How I tuned my water/methanol injection

K-Series Programmable ECU installation questions / support issues
Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

How I tuned my water/methanol injection

Post by Razathorn » Sun Aug 20, 2006 10:00 am

This thread is for those of you who are about to tune water/methanol injection. I'm not a professional tuner. I used "trouble free" windshield washer fluid -- it's suposed to be around 40%+ methanol and has a year round formula that a lot of the subaru folks use.

My car has the jackson racing supercharger kit with the 9psi pulley. The temperatures around here have been > 95 degrees F for about, two months now. Combine that with the fact that the best fuel we can get here is 92 octane and it knocks *no less* than the quick trip 91 and doesn't run near as well, you can imagine the knocking issues I've had at near 100 degree weather with run of the mill 91 octane (the a+b/2 method). I am using the 650cc injectors and thus the k20a2-rsx-650cc-jrsc-10-psi base map as a starting point.

After having to retard ignition *multiple* times across multiple days on the 650cc base map, I ended up being 4-5 degrees retarded on ignition in all boost columns. I previously had the 7psi pulley, and when I was done retarding ignition, the car wasn't too much quicker than it was at 7psi, it just wouldn't stop knocking!

What I did tuning wise was to take the map I had been running and (with edit all tables checked) retard all of the boost columns (using 8psi column for reference) so they were exactly 5 degrees colder than the base map, which according to hondata is tuned to max power, not limited by detonation. This means that off spray, I will be running 5 degrees off the base map, which is the most ignition I had knock tuned out from the base map, just applied to the whole boost map as a safe default since I knew based on the base map that the ignition would probably want to remain the same through the entire rpm range. I set the w/i to spray at 4500rpm at 98% throttle above 5psi (basically WOT after 4500 on the charger). I then removed fuel units in the nitrous control panel in increments of 100 or 50 fuel units (works out to be 1.3 - 2.7% fuel at a time) until the a/f curve started leaning out to around 11.8 - 12.0 on pulls w/ spray. As I did this, I added fuel to the actual high cam fuel map where it lean spiked. This way the over all a/f was around 11.8 - 12.0 and the places ti wanted to lean out a bit I added fuel to the map itself so I could lean the rest of the rpm range out to the same number. When I was all done with fuel, I had added like 2% fuel in one place and had removed 300 fuel units in the nitrous control panel for an a/f of between 11.8 - 12.0 on average. This made it so off the spray it might be a bit rich in that one place I added fuel on the base map.

Next I moved to ignition. Ignition in the base map, as stated earlier, was sitting at 16 degrees in the 8psi column (5 degrees off of 21 degrees from the base map). I started adding 2 degrees ignition at a time via the nitrous control panel. I saw a power increase all the way up to 8 degrees advance (which would work out to 24 degrees at 8psi). There was a 'minor' power gain with 8 degrees advance, but nothing like from 0-2, 2-4, or 4-6, but the car started knocking in one place. I left the ignition advance at 6 degrees advanced which is one degree hotter than the base map, knock free, and slightly below max power.

The car pulls pretty hard, especially compared to being retarded 5 degrees from the base map. It's not nitrous, but it allows you to run full best timing for your setup like you're running good fuel. Given that I'm removing 300 fuel units, that works out to be around 8.3% fuel removed with the nitrous control panel. I have lean protection set at 12.5 for like 250ms, so hopefully if the spray stops, it will go from the 11.8 - 12.0 past 12.5 pretty quickly and cut me. It's worth noting also that the ignition is only one degree hotter than the base map, so it's not really that advanced if spray stopped. All in all, I'm pretty pleased with how easy k-pro made this.
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Mon Aug 21, 2006 11:46 am

Hondata: Can you comment on the tuning method I used in regards to fuel tuning -- was it safe to pull fuel out so that I'm at high 11 - 12.0 on water/methanol or should I have tuned another way. I'm very interested in hearing about fuel tuning with methanol injection -- I realize it doesnt' work exactly the same way with the wideband, any help would be greatly appreciated.

Wayne.
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

User avatar
Hondata
Site Admin
Posts: 8815
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Mon Aug 21, 2006 2:10 pm

My experience was similar to yours - adding water means that you can advance timing and reduce fuel while improving detonation resistance, but it is fairly minor. I advanced 2 degrees and pulled out fuel. Of course, the main problem is that you don't want the water injection to fail (under boost I can smell the methanol, so no problem).

I believe that the water droplet size has a large effect on detonation resistance, and also the position of the injection and water flow volume, so it may be worth playing around with different sized nozzles and re-testing. I use a 50% mix of distilled water and methanol which I get from a hot rod shop up the road, so the water/methanol mix may also make a difference. Again, more testing required.
Hondata

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Mon Aug 21, 2006 3:02 pm

Hondata wrote:My experience was similar to yours - adding water means that you can advance timing and reduce fuel while improving detonation resistance, but it is fairly minor. I advanced 2 degrees and pulled out fuel. Of course, the main problem is that you don't want the water injection to fail (under boost I can smell the methanol, so no problem).

I believe that the water droplet size has a large effect on detonation resistance, and also the position of the injection and water flow volume, so it may be worth playing around with different sized nozzles and re-testing. I use a 50% mix of distilled water and methanol which I get from a hot rod shop up the road, so the water/methanol mix may also make a difference. Again, more testing required.
My primary concern is with running the right afr. I am not detonating until what I think is right at or over MBT, to which I back off 2 degrees.

Seeing as you're a physics guru, I was hoping to get some information on how to determine how much fuel to pull. I'm quite happy with the results I have -- no detonation and ignition advance all the way to full power on 91 octane at 9+psi in near 100 degree weather -- I just want to determine if my end afr is safe, all things considered.

Genreally speaking, even if I wasn't doing injection, 12.0, while a bit on the lean side in my opinion, is still safe -- but when you add methanol into the mix, the stoich point of the mixture lowers (granted, it's not that much with the smal mixture to fuel ratio I'm running). This means that the afr reported in k-pro isn't really the actual air fuel ratio as I understand it, it's the result of the actual lambda combined with the constant for petrol fuel -- correct? I just want to determine if I'm too lean around 12.0 in k-pro considering I've removed 8% fuel and seemingly replaced most of it with methanol.

I've heard that in this case, it's best to tune with the lamda number -- but given that it's a simple calculation to get from labda to petrol afr, doesn't that mean that since I'm at 12.0, I do indeed have a fairly safe lambda, all things considered, regardless of if the 12.0 represents actual end afr.

That's what I'm confused / interested in figuring out.

The snow performance folks suggested tuning to 12.5 and then adding injection... which is just nuts for our motor in my opinion -- you would have next to nothing for 'cooling fuel' under boost.

Any light shed on this subject would be a great help -- I have yet to find a good resource with a 'how to go about tuning methanol injection'. Everyone has their opinion, and hopefully, following these opinions doesn't result in damage. One person who has methanol injection, that I hold in high regard, lives down the street and said that he simply tuned to the right afr, sprayed water/meth, then leaned back to his original afr -- he's been running strong for a long time. Just seems like there's not a good source of information readily available.

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Mon Aug 21, 2006 6:40 pm

I did a little more research and I think I have a better grasp on what's going on with the AFR with methanol in the mix. Please let me know if the following is correct:

When you add methanol to the mix, your end AFR, as reported by k-pro, is NOT actually your air/fuel ratio anymore -- it is what your AFR *would* be if your fuel source was petrol, based on the lambda reading from the wideband o2 sensor. Basically, the o2 sensor reads lambda and kmanager tells you petrol AFR based on the known stoich ratio of petrol and the lambda reading. It seems that regardless of the fuel you use (to a certian degree), you want to be running in the range of .78- .81 lambda (which is around 11.5 - 12.0 petrol afr) on a light to moderate boosted engine like mine. Given that, on methanol injection, a reading of say 11.9 AFR litereally translates into about .81 lambda, acceptable lambda for the engine to run safely, however it does *NOT* litereally mean that there is a air fuel ratio of 11.9:1 any longer since methanol is part of the mixture and it's stoich point is different.

It seems what this leaves us with is the petrol afr range of 11.5 - 12.0 as a 'reference scale' for tuning safe lambda -- even though as we change fuel type, it's no longer representing the actual ratio of air and fuel.

Is that correct?

Second -- since methanol (and the water) both have greater latent heat than that of petrol fuel, you should be able to run a leaner mixture as the methanol fuel and water vapor you are shoving through the engine will remove far more combustion heat. I was not brave enough to lean beyond around 12.0 as I figured with my small spray amount, probably around 15% or less mixture to fuel, it didn't warrant anything much leaner than 12:1.

Very much looking forward to hearing if I have things right -- the internet seems to be filled with misinformation on this subject.

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Mon Aug 21, 2006 7:22 pm

Hondata wrote:My experience was similar to yours - adding water means that you can advance timing and reduce fuel while improving detonation resistance, but it is fairly minor.
My experience was a bit different seeing as how I could use 6 more degrees ignition and make good power. I just simply could not run 9psi on our fuel without pulling 4-5 degrees everywhere. Kansas fuel availability and quality is a joke. All the exotics we have around here go to ONE 100 octane unleaded pump at ONE station that is making a killing off the stuff. While I don't want cali's emissions laws or fuel prices, I could deal with a few more octane points ;).
Of course, the main problem is that you don't want the water injection to fail (under boost I can smell the methanol, so no problem).
I've got lean protection set at 12.5 so hopefully if spray stops, I will get cut. I think I might do additional fuel tuning and get the lean peaks nailed down so I can lower the lean protection to 12.3 or so and be really anal about it.

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

User avatar
Hondata
Site Admin
Posts: 8815
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Mon Aug 21, 2006 9:44 pm

With my RSX I stuggle to stop detonation with the CA pump gas, so I run it at 11.0:1 - 11.4:1 with water/methanol injection.

At first is seems that running a fuel / methanol mixture is going to be a pain when it comes to air fuel ratios (methanol has a 6.2:1 stoichiometric ratio; gasoline 14:6:1 *). The solution is a little understanding on how the o2 sensor works. The o2 sensor does not sense mixture; it senses residual exhaust gas oxygen. There is some translation table which equates a lambda ratio to an oxygen content. Not that the lambda ratio is expressed relative to stoichiometric - so lambda 1.00 is stoichiometric, lambda 0.9 is rich (the same as 13.14:1 for gasoline), and lambda 1.1 is lean. Your AF gauge then takes the lambda value, and multiplies it by 14.7 for gasoline (yes, I know), 9.0 for ethanol, 6.4 for methanol etc.

The trick is realize that the AF ratio is based on the lambda value, which is based on exhaust gas oxygen content, which does not care what molecule is being burn to consume the oxygen.

So when running methanol, if you don't touch the AF gauge, it will read the equivalent gasoline AF ratio. I do this when tuning methanol and ethanol powered cars. You tune a methanol NA car to 12.0-12.5:1 AF, even though the actual AF is in the 5s. The point is that the AF is just a number and the lambda value will be accurate in terms of gasoline AF ratio, so you can continue thinking gasoline AFs no matter what you are actually running. BTW, ethanol cars smell like baking bread for some reason.

So, if you dataloggings shows 12.0:1 with methanol, then it is like an 'equivalent gasoline ratio', and you can continue to use and think in the units which you are familar with without any problems.

* Note the gasoline stoichiometric ratio is not 14.7:1 as commonly quoted. In any case this 14.6:1 is an approximation, as gasoline contains many different hydrocarbon molecules, so 14.6:1 is only for unoxygenated gasoline. Pump gasoline is probably between 14.3 and 14.5:1, not counting the water and sludge.

No more questions which involve a long answer.
Hondata

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Mon Aug 21, 2006 11:43 pm

Hondata wrote:With my RSX I stuggle to stop detonation with the CA pump gas, so I run it at 11.0:1 - 11.4:1 with water/methanol injection.
Did it detonate leaner than 11.4 on methanol spray?
Hondata wrote:You tune a methanol NA car to 12.0-12.5:1 AF, even though the actual AF is in the 5s.
I thought N/A was low 13s, perhaps high 12s, a/f for petrol -- is peak power at a different lambda for n/a methanol car?[/quote]

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

User avatar
Hondata
Site Admin
Posts: 8815
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Tue Aug 22, 2006 11:33 am

Yes, the engine would detonate at much over 11.5:1 AF. This was at around 16 lbs boost on 91 octane fuel.

The peak power AF varies with the lambda sensor, but generally 13.0 - 13.25:1 makes peak power. As a lambda value this is 0.89 - 0.91. With a methanol NA I think we were at 12.5:1 reported gasoline AF, 0.86 lambda and 5.5:1 actual AF. So yes, methanol needs to be richer than gasoline (0.86 vs 0.90). Also you have to be more accurate with methanol fueling as it doesn't like running rich or lean like gasloine.
Hondata

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Sat Sep 02, 2006 10:38 am

Update:

I installed a cooling mist high pressure check valve between the pump and nozzle, re-tuned for DRASTIC humidity changes, ended up with an a/f of 11.9 - 12.0 on spray, and took one degree of ignition out after checking advance and determining there was no value in 22@8psi over 21@8psi, which is what the base map 650cc 10psi map has... so now I have *no* ignition advance over the base map, and off spray, *my* map is at 16@8psi ignition and I advance 5 on spray.

Wayne
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

asutherland
Posts: 8
Joined: Thu Jun 12, 2008 9:30 am

Post by asutherland » Sat Aug 09, 2008 3:05 pm

Hey Wayne, was wondering if you could spare a few minutes to discuss this with me. I'm running a JRSC @ 9PSI on my ITR and having some high IAT and detonation problems...

thx
Andrew

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Sat Aug 09, 2008 6:31 pm

asutherland wrote:Hey Wayne, was wondering if you could spare a few minutes to discuss this with me. I'm running a JRSC @ 9PSI on my ITR and having some high IAT and detonation problems...

thx
Andrew
What would you like to discuss?
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

asutherland
Posts: 8
Joined: Thu Jun 12, 2008 9:30 am

Post by asutherland » Sat Aug 09, 2008 8:37 pm

Razathorn wrote: What would you like to discuss?
Just curious which injection kit you went with, and if you have any suggestions other than/with the injection for reducing my detonation. I'm running pretty decent 94 octane gas, but it seems that if the IAT is even a little warm (around 130F?) it will start to ping after about 4,000 rpm if I have any more than 1 or 2 degrees adv timing.

Also, just wanted to clarify - you're using the NOS control in Hondata to control the meth/water spray? What connection on the ECU do you hook it to?

thanks :) :)
Andrew

PS. PM me for email if you'd prefer that to this.

Razathorn
Posts: 469
Joined: Tue Jun 14, 2005 12:17 am
Location: Lenexa, Kansas, USA
Contact:

Post by Razathorn » Sun Aug 10, 2008 1:03 am

asutherland wrote:
Razathorn wrote: What would you like to discuss?
Just curious which injection kit you went with, and if you have any suggestions other than/with the injection for reducing my detonation. I'm running pretty decent 94 octane gas, but it seems that if the IAT is even a little warm (around 130F?) it will start to ping after about 4,000 rpm if I have any more than 1 or 2 degrees adv timing.

Also, just wanted to clarify - you're using the NOS control in Hondata to control the meth/water spray? What connection on the ECU do you hook it to?

thanks :) :)
Andrew

PS. PM me for email if you'd prefer that to this.
130 is pretty warm, and without water/meth, my car detonates pretty decently as well. I used a snow performance kit, but I would probably go with aquamist next time.

I'm a bit confused by what you mean 1-2 degrees of advance? Like 1-2 TOTAL or over the base map?
2003 black RSX type-s / jrsc @ 12psi / methanol injection / aftercooled

asutherland
Posts: 8
Joined: Thu Jun 12, 2008 9:30 am

Post by asutherland » Sun Aug 10, 2008 8:54 pm

Razathorn wrote: 130 is pretty warm, and without water/meth, my car detonates pretty decently as well. I used a snow performance kit, but I would probably go with aquamist next time.

I'm a bit confused by what you mean 1-2 degrees of advance? Like 1-2 TOTAL or over the base map?
Over base map. Today going up the highway for a couple hours drive I saw temps of 160 F. :( (without even being on it hard)

So the water/meth addition was well worth it huh? Do you know roughly how much lower its made the temp?

thx again
Andrew

Locked