18:1 AFR is "cleaner"?
 

So, assuming the catalyst is fully warmed-up, doesn't a graph such as this indicate that an AFR of 18:1 is better for emissions than stoich? I realize we can only be in lean burn under light throttle and load conditions. Nonetheless, at 18:1 only 02 is higher than at 14.7:1. That seems gooder for everybody. So 18:1 is simply better, no? Why aren't all cars in lean burn under light load and part throttle and steady cruise on freeways?

https://ecomodder.com/forum/member-c...-4gaschart.gif

Probably because of the nox spike at 16:1. They appear to be concerned about the nox most.

That makes sense, but at 18:1 NOx is lower than at stoich and much lower than 16:1. So assuming an efficient catalyst, less NOx going into the catalyst means less coming out the tailpipe, no?

The EPA declared war on NOx and the government want us all to buy more gallons of fuel.

Again though, at 18:1 AFR, this chart suggests NOx is lower than at 14.7:1 and so is CO2. Why isn't 18:1 lean burn common place in contemporary cars? Or is it maybe common and not touted like in the old days of the 90s and early 00s?

That graph is assuming you don’t have a cat,

Given the circumstances today it’s likely because of two things

1. We never did it so it’s impossible (hear that a lot)
2. Emissions controls (not represented by the graph) reduce Nox below the graph,

the unfortunate part is because we are laser focused are a single pollution percentage we rather burn more fuel and make more pollution in grams per mile.

Oh its too difficult for obsolete binary O2 sensors to detect and maintain 18:1 and it was decided that stoich was cleaner not with actual science, but with politics.
And most importantly it burns more fuel.

Doesn't the cat have overtemp isses when too much air is introduced?

I am aware that the graph assumes no CAT. But if the exhaust in cleaner going in, it'll be cleaner coming out, assuming a properly efficient CAT, right? Am I wrong about that? Cleaner in, means cleaner out?

This all would suggest that modifying a car to run lean at 18:1 is likely cleaner than stock. Hondas like the VX and HX and the Insight all did lean burn from 1992-2006 and they had the best or almost best emissions ratings (LEV, ULEV, and such). Those lean burn systems relied merely on ECU tuning, a wideband O2, and a knock sensor, IIRC. Not a super high bar.

I guess I am surprised because I thought the LB engines had higher NOx, but that is no necessarily true, unless something is still missing from this picture.

That is the kind of thing I cannot find clear answers for online. I see contradictory things... some say lean burn makes the catalyst cooler.

It runs out of fuel and saturates the catalyst over time (years) it can become brittle and fail

The only time I have ever seen a confirmed melted catalytic converter was on an extended dyno run on a car that was definitely not running lean.
The fix for the melted cat was "run a bigger cat", as apparently this is a fairly common problem.

That would be a problem. But Hondas such as the Gen1 Insight, Civic VX, and CIVIC HX have not had that problem & they run as lean as 22:1, according to Wikipedia. 18:1 is not nearly as lean and it would produce a modest improvement in emissions and in fuel economy. I don't understand why it was ever phased out, if it was phased out, for light load cruising and moderately low engine speeds. I gotta think it is still programmed into Honda iVEC systems but not promoted.

Older cars (pre 2000) had 2 way CATS or modified 3 way
which aren’t affected by that issue, they also more importantly didn’t need to meet the more stringent NOx limits of later years so some NOx could pass a full trap without issue.


The Insight and newer had a Nox trap and would cycle lean to rich to make sure the cat stayed lit and the NOx levels didn’t increase after the NOx trap filled.

So, perhaps as the mixture goes lean, the CAT cools, and its efficiency decreases. That would mean a DIY lean burn system would inevitably cough out more NOx, which was a major culprit in SoCal smog back in the bad old days of fog banks in the desert.

https://ecomodder.com/forum/member-c...emperature.gif

You just about have it.
 

Look back at the graph that was posted at the start of the thread. Notice the line for HC. It is climbing quickly. This must be dealt with if you want to run clean and lean. The problem is, as you correctly pointed out, the cooling exhaust gas leaves the catalytic converter unable to efficiently eliminate this HC, not the NOx.

Why is HC climbing? Why is it not lowered like CO in an oxygen rich environment? As the engine runs past 17:1 AFR, the combustion temperatures cool. You surmise this from the cooler exhaust temperatures. This implies that the combustion chamber event is also cooler, and it is. This results in the quench area thickness growing in proportion to the reduced combustion temperature. You can sweep the chamber walls with increased turbulence, but you can only do so much. You can also increase intake heat via larger volumes of exhaust gas re-circulation (EGR) noting that EGR in lean-burn mode has more reactivity in the form of OH ions due to free oxygen.

All the above was done by Honda to achieve 22:1 AFRs, but the cost of the system as well as it's finicky nature lead to it's abandonment.

Running leaner than 22:1 AFR leads to excessive combustion variability. This increasing Co-efficient of Variability (COV) results in increasing pollution as CO comes back into play.

In summary, NOx at ultra lean fuel mixtures is not the culprit. HC and CO are.

The cat should burn off the hc and co using the excess NOx

Would that be true even if the lean mixture causes cooling of the catalyst to a less efficient temp? Maybe that's what you are saying. Not certain.

Yes those compounds still get burned up ,
the percentage of those compounds burned decreases though as temperatures decrease.

I like how they just ignore increasing fuel economy as a way to fight pollution, as if the gasoline just miracles it's self into your gas tank tail pipeemission are the only way a car can damage the environment.

I like how that we are laser focused on 1 specific kind of pollution as a percentage (that has been very good vehicle emissions wise for over a decade)
while ignoring others altogether and not caring about grams per mile
all the while ignoring massive sources of that type of pollution from industrial sources.

Then also forgetting that supply side pollution
groundwater, heavy metals and outright spills dwarf the associated vehicle pollution .

At some point you would think we would look at the other sources along with combating a recent trend on lax maintenance and inspection of industrial systems sometimes close to 70 years old.

Easier to mandate to the individual than to a large company that can afford to fight slap lawsuits.

lack of heat is the issue.
 

The catalytic reaction needs a minimum heat/energy level. Without this, the reaction is either too slow or non-existent. We can program the engine run regimes to oscillate between lean-burn and stoichiometric to keep the catalysts hot enough to function.

Also, if you want to use the oxidation potential of NOx to convert the growth in HC and CO in ultra-lean mixtures, you will not have enough to complete the reaction equation. Look at the original graph - there is exceedingly small amounts of NOx the further lean you operate.

The trick to running ultra - lean is in reducing the COV and reducing the quench thickness. This gives you more power to use and reduces HC and CO respectively.

I'll probably try running 18:1 in a few days. However, there's something to consider with leaner mixtures: ignition timing needs to change quite a bit.

When the mix is leaner, flame spreads more slowly. When timing is adjusted such that peak pressure is where it needs to be, a larger percent of burn will be before TDC or where piston angle is poor. The more I've read about it, the more it surprises me Honda was able to improve economy with ultra-lean burn. Maybe it was the lack of infinitely variable gearing, and lean AFRs were a way to virtually reduce displacement on demand.

In order to get ignition timing right without a dyno, I'll likely need to buy an exhaust gas temperature sensor.

How important are such timing changes under light load? You are at like 20% load at 60 mph, right? Seems slight risk of problems under such conditions, no?

You're right in that at light load I'm not going to hurt anything, it's more about efficiency.

If the timing is too retarded, cylinder pressure won't peak until the piston is already well down the cylinder, and a lot of potential work will go out the exhaust. Advance it too much and a larger portion of the work will happen while the piston is still rising on the compression stroke, and will be essentially trying to turn the engine backwards. Either way, fuel economy and power both suffer. Even been a couple of degrees off and the differences will outweigh and gains.

I want to say the Insight's stock engine advances ignition timing 20+ degrees when it goes into lean burn, and being off 3-4° is significant, so simply leaning out without adjusting timing isn't likely to do any good.

I *think* it can be done with an EGT sensor, but may require dyno time.

One of the bummers of Cali smog regs is that exploring such modifications of an engine makes your daily driver illegal. I would not pass smog with such mods. They make me remove my brake booster canister intake tube. Go to the wrong shop, and I get rejected for my gearing. Extra sensors and a non-standard ECU sound like instant fail territory.

Extra sensors maybe not, haven't had a Cali smog done in 15 years so I can't say how deep the software queries your OBD2 only ECU which is the only way they know it was changed. I do know on the F250 ECU, they can't detect an aftermarket box but make notice the software version installed, unsure about a eprom chip . On the VW TDI, they definitely check software revision against a published list. I also know I couldn't add EFI to my 72 superbug which imho was flat stoopid..

Then there's the friend of a friend smog guys.............

By the book
Making a Hybrid into a plug in is illegal because it modifies emissions equipment to make fewer emissions

Pure stupidity of cookie cutter protectionist law

I see as high as 30-31 degrees in lean burn on my ScanGauge.

And if you drop out of lean burn with the same throttle position?

I'll check to be sure, but maybe 12 or so. Any more throttle and it quickly goes negative.

An interesting tidbid.
 

When you look at the graph on the first post, you notice the NOx spike just past the 16:1 AFR? That air to fuel ratio is your most economical mixture. It is the mixture that provides the most useful cylinder pressure for the least fuel burned. Why is that? It is because flame speed is still high and there is the additional "fuel" value of some of the nitrogen oxidizing! More power through pollution!

If nitrogen was not such a "lazy" reactant and required temperatures approaching 2300 deg C to oxidize, we would be able to use it as a fuel!

On the other hand, some of the builders of the first nuclear bomb had a fear that the fission explosion would set the atmosphere on fire. The nitrogen all oxidizing into NOx due to the temperature extremes caused by the nuclear fire. The physicists knew better and were not bothered, but lesser minds ran with the idea.

With California smog you're supposed to put all the stock parts back on, duh :P

At 18:1, you're losing around 20% of laminar flame velocity which sounds like a lot but it's not that many crank degrees. One idea is to look at the timing at a lower load value: if 50% load is -30 deg and 40% load is -35, I think -35 would be approximately in the ballpark. A lean mixture has higher thermal mass, but also a higher reaction rate due to more O2, and the two factors should cancel each other out in theory.

Can you clarify some statements?
 

I agree that flame velocity is slower with increasingly lean mixtures but you then say it has higher thermal mass? And higher reaction rate? Remember, you have less fuel to provide the heat.

Yes. This is comparing the same quantity of fuel, but more air (higher load, higher MAP).

https://wikimedia.org/api/rest_v1/me...88cf4510018a67

There is excess air that absorbs heat. Same quantity of fuel => same quantity of heat, but more air. That decreases alpha.

There is excess oxygen. Reaction rate is proportional to concentration^order, you're increasing oxygen concentration while keeping fuel concentration the same, since the cylinder is still the same volume but you have higher partial pressure of oxygen. That increases omega (with the dot over it).

Going from 1.0 to 1.2 lambda isn't a massive change all things considered, so assuming the effects cancel out is approximately correct. You can always use less timing advance to be safe.

Side note: In the case of adding EGR, reaction rate is not affected much since the concentration of CO and NO is tiny, but you're also raising the temperature considerably which speeds up the reaction.

I have a question about a definition.
 

Concentration.

Omega naught is proportional to the concentration of reactants. In one instance, you are reducing one reactant ( gasoline ). On the other hand you are keeping the oxidizer concentration ( oxygen ) constant. Omega naught will be smaller.

Also, you are ignoring the other variable. Temperature. With less fuel and more air, T(sub b) is smaller ( the flame front burns at a lower temperature ). The temperature relation in the parentheses will be smaller.

At just this side of lean lambda, it is known that burn rate increases. But it rapidly falls off. I worked with three Chrysler vehicles and their lean burn systems. A four, six and V8. At 18:1 AFR, all of them needed quite a bit of ignition advance (60 DCA before TDC, average). However, you are correct in that the addition of EGR increased flame speed and ignition advance could be pulled back to 35-45 BTDC.