E85: A word to the wise

[Xero-Limit]

Good info and some valid points here, but in context of the BRZ/FR-S, it is not as crucial to have the exact mixture be known.

The FA20 makes best torque at xx amount of degrees, the problem is you run into the knock cycle on pump gas there. So you back down about 2-3 degrees. OEM backs down even more. With ethanol your octane will go up progressively with the mixture, and at about 50-70% you can hit MBT without any knock. So at that point any further advance will actually cost you power as it is combusting while the piston is traveling up. Not ideal but not instant carnage. Add E85 and you're not really making much more compared to E70, but you'll have less likely knock sensor activity when it gets hot or heat soaked and more prone to knock. So that takes care of the timing.

Fuel is another story. You need to compensate for the additional demand, and while you can use injector value, MAF, and other means; the ideal is by using fuel compensation on the FUEL side (NOT MAF) so the load values don't change. With EcuTek this is done via custom mapping.

The stock ecu can trim about 40%, more when you combine them all. However as stated above it takes time to get there and not ideal. But if you start off relatively close, the trimming will be fairly quick.

Back to fuel, max power on E85 normally aspirated will happen somewhere around .75-.85 lambda. Difference between those will not be very big as we found out on the dyno. So with map switching what we do is set an E85 map, target a lambda of .8, and time so it learns towards MBT. Say you get E70, it will richen up more to .75 until it trims back to .8. Then the learning will take to MBT as long as there is limited knock activity. This way you're safe, it is within a couple of HP whether you use E70 or E90, and no issues or flex fuel kit to worry about (and I do alter the learning so it is a bit quicker)

For E50 mixes we do the same thing but target E50. Pull back a few degrees of timing, drop the fuel increase in half, and this allows the ECU to trim the final 10% that differs between E40-E60. Timing targets are in a safe range and the ECU will learn the same way it does stock.

Then the pump gas map can be used with E20 and lower.

All of this applies to NA setups ONLY. At 12.5 CR with DI you don't need much of an octane boost to reach MBT, but with boost you'll want flex fuel. This is where MBT becomes highly dependent on the AFR and the boost level, and you have to make custom maps using the flex fuel sensor input (Zeitronix works fine with EcuTek).

This at least has been the experience here with several cars and datalogs. I purposefully ran the car with E30-E70 mixtures to see how it reacted, and as long as you either mix 50-50 or use either pump or full e70-e85, flex fuel is not a necessity IMO. Not that it won't be better (and it will), but since it is still expensive and not legal in STX (SCCA) map switching can work great too if setup appropriately.

[RehabJeff86]

I see that FA20Club has the flex fuel setup available on their website, but no one has it yet??? Did a search and no user feedback...

[RehabJeff86]

Originally Posted by moto-mike

Good info and some valid points here, but in context of the BRZ/FR-S, it is not as crucial to have the exact mixture be known.

The FA20 makes best torque at xx amount of degrees, the problem is you run into the knock cycle on pump gas there. So you back down about 2-3 degrees. OEM backs down even more. With ethanol your octane will go up progressively with the mixture, and at about 50-70% you can hit MBT without any knock. So at that point any further advance will actually cost you power as it is combusting while the piston is traveling up. Not ideal but not instant carnage. Add E85 and you're not really making much more compared to E70, but you'll have less likely knock sensor activity when it gets hot or heat soaked and more prone to knock. So that takes care of the timing.

Fuel is another story. You need to compensate for the additional demand, and while you can use injector value, MAF, and other means; the ideal is by using fuel compensation on the FUEL side (NOT MAF) so the load values don't change. With EcuTek this is done via custom mapping.

The stock ecu can trim about 40%, more when you combine them all. However as stated above it takes time to get there and not ideal. But if you start off relatively close, the trimming will be fairly quick.

Back to fuel, max power on E85 normally aspirated will happen somewhere around .75-.85 lambda. Difference between those will not be very big as we found out on the dyno. So with map switching what we do is set an E85 map, target a lambda of .8, and time so it learns towards MBT. Say you get E70, it will richen up more to .75 until it trims back to .8. Then the learning will take to MBT as long as there is limited knock activity. This way you're safe, it is within a couple of HP whether you use E70 or E90, and no issues or flex fuel kit to worry about (and I do alter the learning so it is a bit quicker)

For E50 mixes we do the same thing but target E50. Pull back a few degrees of timing, drop the fuel increase in half, and this allows the ECU to trim the final 10% that differs between E40-E60. Timing targets are in a safe range and the ECU will learn the same way it does stock.

Then the pump gas map can be used with E20 and lower.

All of this applies to NA setups ONLY. At 12.5 CR with DI you don't need much of an octane boost to reach MBT, but with boost you'll want flex fuel. This is where MBT becomes highly dependent on the AFR and the boost level, and you have to make custom maps using the flex fuel sensor input (Zeitronix works fine with EcuTek).

This at least has been the experience here with several cars and datalogs. I purposefully ran the car with E30-E70 mixtures to see how it reacted, and as long as you either mix 50-50 or use either pump or full e70-e85, flex fuel is not a necessity IMO. Not that it won't be better (and it will), but since it is still expensive and not legal in STX (SCCA) map switching can work great too if setup appropriately.

Hi, does this translates into with your tune, I dont need a ethanol content sensor, your tune will adjust and learn by itself?

[tech4pdx]

Originally Posted by moto-mike

Good info and some valid points here, but in context of the BRZ/FR-S, it is not as crucial to have the exact mixture be known.

The FA20 makes best torque at xx amount of degrees, the problem is you run into the knock cycle on pump gas there. So you back down about 2-3 degrees. OEM backs down even more. With ethanol your octane will go up progressively with the mixture, and at about 50-70% you can hit MBT without any knock. So at that point any further advance will actually cost you power as it is combusting while the piston is traveling up. Not ideal but not instant carnage. Add E85 and you're not really making much more compared to E70, but you'll have less likely knock sensor activity when it gets hot or heat soaked and more prone to knock. So that takes care of the timing.

Fuel is another story. You need to compensate for the additional demand, and while you can use injector value, MAF, and other means; the ideal is by using fuel compensation on the FUEL side (NOT MAF) so the load values don't change. With EcuTek this is done via custom mapping.

The stock ecu can trim about 40%, more when you combine them all. However as stated above it takes time to get there and not ideal. But if you start off relatively close, the trimming will be fairly quick.

Back to fuel, max power on E85 normally aspirated will happen somewhere around .75-.85 lambda. Difference between those will not be very big as we found out on the dyno. So with map switching what we do is set an E85 map, target a lambda of .8, and time so it learns towards MBT. Say you get E70, it will richen up more to .75 until it trims back to .8. Then the learning will take to MBT as long as there is limited knock activity. This way you're safe, it is within a couple of HP whether you use E70 or E90, and no issues or flex fuel kit to worry about (and I do alter the learning so it is a bit quicker)

For E50 mixes we do the same thing but target E50. Pull back a few degrees of timing, drop the fuel increase in half, and this allows the ECU to trim the final 10% that differs between E40-E60. Timing targets are in a safe range and the ECU will learn the same way it does stock.

Then the pump gas map can be used with E20 and lower.

All of this applies to NA setups ONLY. At 12.5 CR with DI you don't need much of an octane boost to reach MBT, but with boost you'll want flex fuel. This is where MBT becomes highly dependent on the AFR and the boost level, and you have to make custom maps using the flex fuel sensor input (Zeitronix works fine with EcuTek).

This at least has been the experience here with several cars and datalogs. I purposefully ran the car with E30-E70 mixtures to see how it reacted, and as long as you either mix 50-50 or use either pump or full e70-e85, flex fuel is not a necessity IMO. Not that it won't be better (and it will), but since it is still expensive and not legal in STX (SCCA) map switching can work great too if setup appropriately.

Wow. Excellent info. Thanks for the details!

[feldy]

as a heads up I just placed a order for this tonight. not a bad little price. http://www.summitracing.com/parts/qft-36-e85