so Brian, the exec summary is:

24 psi on pump with a little methanol.

i'm thinking: fast.

 

Yep 24psi of boost on a half-bridgeported S5 block (S4 rotating assy and 8.5:1 turbo rotors), stock top mounted intercooler, 60-1 HIFI Turbonetics turbo, small-shaft P-Trim setup, with a 0.96 A/R turbine housing. 3" exhaust. ~50% duty cycle (on the high end) on 720/1680 injectors w/ a base pressure of 40psi. Single Walbro GSS341 pump running although I have two in-tank (secondary pump is disabled). EGT's still hovering between 1250*F and 1310*F w/ the probe located just past the turbo in the downpipe. 93 octane pump gas. Air/Fuel Ratios were high 10's to about 11.30's:1. Hammered it in 3rd and 4th gear, no problems. Ran just fine, made a truckload of power, and ran nice and cold. BUR9EQ's all around. Stock NGK plug wires (a new set I just put on). Running 15* leading advance with an 8* split at 15psi and above.

B

 

Damn, that sounds like a hell of a lot of fun

 

brian, you were originally running a 12* split correct? did you notice a difference from just bringing that down to 8?

 

I took a slightly different approach to the methanol injection, mixing it directly into my gas for a 20% meth to 80% pump 91 ratio (by volume). The downside of this is that it requires higher injector duty cycles. Also it can be corrosive to some fuel system components, but my fuel system is pretty much fully aftermarket and alcohol compatible. I am running dual walbros to a -8 feed to 720/1680 combo. This setup does have the advantage that it doesn't require extra tanks or pumps, and also since the fuel is being injected so close to the engine, more of it is going to cooling the charge, rather than being evaporated by contact with manifold or throttle body. However, next season I will most likely ditch this and run a separate methanol injection system. This was more just to experiment with.

My engine setup is a half bridge s4 turbo motor, with Turbonetics T66 turbo at 16 psi, and a stock top mount intercooler like BDC's. I've attached a datalog of a full throttle run from 2nd to fourth. Note the extremely high intake temps by the end, causing my compensation maps to kick in and retard the timing(this is with ambient temps in the 50F range). The car runs perfectly fine on the methanol mixture and seems like it can be tuned slightly leaner than one would on pump.

One thing i've noticed with running a methanol fraction continuously is how clean the O2 sensor comes out. Usually its a nice light brown color, but now you can barely tell it has been used save for the melted anti-seize compound on the threads.

 

Yep. I'm at an 8 degree difference now. It seems to be a bit peppier but only a dyno would prove it.

B

 

Passed 25psi tonight; almost hit 26. Still running cold EGT's on the stock top mounted intercooler. Hello, paradigm shift!

http://forum.teamfc3s.org/showthread...883#post547883

 

Brian
Is there a reason why the datalogs are not going to redline?

 

Nik,

thanks for sharing your meth project with us. i was scratching my head looking at 10 degrees lead timing until i re-read your post and realized it was due to your an automatic correction due to AIT.

one of the problems w mixing alcohol and pump, besides the corrosive nature of meth, is that there aren't as many BTUs in alcohol per volume so you generally have to add more volume than you remove. most of us are up against the injector duty cycle wall as it is, and that would increase the problem.

the other major problem is that alcohol and gasoline do not mix which makes for an interesting situation as to constant delivery of the two?

injecting alcohol separately gets you on the plus side of the issue.

it appears i am going from close to 90% duty cycle (850/1600) on pump to around 70% as i tune up to mid/high 11 AFRs from 10.0 where i am currently.

the addition of alcohol may change what we run componentwise on our cars...

it will significantly change the rotary performance envelope.

just look at BDC... i asked him what boost he was Comfortable running on the street on pump. his answer was 14-15 psi.

he has now run 25-26 with alcohol. (exactly like the buick experience BTW)

with a stock T2 topmount intercooler!

i can appreciate his comment as to "new paradigm."

given the magic of alcohol many of us will be initially airflow-limited by our turbos rather than boost-limited due to knock.

if you look at the 60-1 compressor map you will find BDC is now at the absolute NE corner of the map. no more airflow (60 lb) from that turbo. no more rpm 120,600. time to switch turbos.

there will be lots of turbo-switching.... up your airflow w alcohol.

howard coleman

 

Wow. I'm pretty happy that the turbo I decided on is good up to 30 psi

Mr. Cain, when you come out to Brooklyn, make sure you're prepared to install one of these setups

 

Before I store my car for winter here, i'm going to go to 4x1600 cc injectors and see if i can make it up to 1.5 bar. My turbo should be more efficient than BDC's in that range also. I am doing this more for proof-of concept/testing then a permanent solution.

Over the winter I plan on making my own 100% meth injection system, consisting of 2 1600cc injectors fed by a single walbro (in a standard fuel system configuration). Then I will get a fuel only computer(perhaps a cheap megasquirt) that way I can map out the methanol fraction for various points on the map.

 

Yep. I'm on the highway doing all of this by myself. I don't feel comfortable going mach speed down the highway; I never have. I've done a few top-out-3rd gear runs up to about 7krpm. That's part of the reason for going to the dyno tomorrow afternoon; being able to go another 1000rpm out of it.

B

 

I had concidered doing a similar thing but with 850cc injectors and piggys backing them from the primary injectors. But decided against it in the end I will just wait for the next gen kit to finalise its self before i go any further with my experiments

TBH i was shocked at the data i have got back from my kit. This data forced me to pull my race fuel map and run my road map for safty ( which is out of character )

 

I appreciate the progress being made here.

I do appreciate Brian's use of the word of "paradigm shift" here. This is indeed what is happening.

I am adding the idea of a more of a turbo (GR4042r or GT4294r) to the project list for the winter.

With 860/1680's was hitting 98% duty cycles at 7000rpm with 11.0AFR so this is a much more palatable change for me than more fuel through bigger injectors and higher pressures.

Tony

 

... hence the use of "paradign shift", atleast in this one sense.

I am looking at mid-high 50's% duty cycle TOPS on my current setup with the 60-1HIFI totally maxed out, using 720/1680 combination. I have two fuel pumps but I'm not even using one of them. Funny, huh?

B

 

Based on what I have seen and experienced to date....that is just creepy injector duty. I hope for the same experience soon!

 

I know that with my M7 nozzle, and my Whinebro pump turned WAAAY down, I was only seeing 68% IDCs..

 

there is a thread re delivery sizing...

exec summary.

water approx 300 to 400 cc/min

50/50 water/alcohol aprox 500-600 cc/min

100% alcohol... variable based on objective. i run 1200 cc/min

we should all try to talk cc/minute versus other metrics to make AI more simplified.

M7 denotes.... 7 is gallons per hour. that rate is 63 cc/minute.

so a M7 is 441 cc/minute.

howard coleman

 

Dont forget that the rating of the jet (cc/1) will be taken at full pump pressure. usually 100 psi or 150 psi. IF you vary the supply to the pump for speed control/flow control you will not achieve these figures.

TBH the best way / safest way to tune this is to run the pump at full duty cycle and manipulate the fuel map to get the desired figures.

I am going to be bold and say that claims that you can run a large jet and trim the flow back to achieve the desired flow is utter rubbish and is very dangerous advice. For safe uniform and repeatable result you need to run the pump flat out.
Scott

 

flow rate is generally at 100 psi. obviously, the pressure effects the delivery. we are just ballparking here and maybe a few months out we can split hairs.

most people reading this thread haven't, as yet, done AI and are looking at the option on an initial basis.

the question of course is: what can AI do for me?

the answer is:

it will allow you to run your turbo on pump to the max airflow it can make, forget the one bar stuff........

scott, i respectfully disagree w you about trimming the pump. for instance, my deliverability is just over 1200 cc/min. i start my alcohol at 5.5 psi. if i were to hit the motor w the whole enchilada at 5.5 psi it would drown in alcohol. progressive programs are essential. further, my pump starts initial flow within .2 of a psi. the progressive slope is adjustable. it works just fine, as does BDC's.

i absolutely believe you can trim base fuel around whatever alcohol delivery you set and end up w a highly precise AFR.

my amps arrived and are now in the car so after a day or so for crummy weather i should be back out testing monday. it is supposed to move into the high 50s by thursday so i should be doing a bunch of pulls this week. i expect to dyno my car later this coming week.

stay tuned, we are leaning it out

howard coleman

 

Sorry Howard i need to explain. I have also spoke about this with Richard L.

Unless you have some kind of pulsation damper on the pump system the flo rate will vary quite conciderably as the pump ramps up.

If you could view the pump delivery through the ranges you would see that the flo rate would pulse. In fact the pulsation will reduce flow as much as 100- 60 cc/1 at given points in the flo of the system this is based on a 330cc/1 jet. This is in addition to the amount the pump has already trimed off due to the method of flow control. ie the lower the pump duty the more random the delivery will be and vice versa
I know this to be fact as i have seen it with my own eyes ( or shoked eyes at that )
If you run the pump at 100% duty ie with a staged system you will run the pump at full speed the whole time. If you slow the pump down you will create a random flo rate.

I was warned not to push my tune. After seeing this and on the advice i pulled my race map and went with a road map..


No disrespect meant Howard but unless you have a flow meter then you will be unable to view this in real life and will be in ignorant bliss ( for want of a better phrase )
Scott

 

Morning Scott,

Can see the point your making with that and its quite scary reading! Can see why you switched maps.

I was just wondering how your planning to alter the flow on the new set-up, will it be a switch to a normal fuel injector type injector? So pump will always be on full flow and injector will open close at different rates to vary the amount injected?

edit: sorry just re-read your post above, so its going to be injecting full amount when active but you'll adjust the fuelling to compensate

 

I cant propose any solutions to the problems yet but i am trying to stop others from making a costly error due to the fog created by incorrect info regarding some current systems

SCott

 

scott,

as usual you raise a valid and interesting point.

while the commonly used pump does pulse there are a number of dynamics that enter the mix that help smooth things out.

consider the intake runner.

as the rotor uncovers the runner, vacuum sucks and boost pressure slams the runner volume into the chamber... then the rotor slams shut the entry and the mass of air and AI injectant goes from moving very fast to a complete stop and reverses direction. think mixing.

i have built alot (31) of full tilt 2 liter overhead cam twin Weber DCOE 45 engines for my SCCA road racers (pre rotary days). running them on a engine dyno and putting the load to them and watching the reversion/standoff creep out of the carb velocity stack. the cloud of reversion is of course gasoline fog. it traveled from the carb jet toward the inake valve, bounced off the backside of the valve and traveled back outside the velocity stack of the carb!

reversion is a powerful force and works on our motors to mix the AI injectant.

that's probably why when i look very carefully at my air/fuel ratio logs i see NO variation which would relate to the pump pulsation.

i think of what's going on in our runners sort of like a blender and i do not worry about AI dispersion.

BTW, this dynamic mixing may be a reason to locate our AI nozzles upstream from the butterfly rather than in the LIM... the longer the runner length the greater the mix.

howard coleman

 

the personal experience i've had with meth injection is we always put the nozzle right off the pipe leaving the intercooler and have had great success on the 2jz's. haven't tried a rotary yet will in about a month tho.