Water Injection - Intercooler Pipe or Manifold?

[cheekychimp]

My Forrester Intake Manifold has bosses for direct port injection.

Is it worth taking advantage of? I've been giving a lot of thought to failsafes on an injection system.

It seems that whilst more complicated, using an accumulator and multiple nozzles is safer since it gives you maybe 10 seconds of injection even after a pump failure and the likelihood of multiple nozzles clogging simultaneously is minimal.

But if you go with direct port injection and a nozzle fails does that mean less water from the other nozzles will make it's way into the affected cyinder than if the four nozzles were mounterd in the intercooler piping pre throttle body?

Paul.

 

[JNR]

I had thought about this awhile back, but decided to go with a single nozzle on the throttle body flange (I am gonna build a custom SS unit to replace this POS alum and will thread it perpendicular to the opening; prob. from the bottom, for stealthness).

Part of the reason I personally decided against the intake manifold is that it 'seems' it would not have enough time/heat to fully atomize into steam, thus pulling the energy of heat out. It may work, but I am nervous about getting any sort of water buildup, for when it's at lower rpms, etc...

I bet it could work, but call me safe and I'd like to try and get the full benefits /ubbthreads/images/graemlins/smile.gif

 

[Terry Posten]

I was told to install the nossles as far away from the TB as possible but after the intercooler. That way the water has more time to become vapor to take full advantage of heat transfer and to prevent hydrolock...

Good luck.

 

There are two main reasons to run water injection:

1) Detonation Supression
2) Intake Charge Cooling

Placing the injection nozzle(s) closer to the throttle body, or in the individual runners, will aid in detonation supression as this will steam clean the combustion chambers and prevent any hotspots on the head and/or piston crowns. After seeing the results of this from my first system on my old GVR-4 I was AMAZED at how clean everything came out. Carbon, oil, gunk (technical term /ubbthreads/images/graemlins/wink.gif ), etc. was completely gone with nothing but shining, clean aluminum present. [guinness]BRILLIANT![/guinness]

Placing the injection nozzle(s) as far as possible from the intake manifold will aid in charge temp cooling. A good, fine mist and a longer path to the engine will allow the injected water to turn to steam and remove heat from the intake charge.

On the system in my Talon I've got four small Aquamist nozzles in the runners of my Cyclone intake for detonation supression and one Bete PJ nozzle spraying directly into the compressor of my Green for higher boost charge cooling. A varying mix of methanol and water is being used. /ubbthreads/images/graemlins/devil.gif

 

[cheekychimp]

Pisces,

How long have you been running this? I have been following the thread on one of the WI forums about pre compressor injection for a considerable period of time and the results have been extremely impressive. The only side effect is possible damage to the impeller if the water is not atomized correctly. Have you checked the turbo for any damage. I am extremely interested in this, but have not been able to decide on which way round to inject. Propane pre compressor and water at the intake or water/alcohol pre compressor and propane at the intake.

Paul

 

Paul,

I'm not running pre-compressor injection yet. Waiting on a piece to get machined which will hold a water nozzle right in the middle of the intake pipe pointed directly at the compressor wheel nut, very similar to this piece that Richard at Aquamist fabbed up for his GT-R. I'll be running a different style nozzle, the feed line will run up through one of the support arms and there are only two support arms as opposed to the three in that picture, but the general idea is the same.

I'm not overly concerned about compressor wheel wear on this setup.

First off, the Aquamist nozzles aren't ideal for spraying pre-compressor. While these nozzles do produce a nice spray pattern, they also produce some droplets that are much too big and can damage the compressor blades. I'm running a Bete PJ nozzle which was designed for fogging, not spraying or misting. I've compared this nozzle to an Aquamist nozzle side by side and it produces a much finer mist at 100 psi. Hands down, the Bete nozzle is the winner for this application. Don't get too wrapped up in droplet size and misting/fogging though, it's a different jet for a different purpose.

Second off, I'm running the nozzle directly in the middle of the airstream injecting parallel to it, not at a 90* angle to airflow injecting perpendicular to it. This, along with the no-so-ideal Aquamist jet, is in my opinion the reason that many have seen premature wear on the compressor wheel when running pre-compressor injection. The spray pattern will go directly into the compressor wheel, as opposed to being slammed against the far wall of the intake pipe where it can fall out of suspension and run along the edges as very large droplets. Yeah, the compressor wheel doesn't like to injest big drops of water at 100,000 RPMs!

And finally, the price of Garrett 50-trim wheels (FP Green) isn't astronomical. If I've got to yank the center section/compressor cover and have FP put on a fresh wheel and rebalance it after 10K miles it surely won't break my heart. Now, if you're running one of FP's 30-Series turbos or a nice full Garrett GT-Series turbo setup you may want to think twice about compressor wheel replacement... /ubbthreads/images/graemlins/shocked.gif

This will start out with a 2/3 methanol, 1/3 water mix, but I may work my way up to straight methanol and a dash of Top Lube depending upon the results. Haven't even looked into or considered propane injection.

EDIT:
The pre-compressor nozzle ring I'm having machined up is not cheap, but if someone else is interested in this and is running a 3" intake pipe an order of 2-3 units would drop the price to 1/2-1/3 of the price of a single unit. Yes it's a very unique setup and a long shot for others who may be interested, but I thought I'd throw it out there anyway.

 

Since my last post was quite long I thought I'd space things out a bit.

I've got a ton more research and info on single-port, direct-port, pre-IC and pre-compressor injection. And though I know Ken Inn hates complex W/I systems /ubbthreads/images/graemlins/wink.gif I've got a lot of research also into the system as a whole, safeguards, injection methods, pumps & accumulators, etc.

I love discussing this stuff so feel free to pick my brain a bit.

 

[cheekychimp]

Pisces, I'll PM you, we have a lot to talk about.

/ubbthreads/images/graemlins/grin.gif

I'd also like that piece but I'm going to need a 4 inch one could that be done?

 

Quote:
Pisces, I'll PM you, we have a lot to talk about.


Unless it's ultra-double-secret info, I'm sure others would like it to be public. I've got nothing to hide with my setup. If my research can benefit others I'd be happy to share.

Quote:
I'd also like that piece but I'm going to need a 4 inch one could that be done?


The entire design and machining program would have to be rescaled thus throwing out the window any cost savings on bulk orders. DOH! I can, however, get you started in the right direction. PM me and I'll get you the details (this is ultra-double-secret info... /ubbthreads/images/graemlins/wink.gif ).

 

[cheekychimp]

Quote:
The entire design and machining program would have to be rescaled thus throwing out the window any cost savings on bulk orders. DOH! I can, however, get you started in the right direction. PM me and I'll get you the details (this is ultra-double-secret info... ).



I realized that I wasn't going to be much help for your cost sharing program, but I don't want to go back to a three inch intake. I'll PM you on this aspect.

As for the other stuff, you sound like you have been knocking around on the WI forums as well, so what have you gone with a full aquamist kit or something you designed yourself using a Shurflo. I kind of like the simplicity of the Shurflo but I know the diagnostics of the aquamist kit doesn't work so well with a high pressure pump. A lot of guys have had issues with aquamist pumps breaking but Richard says that whilst most were replaced, a lot of damage occurred because they were mounted in ways or locations that weren't recommended or allowed to run dry. If that's the case it seems a bit unfair to blame the pump design. Would you blame a bearing or piston manufacturer if you ran your engine without oil in it and it seized?

As for direct port injection, how do you have you failsafes setup. Are you just monitoring flow in the main supply line until it branches into the four ports or do you have a flow detector on each nozzle line at the ports?

Finally how is your system set up with the 5 nozzles. Are you going to spray simultaneously or at different points for the compressor and manifold? And are the four port nozzles the same size?

Paul

P.S. A third reason for injecting pre compressor you haven't mentioned is that it has actually been shown to extend or shift the compressor map giving rise to increased flow at the top end.

 

Quote:
I realized that I wasn't going to be much help for your cost sharing program, but I don't want to go back to a three inch intake. I'll PM you on this aspect.


Drop me a line, I'll set ya up.

Quote:
As for the other stuff, you sound like you have been knocking around on the WI forums as well, so what have you gone with a full aquamist kit or something you designed yourself using a Shurflo. I kind of like the simplicity of the Shurflo but I know the diagnostics of the aquamist kit doesn't work so well with a high pressure pump. A lot of guys have had issues with aquamist pumps breaking but Richard says that whilst most were replaced, a lot of damage occurred because they were mounted in ways or locations that weren't recommended or allowed to run dry. If that's the case it seems a bit unfair to blame the pump design. Would you blame a bearing or piston manufacturer if you ran your engine without oil in it and it seized?


Yep, I've been on Aquamist's forum for quite some time now. The pre-compressor injection thread REALLY got the gears turning in my head and spurred all this on.

As for my system, right now the only Aquamist parts are the nozzles in the manifold. I really like the FIA2 and HSV, and as the funds allow I'm going to integrate these into my system. And yes, my system is kind of Frankensteined together, but it works great as-is and all the parts and pieces are dead-reliable.

I like all of the Aquamist parts, except for the pump. I wouldn't run that thing if you paid me to. A buddy with a Stealth TT had one fail on him and he popped a head gasket because of it. Not catastrophic, but could have been. It's almost as if you look at those things the wrong way they'll fail. The Shurflo pump gets my nod, it's a beast of a pump! It can flow all the water I'll ever need it to, can be run dry for short periods of time and it repressurizes the system in under a second. The only downside is that it doesn't work with the blocked flow diagnostics of the Aquamist system. For the reliability of the Shurflo pump I'll give up that feature.

Quote:
As for direct port injection, how do you have you failsafes setup. Are you just monitoring flow in the main supply line until it branches into the four ports or do you have a flow detector on each nozzle line at the ports?


For failsafes on the nozzles going into the manifold, I've got a stainless solenoid and Aquamist check valve right before the line branches off to the four jets. Again, not cheap but I didn't want to cut any corners or risk a methanol backfire.

Also, I've kicked around the idea of integrating a flow sensor and Aquamist DDS3 system, but again it's funds-limited.

Quote:
Finally how is your system set up with the 5 nozzles. Are you going to spray simultaneously or at different points for the compressor and manifold? And are the four port nozzles the same size?


The current setup via my pimp MS Paint skillz:


The reservoir is a 3 gallon fuel cell from Summit and has a GM 0-90 ohm level sensor and gauge integrated into it. The pump is a Shurflo 100 psi unit and the 24oz accumulator is from Shurflo also. I had kicked around the idea of a smaller "water hammer," but I like the larger Shurflo accumulator because it prevents the pump from constantly running and the fact that it's designed to work with the Shurflo pump. The pressure switch on the pump is set so it turns off at 105 psi and kick back on at 88-90 psi. A few strokes of the pump and the system is right back up to pressure.

The manual on-off valve is a simple ball-valve from the local hardware store. I believe it was designed for propane use, but it works great for this application, seals perfectly and is rated to 600 psi. Don't think I'll be going anywhere near that. /ubbthreads/images/graemlins/wink.gif The filters are 40 & 90 micron inline units from McMaster-Carr (part #s 98355K841 & 98355K842 respectively). The solenoids are from McMaster-Carr also and are wired in to two Hobbs switches set to 15 psi for the direct port nozzles and 22 psi for the pre-compressor nozzle (haven't activated this one yet).

The check valves are from Aquamist and have a crack pressure of ~15 psi. Though they're not 110% needed, I prefer to have them in there to prevent any water/methanol from getting sucked into the engine while under vacuum.

The nozzles in the manifold are from Aquamist also. Right now I'm running four .4mm nozzles, but may up that as boost increases. The nozzle in front of the compressor is a PJ15 from Bete. Like I said, it was meant for fogging purposes but it works great for this.

Quote:
P.S. A third reason for injecting pre compressor you haven't mentioned is that it has actually been shown to extend or shift the compressor map giving rise to increased flow at the top end.


Yep, this is my main reason for injecting pre-compressor and what got my mind going on all this craziness. Running higher boost levels and injecting pre-compressor will bring the efficiency back to the left on the compressor map. I can only wring so much out of this 50-trim wheel before it starts to blow really hot air. And once I've tapped that resource I guess there'll be no choice but to sneak a Red in there... /ubbthreads/images/graemlins/devil.gif

 

[cheekychimp]

That's actually a really neat setup. I'd never seen the Shurflo accumulators before but that actually makes a great deal of sense to use one with the Shurflo pump. What are you using for hose and fittings?

And where is the manual on/off valve located, in the cabin or at the trunk?

I also found these on the cooling mist site

Flow/Clog detectors

They are supposed to detect blocked nozzles even on the Shurflo pump. Expensive at $50 a throw but if they can be rigged up with the DDS3 gauge they might be worth a look. I really need to decide how I am going to hook my stuff up and put together a kit. There are quite a few components on the cooling mist site. Also what is the advantage of aquamist's HSV over other solenoids?

Paul

 

Will fuel injectors flow water? Would it be possible to use them for water injection that way they could follow the fuel curve?

Jay

 

[cheekychimp]

There's no doubt that they work, the problem generally is for how long. Most injectors don't suffer from oxidization (rust) if they have petrol run through them. Consequently only limited injectors are built with stainless or rust proof internals. If you can find such an injector or injectors it could be set up to work extremely well, if not it will work extremely well until rust builds up then performance will become sporadic and eventually it will fail.

Finding another injector to replace the damaged one probably wouldn't be a problem, but the possible engine damage resulting from an injection failure during a lean condition would be.

But before you give up on the idea, think of it like this. You trust your injectors to avoid detonation by pumping gas into your combustion chambers every single day you run your engine. If you can find an injector that is impervious to damage from water and/or methanol you would be halfway to a system basically as reliable as your current fuel system.

As an interim or alternative probably (given the cost) Aquamist does do a kit that is mappable and actually uses an injector driver to inject water/alcohol dependent on load but still does it using a pump and nozzles.

And that's all I've got to say about that !!!

 

[cheekychimp]

Also, can you use a fuel cell with foam in it as a water injection tank? I thought I heard somewhere that the water reacts with the foam so you have to remove it.

 

Quote:
That's actually a really neat setup. I'd never seen the Shurflo accumulators before but that actually makes a great deal of sense to use one with the Shurflo pump.


The Shurflo accumulators are quite nice. It's actually just a tad bigger than the pump itself and can hold a great deal of water. The upside to this is that the pump doesn't have to cycle on and off as often. Stays cooler and pump life is extended a touch. The downside is that if there is a solenoid failure a good deal of pressurized water will get injected when it's not supposed to. The moral of the story? Buy quality solenoids! /ubbthreads/images/graemlins/grin.gif

The only other downside to the Shurflo accumulator is that it's only rated to 125 psi. Limits the Shurflo pump selection to the 100 psi and lower range. Not that the Shurflo 150 psi pump is absolutely necessary, but the additional pressure, flow and atomization ability would be nice. Guess you can't have your cake and eat it too...

Quote:
What are you using for hose and fittings?


All the fittings from the fuel cell in the hatch up to the solenoids in the engine bay are brass NPT and compression fittings. From the solenoids to the nozzles I used "Instant-Tube"/push-lock fittings for ease of disassembly if needed. The line is simple 1/4" OD plastic tubing sourced from the local Home Depot. Nothing fancy, but it works great.

Quote:
And where is the manual on/off valve located, in the cabin or at the trunk?


The on-off valve, cell, pump and accumulator, are all located in the hatch. I wanted to retain the spare tire for use on the occasional longer trip, and it makes for quite a tight fit in there but everything is tucked in nicely.

I've got to imagine that with a little creative plumbing the same could be done in a GVR-4.

Quote:
I also found these on the cooling mist site

Flow/Clog detectors

They are supposed to detect blocked nozzles even on the Shurflo pump. Expensive at $50 a throw but if they can be rigged up with the DDS3 gauge they might be worth a look. I really need to decide how I am going to hook my stuff up and put together a kit. There are quite a few components on the cooling mist site.


Those flow detectors are quite nice! They'd definitely work well to show when the nozzles are injecting. Yes, they're a bit spendy but in the long run I imagine it'd be worth the cost. I'm not 110% positive, but I think the DDS3 gauge comes with a flow detector from Aquamist. Wouldn't have to buy the Coolingmist piece.

Quote:
Also what is the advantage of aquamist's HSV over other solenoids?


Aquamist's HSV (High-speed Solenoid Valve) is very similar to a fuel injector in that it's designed to pulse on and off very rapidly. Think duty cycle, similar to a fuel injector. A regular solenoid, such as the ones I'm using from McMaster-Carr, is meant to basically be an open/close switch and can't pulse nearly as rapidly as the HSV. When used with Aquamist's FIA2 or a capable standalone ECU the HSV will deliver water at the same rate that fuel is being injected. A nice setup for precise water delivery.

Quote:
Also, can you use a fuel cell with foam in it as a water injection tank? I thought I heard somewhere that the water reacts with the foam so you have to remove it.


I'm sure you can use the foam in the cell with a water/meth setup. When I received the fuel cell there was a little sheet of paper inside the box stating that the foam was compatable with gasoline, methanol, ethanol and most all other alcohols. I can't see water being a problem. /ubbthreads/images/graemlins/wink.gif

Me, I took the foam out. Mainly this was to accomodate the fuel level sender. In hindsight I probably should have kept it in there, but oh well...

 

[cheekychimp]

What's this guy using to connect the Aquamist quick connect fittings to the AN fittings of the fuel cell?

 

Quote:
What's this guy using to connect the Aquamist quick connect fittings to the AN fittings of the fuel cell?


Not 110% sure what kind of adapter fitting is being run there. Looks like it could be an AN-8 female to 1/8" BSPT female adapter, but that'd sure be a tough one to dig up! The bulkhead AN fittings are just screwed into the fuel cell, though, and can be easily removed and replaced with the proper bulkhead fitting for your application.

I must admit that I did lie (omit info) a bit about my setup. The short section of hose coming out of the fuel cell and into the pump is AN-8 braided hose. The fitting on the fuel cell was AN-8, I had a straight female fitting and some braided hose lying around the garage from a prior project so rather than reinventing the wheel and using a different fitting and line I ran the stainless hose from the fuel cell to the pump. Totally glossed over this in my head.

Also, I find it interesting that the Aquamist pump is mounted in the trunk on that setup. Aquamist recommends running the pump as close to the nozzle as possible as it's not designed to be a "pusher" pump. It can be done with a dual pump setup or a voltage increase, but it's not the ideal way they recommend setting it up. Yet another reason to run the Shurflo pump. Mount that hoss in the hatch/trunk, plumb it and it'll happily pump away no matter what the conditions.

 

Just to add something on a side note, I emailed the guys at DSMlink who said v.3 will have generic output controls. These will enable you to run water injection with a flow map. I also have someone who works for a fuel injector and turbo shop in Redding searching for the ideal injector for this application.

Jay

 

[cheekychimp]

Quote:
The Shurflo accumulators are quite nice. It's actually just a tad bigger than the pump itself and can hold a great deal of water. The upside to this is that the pump doesn't have to cycle on and off as often. Stays cooler and pump life is extended a touch. The downside is that if there is a solenoid failure a good deal of pressurized water will get injected when it's not supposed to.



I was thinking about this today. Given that we activate the system with a pressure switch, why not have a normally closed solenoid right after the accumulator that opens at a lower boost pressure than the main one, say 4-5 psi. With no boost and when the car is off, this low boost solenoid would be locked shut so that water couldn't leak into the engine, but it would always open and pressurize the system before the injection was activated by the main solenoid.