exhaust reactions

[steveo]

this is more curiosity than a problem to solve

I have a car with no cats, a cam with not a ton of overlap, etc. a fairly mild build, but all emissions removed.

I've confirmed that I don't have a blown head gasket (plug read, press test, and test strip for combustion byproducts in coolant... plus I'm only down 50ml after 10,000km) or burning oil, and my nb afr is within a good range

but I have very thick white smoke going on, that smells like chemicals n fuel. its worst at idle, and persists with a warm car, even if the ambient temperature is high.

I can reduce it by messing with timing but not by much. Im currently running 24 degrees @ idle, 900rpm.

if I'm in traffic with a bunch of cars on a cold day, i'm definitely producing way more steam than anyone else.

my understanding is that the byproduct of fairly good combustion can be quite a bit of water, and if egt is abnormal, things can pretty steamy, egt being a function of incomplete combustion reacting in the exhaust, and/or afr.

what I want is a chemistry lesson on what's going on with this steam, or any other theories.

[steveo]

just realized I have no sig, here the car in question http://resfilter.net/files/carstuff/97build.html

[RobertISaar]

a car with no cats with ALWAYS have exhaust that smells of raw fuel, that's just how it is.

i know a cat will convert certain combustion byproducts into waper vapor, but i don't know if they're supposed to be produced in normal, uncatalyzed exhaust gasses.

http://en.wikipedia.org/wiki/Catalytic_converter

a 3-way cat turns:
NOX into nitrogen and oxygen
CO into CO2 (requires free oxygen in the exhaust stream)
HC into CO2 and water (requires free oxygen in the exhaust stream)

[1project2many]

Water is a byproduct of hydrocarbon fuel combustion in oxygen. One of the simplest hydrocarbons is methane, CH4. CH4 + (2) O2 yields carbon dioxide and water, CO2 + (2)H2O . The amount of oxygen needed and the amount of water produced depend on the particular hydrocarbon involved. The more complex the fuel, the more O2 required and the more water produced. In a perfect world the reaction is always complete. Outside this theoretical land we have incomplete reactions with intake air being a gas mixture instead of pure oxygen and a fuel that may not be pure hydrocarbon. CO forms instead of CO2, nitrogen and oxygen bond to form NO and NO2, CO2 and water mix to make carbonic acid. Sulphur, hydrogen, and oxygen mix to form sulphuric acid. What's produced at the exhaust depends on many factors, the most important being the ratio of reactants going in, the temperature of the reaction, and the time allowed for the reaction.

When burning fuel and air at a stoichiometric ratio you're mixing enough air to completely react all the fuel. What you're getting at the exhaust port is the result of how well the fuel and air react based on temperature, combustion time, and how well the gases intermix before and during combustion. Balanced in does not guarantee perfect out. If you consider the target to be complete combustion of fuel, the following are responsible for inhibiting the reaction:
1) Time. The fuel/air reaction generally takes more time to reach completion than what's available before the end of the power stroke.
2) Temperature. The combustion reaction happens faster where heat is greater. It is hampered around it's edges where cylinder walls, pistons, and the head never achieve the same heat as the reacting gases.
3) Distribution. Air / fuel distribution as a rule is poorer where large chambers and low compression are involved, where ports create little turbulence, where larger diameter cylinders are involved.
4) Mixture purity. Any exhaust reversion or incomplete emptying on exhaust stroke puts inert gas into the fuel/air mix which slows combustion.
5) Fuel form. Liquid fuel droplets cannot burn. It takes time and energy to produce gaseous fuel from liquid, which increases the time required for complete combustion.

The combustion reaction, once started, wants to go to completion. There is heat energy available to help sustain the reaction and there are unreacted components which would love to be reformed in simpler, more stable molecules. If you open the exhaust valve while this reaction is happening and push the mixture out into the exhaust pipe you are not ending the reaction. You are simply changing the factors which inhibit complete combustion. You're spreading gases out physically which changes distribution. You're introducing gas into thin wall pipes which allow heat energy to transfer away from the reaction. And you're reducing the chances for any liquid fuel to be converted to gaseous fuel so it can react. In a system with a converter the heat contained in the exhaust causes unreacted oxygen and fuel passing over the platinum catalyst to begin reacting with vigor. Without a cat, there's little chance you'll accidentally get anything like what the converter does. So instead of complete combustion products you get a mix of everything I listed above plus unreacted fuel and air.

I would disagree strongly with Robert's assertion that all vehicles without a cat will always smell like fuel at the tailpipe though. Many of the basics of combustion were understood in the '30s and '40s and engineers tried many ways to produce more efficient and cleaner vehicles even back then. The placement of intake and exhaust manifolds on most inline engines provided substantial amounts of heat to ensure vaporization of the intake charge was possible. Small ports under a small carb with an efficient venturi ensured fuel was sheared into small droplets before being driven against heating plates cast into the base of the intake while large amounts of cast iron provided plenty of heat, once warm, to keep the fuel vaporized on its way to the cylinder. Small bores and conservative cam timing helped create prime conditions for best mixture with minimum exhaust dilution. It wasn't until after WWII, and really after the '50s, when the strength of the US economy made wasting fuel through poor combustion less of a big deal. Many of those early engines did a great job of burning fuel when tuned correctly and run within design parameters. Cruise conditions often allowed ratios approaching 17:1 and high numeric rear gear ratios allowed lower torque. NOX wasn't a concern back then so engineers would try to create as much heat as possible without inflicting damage to the engine. Although it was made 35 years before the invention of the catalytic converter, you can stand behind my '36 Plymouth while it's running without smelling unburned fuel.

[RobertISaar]

perhaps raw fuel was the wrong choice of term...... but traveling behind the average cat removed vehicle around here, you can smell the difference quite easily. perhaps it isn't such an issue with certain engines, but i guess i get gassed/smoked out by enough of them to generalize.

[1project2many]

I think it illustrates the idea that in some respects we've lost ground. Computerized controls and EFI allow more flexibility than ever before but there were some very strong advantages to properly designed carbureted systems that don't come as easily to a typical EFI implementation. Since cats are mandated on all vehicles, and since they need oxygen and fuel to work, doesn't it make sense that engines that don't emit enough of both might not pass emissions due to an inoperative converter? Why would engineers build an engine that might not light the cats when needed?

[steveo]

Since cats are mandated on all vehicles, and since they need oxygen and fuel to work, doesn't it make sense that engines that don't emit enough of both might not pass emissions due to an inoperative converter? Why would engineers build an engine that might not light the cats when needed?

i thought that's why they added air injection in the first place

thanks a lot for that write-up by the way, that's a lot of food for thought

[1project2many]

AIR pumps were originally added to encourage the combustion reaction to continue in the exhaust manifold. The converter was filled with platinum beads. Baffles directed exhaust gas through and around the beads. These cats took much longer to come up to temp than today's converters. So the pump would operate at low temperatures before the cat lit off. The heat would be carried to the cat to warm up the beads while the exhaust was cleaner than if there was no AIR system. As the beads warmed, the fuel air reaction would begin in the converter and after a time, the AIR system would be deactivated. I have to admit that I don't know if AIR is activated today to supply O2 during rich conditions. I do know that it's still used after startup, primarily when engine temp is low, just as it was years ago. I also know that activating AIR when the exhaust is extremely rich and warm can result in sudden, violent reaction throughout the exhaust system, also known as "Boom!"

[delcowizzid]

tuning to 14.7:1 fuel ratios makes the most CO2 and H20 and also the most nox.if you tuned it to run at 14:1 ish with no cat you would lower the odour and reduce h2o and co2 quiet a bit and wouldnt steam so much and would probably live a lot longer if you sit in traffic a lot.tuned at 14.7 and no cat with exhaust dumped under the car is terrible for passenger compfort ive found its unbearable richer is so much nicer

[steveo]

i'll try that. i do sit in traffic a lot these days, and my afr is definitely not on the rich side of things.

i'm closed loop on an LT1, running basically 407mv at idle and 425mv off idle, haven't actually checked it with a wideband to see what that means yet... wonder how far i should bump it up to try to hit closer to 14.1.

[RobertISaar]

my notes show 407 as being ~14.93 and 425 as being ~14.89. i've had a theoretical chart compared against someone's wideband and it tracked very well, though i don't know if i would use it for anything other than a guess at AFR.

14.1 i have as being 800mV.

[steveo]

haha yeah thats totally nuts, no narrowband could be accurate with an 800mv midpoint. i dont really want to lose closed loop idle either.

i'll try 550mv or something to see if it even makes a diff, just for kicks, and bump the idle trim up a bit.

i've always been a big fan of shoving any tune towards the rich side to see what happens anyway.. it's usually a great choice just to see how it reacts.

[RobertISaar]

i tend to force 450 mean at all times, or at least 450 mean minimum(with a very small gap above and below for rich/lean)... not saying it's the best way of going about it, but some calibrations can cause endless surge/instability problems with the way they're setup from the factory. the one i'm looking at right now uses scalars at idle and tables at non-idle(vs airflow).

idle mean: 442
idle rich: 495
idle lean: 389

table mean: 424-517
table rich: 438-530
table lean: 380-509

idle stability was GREATLY improved after some tweaking.

[steveo]

yet another part of EE i swear that there MUST be more parameters for, but nobody has ever discovered them

[RobertISaar]

likely. how important is it to you? i have a good understanding of the O2 logic in the MPFI 60V6 applications and used it to understand large parts of the 94-95 SFI 60V6 O2 calibration. i would imagine the O2 logic for EE would be most like DA2/3.