Water/meth injection (was: new to ms)

[subydude]

well im new to ms but i was recently checking out the features and i was wondering why there is no form of meth/ water injection control

[jsmcortina]

What did you have in mind?

EDIT:
I think this support should be added - keen to hear how users would like it to work so we can create the code to match.

James

[UnaClocker]

Well MS3 has sooo many injector channels.. Seems like it'd be easy enough to use 1 or 2 of them optionally for driving a set of injectors that have a water/meth mix fed to them. Maybe combined with a generic output that could trigger the fuel pump for that system just before the water injection comes alive. Seems like MS1 had a water injection system more like this, and it got left behind with MS2..

[Paul_VR6]

MS1/Extra had provisions for pump on/off as well as the Aquamist fast acting valve.

[jsmcortina]

I know that MS1 had WI and replicating that feature should be east enough, but what else?

A table of WI pulsewidths? duties? anything?

James

[UnaClocker]

A mapable table would be great, duty cycles like the boost target table would probably work well for that.

[Tjabo]

What you mentioned there James would be great for the people running the High Speed Valve from Aquamist. The unfortunate part is that the valve itself is around $250 if I recall correctly, and there is no viable alternative for a valve that can really do a proper PWM approach to the injection events for water.

The alternative that seems to actually work fine is to run one of the many progressive controllers that PWM the pump to vary pressure. In that case, what I would like is the ability to have the MS3 recognize the output from my flow valve, and then provide a provision to either perform a boost-cut function if flow falls out of range (too low or too high, since too high can indicate a hose fitting blown apart or something) at defined points.

OR, for the people who are running MS controlled turbo boost via a solenoid, the boost level could be limited to a safe level (as opposed to a total boost-cut ala overboost protection) if the flow amount went out of range.

The other failsafe feature that I think might work well for people running large percentages of methanol (and a wideband all the time) is a simple table of AFR @ kPa. Simply specify an AFR at a given kPa, and if the AFR goes above that number, boost-cut kicks in to save your motor and let you know you need to investigate what is going on.

There has been some concern expressed about this proposed method, but my AFR readings in boost are very solid on my datalogs, and if it's the same device I used to tune the car, then it should work fine for a failsafe also. Potentially for people who have located the WB sensor bung in high temp locations close to the turbine outlet, or high pressure locations before a catalytic converter or before a restrictive muffler there could be some issues. For mine though, and I suspect for many others, I think this would work fine to give me a crude but simple and effective failsafe.

What do you think?

Thanks!

Thad

[Black99rt]

I plan to use the ms1e/msIIe boost duty cycle circuit to pwm my pump. I was going to have somebody edit the open loop boost duty portion of the code for MAP on the y-axis of the table, then map out delivery that way. I also need to open an anti siphon solenoid anytime I'm injecting.

Once it was setup, I was going to take a few flow measurements at different duty cycles to help with mapping delivery to be the desired percent of fueling. MS knows approximately how much fuel you are flowing if reqfuel is correct, so eventually it might make sence to use a formula instead of a lookup for this.

It would be really slick if this style of implementation could be added to msIIextra code and hence the msIII code. We already have vendors selling the boost duty mod circuit, it is already documented, and the fet included is beefy enough for driving this kind of pump. (the 150psi ones at least with the current sunk to a seperate ground wire) All we need is the code support.

My particular application can not use a standard pump controller. The twin screw blower is positive displacement so I can't use boost referenced or I'll over inject at low rpm and under inject at high rpm. I can't use MAF referenced because I don't and never did have one. (dodge used speed density)

The failsafe plan was to use the low washer level light on my dashboard to trigger a relay to shut off my water injection system. This way the pump would never run dry as the system would be shut off early. It would also set the table switch input on the MSII to drop back to a more conservative timing table. A manual switch would be wired in series turn the system off if for example I wanted to save meth for later and run low timing for a while. A more advanced version of this implementation would be use use an off the shelf flowmeter failsafe availble from some of the W/I companies to trigger the table switch.

An AFR boost cut failsafe isn't a bad idea regardless of other features or integration with W/I. It would have to have some hysteresis to avoid stuttering on a lean tip.

[JET]

What did you have in mind?

EDIT:
I think this support should be added - keen to hear how users would like it to work so we can create the code to match.

James

Since staged injection adjusts the pw to account for the secondary injectors, is it a possibility to "sum" the difference between primary/secondary injectors when using staged injection? Ms1 IIRC uses bank2 as a reference to pulse the high speed water injector. Aquamist did have a limited run add-on hardware solution (mainly for RX7 users) for their kits, but it would be a nice feature if it could be implemented in software.

[Tjabo]

Hey guys, just a quick note, the AFRsafety system works so well that you can really lean on your water/methanol injection system of whatever type you choose, as long as you run a significant proportion of methanol in the mix, and adjust your VE table values accordingly. James really outdid himself on this chunk of coding!

Thad

[Tjabo]

Hey guys,

I think I have developed an "in" with the Aquatec company that makes the higher pressure pumps being used by many of the kit makers.

The item currently standing in the way of us having a truly affordable high performance water/methanol injection system capable of mapped operation is a proper solenoid. We need one with seals capable of withstanding methanol exposure, capable of holding high pressures (up to 250psi), capable of maintaining a good response time at high pressures, and capable of flowing an appropriate amount of fluid. I'm currently running close to 1 liter per minute into my little four banger, so I think we're going to need something that can flow more than that, or maybe run more than one.

Any thoughts?

[racingmini_mtl]

I have no experience with it but have you looked at FJO's product? Here's a link: http://www.fjoracing.com/products/waterinjection/.

Jean

[Tjabo]

Thanks Jean!

That is a good looking setup, but I think it is pretty much like the rest of the non-Aquamist ones that attempt to control their flow volume through use of pump speed, which isn't all that effective because it doesn't have that much dynamic range. The nice thing from what I think I'm seeing on that FJO product is that like Labonte, they have factored RPM into their tables for fluid delivery. Still, without good range afforded by a PWM style solenoid or injector, it's effect may be more wished for than real.

With the MS3, and the ability that James is building in to be able to PWM a solenoid at different frequencies and in concert with the fuel injector duty cycle, we have the potential to have a high pressure, high delivery system with mappable broad range. AND all at a very affordable price!

The standard methanol compatible 250psi pump is generally available for around $129, and then on top of that all you really need is nozzle(s), fittings, tubing, and the tough part, the solenoid. If a proper solenoid can be sourced, it is possible to assemble all of this stuff for just over $200, and have a system that is better than most (or maybe any other) on the market.

Aquamist is the standard in truly mappable PWM style systems, but they run into the $1,000 range real quick, and they are limited in volume and pressure.

Sorry for the long post!

Thad

EDIT: Just to clarify, the solenoid in most of these kits like my Coolingmist one, and probably this FJO one is just a safeguard against having the water drain into the intake and hydrolock the engine. It is a simple on/off with the pump coming on or turning off, no PWM function.

[racingmini_mtl]

You may be right but I was under the impression that they did use PWM on the solenoid. At least the mention that they're high speed solenoids. But there is almost no technical information and no specs so it's difficult to say one way or another.

Jean

[tk_tech]

hey guys,

i have tested out the cool mist fast acting valve and it seams to work well, i have tried running it two ways. one connected right up to the injector channels, and the second with using the boost control settings and changing the HZ to 400 like asked for by the manufacturer.

this seems to have worked but, i am trying to do this while i am racing doesn't work. unfortunitly i keep picking up time when just tuning on gas, this is good in one way but bad in the fact that i haven't been running the alcohol system because i have wanted to max out on gas first.

in the testing i have done, i able to get a progressive control over the fuel injected, and did see excellent drops in my intake temps by as much as 60-70 degrees. if your using methonol when racing make sure that your running the system opening early enough to see the cooling effect during the pass. i found when running the system only above 4000 rpm, i only saw the temp decrease in the last half of the track.

sorry i dont have much more, for you. everytime i started running the system i was running way to rich. and this doesn't work out to well when you trying to win races. i dont have a home track so testing is at a premium

so if possible can we please get some type of control for this type of system, i still feel like i hacked system to get it to run.

dont know if this helps anyone, one question i have is does anyone know what i should run for afr when running 15% methonol added to my feul curve.

there should be a formula to figure this out, no!!!!!!

thanks
sean

[Tjabo]

Conceivably you can run a little less excess fueling that you would otherwise need for combustion chamber cooling, since the methanol and/or better yet water/methanol suppresses detonation. I've found that with a large flow of 30%water/70%methanol (~850cc/min into a 2 liter turbocharged setup) my engine ran like mad in the mid to high 11s to 1 AFR range. I may have to lean it out to that range again now that ambient temps are down some here in Michigan.

Thad

[Kurt A]

Interesting thread! My thoughts on this are to run a standard 43 PSI boost referanced regulator, and in tank pump. Bosch EV1 style injectors (as many as required) mounted in the intake plenum and be controlled by MS through staged injection. Fluid to be 70 % Meth, 25% water and 5% ATF to lubricate the injectors.
Thoughts, Concerns?
Kurt

[Tjabo]

So you're talking about an in-tank (regular automotive style) pump in a supplementary tank for you water/methanol/ATF mixture?

Assuming I'm understanding correctly, I really want to see how this works out. My only suggestion would be to think about putting your supplementary injectors in the runners as opposed to the plenum. I think that way you probably only lose a tiny bit of the intake charge cooling potential, but you ensure good distribution (I think). Actually I need to think about that last part, as I am already second guessing it based on the unequal airflow into the runners. . . . Hmmmmm. . . .

[Kurt A]

So you're talking about an in-tank (regular automotive style) pump in a supplementary tank for you water/methanol/ATF mixture?

That would be correct.

Assuming I'm understanding correctly, I really want to see how this works out. My only suggestion would be to think about putting your supplementary injectors in the runners as opposed to the plenum. I think that way you probably only lose a tiny bit of the intake charge cooling potential, but you ensure good distribution (I think). Actually I need to think about that last part, as I am already second guessing it based on the unequal airflow into the runners. . . . Hmmmmm. . . .

I have pondered this as well and do not have a decision based on any facts yet. I have also considered injecting the WMA solution prior to the intercooler thinking that a wet intercooler core may allow for even better heat transfer.
Kurt

[racingmini_mtl]

Water injection will be more efficient after the intercooler. The intercooler will get rid of more heat in the air with the highest temperature difference between the compressed air and the ambient air. That's basic physics: higher temperature difference means higher heat transfer. If you inject water before, you will cool down the compressed air which will reduce the intercooler heat transfer.

Jean