Air fuel ratio for max power
 

Assuming i could get my air/fuel ratios to exactly the right mix thru the entire rpm range, what air to fuel ratio would give me the quickest acceleration and time? Starting from 1000 rpms till redline, would it be 12.5-1 or would you only want that ratio once on full boost? Is their any part of the rpm range where you wouldn't want the 12.5-1 ratio? Lets assume my turbo hits full boost at 3500 rpm's and can hold 25 psi to redline, most dyno charts dont start out at the 12.5-1 ratio so im wondering, if it was fully controllable, at what time in rpm range would i want it to hit and when would i want it to end assuming i want max power? Thank you

Here's part of the answer with graphs & pictures:

<a href="http://forums.rennlist.com/cgi-bin/rennforums/ultimatebb.cgi?ubb=get_topic&f=18&t=001921" target="_blank">Topic: Air Fuel Ratio under boost Question : </a>

Front the BMEP chart there, 12.5:1 is where maximum cylinder pressures are created. Highest pressure comes from highest heat as well (combustion). An engine that's designed to run at maximum power should have safeguards to deal with the heat such as sufficient cooling and materials chosen to deal with this heat.

Our 951s appears to have had this design in mind with a decent cooling system with oil-cooler. Ceramic exhaust ports keep the heat out of the head and heat-shields on the headers and exhaust help keep the flow moving quickly towards the turbo. The sodium-filled valves help to deal with the high temperatures. Given all of this, a 12.5:1 ratio would appear to be OK for a street car. However, on a racecar that sees continuous operation at full-throttle, a 12.0:1 ratio may be better. You lose a little power, but you gain durability from lower exhaust-gas temperatures and you'll have a little safety margin in case you run across bad gas or high-temperature days at the track.

Now the question of when it would be best to have 12.5:1 isn't at what boost level, but rather at what load level.

For example, under light-load partial-throttle operation at high-RPM say... 5000RPM, you may be building up 2-3psi of boost. Since you're still under close-loop operation with O2-sensor feedback, your mixture will average out to be 14.7:1 through dithering between 13:1 and 16:1.

However, if you're under full-throttle high-load conditions with the gas-pedal completely floored you can develop the same 2-3psi boost at 3000rpm. In this scenario, the DME will switch over to open-loop mode you'll want to have 12.5:1 for maximum combustion-pressure and power. The more combustion-pressure you can build, the more power you get and the hotter and faster the exhaust you'll get to spool up the turbo.

That's partly why the APE Stg.2 chips are so lethargic in the low/mid-range. They're so rich that they kill off the combustion and the resultant cooler exhaust doesn't help to spool up the turbo as quickly as it can if they had a leaner mixture.

However at no point under full-throttle should you get any leaner than 12.5:1 either. This leads you closer to the detonation territory and harmful damage may result. Especially at high-RPMs where knock & detonation can't be heard and the knock-sensor is useless. So if you've got one of those tidal-wave air-fuel ratio curves, you'll want to at least adjust fuel-pressure or boost such that the leanest spot of the curve doesn't get leaner than 12.5:1.... :D

I'd post a pic of my fuel curve off the dyno if I had it lying around... essentially, you want it to start pretty lean (14.x:1) then richen up when the boost comes on (to around 11.5:1 (10.5:1 maybe?) or even richer if your car starts to ping). The lean beginning will give you better acceleration (less fuel so it doesn't get bogged down), but its VERY unsafe to run that mixture under boost (and it makes more power with more fuel under boost anyway), so that's what my mixture is... We did over 100 runs tuning my fuel curve, and this just plain worked best (again, I'd have to go hunt down a chart to see my exact a/f -- unless Derrek from Huntley has one in his dyno that he wants to print out and post here ;) )

The better solution to all this is to run closed-loop EGT... hope you got some bucks for that setup, though.