PDA

View Full Version : Replacing IC with water - Request For Comments (longish)


robbilau
20-02-2004, 09:22 PM
I am interested in replacing my FMIC with water injection and I wanted to open discussion on how best to do it. I have a turbo Miata with MoTeC M4 Pro ECU and Aquamist 2c.

I have some flexibility in driving the injection as the M4 allows me to choose the axes of a custom table. Currently I use MAP/RPM but I have the following choices:

X Axis (Horizontal):
1 ? RPM
2 ? ET (Engine Temp)
3 ? AT (Air Temp)
4 ? Aux V (Auxiliary Volts)
5 ? Aux T (Auxiliary Temp)
6 ? G Spd (Ground Speed)

Y Axis (Vertical):
1 ? TP (Throttle Position)
2 ? Eff (Efficiency Point)
3 ? Load (Load Point)
4 ? MAP (Manifold Pressure)
5 ? Aux V (Auxiliary Volts)
6 ? Aux T (Auxiliary Temp)
7 ? G Spd (Ground Speed)
8 ? Duty (Injector Duty)


IMO the best option would be to use a second temp sensor (Aux T) and use AuxT and Duty as the axes. Duty will likely be the best indicator of actual air flow through the motor and AuxT will advise how much intercooling is necessary. However it might be possible to use a dedicated PID controller of some sort for this, similar to a boost controller.

Assuming I use the MoTeC map I'm thinking AuxT is at the compressor outlet with a jet immediately after so that the system is reacting to heat before that heat gets to the jet so it has no lag, preventing a burst of hot air from hitting the manifold. This assumes a constant rate of cooling and can't compensate for changes in that rate due to environmental humidity or water temperature. Any idea regarding the degree to which these environmental differences will play a part? I also have one more output so I can add a second jet somewhere else. Will I get better atomization with multiple jets or will I simply be complicating the system? If I do use an additional should they work in sync or be staged with the onset of heat/boost? I could even inject pre-turbo with a second jet (understanding the issues with compressor wheel wear).

Last is the actual piping. I have access to a machine shop so I am not bound to a simple cossover tube from turbo to manifold. I can increase the diameter for example to allow more volume, recess the jets, change the shape of the tube. This is assuming I use metal of some sort - perhaps another material will be better or simple insulation (Thermo-Tec wrap etc.) would suffice to reduce heat soak from the engine bay.

I think that's enough to start the ball rolling. All comments and ideas are very welcome!

Charged Performance
25-02-2004, 12:27 AM
Deleting an intercooler and its related piping can be done.

Your approach with the MoTec sounds appropriate.

If you don't mind a little more complication and it could be done, my intial thoughts would be to engage the system (start injecting) based on crossing a threshhold intake temperature of say 110* F. Then preferably meter the water flow based on duty cycle and EGT.

What this will do is always inject when you are getting high intake temps but at the same time ensure you are not bogging the motor down with too much water based on EGTs.

Could you have that matrix only be active based on a binary input from the intake temp? Do you have an EGT input?

Again my comment adds complexity which may not be needed - your initial thoughts would work very well also.

robbilau
25-02-2004, 07:06 PM
Unfortunately I can't get EGT into that mix as I can only use one temp input, either air or EGT, at a time. Also, I don't have an EGT probe (yet). But I think I will be able to get a decent result this way

I should have a simple crossover tube in the next few weeks to start playing with - Flyin Miata is putting them into production imminently. I have a simple map prepared to start from which should allow a max of about 100 deg C cooling with some room to grow if the data on the ERL site is correct.

I'll report back as I make progress.

robbilau
26-02-2004, 11:06 PM
OK, I just ordered a second HSV and other bits so I'll be doing two jets to start: one 0.4mm 150cc jet before the throttle body for det control. I'll run this off of MAP/RPM.

The second jet will be after the turbo and compressor outlet temp sensor as far from the throttle body as possible. This jet will be the smallest I can get by with, probably 0.7mm and will run off of inj duty and compressor temp with a ceiling of around 18% of fuel flow. This map will be tuned rather like a fuel map but for temps.

I ordered an accumulator and a check valve for each of the jets and a Y splitter after the accumulator to route to the two HSV's. Hopefully will be enough to run just the one pump for now.

Richard L
27-02-2004, 01:11 AM
robbilau,

Just in case you require more water on your quest for more cooling, I have drawn a simple circuit for you below. The circuit basically reads the pulses from the grey wire and extending it to enable you to power up a small priming pump to increase the flow capacity of the aquamist pump.

Any standard windshield pump will do, it will boost the output capacity to approximately 600cc/min at 6 bar+.


http://www.aquamist.co.uk/forum/stretcher.GIF

Charged Performance
27-02-2004, 02:01 AM
This circuit Richard developed works very well - we have used it on 2 STi's to keep the priming pump going between the pressure switch pulses. Make sure the priming pump is a type that doesn't hold pressure when it is off though.

robbilau
27-02-2004, 02:26 AM
Thanks, Richard!

robbilau
01-03-2004, 04:36 AM
Here is a picture of the engine bay and turbo crosspipe.

http://www.gvii.net/dlralt/installation_photos/PIC00006a.jpg

I know for cooling I want to inject as far from the throttle as possible. I need to put an air temp sensor after the turbo outlet, and then some distance after that the jet.

I want to make sure that I do not put the jet close enough to the temp sensor to affect it so I need adequate separation. How far from the temp sensor should the jet be?

My other concern is about the BOV. I don't know if a jet close to it will affect it badly. Cleaning it regularly is no big deal.

I will have a small jet right before the TB to take car of detonation so this is purely for cooling a non-intercooled turbo. Where would you put the temp sensor and the jet?

Richard L
03-03-2004, 12:49 AM
robbilau,

Is it possible for you to superimpose a few arrow heads onto your image so that I know exactly where you intent to place the jets and ATS.

robbilau
03-03-2004, 03:30 AM
Here you go:

http://robbi.laurenson.com/users/robbi/PIC00006a.jpg

Richard L
04-03-2004, 01:19 AM
Saw the picture, your proposed position seemed to be very logical. I can not see any probelm reading the correct temperatures from your two sensors.

robbilau
04-03-2004, 04:13 AM
Thanks, Richard. Now all I have to do is hurry up and wait for parts. Aaaargh! :lol:

robbilau
21-04-2004, 07:04 PM
Woohoo! My parts finally shipped yesterday from Flyin' Miata, so they should be in Monday or maybe even Friday if I'm really lucky.