PDA

View Full Version : When to activate Water injection?


Philip
02-11-2003, 02:10 PM
Hi all,

Ive just about finished installing a Paxton supercharger onto my rover powered cobra replica. Im going to be boosting it by around 6 psi and want to use water injection purely as a precaution. I dont think i'll need it but better safe than sorry. My question is that im going to reprogram my ecu to control the switch on point of the water. As the ecu uses manifold air temp, revs, manifold pressure etc, i can use these to form a switch on point. So what sort of temp and boost level should i be looking for the water to come on at? My comp ratio is 9.75, its a 3.5L V8 if thats any help,
Phil

Forum Admin
03-11-2003, 03:13 AM
I can not provide you with an answer on the specific temperature but will point to some issues to consider as you implement.

Where is the sensor taking the temperature from? This may seem obvious but to be helpful it needs to be after compression. It probably is, but just want to be clear for others who come accross this since some sensors are taking pre-compressed temps.

If you base it on temperatures before the jet - you may not know if you are using too little or too much water for your intent.

For purely an intercooler function rather than utilizing all the benefits of water injection, you don't want to inject too much water because if you are leaving your fuel settings rich so that the engine would run safely even without water injection you would likely lose power with water and rich fuel settings.

I think one application that would be very interesting if your ECU can handle this would be to have the ECU not only control the on/off function of the water injection but also the flow of water. This would require the jet to be before the temperature sensor - the higher the temperature the more water is injected - the lower the temperature the less water is injected. Will your ECU provide a pulse width signal from a map based on intake temperature as the load?

Philip
03-11-2003, 09:16 AM
Hi,
thanks for that,
yes the sensor is after both where the water is going in and after the blower. (In the plenum) I havent got the room for an intercooler so i was hoping to use water for this purpose. I can reprogram the ecu to do what ever i want, im currently making it turn water on at 4psi boost as long as revs are over 4200 or manifold air temp is above 50degrees C. I havent got it running yet so these are all guessed figures. I could do as you say and start it at 50% duty cycle and work up to 100% as temp increases. Wouldnt be too difficult, just means a bit of programing and a driver pcb making up. DO these figures look ok to you or are they too low? Ive read that water shouldnt be turned on below 6psi, but i might not reach that level of boost,
Phil

Forum Admin
03-11-2003, 01:08 PM
For this intercooling purpose if you have a valve in the WI system that can be driven as you describe based on temperature then you will be sure just to inject enough water to maintain the desired intake temperature. That would give you good reference for intake temperatures when programming your fuel without worrying about too much water. The temperature you are seeking is up to you really and may be a matter or trial and error. Also the exact IDCs and flow rates will take some trial and error. Without mapped water injection you may be injecting going well above and well below your target temperature rapidly and the on/off cycling of the WI may in that case cause inconsistent engine performance if it is more water than needed. I hope I have explained this in a way that you understand what I mean.

There is not solid rule on what boost level to inject on. The 6 psi reference may be for turbos and that level would be for a couple reasons. Turbo engines frequently have lower compression ratios and 6 psi would be a good measure to prevent running the WI during relatively low loads when it may not be necessary. Some GTP users for instance do not run much more than 6 psi of boost and are using fixed pressure switches that engage WI at 3 psi.

Philip
03-11-2003, 02:00 PM
Thanks, I'll have to play with it when i get it running and see whats best for my application,
Phil

M3Turboca
03-11-2003, 06:00 PM
I would say around 3 to 4 psi if you get no detonation and your air/fuel is fine then 3 to 4 psi is a good number.

What is your redline on your motor and at what rpm do you build max torque. Where do you build 3 to 4 psi at what rpm.

To trigger it by duty cycle will is nice but not necessary for the setup you have.

What water injection setup are you using if aquamist what size nozzle ?

Philip
03-11-2003, 08:52 PM
Hi,
thanks for that,
my engine produces max torque at 4250 rpm and im limiting max to 5700rpm. Its a centrifugal blower so max psi will be at max rpm. Im going to use a 0.4mm aquamist nozzle but im using an adjustable pressure water pump ( 40-120 psi) and will adjust to get around 10% of the fuel usage at max rpm. My guess is the engine will use 16 gph at max rpm so i will adjust pressure for 1.6 gph from the nozzle. Does all this sound OK?
I haven't got the engine running with the blower yet, but as water screws up the readings from my Wideband Lambda sensor iv got to turn the correction off in the ecu when the water is turned on. So i might as well get the ecu to turn the water on. If Im doing that i might as well have a good control for the switch on i.e. Manifold air temp, Revs, Boost.
Phil
www.replica-cobra.co.uk

Brad
18-11-2003, 02:59 AM
Philip
As your engine is using a Paxton supercharger boost is about the only trigger you need for activation. Boost will come on consistant with rpm. If you are only going to use 6psi turn water flow on at 2psi increase water flow propotional to fuel.
On a 3.5 liter engine with 6psi boost the normal jet size is 0.7 to 0.8 when increaseing boost to 10-11psi you may have to use two 0.8's and enough flow to keep up. You should alway make sure your system can put too much in. If it can not make the car stumble there is not enough water volume for proper tuning. All pump lose flow with time so make sure it is over built.
One of the great advantages of water injection is you can get very aggresive with the timing at 6psi. You should not have to retard but one or two degrees from stock if things are tuned correctly
One method of tuning is to not use water. get your fuel map and timing correct for the setup, then add water and increase timing.
Increasing timing will show a slight increase in O2 reading.

Philip
18-11-2003, 04:10 PM
Thanks for that,
Ive redesigned the switching points to come on with boost (3psi) and Manifold air temp, probably dont need temp bit but its done now. Ive also added another switch point for around 5 psi. My idea is to start the pump with around 10-15% water to fuel at 3psi by switching the pump on with a lower voltage then increase it at 5 psi to get 10-15% water to fuel at that pressure,
Phil

Brad
18-11-2003, 07:23 PM
Philip that is a good way to ccontrol flow. I have made many stepped over the years. They left me with uneven performance. I now use the Auamist high speed valve tied into fuel injection duty cycle. It gives a smooth progression in flow based on engine load. Fuel flow is based on engine load at every rpm. This has proven to be the easiest and smooth operation. We started tieing the high speed valve into injector duty cycle about 1999 have has been involed in over a 100 installs and upgrades.
The biggest thing is to use the correct amount for your engine, do not get too hung up on the math. Work with engine performance.

Hemi T04
19-11-2003, 09:58 AM
I know that swtched - or staged water injection systems are not very smooth in their operation, so i was wondering along the line of the above post- adjusting the voltage to the water pump to controll the flow- in reference to boost pressure using a pressure transducer and an electronic circuit. Say for example the pump turns on at 2 psi- with 9 v dc- then inceases to full voltage at maximium boost to13.8 v. Would this work ok- or would it kill the pump over time. Or will it need a refernce to rpm signal as well to give good water flow control. this could be done off the coil probably.
What do people think? The initial sytem could be pretty easy to setup.

Philip
19-11-2003, 02:15 PM
Brad,
I had wondered wether it would be smooth enough as Ive no experience with this sort of thing. Im currently adding some code to my ecu to switch a valve from 80% duty cycle to 100% as boost increases from 3-6 psi. It will still be stepped but in 0.5psi steps. This will need some work before it is finished so will end up with my first idea to start with then change to this later. One question for you, any idea what sort of frequency the Aquamist fast shutoff valve works well at?
Thanks,
Phil

Brad
19-11-2003, 02:22 PM
Philip, The high speed valve will operated at the same frequency as a fuel injector. It wil operate up to 250 cycles per second. It can be driven with the same map as the fuel to start with then modified for the most perfect fit.

Philip
19-11-2003, 03:30 PM
Brad,
hadn't considered doing it that way, that would have been easier for me. Can soon change code for that. Do you know what sort of drive I would need for the valve? i.e. is it the same as turning an injector on/off, or can I just drive it with a standard power transistor?
Phil

Forum Admin
19-11-2003, 05:25 PM
Yes it can be driven in the same way as an injector on/off. It is a solenoid valve - 13ohm coil resistance. Normally closed that opens when the circuit is charged. A fused 12v+ source with a switching ground will drive the solenoid. It has an transition change time of about 1ms from fully closed to fully open and vice-versa.

Brad
19-11-2003, 05:29 PM
The Aquamist highspeed valve 806-244 is a 13ohm resistance valve. If you driver will do high impedence you are done. If you have a low impedence driver you would have to a FIA2 amp to convert the low impedence drive to drive the high impedence highspeed valve.
If you are using peak and hold injection the FIA2 will convert that signal into a standard duty cycle to drive the highspeed valve.
The highspeed valve with a constant water pressure supply is the most tuneable setup.

Philip
19-11-2003, 06:11 PM
Hi,
no im not using peak and hold as my injectors have external resistors so this is not needed. So what your saying is fire the water valve when the injectors are fired? So do you then increase water pressure to tune the amount of water or do you need to increase/decrease the Pulse Width to tune? Just had a quick calculation and all my injectors will fire 1500cc/min of fuel, so If I fired a 150cc/min nozzle at the same time as the injectors then I would get 10%, Yes?
thanks for all your help,
Phil

Brad
19-11-2003, 06:43 PM
Yes you will fire the highspeed valve with the injectors. Changing the pulse width changes the flow, use a 1.0mm nozzle or you may have to move up to 2 0.9mm nozzles for high horse power. The high speed valve will not require a resistor. Keep water pressure constant.
You will be running the water injection just like a fuel system.
If you elect to not use a different map for the water then you will change jet size to create the balance between fuel, air and water.
You math is correct on water fuel ratio.

Philip
19-11-2003, 07:24 PM
Brad,
that sounds too easy, will order a few different sized nozzles and try them out. How do you know when youve got it right? Obviously water/fuel ratio is easy enough to do but when do you know that you carn't add any more water?
thanks,
Phil

Brad
19-11-2003, 08:11 PM
The engine will get a little quiter when correct and power will be better. At some point you need to put too much in. Increase flow until the car stubles, that will be too much. Then work down to best power. The main reason for making the car run bad is so you know what that feels like. If you tune the car on a hot day and then the weather turns real cold the car will run bad.
Basicly it is by feel or on a rolling road.

Philip
19-11-2003, 08:19 PM
Brad,
beginning to understand this all now thanks. One more question: If im pulsing water then surely the air will go hot - cold - hot - cold? Thats why I assumed it best to adjust flow rather than pulse. I take it this isnt the case?
Phil

Brad
19-11-2003, 10:17 PM
The cycle at which the valve cycles is very fast and does not create the issue you discribe. The tuneablity of using pulse width is much more usefull than a constant flow.

Philip
19-11-2003, 11:01 PM
Brad,
Thanks for your help
Phil :)

TurboGTi
22-11-2003, 01:13 AM
Phlip hi,
Tell me something what ECU are you using on your project.

U seem to have a pretty awsome ride there and the guys in this forum are quite helpfull keep it up guys ..!

Philip
22-11-2003, 02:02 PM
Hi,
its a Megasquirt ECU, you build it yourself and make it fit your application. Sounds harder than it is, there are hundreds of people doing it all over the world now thanks to their forum, who have no experience with electronics or engines. I knew next to nothing about engine management before I started, still dont know a lot, but the engine runs so much better than before.
All the details are on my site, dont be put off if you feel you wont be able to do it as there is a great deal of help from the forum and Ive not heared of anyone not getting there eventually,
Phil
http://www.replica-cobra.co.uk