I haven't asked a technical question regarding the 420a in years, figure it's time to stoke the coals a bit. If this project becomes successful, expect a wiki with parts list all for <$100.

I'm designing a water injection system based off of tidbits I gathered from autospeed.com. The golden piece of information off of that site was a suggested pump that works like an aquamist pump at a fraction of the price (~$50-60). I'm going to use the stock windshield wiper tank, and set my emanage to turn on a relay for the pump when boost begins. The nozzle will be placed after the intercooler, some distance away from the TB.

Goal is to run water only for detonation control. No methanol. I won't need any alcohol in the mix as the car will be in cali and freezing temps are not a worry. I want to maximize boost available on a s16g with 91 octane, intercooled, 8.8 pistons, stock ignition. Without water injection, some have reported seeing pinging at 15 psi on this setup. I want to get a few more psi if possible.

I'm looking at the following nozzles off mcmaster: 3178K87, or perhaps 4759T15. The fogging nozzle appears to atomize the water the most, based on the description. See http://www.mcmaster.com/#catalog/117/2072/=aunli4 . Any suggestions here?

If you run water injection can you chime in on your experiences with details of your setup? Like what is the volume of your water expressed as a %age of your fuel volume? Maybe an old schooler versed in water injection will see this thread (turby, are you reading this)?

Once you've implemented water injection, what do you lean your AFRs out to now, since the water is putting a clamp on combustion temps instead of the extra fuel?

Also, can anyone recommend a small/light weight 12V to 220V inverter that will survive underhood temps?

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

With said pump, is there a controller as well? Is there a voltage regulator to feed the pump or will the output vary with system voltages? Does the system require a pressure regulator of any kind? What calculations are you using/will you use to establish flow volume & pressure?

I am interested in H2O injection as well in FL for the same reasons. A DIY kit is well worth the R&D.

From the MS Manual:

http://msextra.com/doc/ms1extra/MS_Extra_Hardware_Manual.htm#waterinj

When the manifold boost level (detected by the MAP sensor) and engine RPM are above the set values in Megatune and the Manifold Air Temperature is above the set value, the Pump is turned on via X2 for a V2.2 pcb or JS0 for a V3 pcb. X3 (JS1 for a V3 pcb) is pulsed at the same rate as injector CH #2 (this can be used to drive a fast acting solenoid). The theory being that if a water nozzle is selected to give a 15% flow rate of the total of the injectors flow rate then it follows that it should naturally give the right amount of water. So if your total injector flow is 2000cc/min (all of your injector flow rates added together) then a 300cc/min nozzle would give you 15% water to fuel ratio, which is a good starting point. i.e. 15% of 2000cc of fuel = 300cc of water.
The Manifold Air Temp is ignored once the water has turned on as this should reduce vertually straight away with the water evaperating in the air. The water system doesn't turn off again untill boost pressure or RPM has dropped below the set levels.

---

2012 2500HD LML

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Check this thread: http://www.alcohol-injection.com/forum/nozzle-size-calculator/how-convert-gph-ccs-nozzle-size-427.html

63 cc/min = 1 gph (we will need to back this up somewhere)

1 gph = 0.01666 gpm (if needed for certain products using that measurement)

Using the nozzle 4759T15 form McMaster, we see that the volume output changes significantly with input pressure. I have not looked too closely at the pump specs, but I am not sure how we would go about calculating the line pressure with no regulator. What kind of pressure would we see in the line to the nozzle upon power-up?

Using the same pump, at 100 psi (for example) the nozzle flows 26.4 GPH, which by using the above formula equals 1663 cc/min flow through the nozzle. That is quite a bit of water being injected. @40 psi the nozzle will flow 1058 cc/min. Clearly we need to step down the orifice based on our fuel requirements.

---

2012 2500HD LML

Good info. I've looked a little farther into this too.

I combined the data available at spray.com (page 17 of their unijet fine atomizing nozzle data sheet) with actual measured ml/min experimental data gathered by autospeed for various nozzles. I used this to estimate pressures that are seen in the line for nozzles ranging from 0.6 to 4.


For a size 0.6 unijet, it flows 65 ml/min on this pump. I estimate that the line pressure is therefore around 100psi.

For a size 3 unijet, it flows 230 ml/min on the pump, so I estimate the line pressure drops to around 60 psi.

For a size 4 unijet, it flows 260 ml/min, so the pressure drops even further to around 50 psi.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Ive run meth injection for over a year now. Built a kit from devils own. I run mine threw MS and activate it at about 13psi. You should think about running a small amount of meth in your mixture to help control knock better. The water helps cool the intake charge while the meth increases octane to help with knock. Obviously cooler intake temps will also help with knock but it will be minimal gain. I just run blue -20 washer fluid. Its ab $2 a gallon and contains 35-40% meth. You should be able to max that turbo out on the stock ignition no problem. I use to run a 20g at 20psi with meth on the stock ignition

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

bookmarked

___________________________________________



http://www.2gnt.com/quote_db.php?id=278

Outside of Corona: "That sucks, peace out..." (jamie walks away)... "Sweeet, what we doin?" -Zac
Tim's garage (prior to motor install): "Im gonna smoke real quick, that way if you fuck up I am already nicotined up" -Tim

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Im running a M12 nozzel and the devilsown silver pump, I think its around 115psi IIRC. I cant remember the flow rate of the system off the top of my head.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

Damn, that's the 756 cc/min nozzle.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

I just went and looked Im running the M10 nozzle. Its been a while since I have messed around with this. The nozzle is rated at 630ml/min. I have my 650s about maxed out so Im really only getting 1/4 the nozzle per cylinder.

ml/min = cc/min
.25 * 630 = 158cc per cylinder of the water/meth mixture. Since my mixture is approx. 40% meth I am really only getting 63cc of methanol and 95cc of water. If you plan on running straight water then you should probably run a nozzle less than 400ml/min.

But like I said I would run the larger nozzle and add some meth into the mix. It will let you run more timing and also let you pull some more fuel out of your injectors to lower the duty cycle. However, your setup will now rely on the meth for fuel and if you ever run out your motor will be toast if you boost it. With just water you wouldn't have this problem.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

If I remember correct my duty cycles decreased by about 10% but I can also run a little lean and still not see any knock. As far as octane ratings go it may only raise pump gas to 94 octane or so, not really any significant amount. If you have a way of logging intake temps and knock it would be interesting to see if there is much of a difference between running straight water or water meth mixture.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

Interesting. Yes I can log IATs, and was planning to do this with the setup.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Bringing this back as I was thinking abut it the other day.

Provisions for a check valve mounted as close to the nozzle as possible? Necessary for applications where water injection does not occur until a medium boost level (say max boost is 30 psi, and water injection does not occur until 10 or 15 psi).

---

2012 2500HD LML

Yes. We can use a check valve that is part of the spray nozzle/screen strainer assembly, supplied by Spraying Systems Company.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

I was at Northern Tool today and saw this on a shelf:



http://www.northerntool.com/shop/tools/product_200355686_200355686

12 volt water pump, 70psi capable, and 2.2 GPM. Is that close to our needs? I need to revisit some of the calculations we made in this thread. The use of such a pump would eliminate the need for a power inverter. The pricepoint is not bad either, and they come in varying models/flowrates.

---

2012 2500HD LML

You can use a pump like that, and its comparable to all the kits out there that use sureflo. The point of my first post was that a huge clunky pump could be avoided, and we could get something that functioned almost as well as an aquamist pump at a fraction of the cost.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Recording this for future use:

Check valve: 8567T12 | Type 303 Stainless Steel | 1/8" NPT Male/Male
Temperature Range: NPT Male: 35° to 300° F
Maximum Pressure: NPT Male: 200 psi @ 300° F
Cracking Pressure: 2 psi
$21.16

Coupler, to be welded in UICP: 4443K661 | Type 316 Stainless Steel | 1/8" NPT female x NPT female
Maximum Pressure @ 72° F 3000 psi
Max Temperature: excess of 350F
$4.61

Nozzle: 4759T15 | Brass fogging nozzle | 1/8" NPT male
Maximum pressure is 3,000 psi
Maximum temperature is 450° F
$24.39

Nozzle selection is still TBD depending on fueling requirements. The majority of the selctions are NPT male however, and will require some sort of bulkhead fitting to join the NPT male of the nozzle to the plumbing hose of check valve if installed at the IC pipe. My current thoughts are welding a female/female NPT coupler at the IC pipe before the throttle body as discussed earlier. Installing the nozzle from the inside, the check valve on the outside, and routing the plumbing to the check valve from the pump assembly.

---

2012 2500HD LML

It seems like it would be easier to just piece a kit together from Devils own or some other water injection site. I built my whole kit for around 200 if IRC. And that was with a tank and pump also. It would also be bolt on and not have to worry about welding.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

---

2012 2500HD LML

---

2012 2500HD LML

http://www.iprocessmart.com/techsmart/conversions.htm

Yes that conversion is correct.

Also the only issue I can see with that pump is the atomization of the water through the nozzle. Most water injection pumps run up to 150psi. Remember you have to subtract the amount of boost from the pump pressure to get the final injection pressure. So if he plans on running 20-30psi then the nozzle will only see 40-50psi of pressure to atomize the water.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

According to that link, CC = mL. Not multiplied by a factor of .25.

Indeed what you mean about the pressure though. Subtracting boost pressure will inevitably net a lower nozzle pressure. It seems we need to rely on a 120 or 240 volt pump to provide what we need.

Edit: I see where that .25 factor came from. I was just reading it wrong in your original post.

---

2012 2500HD LML

Sry if I didn't make that clear the .25 was to figure out per cylinder how much water/meth was being injected.

98 Eclipse GS Turbo
Built motor - MS2 - Holset hy35 - Zoom Stg 4 Clutch - 57.5mm TB - ARP Headstuds - Turbonetics wastegate - 3" Turbo back - FMIC - Greddy RZ BOV - Walbro 255 - 650cc Injectors - Fidanza Flywheel - Cobra calipers w/ 13" Rotors - Oil Cooler - DevilsOwn meth injection

1981 Chevy Scottsdale
1987 Nissan 300zx - Chump Car
2001 BMW 325i

So I guess no one ever came up with a DIY kit?

Anyone know anyone who can help me tune this in the Pittsburgh area (Greensburg)?

Yea man I'm sorry, I don't have a very long attention span. I picked up several other projects last year and all I've been doing on the 2g is basic maintenance.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO

Can you use this?


The circuit will convert 12V DC to 120V AC. This circuit can handle up to 1000Watts supply depends the T1, T2 and transformer used. Please see the note.



Components list:



Part......Total Qty....Description

C1, C2....2............68 uf, 25 V Tantalum Capacitor

R1, R2....2............10 Ohm, 5 Watt Resistor

R3, R4....2............180 Ohm, 1 Watt Resistor

D1, D2....2............HEP 154 Silicon Diode

Q1, Q2....2............2N3055 NPN Transistor (see “Notes” )

T1..........1............24V, Center Tapped Transformer (see “Notes” )

MISC........1............Wire, Case, Receptical (For Output)


Notes:

1. Q1 and Q2, as well as T1, determine how much wattage the inverter can supply. With Q1,Q2=2N3055 and T1= 15 A, the inverter can supply about 300 watts. Larger transformers and more powerful transistors can be substituted for T1, Q1 and Q2 for more power.

A candle burning on both ends is twice as bright but lasts half as long.

Ok, so I just re-read your post and then went to Autospeed and read up on that pump they used.

Why not just use a 12v/110v inverter like this?

HERE

A candle burning on both ends is twice as bright but lasts half as long.

Awesome. Yes that would work great, but it looks bulky. I've heard about 12v inverters that were the size of a deck of cards.

After re-evaluating my setup I honestly need to get a better fuel setup before I do this. Gotta keep my eyes open for portfuelers in the WTB section.

2gnt: '99 RS-T, killed by a toyota, pending rebuild...
Daily: Volt
Daily #2: '99 EVG ebike- STOLEN by PEDRO