Pre and post turbo injection in an audi b5 s4

[cinesnow]

Hello Everyone, this is my first post on the forum but have been lurking for quite a few years. First I have to say thanks to everyone who has contributed to this forum over the years and helped the development of water and alcohol injection.

Seeing as there is a limited amount of information being shared I want to provide some insight on what I have put together in this application in the event anyone else can benifit from it and ask a question or two.

The car is a 2001.5 audi b5 s4, I raised the compression rato to 10.5:1 and have bored the cylinders out to around 2.9 liters of displacement. The car is currently set up with 1000cc injectors for the fuel circut and the turbos are K24's. I'm currently running without intercoolers but plan to re-install them in the summer months and log the differences. The car is almost complete and I should have it on the dyno in a few weeks.

As for the water injection, I plan on running de-ionised water and isopropanol.*
The pump is the same aquatec unit that aquamist, aem and devilsown provide. The lines are all ptfe lined, stainless braided *in -4 with repect to the pre turbo T which goes from -4 to -2. Also worth noting, all of the -4 fittings are crimped, not the rebuildable type. I have placed two injectors, one for each turbo about 6" before the throttlebody using 550cc jets on each side. I gave a call to Jeff at howerton engineering who got me set up with a pair of the beautiful axial brackets which I first read about here (thank you Richard) and am using 150cc jets in each bracket.*

I originally set up the system with a solenoid consisting of two ports, one in and one out. The pre turbo injectors are aquamist and the post turbo ones are aem and have check valves integrated into the injector bodies.
I decided after bench testing the system that i was not comfortable with the way the line pressure diminished in all four injection points, the way it tapered down pre turbo specifically after the solenoid had shut. After searching around and making some calls I was able to have a custom solenoid made that consists of two seperate coils and plungers with the capability of controlling two injection points independently. The port configuration has one in, two outs and a common exhaust port shared by both outs in the normally closed state. My intent is to have the line pressure pre turbo and tb diminish as soon as the solenoid closes helping the check valves to actually shut instead of taper down leaving solid drops where they shouldn't be.

Just to elaborate alittle further,*
port 1 is the pressure source
port 2 is the exhaust port, which will run to the washer reservoir
port 3 is pre throttle body
port 4 is pre turbo

Question 1, Has anyone else set up a system similar to this?

Question 2, I am considering swapping out my manual transmission radiator and using one with a built in oil cooler with the intent of running the alcohol through it. I'm hoping the extra heat transfer will be of some benefit in regards to line pressure, atomization and poential ratios of water to isopropanol. Has anyone tried this?

Thanks in advance, Marc

Pics for the curious:







-2 an line to pre turbo injectors was a perfect fit in the sai ports with a little massaging:


I havent put back in the pre turbo and exhaust lines but you can get an idea of the setup:

[HYDE16]

Nice work on the setup, I'll be testing a small .3mm Aquamist jet pre-turbo on a TSI engine in my MK6 GTI. I don't have access to the turbine wheel like you do with my K04 turbo so I decided to set the jet way farther back right at post MAF. I'll trigger it at high boost with a boost switch a separate solenoid fed from the line pre-FAV (non-progressive basically).

[parmas]

Hi Marc,

Nice work indeed and I am sure you thought much about it.

I didn't understood well the solenoid system although it seems a little complicated and prone to fail or else not be precise as it should.

I am not arguing about your system but maybe I could suggest you a better way of what I have in my mind for my current setup.

I am running a 1.5L GT28 turbo target of 340bhp @ 22psi. Currently using post water (only) injection but considering pre-turbo.

I have the AEM injection system running 350cc post turbo and thinking of 150cc pre-turbo. Now since I have no intercooler, I am starting injection about 3psi with a gradual increase until full flow @ 150psi pump pressure till 18psi boost. Now I wanted pre-turbo to begin at least @ 10psi boost.

I intend to use a variable cracking pressure check valve mounted directly with the AEM nozzle and thinking of setting 100psi cracking pressure. This way you simply make sure that injection is pressurized enough to mist preventing blade damage and maximum atmoization. You only need to adjust correctly the check valve.

Here is the check valve in subject : http://www.ham-let.com/catalogue/product/7853

Regarding question 2 :

I tried to use the same theory but instead of radiator heat turbine exhaust heat !

I used a braided pipe from pump to engine bay and converted to a stainless brake line. Welded bracket to the brake line to a spacer that fits into three bolts of the turbine housing.
After testing, it eventually gave no positive results but instead exhaust heat was actually heating the intake through the line and nozzle!

To make sure the thoery of heating the liquid could make a difference .... try this as it was eventually easier to do !

Boil the distilled water using a kettle and pour nto the empty tank. Heat enough water and go for a ride immidiately after to test it.

I did it and eventually gave negative results of increased air temps by 10C !

Good luck

[cinesnow]

Parmas,
Thanks for the input, I was thinking a check valve would be essential but wasn't sure on cracking pressure yet, i'll order some of those check valves and see how it functions again soon. With the check valves, the line sould keep filled and only discharge the line pressure back through the solenoid and into the washer res. To make sure it donen't drain the line between the solenoid and injectors, introducing air - I made sure to have the exhaust line higher than the rest of the circut.

Today I decided to do a threshold test on a stock audi radiator with an integrated gearbox cooler. I cut down the flanges and tapped the ports for -6 and put it on the test bench, It held up to 700 psi and ruptured internally at 800.
I don't expect that the trans cooler portion of the radiator would ever see more than 300'ish psi in operation, but we'll see what happens once it's set up.

Very interesting setup with the braided line, do you have any pics? 10*c is a huge rise!

More to come soon...

[parmas]

Originally Posted by cinesnow

Parmas,
Thanks for the input, I was thinking a check valve would be essential but wasn't sure on cracking pressure yet, i'll order some of those check valves and see how it functions again soon. With the check valves, the line sould keep filled and only discharge the line pressure back through the solenoid and into the washer res. To make sure it donen't drain the line between the solenoid and injectors, introducing air - I made sure to have the exhaust line higher than the rest of the circut.

Marc, a check valve with high cracking pressure should be good enough to close before leaving solid drops. Stock AEM check valve is 15psi cracking pressure while Aquamist I do not know. My intension is to leave 15psi cracking pressure on PRE-Throttle Body while setting cracking pressure of 100psi on the PRE-turbo check valve. This way the pre-turbo always activates far after from the pre-throttle body injection.

Let's make an example on mine:

I do not have any intercooler but only a 20cm pipe between turbo and engine. Because of my setup, it is critical that water injection is initiated as low as possible to prevent pre-ignition during spoolup.

- Set the AEM water injection to start at 0psi boost. Eventually it takes till 2psi boost to eventually crack that 15psi check valve.

- On the AEM, the FULL injection (= 150psi AEM pump pressure) is set to 20psi boost.

- Set PRE-Turbo cracking pressure of 100psi. To roughly calculate pre-turbo injection start let's make FULL PUMP PRESSURE \ DIV BY \ FULL BOOST = 150psi \ 20psi = 7.5pump_psi per 1boost_psi.

- With this setup, for every 1psi boost you get 7.5psi pump. The fact that the injection eventually starts at 2psi boost although set to 0psi is the fact that it needs to pressure the valve 7.5psi x 2psi setting.

- Now if we know how much pump pressure we get per boost we can know that @ 100psi pump / 7.5 = 13.3psi boost

- Conclusion is that Injection throttle body starts at 2psi boost - turbo injection starts at 13psi boost - full injection obtained at 20psi.

Today I decided to do a threshold test on a stock audi radiator with an integrated gearbox cooler. I cut down the flanges and tapped the ports for -6 and put it on the test bench, It held up to 700 psi and ruptured internally at 800.
I don't expect that the trans cooler portion of the radiator would ever see more than 300'ish psi in operation, but we'll see what happens once it's set up.

To be sure, what is the purpose of the radiatior to heat the water or to cool the water ?

Very interesting setup with the braided line, do you have any pics? 10*c is a huge rise!

More to come soon...

............

[cinesnow]

Parmas,

Thanks for the formula and That car of your looks like a rocket, what does it weigh?
I'll be sure to set everything up with the check valves once Iv'e optimised the injection start points. The ham-let check valves are really nice, just waiting for them to email back so I can get them ordered.

The radiator would be used to heat the alcohol line, I'm going to conduct my first set of tests on the dyno without it and come back to this concept after everything else is dialed in.

[parmas]

Originally Posted by cinesnow

Parmas,

Thanks for the formula and That car of your looks like a rocket, what does it weigh?

You mean how much? I am thinking something more than 1Ton but next month will weigh it precisely

I'll be sure to set everything up with the check valves once Iv'e optimised the injection start points. The ham-let check valves are really nice, just waiting for them to email back so I can get them ordered.

If you don't get a response you might actually need a distributor. I tried to communicate with them with no response and eventually got a distributor to order them. Make sure you order the Perflour seal as they are rated above 300C !

The radiator would be used to heat the alcohol line, I'm going to conduct my first set of tests on the dyno without it and come back to this concept after everything else is dialed in.

I know what you feel right now but I think that is gonna be a waste of money and time. Please trust me on this, try with the kettle only distilled water and compare to ambient water temps. Also heating alcohol into a car is a timing bomb that could actually explode. I only suggest to use water only in that case.

...............

[cinesnow]

Parmas,

I got abold of somone over there, will see them in about five weeks or so.

You were right about the water and kettle trick, too close to the boiling point was a negative effect.
Wasn't sure about having an alcohol mix in it but I definately agree that heating iso would also be a very bad idea.

[parmas]

Originally Posted by cinesnow

Parmas,

I got abold of somone over there, will see them in about five weeks or so.

You were right about the water and kettle trick, too close to the boiling point was a negative effect.
Wasn't sure about having an alcohol mix in it but I definately agree that heating iso would also be a very bad idea.

Thank you for trying my suggestion. Now i am sure the theory is like that.

Also do not try to cool the water as I tried it aswell and was a negative result!

For once it seems that better leave it to whatever the ambient temp is

[rotrex]

I have tried precompression injection with a Rotrex C30-74. I took pictures of the turbine wheel before and after some use and could not spot a single difference except the blades looked more clean.

Powerwise I did not make any feelable difference comparted to direct port injection with similar flow. I injected 300ml/min.
The sound of the supercharger changes a bit. You can hear more of a noise. Intake asir temps drop to ambient using 50/50 water methanol.
The nozzle was positioned about 5" away from the intake that was S-shaped. I positioned in a way it would not pool on a wall.