I took my car out for a spin with a automotive gas analyzer plugged into the tail. My original intention was to measure if there was significant air moving from the intake plenum directly out the exhaust when the high-overlap VTEC cam was engaged. My expectation was that there would be some blow-through, which would show up as unburned HC in the exhaust over 6500 RPM. This is a stock 2002 S2000.

The experiment was to drive the car at 2000, 5000, 7000, and 8500 RPM under no load (in neutral), some load (driving down the freeway), and full throttle (driving down the freeway at full throttle with the brakes on to keep the speed constant). After the experiment I stayed on the freeway for a while to let the brakes cool off!

I measured NOx, O2, CO, and HC. The gas analyzer calculated lambda for me, which is the fraction of the stochiometric air/fuel ratio that the car is actually running. Generally, lambda of 1.0 is perfect, lambda less than 1.0 is rich, and lambda greater than 1.0 is lean.

Anyway, the results surprised me.

• There was no unburned HC or NOx at any RPM/load combination.
  
• For nearly all RPM/load combinations, lambda was 1.01 and CO was less than 0.01%. There were two exceptions:
  
• Off throttle, downhill, lambda jumped wildly to 6 and even peaked at 20.
  
• At 7000 and 8500 RPM, at full throttle, lambda dropped to 0.85 and CO output jumped to 8.5%! Lambda was 1.01 at partial throttle at these RPMs, and it was 1.01 at full throttle at lower RPMs.

My conclusions:

• The catalyst on this car works very well. It was actually a bit dirty when starting up, but after that it was running less than two orders of magnitude less pollutants than allowed by California law. This is generally a clean engine.
  
• The torque kick I feel between 6500 and 9000 RPM is due to the engine management computer running the car rich, rather than some VTEC effect. This air/fuel map is the real secret to Honda's spectacular HP/litre numbers, and it's kinda cheating, since the car is sold as an ultra-low emissions vehicle.

Here's why running rich increases power so much: For a given amount of air moved into the engine, you have a fixed amount of O2. If you burn that O2 into H2O, you get four hydrogens burned for every O2. If instead you burn that O2 into CO2, you get one carbon burned for every O2. Four hydrogens burned releases (quite a bit) more energy than one carbon burned. The hydrogens burn FIRST. So if you make the mix rich, some of the fuel goes partially unburned, and that part is the CARBON. You end up with partially burned carbon, or CO, and more energy per O2.

My recommendations: I suspect the VTEC cam keeps the car breathing relatively well out to redline. If you want to start that torque bump at a lower RPM, never mind about engaging the VTEC earlier, run the car rich earlier instead. Of course, that means screwing with the engine management computer instead of dropping in a cam, which may be more expensive!

My question: Why doesn't Honda pump enough air into the exhaust to let the catalytic converter burn the rest of the CO? 8.5% CO would need less than 30 litres/sec of air to become CO2! (Imagining 30 litre/sec air pump on exhaust system.)

 

sounds like you know what you're talking about, and i followed it for a good part of the time, but now that i'm lost i'll be interested as to what some of the other torque psyco's have to say. Especially, whoever it is that has "torque is for fat people" in their sig

 

Very interesting. Thanks for posting the data.

It would be great to check other ecu's results, or even after some have adjusted their VAFCs. I would imagine checking exhaust downstream of the cat might also hide a little of what the engine is putting out upstream from the cat.

 

I think I understand most of what you said. It confused me at first cause I was fixated on this equation.

2CH3(CH2)6CH3+17O2 --> 18H2O + 8CO2

I always thought that most of the chemical energy in hydrocarbons came from breaking the bonds between hydrogen and carbon and that the excess hydrocarbon molecules were converted to CO2 and H2O in the catalytic converter.

It never occured to me that excess hydrocarbon would seperate into carbon and hydrogen inside the combustion chamber, resulting in hydrogen+oxygen reaction.

And all this time I was wondering where CO came from.

 

That's really interesting data, and a valuable experiment. Thanks for posting!

I'm not ready to agree with your conclusions about running richer, though. I see what you're saying about the richer mixture in VTEC, but it's CLEARLY the extra breathing capacity afforded by the high-overlap cam lobes that provides the torque boost. No VTEC cams = no torque boost, regardless of the A/F ratio. In fact, V-AFC owners have shown uniformly that the path to greater power lies in leaning out the mixture. (I see you're fairly new here -- do a search on V-AFC.) Our best guess is that Honda richens the mixture during WOT VTEC primarily to aid in cooling, taking advantage of the fact that there's no emissions requirement in that scenario.

Oh, and the car's only a LEV, not a ULEV.

Cheers,
John

 

It's generally accepted that running leaner will produce more power, but that comes w/ a price. The closer you get to stoichiometric (14.7:1 A/F), the more power, but the hotter the burnt charge becomes. If I remember my chemistry, carbon's flame temperature is about 3,000 fahrenheit. If any of you have ever played around in class, you'll know that aluminum turns into a flabby goo at that temperature. Unfortunately, that's exactly what the block and head are cast out of. Additionally, iron's melting point is a few hundred degrees below that as well, and headers don't have the protection of the block's liquid cooling jacket.

Going over stoichiometric introduces other problems. Nitrogen combusts at temperatures that high, and since there's no more carbon or hydrogen to burn, the nitrogen readily combines w/ the excess oxygen, producing whole bunches of nitrogen compounds that are quite toxic.

oh, and running lean will produce more NOx compounds as you lean the mix, just not as much as going over 14.7:1...

 

[QUOTE]Originally posted by Ray
[B]I think I understand most of what you said.

 

The combustion equation for gasoline (octane) and dry air at stoichiometry is:

C8H18 + 12O2 + 45.12N2 > 8CO2 + 9H2O + 45.12N2

During the reaction heat is released. The reactants contain approximately -208000 kJ/kg of energy and the products of the reaction contain approximately -581000 kJ/kg of energy. The remaining energy (reactants minus products) is released in the form of "heat". The heat of reaction is approximately 2400 degrees Kelvin at 10 atmospheres of pressure in the cylinder. There's no other "magic" here. It's just chemistry.

The reason that a rich mixture "typically" makes more power is due to the inability to achieve adequate mixing of air an fuel in the cylinder, so in order to take advantage of all of the oxygen in the combustion process you add excess fuel. The excess fuel in more likely to "find" some oxygen hanging around and team up with it. The heat of combustion is greatest when all of the oxygen and all of the fuel is burned, and the heat is what does the work on the piston (consider PV=nRT from chemistry). As the temperature goes up, the cylinder pressure goes up.

The fact that the piston top, the cylinder walls, and the combustion chamber are all being cooled keeps their temperatures below the melting point.

Improving combustion efficiency, cylinder filling/evacuation, and controlling A/F ratios is the key to making more power. Keeping the speed of the reaction low enough to "burn" rather than "detonate" is the next trick, typically controlled by the "octane" rating of the fuel (high octane rating equals slowly "burn" rate) along with the ignition timing.

I hope this helps you to understand what's going on.

 

Greenlight's explanation of why you generally run richer than stoich is dead on. In fact, for a well designed modern engine (normally aspirated) we typically find that A/F ratios in the 12.5-13.5:1 range produce optimum power. When tuning a nomrally aspirated engine I usually shoot for a starting target of 13:1 and then vary from there to see what happens to power.

The F20C runs way too rich at high rpm (as evidenced by the lambda readings) with A/F ratios approaching 10-11:1 - about as rich as you'll see on some older turbo cars from the factory! This reduces power, cylinder temps, and reliability concerns. However, past experience with other high output Honda motors suggest thisengine will be quite happy to live at 12.5-13:1 ratios all day long (and the results for people using the Mugen ECU suggest this as well). Thos who have tuned their VAFC's with an A/F meter have found that 12.5-13:! is a good range for optimal power as well.

The boost you feel at VTEC is due to several things, including a change in ignition timing, fueling and airflow. But make no mistake, the change in cams is the major factor in creating that boost. The transition could be seamless, but Honda seems to have a prediliction for creating a spike to let people know they have VTEC. On the old Integra GS-R the VTEC transition was seamless stock, but it had secondary intakaes that opened up to give you the perceived boost. Smoothing the transition usually just involves moving the VTEC point.

Oh, and on the closed throttle lambda, the ECu shuts off the injectors at closed throttle above a certain speed/rpm. This preserves fuel of course.

UL