O2 sensor (lambda probe)

[silver123]

hi,

I was thinking what the O2 values in measuring blocks mean and what can we do with it.
lambda probe is O2 sensitive sensor and it measure O2 percentage in exhaust gas - exhaust manifold.

So if we multiply value of air mass sensor (g/s) with O2 sensor value (%) we should get the amout of air that reacted with the fuel and if we know the fuel consumption we can calculate the amount of unburned fuel.

But since I newer trace the "fuel consumption" data in engine measuring blocks, I gues that it can be calculated by air mass and O2 sensor?

on the Wiki I found these equations about air to fuel ratio;

λ=AFR/ideal AFR
AFR=air mass/fuel mass

so according to these two equations... AFR=λ*14.7 and fuel mass=air mass/AFR
so fuel mass=air mass/(λ*14.7)

but I do not know if the λ is O2 value, since the O2 value should be voltages transfered to the % (O2% in exhaust).

 

[Uwe@Ross-Tech]

There are two kinds of oxygen sensors:

1 - Old-school switching ones, which basically tell the ECU only whether the mixture is rich or lean. Under normal operation conditions (i.e. less than full throttle) the ECU tries to adjust fueling so that the half the time the sensor says lean, and the other half it says rich, which results in an A/F ratio that averages stoichiometric. Under full throttle, the ECU will then add more fuel (because maximum power is made at a somewhat richer mixture, and on a turbo-charged engine a somewhat richer mixture is needed to keep the EGT within the limits of what the turbo will tolerate). Unfortunately, with these old-school sensors, there is no way to accurately determine the A/F ratio in real-time. You can identify engines using these old-school sensors by the fact that they show voltages in measuring block group 031.

2 -- Newer style linear sensors. These do produce a voltage that is related directly to the current A/F ratio, but it's difficult to correlate that voltage to A/F ratio unless you have a spec sheet for the sensor from the manufacturer. Engines using this type of sensor can be identified by the fact that they show Lambda readings in measuring block group 031. Field 1 is the actual current Lambda, while field 2 is the Lambda that the ECU is trying to achieve at the moment. If you have this type of engine, calculating A/F is trivial: A/F = Lambda * 14.7.

-Uwe-

 

[silver123]

There are two kinds of oxygen sensors:-Uwe-

WIKI
Wideband zirconia sensor
A variation on the zirconia sensor, called the 'wideband' sensor, was introduced by Robert Bosch in 1994 but is (as of 2006) used in only a few vehicles. It is based on a planar zirconia element, but also incorporates an electrochemical gas pump. An electronic circuit containing a feedback loop controls the gas pump current to keep the output of the electrochemical cell constant, so that the pump current directly indicates the oxygen content of the exhaust gas. This sensor eliminates the averaging delay inherent in narrow band sensors, allowing the control unit to adjust the fuel delivery and ignition timing of the engine much more rapidly. In the automotive industry this sensor is also called a UEGO (for Universal Exhaust Gas Oxygen) sensor.

OK, that is about the type of the probe, but the measured value (voltage or electrical resistance transfered to voltage) is in both types of probes the volumetric amount of O2 in the exhaust gas.
So ECU does not read lambda directly but it is calculated.

The question is how ecu calculate lambda (equation wanted) or how the AFR is calculated and how can we calculate total fuel injection mass flow in g/sec.
I only found these equation:

AFR=100*inj time*rpm/120,000
INJ MASS(g/sec)=100*inj time*rpm/120,000/14.7
RELEASE HEAT-durring combustion(kJ/s)=air mass*(1-O2%)/14.7*fuel energy density


I need to know this becuase I am working on a calculator that will calculate engine power by knowing how much heat are generated durig combustion process and how much heat flow is leaving the engine by exhaust manifold (exhaust temp sens).

Of course there are some heat losses through engine walls, because process is not perfect polytropic, but heat losses through radiator are not constant and are not proportional with heat flow through engine wals so they can't be included in the equation and I also think that radiator heat losses are relative small.

BTW: air mass sensor can't determine water vapor in the air, so how the ecu know the exact volume of air? If there are high relative humidity the ecu think that there is more air

in the cylinder than real, so I presume that O2 sensor brake in to tell ecu that there is not enough O2 in the exhaust stem.

 

[Uwe@Ross-Tech]

OK, that is about the type of the probe, but the measured value (voltage or electrical resistance transfered to voltage) is in both types of probes the volumetric amount of O2 in the exhaust gas.

No, from the same article:

Zirconia sensor
...
The voltage produced by the sensor is so nonlinear with respect to oxygen concentration that it is impractical for the electronic control unit (ECU) to measure intermediate values - it merely registers "lean" or "rich", and adjusts the fuel/air mixture to keep the output of the sensor alternating equally between these two values.

The question is how ecu calculate lambda (equation wanted) or how the AFR is calculated and how can we calculate total fuel injection mass flow in g/sec.

I don't know how the ECU calculates this stuff. We make diagnostic software, not engine management software...;-)

Some of the Gasoline Engine Management Books from Bosch might have the info you seek, but I can't guarantee it.

If there are high relative humidity the ecu think that there is more air

Nope, increasing humidity actually makes air less dense because a molecule of H20 is lighter than either a molecule of either N2 or O2 (which together make up 99% of air.

At 40 degrees C, 100% humidity, you've got about 50 grams of water per kg of air -- and this value drops quickly at lower temperatures. That's 5%, not a huge amount. So you design an engine management system to to provide nominal fueling at 50% humidity and you're really only talking about a 2.5% error at minimum or maximum humidity. If you test a bunch of brand new MAFs, I doubt they are more accurate than 2.5%, and uncertainties like this are why the ECU still needs an oxygen sensor to be able to fine-tune the mixture.

Basically, ECU uses the various sensors (MAF, MAP, Air temp) and parameters (RPM, throttle opening) on the intake side to estimate the amount of air entering the engine, and squirts a corresponding amount of fuel. This is plenty accurate enough to have an engine run well, but not accurate enough to meet modern emissions standards. So the Oxygen Sensor is used as a feedback element to further refine the amount of fuel provided to the engine.

As far as estimating the amount of power being produced based on heat losses, I think you're tilting at windmills. I do not think the heat loss through the cooling system is negligible, and without knowing this, your calculations will be off my more than other, far more simplistic ways of calculating the power produced.

-Uwe-

 

[silver123]

As far as estimating the amount of power being produced based on heat losses, I think you're tilting at windmills. I do not think the heat loss through the cooling system is negligible, and without knowing this, your calculations will be off my more than other, far more simplistic ways of calculating the power produced.-Uwe-

I agree there are more simplistic and more accurate power measurment methods like acceleration method (by 2nd Newton's law), but it is a bit inpractical as you need to find realy flat surface (withough bumbs) and this can't be road as you will obstruct the traffic (when coasting down).

On the other hand heat difference flow measurment (release heat minus exhaust heat) could be done withought the need of coasting down. But as you said radiator heat flow is not negligible.
Only as I know the "ENGINE LOAD" value in measuring blocks is also a calculation of how much load the engine is overcoming in proportion of how much load could overcome at the moment. So engine load is basicaly torque calculation.

I don't know how the ECU calculates this stuff. We make diagnostic software, not engine management software...;-)

with full respect - calm down, there are also others that could answer this!

But anyway, do you maybe know anything about fuel mass flow calculation. On your site there is some excel sheet calculator about injector mass=>http://www.ross-tech.net/vag-com/examples/air_fuel_EXAMPLE.xls

 

[Uwe@Ross-Tech]

with full respect - calm down, there are also others that could answer this!

I'm perfectly calm. I was simply explaining that I didn't have an answer your question, as opposed to ignoring it.

But anyway, do you maybe know anything about fuel mass flow calculation. On your site there is some excel sheet calculator about injector mass=>http://www.ross-tech.net/vag-com/examples/air_fuel_EXAMPLE.xls

To be honest, I don't think the results produced by those calculations are very accurate. There are too many things it doesn't take into account, such as the the fact that the injectors cannot open and close instantly, and that the density of fuel varies with temperature and the temperature of the fuel in the rail can be quite high.

-Uwe-