Lets Talk Water Meth Injection

[joshuaho96]

On 9/21/2021 at 12:17 AM, fullboost33 said:

Hopefully a bit of a thread revival!

About to purchase a WMI set up for my r33 gtst. Plan is to run 100% water rather than a mix. 

Was hoping to hear from guys running just a post intercooler nozzle and see if there was any feedback on distribution amongst cylinders?

Hoping to get away with a post cooler nozzle rather than the added complexity of direct port - however if that’s the way to go then so be it!

 

The intake is not at all designed to have equal liquid distribution, so I would expect that pretty much all of the water that doesn't evaporate in time is going to end up in the rear few cylinders. So a post-intercooler nozzle would have to inject very little water to ensure it all evaporates before reaching the intake plenum.

I had an interest in water injection at one point but the sheer complexity involved in doing it right is a huge headache. You need to control it effectively as a separate fuel injection system. The injectors need to be much higher pressure than port fuel injectors because you need to atomize it finely for good evaporation. Water is corrosive which means everything in the system has to be designed around that. Water also freezes which could damage the system if you leave the car outside and it falls below freezing overnight. Water can also grow things which is yet another problem to deal with that can clog injectors/lines/etc. To do it right you would have to run AFRs that are normally considered incredibly dangerous, like lambda 1 for the entire operating table at operating temperatures. Then instead of using enrichment to keep the engine from detonating to death you inject enough water to absorb the heat of compression and initial flamefront development. So if the water injection system fails for any reason then you destroy your engine, basically instantly. On top of this water displaces oil, so any liquid water hitting the valve stem will contribute to valve stem wear, any water hitting the cylinder walls will also wash down the walls and potentially contaminate the sump. One journal paper I read where they experimented with water injection they discovered bore scuffing and they also milkshaked the engine oil so it really pays to be cautious.

Nostrum Energy was supposedly working on this on their R35 demo car, I don't know what ever came of it. I bet they ran into issues because despite talking it up I've yet to see them actually ship anything. The Bosch system in the M4 GTS is also known to be troublesome, frequent system failures there as well. Aquamist is basically the only game in town for something you can actually buy today.

[GTSBoy]

On 9/22/2021 at 10:47 AM, joshuaho96 said:

You need to control it effectively as a separate fuel injection system. The injectors need to be much higher pressure than port fuel injectors because you need to atomize it finely for good evaporation. Water is corrosive which means everything in the system has to be designed around that. Water also freezes which could damage the system if you leave the car outside and it falls below freezing overnight. Water can also grow things which is yet another problem to deal with that can clog injectors/lines/etc. To do it right you would have to run AFRs that are normally considered incredibly dangerous, like lambda 1 for the entire operating table at operating temperatures. Then instead of using enrichment to keep the engine from detonating to death you inject enough water to absorb the heat of compression and initial flamefront development. So if the water injection system fails for any reason then you destroy your engine, basically instantly.

Literally every single one of these "problems" has been solved many times over.

[Dose Pipe Sutututu]

Or just run E85 and don't bother with WMI because most of us aren't really competing in WTAC or Texas miles (Australia doesn't even have anything close to that length that's a race prepped surface).

[tridentt150v]

Plus you are saying that there is 'residual' water in your engine....wrong.

WMI isn't used from 0 to 100, you use a management system and only begin injection at a determined threshold.

For me [and from memory] it starts at 8psi then ramps up until it is all in by 14psi......and/or say 50% IDC and all in by say 80% IDC.

So when you are on a trip and you boost, injection begins.  But when you are cruising up your street and parking in your garage etc there is no WMI injected because you are driving well below [or should be] your thresholds.  This means you burn off any residual [if you have any] and also allows the motor time to 'dry out'. 

BTW if you are getting residual at all, then you haven't set the system up properly - you are over injecting.   ie an injector nozzle that is too big for the engine, or thresholds that are way too low to be effective.

I have never had milky oil, my spark plugs aren't rusty wrecks, my bore and valve heads [when using my endoscope on the phone] are clean.  I've been using a 50/50 mix for probably 10 years now...all the bad things should have happened by now if you believe the doomsayers.  But nothing.

And you don't really need separate injectors etc, just one - KISS.  The swirl of the air inducted through the inlet is all you need for an even distribution if you position your nozzle correctly.  Maybe for a drag or time attack etc you will see minute gains, but for the street...forget it.

I have no access to E85, hence the WMI.

And the final word...my WMI system works, it lowers temps, allowing more of everything.

 

 

[tridentt150v]

Link to my old old thread:

 

[joshuaho96]

On 9/21/2021 at 9:19 PM, GTSBoy said:

Literally every single one of these "problems" has been solved many times over.

I know there are flow sensors that you're supposed to use, Aquamist has nozzles that are PWM instead of the old style pump duty cycle, Bosch's solution to water freezing is to pump all water back into the trunk after you turn the key off to empty the lines, Aquamist tells you to add methanol to keep it from freezing/keeping stuff from growing. General point is to get it to truly OEM levels of integration like what you get in an M4 GTS is just such a huge headache. As I said, even the Bosch system in the M4 GTS has a known issue with valve block corrosion, lots of people reporting the filter clogging easily, etc. 

I still think about doing it sometimes because it hurts my soul to think about how perfectly good fuel is being injected purely for cooling purposes instead of actually burning properly, but it really is not a nice straightforward thing to do. I would want to put the injectors on the bottom of the Nismo plenum because that's where the bosses are for secondary injection. But that would mean even more crap that has to fit under there, a set of injectors and the valve block and the water lines. Putting the injectors on top with a bracket for the valve block could also work, but that's going to be competing with the ITB linkages for clearance. 

On 9/21/2021 at 9:36 PM, tridentt150v said:

Plus you are saying that there is 'residual' water in your engine....wrong.

WMI isn't used from 0 to 100, you use a management system and only begin injection at a determined threshold.

For me [and from memory] it starts at 8psi then ramps up until it is all in by 14psi......and/or say 50% IDC and all in by say 80% IDC.

So when you are on a trip and you boost, injection begins.  But when you are cruising up your street and parking in your garage etc there is no WMI injected because you are driving well below [or should be] your thresholds.  This means you burn off any residual [if you have any] and also allows the motor time to 'dry out'. 

BTW if you are getting residual at all, then you haven't set the system up properly - you are over injecting.   ie an injector nozzle that is too big for the engine, or thresholds that are way too low to be effective.

I have never had milky oil, my spark plugs aren't rusty wrecks, my bore and valve heads [when using my endoscope on the phone] are clean.  I've been using a 50/50 mix for probably 10 years now...all the bad things should have happened by now if you believe the doomsayers.  But nothing.

And you don't really need separate injectors etc, just one - KISS.  The swirl of the air inducted through the inlet is all you need for an even distribution if you position your nozzle correctly.  Maybe for a drag or time attack etc you will see minute gains, but for the street...forget it.

I have no access to E85, hence the WMI.

And the final word...my WMI system works, it lowers temps, allowing more of everything.

 

 

I was just referencing a specific journal paper where they were injecting a pretty crazy amount of water in an attempt to achieve ignition timing on par with say 100 RON with 91 RON fuel + water injection. It's likely that they were completely overdoing it, especially because they found that water injection timing coinciding with IVO was a source of problems. They just show up, warm up the engine, then beat on the engine for hours and hours for months to collect all their data. My goal would be to completely eliminate boost enrichment with water injection, so to me per-cylinder control of water injection is a big deal.

I actually found the paper, turns out it was the Nostrum Energy people supporting some university research, but I don't think they were alone in their findings: nostrum_sae_water_injection.pdf

  

On 9/21/2021 at 9:34 PM, Dose Pipe Sutututu said:

Or just run E85 and don't bother with WMI because most of us aren't really competing in WTAC or Texas miles (Australia doesn't even have anything close to that length that's a race prepped surface).

I think this is probably where I'm headed, just because it's so much easier to run a brushless pump, run higher flow injectors, add a flex fuel sensor to the fuel filter outlet and the rest is entirely ECU.

Edited September 22, 2021 by joshuaho96

[GTSBoy]

No, but seriously, there have been thousands of people in Australia running WMI, with proper ECU integration (giving all the protections to reduce boost and/or timing and/or add fuel back in if the water stops), no freezing (only need a little methanol for that, let alone the typical 50/50) and all the other goodies. It's old science.

You can quite literally keep adding water/meth right up until the point where you quench the spark and the engine will keep gobbling it up and making more power. I can't remember the numbers exactly, but it is in the order of as much water/meth as fuel. Might even be a multiplier on fuel rate.

[fullboost33]

I should have been a bit more specific about the set up sorry, I’m going for an Aquamist kit and wiring the failsafe into my ecu etc.

Running the standard 25 side entry plenum with the plazmaman pipe kit (135° pipe into plenum) and would position the nozzle shortly after the intercooler I'm thinking.

Thanks for the link to your install tridentt150v I’ll check it out! Sounds like your single nozzle set up works perfectly 

[joshuaho96]

On 9/22/2021 at 12:27 AM, GTSBoy said:

No, but seriously, there have been thousands of people in Australia running WMI, with proper ECU integration (giving all the protections to reduce boost and/or timing and/or add fuel back in if the water stops), no freezing (only need a little methanol for that, let alone the typical 50/50) and all the other goodies. It's old science.

You can quite literally keep adding water/meth right up until the point where you quench the spark and the engine will keep gobbling it up and making more power. I can't remember the numbers exactly, but it is in the order of as much water/meth as fuel. Might even be a multiplier on fuel rate.

I agree it's old science, I agree it can make a ton of power. It's more just that I really need to keep the car reliable and after pulling pretty much everything off the intake side of the engine I'm not sure I want to deal with even more stuff crammed in there. I feel like I can barely even do basic mechanical work on this thing so I don't really want to start going nuts when I could get at least substantial thermal efficiency gains just by going to E85.

On 9/22/2021 at 1:56 AM, fullboost33 said:

I should have been a bit more specific about the set up sorry, I’m going for an Aquamist kit and wiring the failsafe into my ecu etc.

Running the standard 25 side entry plenum with the plazmaman pipe kit (135° pipe into plenum) and would position the nozzle shortly after the intercooler I'm thinking.

Thanks for the link to your install tridentt150v I’ll check it out! Sounds like your single nozzle set up works perfectly 

Side entry plenums tend to deliver more liquid to the center cylinders, front entry plenums tend to favor the back cylinders. Methanol helps a lot with getting the mixture to evaporate more readily but I really do think if you plan on injecting more than just a token amount to drop IATs for in-cylinder effects you probably need to do individual injectors.

[tridentt150v]

On 22/09/2021 at 7:58 PM, joshuaho96 said:

ISide entry plenums tend to deliver more liquid to the center cylinders, front entry plenums tend to favor the back cylinders. Methanol helps a lot with getting the mixture to evaporate more readily but I really do think if you plan on injecting more than just a token amount to drop IATs for in-cylinder effects you probably need to do individual injectors.

I'd question this in an FI setup.  You are basically saying that the some cylinders get more air - based on plenum type.   I'd like to see the real life measurements that back this statement up.  Can it even be measured?  Do you have any?

I'd be interested, make for an interesting read.

[joshuaho96]

On 10/2/2021 at 10:20 PM, tridentt150v said:

I'd question this in an FI setup.  You are basically saying that the some cylinders get more air - based on plenum type.   I'd like to see the real life measurements that back this statement up.  Can it even be measured?  Do you have any?

I'd be interested, make for an interesting read.

https://www.facebook.com/STAVTECH/photos/pcb.1278381832338336/1278380939005092

It can be measured, usually with an EGT probe per cylinder. The RB26 is pretty notorious for running cylinder 6 the leanest and the rear turbo the hottest. I plan on putting a Nismo plenum in mine as I button the whole engine back up to maybe help with that issue a little. From what I can tell it probably doesn't totally equalize the airflow between cylinders, just makes the issue less severe than factory. HKS is supposedly working on a dual plenum design which on paper should make it close enough to not matter, their description starts at 4:03 and shows a CFD comparison between a single and dual plenum design: 

 

The other thing to keep in mind is that with pure water injection there is a ton of energy required to actually turn the water into vapor. If you're going for more than just IAT reduction it implies that you're going to have liquid water in the manifold. If most manifolds already struggle to have equal air distribution between cylinders it's not hard to imagine how hard it will be to get liquid water to distribute equally when it really doesn't want to change directions compared to air.

[tridentt150v]

But again...if you have liquid water or over injection, you are doing something wrong.  We are not talking water....we are talking ultra fine mist.  And your injector jet plus the pressure that the WM is under mists/vapourises the mixture, no other energy - given that some engine heat after injection probably doesn't hurt either.

Running No.6 leanest is as much a fuel problem as any air distribution issue.  For air it would need to get more air for lean, not less.  On an RB25 plenum I can't see that happening to the same extent.

I can understand your arguments, but its just not happening in practice.

BTW that link really doesn't have the data you would need to support your statements.  Pity I was looking for a good read.

[tridentt150v]

One thing I didn't point out......if cylinder No. 6 runs lean because its getting more air then a properly running WMI system is a great advantage in helping with it,  because any cylinder that gets more air.....also gets more WM.   

Because atomised spray from the injector jet into the air stream becomes a direct ratio of air volume. 

And visa versa, any cylinders that get less air - and so run richer - will get less WM.  So WM 'smooths out [note I didn't say evens out]' small intake differences.  So two birds, one stone!!!

  

[Piggaz]

Has there been any documented evidence about this number 5/6 running lean thing or is it something that’s been regurgitated online because someone said it and is now become “gospel”?

I bought into this about 12 years ago when buying the Nismo plenum, I do run IEGT sensors BUT the guy that tunes it said he hasn’t seen this issue even with the standard plenums.

 

[joshuaho96]

On 10/3/2021 at 3:36 AM, tridentt150v said:

But again...if you have liquid water or over injection, you are doing something wrong.  We are not talking water....we are talking ultra fine mist.  And your injector jet plus the pressure that the WM is under mists/vapourises the mixture, no other energy - given that some engine heat after injection probably doesn't hurt either.

Running No.6 leanest is as much a fuel problem as any air distribution issue.  For air it would need to get more air for lean, not less.  On an RB25 plenum I can't see that happening to the same extent.

I can understand your arguments, but its just not happening in practice.

BTW that link really doesn't have the data you would need to support your statements.  Pity I was looking for a good read.

Liquid water is water, the injector droplet size primarily affects surface area. Even with good injectors there is always some amount of liquid that remains on the walls of the intake. Gasoline has a much lower enthalpy of vaporization than water and transient fueling control is a big deal because despite that lower enthalpy it still tends to stick around as liquid on the intake manifold walls. Injecting gasoline to coincide with the intake valve just lifting off the seat tends to negatively affect combustion stability and raises the possibility of pre-ignition from the papers I've seen. With port injection it's pretty hard to screw it up but direct injection you can easily cause LSPI from the fuel that ends up in the ringpack/crevice volume. With water the problems I'm talking about are somewhat easier in certain regards but also worse in others. The higher enthalpy of vaporization means more wall wetting, more water makes it into the combustion chamber without turning to vapor but it's not a fuel so you don't worry about a big rich spike causing problems on decel/overrun. If you want to inject a decent amount of liquid well upstream of the individual runners your intake path needs to have balanced wet flow, which is what carb engines had to do back in the day and even then it's pretty well known that an old small block chevy V8 will need to run rich on average because the cylinder distribution of fuel is not balanced. I believe Haltech recently put out a video where Matt mentioned running into these issues. I would not be surprised if you end up with something like 80% of the upstream liquid water in an RB25 manifold going into the center cylinders and very little going elsewhere.

On 10/3/2021 at 10:32 PM, Piggaz said:

Has there been any documented evidence about this number 5/6 running lean thing or is it something that’s been regurgitated online because someone said it and is now become “gospel”?

I bought into this about 12 years ago when buying the Nismo plenum, I do run IEGT sensors BUT the guy that tunes it said he hasn’t seen this issue even with the standard plenums.

 

The lean cylinder 6 thing is more an RB26 issue and seems to be multiple problems in one. The rear cylinders are the last to get coolant flow so they run hotter by default. On top of that it tends to get the most air. The geometry seems compromised in order to clear the clutch master cylinder. Balancing airflow is a real thing and if you search there are a number of guides/CFD experiments talking about it in manifold design, both for dry flow and wet flow. If you're looking for some hard data publicly posted as far as I can tell nobody likes to share hard numbers for anything in general. I can tell you that a lot of people in my area have blown RB26s and when compression tested usually cylinder 6 is the worst, then cylinder 5. @Sydneykid supposedly has the data here: 

I also tend to believe Nismo's claims, usually they don't just make things up. It would be weird to falsely advertise a problem (uneven cylinder airflow), then claim that fixing it with your plenum will net all of 2% more power for some hilariously high price.

[r32-25t]

You haven’t been around these Japanese junk parts for long enough then 

[GTSBoy]

Nismo sell the same snake oil that HKS et al do, just in a more fanboi friendly box.

[niZmO_Man]

On 08/10/2021 at 2:24 PM, GTSBoy said:

Nismo sell the same snake oil that HKS et al do, just in a more fanboi friendly box.

You pay for the prestige of all that heritage/racing pedigree

[RICE RACING]

There is a new alternative to the others mentioned here that will be released soon, been working on it for a while now.

I've done a bit on water injection as have many others > an external link for people interested in some of the details below its worth a read I think as it covers some of the control methods and hardware details not commonly known or discussed. https://www.syvecs.co.uk/forum/viewtopic.php?f=4&t=499&sid=7033ba860615fc3cf3a7de23768d620d&start=40

[s213b]

Has anyone here running WMI had any problems with their setups since the last post here? I’m looking at a setup for my next build, I’m thinking aquarist but they seem hard to find online. 
 

cheers