Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Automotive Air fuel ratio effects in diesel vs. petrol engines

  1. Feb 21, 2017 #1
    I'm a mechanic who is currently doing a study automotive engineering.
    Being someone who worked on both diesel and petrol cars I know what a higher or lower air fuel ratio (AFR) will do for both engines. However I don't fully understand how the following is possible.
    "Adding fuel in a diesel engine will result into more heat, however adding fuel in a petrol engine removes heat. Removing fuel from a diesel results into lower temperatures, however removing fuel from a petrol engine increases heat."

    It's like saying adding wood (fuel) to a fire will make a bigger fire with more heat, but then also saying adding more oxygen a fire increases heat. Both statements seem true, but how is that possible?!

    Here I will sum up some things so they won't have to be mentioned.
    For diesels-
    • Adding fuel in a diesel engine rises EGT's (exhaust gas temperature)
    • Diesel engines always run lean (even though their might be rich pockets)
    • Diesel engines achieve combustion by pressure and heat
    • Diesel engines use their injection system to time their combustion (so no pre-ignition is possible)
    • Diesel engines don't have a real throttle valve
    For petrols-
    • Going rich results into unburned fuel which turns into gas absorbing heat (temperature decreases)
    • Leaning out in a petrol engine rises temperatures.
    • Since leaning out increases temperatures pre-ignition and knocking might occur
    • Petrol engines have to stay within a small range of AFR's
    • Petrol engines have a throttle valve limiting air going into the engine
    I've also got a few questions.
    1. If we would run a diesel engine very rich, for example with a AFR of 13:1 will have the same cooling effects we see in petrol engines running rich mixture? (ignoring practical problems such as soot and NOx production)
    2. If we go lean in a petrol engine it will run hot. But what if we go beyond slightly lean, let's just say we run a AFR of 30:1. Would we see a drop in temperatures?
    3. What is the limiting factor in going lean? Can we simply keep going leaner till the engine starts struggling with combustion? Or is the limiting factor lubrication (from fuel)?
    Do diesel's never run rich enough to experience any cooling effects? And that petrol engines never run lean enough to experience temperature drops in combustion? Is it simply the case that diesel and petrol engines run the hottest between 14.7:1 and 18:1? And ratio's above or below that see lower temperatures because they need more fuel or oxygen to create heat?
  2. jcsd
  3. Feb 22, 2017 #2

    Randy Beikmann

    User Avatar
    Gold Member

    The diesel part here is relatively simple. Without a throttle, the engine always takes in the same amount of air. Adding more fuel creates more heat, and more torque. Because of the very high pressures and temperatures in a diesel engine near TDC, I don't know that there is a "lean limit" to a diesel, where there would be no combustion. My first guess is you can run them as lean as you want (why you would is another question). Because of the high temperatures and excess oxygen, diesels always produce a lot of NOx, but not much unburned fuel. Run too rich, however, and it's unlikely that all the fuel molecules will find oxygen molecules to react with. So you'll have a huge amount of soot before you approach the stoichiometric air/fuel ratio. I imagine you'd get cooling with even more fuel, but no one in the neighborhood would be able to see anymore.

    For the gasoline engine, part of this is simple. Because the temperature and pressures are lower, you do rely on a spark for combustion, and the spark will only ignite the mixture reliably in a limited A/F ratio range (I don't have any of my IC engines books here, so I can't look up what is typical). Also, for the catalyst to clean up both the unburned fuel and CO, and NOx, the air/fuel ratio must be very close to stoichiometric. In other words, you have to burn all the air you bring in (so to control torque, you need to have a throttle to limit the air). As far as why lean engines run hot, I can't remember. Maybe another forum member will.
  4. Feb 22, 2017 #3
    Thank you for your answer. Yes it's more of theoretical question then a real-world one, the amount of soot and NOx would be totally unrunable. And there isn't really a reason why you would want to go even leaner in a petrol engine either.
  5. Feb 23, 2017 #4


    User Avatar
    Science Advisor

    Centrifugal spark advance before TDC is set for a stoichiometric AFR. Speed of combustion is faster in a lean mix so the fuel burns before TDC. That transfers more heat to the piston and cylinder head before the combustion products can expand during the power stroke.
    Development of lean-burn engines, that continuously compute the spark advance required, do not run hot when lean.
  6. Feb 23, 2017 #5
    " however adding fuel in a petrol engine removes heat . . .l however removing fuel from a petrol engine increases heat."

    I once read a study where the temperatures of the exhaust manifold and catalytic convertors were tested on engines operating at temperature. The tested conditions were the temperatures at 30 MPH at 0 and 7% grade and then again at 60 MPH at 0 and 7% grade. The results were surprising. The study was about engine efficiencies and the purpose designing of engines. I think the answer you are asking for was in this study but can you verify what it is your asking first? You write "removes heat," but I get the sense you may be saying the combustion temperature goes down when more fuel is added. Those are two different questions or at least different variable conditions apply.
  7. Feb 24, 2017 #6

    jack action

    User Avatar
    Science Advisor
    Gold Member

    The confusion comes from the definitions of "rich" and "lean" in both cases.

    A gasoline engine is meant to be used a given fuel ratio. On an efficiency point of view, it should be at the fuel ratio burning all the fuel. If you deviate from that "ideal" fuel ratio, some fuel is not burn, thus wasted. If you decrease the fuel amount, the heat will drop accordingly. The thing is that if you increase the fuel amount, the percentage of fuel burn decreases, but there is more fuel and it more than compensate for the expected power loss due to mixing inefficiencies; In fact you get a power gain. This is only true to a certain point though. So you get an effective fuel ratio range going from "most efficient" to "most power" that are label from "lean" to "rich". The limits are around ##\lambda## = 1.12 (lean) and ##\lambda## = 0.875 (rich).

    With a diesel engine, it's different. We actually control the power output with the amount of fuel injected in the mix. To get little power, you inject as little fuel as possible and you increase the fuel amount to get more power. The problem with diesel engine is that the richer it gets, the harder it is to mix the fuel well with the air, thus decreasing efficiency (but not power). There is an upper limit where you cannot compensate the efficiency drop by adding more fuel, just like for a gasoline engine. So you get an effective fuel ratio range going from "low power / high efficiency" to "most power / low efficiency" that are label from "lean" to "rich". The limit is around ##\lambda## = 1.25 (rich) and there is no limit for lean (as little needed to run the engine).

    So the "rich" air-fuel ratio of a diesel engine is actually leaner than the "lean" air-fuel ratio of the gasoline engine.

    The following figure shows the actual effects of air-fuel ratio:


    The top graph relates air-fuel ratio combustion efficiency to the air-fuel equivalence ratio for a spark-ignition (SI) and compression-ignition (CI) engines.

    The other graph represents the actual amount of fuel burn vs the air-fuel equivalence ratio for the same engines.

    For a diesel, anything above ##\lambda## = 2 burns efficiently, but produces less and less power as you increase the value. The maximum power happens at 1.25 - only use in a racing context due to the enormous amount of soot produced (see below); production vehicles usually don't go under 1.65.

  8. Feb 24, 2017 #7
    Jack Action,

    I'm not sure your response is answering Suomi123's question, " . . ., however adding fuel in a petrol engine removes heat. . . ., however removing fuel from a petrol engine increases heat."

    First, ICE efficiencies are fleeting. What would happen to your chart if you took 5 different engines and graphed the effects of each engines AFR's at 5 different RPM's? Additionally what would happen at various loads applied to the engines? Loading aside I think there would be 25 different graphs. Furthermore if this is an efficiencies question is heat actually removed or is it displaced to emissions?

    If this is a question about actually removing heat then is this a question about "unintentional effects" of the outlawed practice in racing of Anti-Detonant Injection? This increases power and would be the rare occasion where a benefit could be found from bad tuning and poor maintenance in the family car.

    In a properly operating, modern emissions system there should be a negligible amount of unburned fuel. Agreed?
  9. Feb 24, 2017 #8
    Yes, I did indeed mean lower combustion temperature which is different. My bad for using the wrong words.

    This is the info I was looking for, thank you very much! Interesting graph's as well.
  10. Feb 24, 2017 #9
    Baluncores's reaction might be the answer I was looking for. It sounds like something that would make sense, I was looking where the extra heat came from.

    I imagine the charts would look somewhat different with different engines in different conditions. Certainly with diesel's if they have older or more modern injection systems. My question was more about why there are higher combustion temps when burning leaner and lower combustion temps when burning richer.
  11. Feb 24, 2017 #10

    Then you have asked about efficiencies and that has been answered. For further discussion the condition you have asked about will exist until a true continuous valve timing system is developed. The closest to succeeding I've seen is Christian Koenigseeg's work with pneumatic valves. He did well with lift and duration along with the efficient timing of open and closing of the valves for moving air. A couple of things yet to be solved are the "continuous" and cold starts. The system is set up through the ECU for the valve timing to have specific characteristics during specific RPM ranges. Efficiencies and inefficiencies have been narrowly isolated but valve timing is still not seamlessly continuous and per cylinder optimization is limited. The issues with cold start is self evident.
  12. Feb 24, 2017 #11

    jack action

    User Avatar
    Science Advisor
    Gold Member

    The point was to show that diesel and gasoline engines are not operating in the same air-fuel ratio range. Technically, a diesel engine air-fuel ratio is never "rich", i.e. below stoichiometric ratio. So you cannot compare a "rich" gasoline air-fuel ratio with a "rich" diesel air-fuel ratio.

    The efficiency I considered was only the portion related to the air-fuel ratio (##\eta_{\lambda}##). The graphs I've shown represent equations found in Design and Simulation of Two-Stroke Engines:
    [tex]q_{in} = \eta_o \eta_{SE} \eta_{\lambda} \frac{q_f}{\lambda AFR_s}[/tex]
    • ##q_{in}## is the heat available for combustion;
    • ##q_f## is the heat available from the fuel;
    • ##\lambda## is the air-fuel equivalence ratio;
    • ##AFR_s## is the stoichiometric air-fuel ratio;
    • ##\eta_o## is an overall efficiency that express the incomplete combustion due to incomplete flame travel into the corners of particular combustion chambers, weak or ineffective ignition systems, poor burning in crevices and flame decay by quenching in most circumstances;
    • ##\eta_{SE}## is defined as the mass of fresh air with respect to the total mass trapped inside the combustion chamber prior to the combustion; since exhaust gas may not have completely escape the cylinder during the previous cycle;
    • ##\eta_{\lambda}## is the combustion efficiency due to the air-fuel equivalence ratio and is defined as:

    For a SI engine with ##0.80 < \lambda < 1.20##:
    [tex]\eta_{\lambda} = −1.6082 + 4.6509 \lambda − 2.0746 \lambda^2[/tex]
    For a CI engine with ##1.00 < \lambda < 2.00##:
    [tex]\eta_{\lambda} = −4.18 + 8.87 \lambda − 5.14 \lambda^2 + \lambda^3[/tex]
    For a CI engine with ##2.00 \le \lambda##:
    [tex]\eta_{\lambda} = 1.00[/tex]
    Out of range values for ##\lambda## will make ignition difficult.

    The top graph represents those equations, the second graph represents the same equations divided by ##\lambda## (from ##q_{in} = \eta_o \eta_{SE} \frac{\eta_{\lambda}}{\lambda} \frac{q_f}{ AFR_s}##).

    Yes, in normal conditions. But when you push the gas pedal to the floor (WOT), the emissions are usually set aside, preferring rich mixtures for maximum power output and you will get unburned fuel. WOT is not considered a normal operating condition. In racing, unless forbidden by some rule, any engine tuner would be crazy to not run rich air-fuel ratio.
  13. Feb 25, 2017 #12
    Jack Action,

    You've answered Suomi123's question with flawless examples and explanations and I concede I took the question asked a step too far. Allow me to explain what I was considering.

    I'm retired now but unchanged from when I was working is the over riding rule; The internal combustion engine is inefficient. Anyone looking to work with IC engines needs to ingrain this into their working souls because there is very little they will ever work on that can't be traced back to an inefficiency problem. The most efficient engines today are only pushing 30% efficiency. When I started it was 20%. Turbo and supercharging improves this all a bit.

    Another concept anyone who plans to work with engines needs to understand, Their issues with solving problems isn't going to be the 30%, it's going to be the 70%. That is what they are going to be chasing. It doesn't matter if your working for GM or McLaren, it's expected that this is understood.

    Anyway, Suomi123's question was about heat and the source of the heat is the exhaust. A modern 5 gas, Gas Analyzer measures Hydrocarbons (HC), Carbon Oxides (CO), Carbon Dioxide (CO2), Oxygen (O2), Oxides of Nitrogen (NOx) and of course Lambda or AFR.

    At AFR of 14.71:1, CO2 is peaking and NOx is almost at it's peak. It's peaks at a slightly higher AFR (or lean) but both go down with either richer or leaner conditions.

    CO and HC are close to their lowest levels at Lambda and O2 is a rising slope as AFR increases.

    As I assume Suomi123 is aware as a mechanic, NOx needs to be read while the engine is under a load or on a dynamometer. This is why my first post suggested looking into the exhaust manifold and catalytic convertor temperatures.

    An important and pertinent point, with a modern engine the amount of water condensation coming out of the tail pipe equals or is pretty close to the amount of fuel combusted. Suomi123 would serve his future well to understand the condensation and where it is coming from. I would also challenge her\him to research the temperature of the gases in an exhaust. Are they all the same temperature until they are out the tailpipe or inside the combustion chamber? If not when are they different? Where are they different? Why are they different? And finally, Nitrogen in should equal the Nitrogen going out. If not, what's happening?

    "Adding fuel in a diesel engine will result into more heat, however adding fuel in a petrol engine removes heat. Removing fuel from a diesel results into lower temperatures, however removing fuel from a petrol engine increases heat."
    Suomi123, remember nothing is ever lost or gained, only converted or displaced. If you ever find yourself pursuing anything that would violate this, you know. Good luck.

    Racing engines running rich? I'm not going there. ;)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted