floatyghosthat

iTrader: (5)

Hey guys,

I have a simple question that has been really bugging me, but it's hard to find anything online about it. Hopefully someone with more advanced tuning/chemistry education can drop some knowledge on me...

Stoichiometeric balance for a gasoline/air reaction (lambda) is 14.68.
The measurement of standard atmosphere is 1 bar or around 14.7 psi.

Is there any reason why these 2 numbers are so similar, or is it merely coincidence?

If anyone can answer that, they win free internet candy.


Thanks!

Verneuil

Stoichiometeric is 14.7 pounds of air to 1 pound of fuel. and atmospheric pressure is 14.7 pounds. But they each represent completely different values. so coincidence

floatyghosthat

iTrader: (5)

That's what I had understood, but I just wanted to make sure. Thank you very much sir.

blmcquig

iTrader: (15)

it doesnt have to be pounds of fuel. its just units. 14.7ish units of air to 1 unit of fuel for stoich.
the lambda term just means how close or far off of stoich it is.

arghx

iTrader: (3)

um google "convert 1 bar to psi" and you will see that it's ~14.5 psi to one atmosphere

14.7:1 or lambda = 1 is the "ideal" combustion in the sense that technically there is enough air and fuel to produce "complete" combustion. that means it should produce water vapor, N2, O2, and CO2 although there is always some improper combustion that has to be cleaned up by a cat.



so at 14.7:1 you produce the maximum amount of CO2 emissions but the overall least amount of everything else. leaner than 14.7:1 and you get lower fuel consumption with less CO2, less HC, less CO, and more NOx. Richer than that and you get less CO2, more HC, more CO, and less NOx.

That's why controlling emissions is such a complex problem. No matter how you tune the engine, you are trading one type of emission for another unless you can reduce overall brake-specific fuel consumption (the amount of fuel needed for a given amount of power output). Even when the engine is running at lambda = 1 (14.7:1) you are producing CO2 greenhouse gas. If you run it super lean like on certain stratified charge direct injected engines, you produce high NOx emissions. Then the NOx has to be trapped in a special converter, and every couple minutes the engine has to dump fuel into it to convert the NOx and change it into to N2, H20, and CO2. and then over time the sulfur in the fuel builds up in the cat and you have yet another emissions problem...

blue87

iTrader: (2)

The units have to be mass/weight (on earth) to be ~14.7 to 1. It will not hold true for example for units of volume.... but what I think you meant is they don'thave to be pounds, could be grams, slugs, ounces, etc as long as they are the same...

blue87

iTrader: (2)

He probably thinks 1 bar = 1 atm....

1 atm is about 14.7 psi. but regardless still has nothing to do with stoich.

No DI engine in the US right now as far as I know is using stratified charge. There are many OEMs that are using stratified charge, just not in the US market and the main reason is emissions regulations. The US regulations are too stringent to allow for stratified operation for the most part.

The generalizations you mentioned are absolutely true... emissions controls are very complex as I think you realize... but it is much more complex even than you elludede to... especailly when you start to consider transient operation, the time above and below stoich, the rate of switching, the O2 storage effect of some rare earth elements such as cerium. Decel fuel cut, fuel cut recovery enrichment, open and closed loop control of all of these events depending on how much money you want to spend on O2 or A/F sensor some have wider ranges than others and different response, different resistence to cracking due to moisture during warm-up etc. Don't forget that all of these systems in modern vehicles have to be monitored via OBD which in of itself can actually cause emissions or driveability concerns during the monitoring period....

We hear about the Platinum, Paladium, and Rhodium (precious metals) used in catalysts all of the time, but there are literaly hundreds of other metas used in the actual washcoats of catalysts which are highly regarded secrtes by the manufacturers. For example Nickel will reduce the Sulfur smell which Americans don't like, but Nickel is regulated in Europe so it can't be used as much. As a result Europeans think the sulfur smell is a hig performance thing because they are used to it for example....

Emissions standards for OEM vehicles pertain to "Full Usefull Life" which can be as high as 150,000 miles depending on the standard the vehicle was certified to. These standards are so low that any little pretubation in the first few seconds of engine operation can completley fail the standard. This is why the OEM do so many "stupid" things during engine start up to ensure the emissions are reduced as much as possible and to protect the catalysts at all costs under high loads.

I am always hearing all of these random theories about enrichment being used to prevent knock etc.... for the most part all enrichment is used to protect the exhaust hardware, Exhaust valves (obviously not in a rotary), exhaust manifolds, catalysts, O2, A/F sensors etc...

I obviously don't have a point it islate and I am just rambling.


Arghx I hope that we can meet up at a meet one of these days, I am getting more invloved in the community when I can, I think we can have some very interesting converstations/ knowledge exchanges...

arghx

iTrader: (3)

The biggest problem has been the EPA cat life requirements and the sulfur in US fuel, which until recently was significantly higher than in Europe. As I mentioned before, the sulfur builds up in the cat during the process of regenerating the NOx trap. The sulfur can be purged but the final result is still deterioration of th converter. Every direct injected engine in the US (and I presume "world" engines) uses homogeneous mixtures at lambda 1 as opposed to lean homogeneous (lambda 1.2ish) or stratified charge at about lambda 1.6. I have a technical manual from VW/Audi which explains these issues.

Decel fuel cut was one of the first advantages achieved by electronic fuel injeciton. Closed loop control and operation in lambda = 1 is used in more and more of the engine operating cycle. With direct injection to my knowledge you don't need asynchronous acceleration enrichment pulses, which hurts fuel economy.

Most modern engines have at least one factory wideband, especially if they are direct injected. The Evo X 4B11T is a notable exception, but it is port injected.

also depends on the substrate design

yup. another deterrent to stratified charge engines

on direct injection engines it's not as big of a deal anymore. First there's a small pilot injection during the intake stroke, then the injector fires late in the compression stroke, about 20 degrees BTDC. The spark plug fires ATDC so you have very retarded fuel injection and spark timing. The hot expanding gases warm up the cat very fast. It's one reason why I expect direct injection on just about every engine within 5 years.

Too many people reading car forums too much. What applies to people hotrodding in their garage doesn't relate that closely to control strategies used by "real" engine calibration/emissions certification engineers. The OEM's use ignition timing to control knock more than anything else. Lots of turbo engines stay at lambda = 1 until 4 or 5 thousand rpms from the factory.

yup

It's interesting stuff. I have a lot of technical manuals and journal articles that I collect. That's how you cut through internet bs, you go right to the source.

I would like to as well. I post all the time but I'm not nearly as involved in meets as I once was. The local rotary scene isn't what it used to be and my financial situation keeps me from going to many "big meets" right now. I've got enough budget to keep the car from completely falling apart but that's it. I'm starting a new job soon so that may help.

blmcquig

iTrader: (15)

Really? I didnt know that cars came from the factory with a wideband.
Ive really enjoyed reading a lot of things written in the PFC and single turbo section. A lot of the stuff i kinda know, but I almost everyday learn something new and interesting.
A friend of mine had a 03 focus SVT with 3 O2 sensors, and they were all 4 or 5 wire. I wonder now if they were wideband. i had always just assumed it was just a heated narrow band. I wonder...

arghx

iTrader: (3)

Rx-8's have a factory wideband. So do basically all Subarus of the past 10 years. You can see the wideband signal calibration in Subaru ECU's if you hook up a proper cable to the diagnostic port and read the calibration with some open source software. The wideband is made by Denso and it uses 4 wires as opposed to 5 on some sensors. That sensor family tends to pig rich at about 11:1 . It's a little different than the Bosch sensor used in most American and German cars, as well as aftermarket systems.

I don't know about your friend's car, but I know the new direct injected Ford engines have them for sure. The Ecoboost twin turbo V6 and the direct injected V6 used on the new Mustangs, Explorers, F-150 etc all have a factory wideband. It's much easier to control emissions with a wideband during normal driving. That's because the ECU can more accurately calculate how much fuel to add or subtract to keep the AFR cycling around 14.7:1 .

blue87

iTrader: (2)

yeah pretty much all modern cars are going to factory wide bands... but most of them are not as wide as say a Bosch/ETAS aftermarket sensor. They generall are used for control near stoich and not suring enrichment.

Some manufacturers use a more expensive A/F wideband sensor and can actually control clised loop during enrichment also... but on a production level this is big money... some OEMs only use this tech on high performance or high dollar cars where there is some true benfit or they need it becuase they are close to an emissions limit or something....

blmcquig

iTrader: (15)

so i had this thought the other day while thinking about this subject...

(i know the voltage isnt right, but you get the idea...its something like that.)

if a stock narrow band is like 0-1v
and a normal wideband is 0-5v
would that make these "but most of them are not as wide as say a Bosch/ETAS aftermarket sensor. They generall are used for control near stoich and not suring enrichment." maybe in the 0-3v range then? so... could they be maybe more like "medium band" sensors?
i dont know, just rambling. but it got me thinking.

arghx

iTrader: (3)

narrowband sensors directly generate a voltage (between 0 and 1 volts) based on the amount of oxygen in the exhaust gas. They are relatively simple devices and don't need a controller (especially if they don't have a heater like on Rx-7's). There is a drastic change in voltage around .45v which is at 14.7:1 or lambda = 1. The ECU knows the mixture is rich or lean of 14.7:1 but it can't easily figure out how far off.



Wideband sensors work a little bit like a hotwire mass airflow sensor. There's a little cell inside the sensor with air that is kept at 14.7:1. As exhaust flows into the rest of the sensor, the wideband has to figure out how far off the exhaust is in comparison to the 14.7:1 cell. It pumps oxygen in if the exhaust is rich and it pumps oxygen out if the exhaust is lean.

The current increases as more oxygen is pumped, and the polarity shifts depending on whether the exhaust is rich or lean of lambda=1. So a lean mixture will have current with positive direction, and the leaner it is the more current will flow from the controller. A rich mixture will have current with a negative direction, and the richer it is the more current will flow from the controller. The controller will take the current signal and might convert it to a voltage for the ECU or whatever to read. It depends on the application. On an aftermarket system, the controller usually converts the current signal to a linear 0-5v where higher voltage = leaner AFR.

Most of the OEM Denso wideband sensors I have seen, like the family of sensors used in the Subarus and the Rx-8, will "peg" rich at about 11:1. Here is a factory wideband o2 sensor calibration for a 2005 Subaru Legacy 2.5 turbo.



You can see that at about 11:1 the signal "bottoms out" and you can't tell just how much richer it is than that. Here is the signal curve of the Bosch LSU 4.2 sensor used on most aftermarket widebands:



This sensor "bottoms out" at lambda = .70 or 10.3:1 as opposed to 11:1. Aftermarket systems using this sensor will show richer values but they are only "guessing." These sensors were never really meant to be used the way we use them. Real engineers use expensive emissions testing equipment rather than production-grade wideband sensors. The nice thing about the Subaru/Denso OEM wideband is that it can withstand a lot of heat. The engineers located their factory wideband pre-turbo on most models (for emissions), whereas the Bosch sensor in an aftermarket application is usually located much farther back in the exhaust piping to keep it from failing prematurely.

blmcquig

iTrader: (15)

so does using a wideband in production cars just make the whole process more accurate/efficient? that is, if both the narrow and wide band are just trying to keep the afr hovering around 14.7:1
so the wideband in a production car just makes it happen faster, and more efficiently?

peejay

The wideband is used in newer cars because it allows closed-loop control under all conditions, not just idle and cruise.

I think the answer to the original question is lost... the answer is yes, it's just a coincidence that the fuel we happen to use happens to have a stoich ratio that is the same as the air pressure at sea level when measured with completely arbitrary units.

It's not a conspiracy like how Fahrenheit has exactly 180 degrees between freezing and boiling water (at standard day)

arghx

iTrader: (3)

Yes. Compared to a narrowband a wideband can more accurately determine how much fuel adjustment is needed to achieve the 14.7:1 target AFR used for emission control.

yes but most engines with factory widebands don't actually use closed-loop control under heavy loads. The R35 GT-R does but that is a very expensive car:



You can see that the R35 has two modes: stoichiometric mode, and full power/enrichment mode. The graph shows the maximum torque allowed in each mode at a given rpm. The R35 is very complex for a port-injected V6 engine. It has two MAF's, two electronic throttles, two intake manifolds, electronically controlled sequential gearbox, and wideband AFR sensors for each bank. The ECU uses the transmission, electronic throttles, boost control, MAF sensor, ignition timing, widebands, etc to achieve the requested torque in each mode. It probably switches between modes based on input from the accelerator pedal position sensor, rpm reading, etc.

blue87

iTrader: (2)

Another fun place where facory widebands are used is during fuel cut recovery. After a deceleration fuel cut the exuast (catalyst specificaly) is loaded with Oxygen from the engine pumping air through it.

If you just start the engine and run at stoich you will see a large spike of NOx (oxides of nitrogen) that pass through the catalyst and will fail emisions standards during transient driving.

To overcome this most OEMs have some sort or oxygen storage model to determine how much oxygen is store in the catalyst. So when they start the engine again they run the engine rich for a shirt period of time. How long and how rich (not very usually) is calculated as they don't want to go so rich for so long that HCs (raw fuel) get through the catalysts, it also hurt fuel economy so enrichment is always minimized as much as possible for that reason also.

Anyway, in this fuel cut recovery region they also generally run in closed loop enrichment since it is close to stoich and you can do so with the "cheap" A/F sensors and since emissions performance is so heavily regulated....

peejay

Actually.... they do.

Aside from the chimeric dual MAF/throttle arrangement like an old BMW V12, you're describing a lot of modern cars. The only difference is, you don't have to take your VW or Ford in to the dealer every 1500 miles to have the induction balance re-checked like an old MG. (And the transmission in the SHO manages to handle full twin-turbo 3.5l V6 power in a 4500lb AWD car running 12s all day long, without "abuse management" or exploding... )