time, it is more cost efective now for me to buy stuff then build.
Good example I was making ethanol and running it in my lawnmower. I made a good product 94% alc. 6% water. It works, but if I included my time it was very expensive. Just going to buy methanol from now on.

I will start a new thread on my fjo setup findings in the next couple of weeks.
Good luck with your experment.

 

I ended up liking dual 10 GPH @100 PSI nozzles with the ShurFlo pump. They spray 250mL/min with a 25ms ON time, 175ms OFF time.
Full ON they spray 1000mL/min. I have a time resolution of 6.25ms and made a 3D lookup table for the spray time vs. RPM vs. KPA. I made the KPA table go from 58KPA up to 207KPA. The RPM goes from 1200 RPM to 6400 RPM.

The spray pattern is good even at the 25ms ON, 175ms OFF PWM. I should be able to start spraying at very low boost with 250mL/min.

I have about 30 minutes of run time PWM on the GM coolant relay. I pulled the cover off of it and there is no signs of arcing at the contacts.