stoichiometric vs. AF sensors

[Paul Workman]

Wide band sensors: gasoline vs. alcohol = The question comes up when ethanol is introduced to the fuel, and what accomodation if any is necessary, far as AF reading goes...

Down to cases:

Ethanol is oxygenated, and therefore requires less air to burn the fuel, i.e., given the "stoich" for gasoline is 14.7 parts air to 1 part fuel (14.7:1) , and for ethanol it is 9.1:1.

The O2 sensors measure free oxygen percentage in the exhaust gas (as I understand it). The oxygen is detected by the wideband sensor and then electronically analyzed and interpreted by a processor before being displayed as the AF ratio.

The thing making me is that the wideband O2 sensor is an indirect reading; measuring O2 level in the exhaust, and not the actual air volume entering the engine. Adding or subtracting gasoline will result in a difference in the exhaust O2 %, which eventually results in an AF value readout - but based on gasoline fuel calibration, methinks.

But, suppose we were to switch the wideband sensor over to an exhaust pipe of a motor running on pure ethanol. Would the proper 9.1:1 (actual) AF mixture going into the engine be interpreted by the same exhaust O2% wideband equipment as a 14.7:1??

I think maybe so (based on what I think I know about the wideband sensors and their attendant processors/calibration). Put another way, the O2 percentage in the exhaust gas would have to result in a reading of 14.7:1 in order for the actual alcohol 9.1:1 stoich to be correct when interpreted by a wideband exhaust sensor calibrated for gasoline...NO?

One of you automotive engineering types prolly know, huh?

P.

[LGAFF]

Paul that is where lambda comes in...

[1990 quasar blue]

So you're saying for example if I strap my car to a dyno and hook up the wideband that it should be calibrated for the 10% ethanol that we are running here as opposed to 100% gas? If so, that makes sense to me. I believe Dom recently said that we can change stoich in the calibration from 14.7 to 14.3 which would compensate for the 10% ethanol. That alone leads me to believe you're right about a wideband needing a calibration change as well.

[Paul Workman]

Originally Posted by 1990 quasar blue

So you're saying for example if I strap my car to a dyno and hook up the wideband that it should be calibrated for the 10% ethanol that we are running here as opposed to 100% gas? If so, that makes sense to me. I believe Dom recently said that we can change stoich in the calibration from 14.7 to 14.3 which would compensate for the 10% ethanol. That alone leads me to believe you're right about a wideband needing a calibration change as well.

Well, I read that too (the 14.3:1) but I'm not at all certain that is correct...is my point.

As far as lamda tables go, that refers to injector dwell and consequently affecting the actual AF ratio according to alcohol %.

Let me rephase the question:

What would the wideband rig indicate for AF if the actuual fuel were alcohol with the INTAKE MIXTUR set at 9.1:1, keeping in mind the wideband jig being a constant??

P.

[1990 quasar blue]

http://m.*************.com/c6-scan-a...r-for-e85.html

According to these guys it looks like they're looking for mid 8s for their afr numbers.
(I hope I understood you better this time.)

[LGAFF]

Paul, LAmbda is current air fuel/stoich

Get a copy of Greg Banishes book......ISBN 978-1-932494-90-7



Lambda (λ) is the ratio of actual AFR to stoichiometry for a given mixture. Lambda of 1.0 is at stoichiometry, rich mixtures are less than 1.0, and lean mixtures are greater than 1.0.

There is a direct relationship between lambda and AFR. To calculate AFR from a given lambda, multiply the measured lambda by the stoichiometric AFR for that fuel. Alternatively, to recover lambda from an AFR, divide AFR by the stoichiometric AFR for that fuel. This last equation is often used as the definition of lambda:


Because the composition of common fuels varies seasonally, and because many modern vehicles can handle different fuels, when tuning, it makes more sense to talk about lambda values rather than AFR.

Most practical AFR devices actually measure the amount of residual oxygen (for lean mixes) or unburnt hydrocarbons (for rich mixtures) in the exhaust gas.

[Paul Workman]

Originally Posted by LGAFF

Paul, LAmbda is current air fuel/stoich

Get a copy of Greg Banishes book......ISBN 978-1-932494-90-7



Lambda (λ) is the ratio of actual AFR to stoichiometry for a given mixture. Lambda of 1.0 is at stoichiometry, rich mixtures are less than 1.0, and lean mixtures are greater than 1.0.

There is a direct relationship between lambda and AFR. To calculate AFR from a given lambda, multiply the measured lambda by the stoichiometric AFR for that fuel. Alternatively, to recover lambda from an AFR, divide AFR by the stoichiometric AFR for that fuel. This last equation is often used as the definition of lambda:


Because the composition of common fuels varies seasonally, and because many modern vehicles can handle different fuels, when tuning, it makes more sense to talk about lambda values rather than AFR.

Most practical AFR devices actually measure the amount of residual oxygen (for lean mixes) or unburnt hydrocarbons (for rich mixtures) in the exhaust gas.

Gotcha. And, I see how easy it is to confuse the actual AFR with that predicated indirectly via the O2% in the exhaust (i.e., the wideband rig).

There's two differnt things afoot:

1) The actual AFR at the injectors, and

2) the other is the O2% measured by the sensors which is used to calculate or predict what the actual AFR would be for a given fuel.

Your last statement receives props from Charles O Probst in his book, "Corvette Fuel Injection" ((pp 18, ch 2) ISBN 0-7386-0861-9). Both statements seem to support what I suspected: - In a closed loop operation, O2% in the exhaust remains the same as long as the appropriate AFR exists for the particular fuel being used.

So, if (as Probst and other turners indicate) if a 14.3 or 14.2:1 is good for gasoline, then I'm wanting to draw to the conclusion the AFR for the respective fuel would be adjusted to the proper ratio, as long as the O2% in the exhaust remains constant. In other words, one does not need to run out to buy a wideband rig with multiple lamda tables in it. Simply pretend the oxygenated fuel was gasoline. Then set the fueling tables according to established AFR wideband "readings" and the actual AFR for the oxygenated fuel will fall into place. (Of course software capable of handling flex fuel would monitor specific gravity and make lamda adustments on the fly. Not possible with our old programming or processors, I guess.)

Or so it seems to me at this point

P.

[LGAFF]

Just started reading up on this stuff....the explantions are not my thoughts....hikacked from the book and the web

[XfireZ51]

A 10% ethanol gasoline blend drops stoich to ~ 14.3:1. That means you need a richer mixture to equal the same energy available from a 100% gasoline mixture. I change the Stoich number in the calibration. The ECM will use that to calc the pulsewidth required to achieve that. So, if I use straight gasoline while leaving the calibration at a 14.3 stoich, in closed loop the BLM will be based on achieving a 14.3 AFR but in fact motor will run slightly rich. There's nothing "hard-wired" about 14.7 AFR. That's why I can manipulate the O2 sensor voltages and basically re-define what the ECM considers a stoich condition.

[Paul Workman]

Originally Posted by XfireZ51

A 10% ethanol gasoline blend drops stoich to ~ 14.3:1. That means you need a richer mixture to equal the same energy available from a 100% gasoline mixture. I change the Stoich number in the calibration. The ECM will use that to calc the pulsewidth required to achieve that. So, if I use straight gasoline while leaving the calibration at a 14.3 stoich, in closed loop the BLM will be based on achieving a 14.3 AFR but in fact motor will run slightly rich. There's nothing "hard-wired" about 14.7 AFR. That's why I can manipulate the O2 sensor voltages and basically re-define what the ECM considers a stoich condition.

Excellent explaination, Dom. I totally "get" what you've said and how you're manipulating the ECM to accomdate the 14.3:1. (Looking forward to your "101" presentation at BG!)

Then the remaining question has to do with the readout of the wideband instrument itself when a % of oxygenated fuel (ethanol) is added:

Fact: Measuring exhaust gas content IS NOT a direct measurement of the actual air and fuel entering the motor! That (indirect) method for AFR reading is derived - estimated based on oxygen levels in the exhaust: two separte things - actual AFR and AFR "predicted", but based on other measurements.

So, with respect to exhaust gas analysis, it comes down to the truth or not of this simple statement:

Is the readout of the wideband exhaust sensors the same for a given oxygen % in the exhaust gasses, regardless of % of alcohol in the fuel?

If the answer is YES, then the next question is:

Will adding alcohol fueling to result in an AFR previously established with gasoline IN FACT result in a ACTUAL AFR comiserate with the alcohol percentage? In other words if the AFR was 14.7 on gasoline, will it now actually be approx 14.3 (using your number) even tho the wideband sensor is reading 14.7:1?

If that is true, then to set the actual fueling AFR in the ECM, to get an ACTUAL (say) 14.3:1 at the injectors, the same exhaust gas analysis wideband rig would have to display the original (say) 14.7:1 by indirect measurement, theoretically.

Granted, actual AFR might eventually need to deviate to accomdate other factors, e.g., water content, vapor pressure, etc . Case in point, the stoich for 100% gasoline might be 14.7:1 compared to 100% ethanol that is aprox. 9.1:1. This being the case, the 10% alcohol AFR calculates closer to 14.14, but conventional indications are ≈14.2 or 14.3 might work better.

(For a .54 AFR correction in closed loop, I suppose the ECM just handles it, no?)

Bottom line:

I'm pretty sure that if I use a wideband monitor and establish a particular AFR using some combination of pure gasoline or a mix (likely), that if for some reason the (alcohol) % changes, that tweaking my ECM's fueling to re-establish the original AFR measurement, that the new ACTUAL AFR into the motor will be essentially correct (allowing for minor tweaks, depending on other factors).

If I'm wrong with reguard to the actual AFR when using a wideband, then something else is afoot that I'd like to learn about.


P.

PS: Gotta run...no time to proof...will do later.