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

Combustion Temp's for an IC Engine

  1. Nov 8, 2005 #1
    This may be the wrong forum, so mods please move if necessary (and forgive my error :) ).

    I was trying to research combustion temperatures (for a typical four stroke engine) as they relate to air/fuel ratio (AFR). I have googled my butt off, and I read many interesting things. But, I still have some questions. My wisdom (or lack thereof) and experience tell me that leaning an AFR from stoichiometric, 14.7:1 (A/F), will cause exhaust gas temperatures (EGT) to decrease. The result being due to the fact that more of the heat is being absorbed in the combustion chamber primarily due to less fuel being available for both 'cooling' and controlled burning. In fact, if one see rising cylinder temperatures and decreasing EGT's it is a sure sign of some other form of ignition (detonation or pre-ignition) in the cylinder. It is also true in my experience that EGT rises when the AFR is made richer. I have always assumed that this was because some of the fuel was 'cooling' (absorbing heat) the mixture, and much of that was carried out through the exhaust stream and dissipated through the exhaust system. A good example of this is turbocharged engines. They need to run much richer than 14.7:1 in high-load situations, and their EGT's are much higher.

    What I am curious about is any information describing actual combustion temperatures related to different AFR's. If the combustion process is complete (or successful) is there any difference in combustion temperature when the AFR is say 11.5:1 versus 15.5:1? I am guessing that an engine would make more power with and AFR of 15.5:1, but that it could only do it once because the heat would not be tolerated by the components.

    Older racers, always thought that leaning out the AFR made EGT's rise. I think this is because cars used to come from the factory with much richer AFR's, and when they leaned them out EGT's actually rose some. However, I have never experienced this (and I am not that old), so I am a little confused.

    More Power = More Heat. At some point though, a really lean mixture say 20.0:1 isn't going to burn very well. So, I assume at that point there is less heat. Is there a curve plotting heat versus AFR?

    Sorry for the long rambling post.
  2. jcsd
  3. Nov 8, 2005 #2


    User Avatar
    Science Advisor
    Gold Member

    I'm currently learning a lot of this stuff at the moment, but with regard to CI engines, and I think you're asking generally about SI engines. I'll ramble on for a bit, extract what you can. I'll have a good read sometime and try and come back with an answer to what you actually asked, or with a bit of luck someone else will come to the rescue. For reference, most of this stuff can be found in John Heywood's 'Internal Combustion Engine Fundamentals', which I've only had for about a week...

    At full load, complete use of the air taken into the engine is the key issue, with regard to developing as much power as possible. At anything other than full load, when the engine is throttled, efficient use of the fuel is the key issue.

    At WOT, a rich mixture is used, with an equivalence ratio of around 1.1. Mixtures even richer than this are occasionally used to effectively cool the charge air by evaporation, thus increasing the density and improving the volumetric efficiency.

    At part-load, it is desirable to dilute the mixture, either by providing an excess of air, or (more recently) by using EGR systems. This leaner mixture causes an increase in thermal efficiency by increasing the work produced by the expanding gas, decreasing the work lost to pumping, and decreasing heat lost to the cylinder walls because of the lower cylinder temperatures.

    Where no regard is to be paid to legislation (notably NOx emissions), excess air is generally used to dilute the charge, and to run the engines with an equivalence ratio of less than unity. When restricted by legislation, a stochiometric mixture is used in combination with catalytic aftertreatment of the exhaust gases, which are then used to dilute the charge rather than fresh air. This naturally leads to a reduction in NOx.
  4. Nov 8, 2005 #3
    Thanks, I will have to check out that book. What is the definition of equivalence ratio? What does 1.1 represent (richer than 14.7:1, leaner)?

    I guess I was thinking along the lines of something like a distillation curve, maybe. For example, what are the CHT's for running an engine with 87 octane for 1 minute at 1500 RPM's with 55kPa manifold pressure for the following AFR's:


    If all other things are constant, how would the CHT or combustion temperature change? As the AFR is continually leaned out, does the temp continue to rise until combustion becomes incomplete?

    Is vaporized fuel really that 'good' at cooling the charge? It must be because turbo cars require much richer AFR's than naturally aspirated ones. e.g. 11.5:1 vs 12.8:1 during wide open throttle
  5. Nov 8, 2005 #4


    User Avatar
    Science Advisor
    Gold Member

    Sorry, equivalence ratio (phi) is defined as the ratio of the fuel/air ratio to the stoichiometric ratio. In other words, if phi=1, the engine is running at stoichiometric AFR. For phi>1, the mixture is lean.

    I can't answer your question with regard to temperatures yet, but it seems sensible to suggest that very lean mixtures would give low combustion temperatures due to a lack of energy supply, rising to higher temperatures somewhere around stoichiometry, and then falling off again when the cooling effect of vapourisation starts to bring cylinder temperatures down again. (It's this 'somewhere around' which I can't give you any insight into!)

    Vapourised fuel itself won't cool the charge, - it's the process of evaporation that will take energy from the charge air and cool it.

    Turbo cars don't necessarily require richer mixtures for cooling, but air/fuel ratio is a factor which is adjusted to control knock in turbocharged spark ignition engines, and the additional cooling effect of the richer mixture will allow higher boost pressures to be developed with a resulting effect on performance and overall efficiency.
  6. Nov 8, 2005 #5
    Understood. It is just the inverse of Lambda which is the ratio Air/Fuel.

    Thanks for that explanation.

    Would you consider 'knock' and temperature to be related? Knock being either pre-ignition or detonation. In either instance, it represents additional flame either pre or post the normal spark. I have just never heard anyone describe it in terms of preventing knock. In fact, I have some information from a test done a Porsche engine. I work for a company that did some testing on a Porsche motor running 20lbs of boost. For endurance racing, they found that the motor could live as long as they wanted/needed it running 11.7:1. For sprint races (20-30 minutes), they found that they could lean it out to approximately 12.2:1 with no problems. For more of a drag racing application (15-20 seconds max), they found that the motor could survive with 12.5:1. However, at 12.7:1 the motor was toast after just one pull on the dyno.

    I will inquire about the 'knock' aspect particularly in terms of pre-ignition, but I have never heard anyone talk about it in those terms.

    Does the highest combustion temperature occur at stoich (14.7:1)?
  7. Nov 8, 2005 #6
    BTW, I would definitely consider pre-ignition and detonation different animals. If you open up your engine and find a big hole in the middle of the piston, it is definitely pre-ignition. If the side of the piston is cracked, it is a result of detonation.
  8. Nov 9, 2005 #7
    Ha finally a topic in my field...combustion.
    Most Fuel/Oxidizer have the highest (adiabatic) flame temperature under stoichiometric conditions. This is because All the high enthaply compounts are being converted into low enthalpy compounds leaving the remaining heat release for increase in temperature.
    So say methane combustion
    CH4 + 2(O2 + 3.76N2) -> CO2 + 2*H2O + 7.52*N2
    You see there is no O2 or CO or anything like that left. the combustion takes place perfectly and the adiabatic temperature will be maximum.
    You can use online programs to calculate the adiabatic flame temperature for you.
    http://astronautics.usc.edu/utility/flame_temperature/" [Broken]
    detonation is a shock wave fed by a combustion process, this can be the case in pre-ignition
    Last edited by a moderator: May 2, 2017
  9. Nov 10, 2005 #8
    Thanks for the calculator link. That is what I was looking for. However, you have explained it as well. I am not familiar with 'enthaply' but have found some reading on the internet. So, I will try to educate myself before asking questions that are simply dumb.

    For the calculator:
    What should I select as my 'Reactants' for typical gasoline? HCO and O2?

    What does the Temperature input represent? Would this be like Intake Air Temperature and the Temperature of the gasoline as entering the chamber? i.e. charge temp?
  10. Nov 10, 2005 #9
    If phi>1, that means that the fuel/air ratio > stochiometric ratio. Which means that fuel in comparision to air is higher than for a stochiometric ratio. Shouldn't then the mixture be rich at phi>1?
  11. Nov 10, 2005 #10
    I found a better one for you.

    http://navier.engr.colostate.edu/tools/equil.html" [Broken]

    Ethalpy is defined as h = u + Pv where u is specific internal energy ent P is pressure and v is specific volume.
    Don't worry about it to much it is just a measure of the amout of energy per species under given conditions.
    Last edited by a moderator: May 2, 2017
  12. Nov 10, 2005 #11


    User Avatar
    Science Advisor
    Gold Member

    Yes, sorry. I usually think in terms of air/fuel ratio rather than fuel/air ratio! Good job somebody's paying attention...

    TexanJohn, I would also point out the fact that high combustion temperatures are not necessarily directly related to high exhaust temperatures, since in the combustion chamber you have heat generated from compression as well as heat generated from chemical release happening at what can be rather different times, and thus a lower in-cylinder temperature (with retarded timing) can produce higher temperatures further on in the engine, ie the exhaust. Not sure if we'd got over that or not, but thought it worth a mention.
  13. Nov 11, 2005 #12
    real world is not perfect
    not all the fuel burns
    because some never becomes vapor
    or the flame front never has time to get there
    in a very turbulent combustion chamber
    in the very short time allowed

    lean the mix tooo far and it will not go bang
    then you need stratifyed charge and precombustion chambers
    just get it to burn

    turbo motors run hotter because there is both more air and fuel
    cramed in by the turbo and the air temps are higher to start with
    before compression adds more heat

    I would guess nobody ever gets a perfect 14.7/1 AFR or even trys to
    and all error to the rich side so when they say lean they mean less rich and closer to perfect 14.7/1
    there for hotter and more power but more likely to blow up
    and nobody goes to the lean side at all as the consequences are well known
    Last edited: Nov 11, 2005
  14. Nov 11, 2005 #13


    User Avatar
    Science Advisor
    Gold Member

    This isn't right. CI engines always run lean, and by lean, I mean a higher AFR than stoichiometric. Typical AFR's are in the range 18-70. SI engines will also run lean under certain circumstances, with typical AFR's in the range 12-18.
  15. Nov 11, 2005 #14
    CI = diesel sure but I was comeing from a hotroders gas/spark end
    as diesels are heavy and SLOW so not of much intrest
    and turbo/gas [SI] cars stay rich or blow up esp in addon turbos
    or hotrods with bigger then stock turbos where the set up is closer to the ragged edge to start with
  16. Nov 11, 2005 #15


    User Avatar
    Science Advisor
    Gold Member

    Again, spark ignition engines do often run lean. I work with gas engines which rountinely run at 60:1 AFR's.

    In any case, who said compression ignition was of no interest? The OP didn't specify whether it was referring to SI or CI cycles. Please bear in mind that there are MANY more applications of internal combustion engines than hotrods.
  17. Nov 11, 2005 #16
    while it is nice to hear about lab results
    and interesting to hear about cutting edge research
    I would think a 60/1 AFR is pure lab and very far from the street
    at this point and time esp in a turboed motor

    I would be thrilled to hear how a higher ratio canbe made to work
    as we all would like to save gas costs
    as long as the motor stays in one piece
    and makes good POWER
  18. Nov 13, 2005 #17


    User Avatar
    Science Advisor
    Gold Member

    We sell something like 10,000 of these engines every year, and they're turbocharged. It's nothing to do with lab results, these have been in production for 15 years or so. We get as much power out of them as diesel engines of the same size.
  19. Nov 14, 2005 #18
    Lots of good discussion now. :)

    I have several questions now. :) Is maximum heat generated at stoich? Is this true for all fuels? I am sure someone here can make more sense of this slide show than me: "[URL [Broken] On page 34, it shows flame temperatures for various fuels. It appears to me that the most heat is generated at stoich except for nitromethane and hydrogen. If true, why do engines 'burn up' when they run lean under load? Does this graph (and all the preceding formulas) not take into account additional pressure from load? However, if the hottest temperature occurs at 14.7:1 (for gasoline) why don't engines burn up when loaded at that fuel ratio? Or do they? Although, I will say that I don't know if any factory computers would allow you to 'load up' a vehicle (say greater than 80kPa in the intake manifold) and still command an AFR of 14.7:1. Most computers would be in some form of Open Loop processing and run rich (command a rich AFR). And when I say rich, most factory computers will command something like 11.5:1.

    So, I am trying to reconcile combustion temperature with load. I know that I can cruise along the highway with little load on the engine and run 16:1 with no problem. I also know that at 16:1 with load, the motor burns up. Why? Does the load create that much more heat? How? How does that relate to the slide show above?

    What kind of engine specs run/operate at 60:1?
    Last edited by a moderator: May 2, 2017
  20. Nov 14, 2005 #19


    User Avatar
    Science Advisor
    Gold Member

    Where's this slide show?!

    As I said, I'm not sure about maximum in-cylinder temperatures coinciding exactly with stoichiometry. Increasing fuelling does cool the engine, since the volatile fuel droplets will consume heat in order to evaporate, and this is a reason why lean mixtures might cause an engine to run hotter. In addition, factors such as ignition/spill timing, dwell timing and even variable cam timing and lift can make a massive difference in cooling the engine, especially when these are used to promote scavenging of the cylinder, when you will have cool fresh intake air passing over the piston at TDC and straight out through the exhaust. To complicate matters, in more technologically advanced engines, these will be adjusted on the fly and relative to mixtures. I really need to do some reading up, sorry!

    The engines I mentioned which run at 60:1 are medium/large (60 odd litres) gas engines used mainly for power generation where fuel economy and engine durability are prime factors. While they're definitely not typical of an automotive application, the principles are identical.
    Last edited: Nov 14, 2005
  21. Nov 14, 2005 #20
    I messed up the link. Here it is: http://me.queensu.ca/courses/MECH43...on%20Theory.ppt" [Broken]

    Doesn't work, try: http://me.queensu.ca/courses/MECH435/5.%20Combustion%20Theory.ppt [Broken] This one works. Right click and 'save as'
    Last edited by a moderator: May 2, 2017
  22. Nov 15, 2005 #21
    Let me ask the question this way:

    Can someone give an approximate difference in temperatures (if there is one) when the cylinder pressure rises but the AFR stays constant? For example, if I am cruising along the highway, manifold pressure is 50kPa, and lets say that cylinder pressure (peak) is 8,000kPa. What is the temperature of this combustion? What if the conditions change so that I am going uphill, manifold pressure is now 85kPa, and cylinder pressure is 10,000kPa. What is the temperature of this combustion? Assume both have an AFR of 14.7:1. How much does the 'load' on the engine (I am assuming that cylinder pressure increases) affect the temperature.

    My fundamental assumption is that the increased load (going up hill) requires greater cylinder pressure (to maintain say the same constant speed), which then increases the temperature. Is this correct, logical, or am I missing something?

    BTW, the link above to the slideshow on pg34 indicates 'Constant Pressure'. I think this is where my question is arising. If pressure increases 25%, how much will the temperature rise for 87 octane mixed at 14.7:1 AFR?
  23. Nov 19, 2005 #22


    User Avatar
    Science Advisor
    Gold Member

    Congrats! Finally you are learning something important :biggrin: .
  24. Dec 2, 2005 #23
    Maximum heat release

    There is a difference in maximum heat release and maximum adiabatic flame temperature. Since a stoichiometric mixture usually does not burn up 100% you can actually get more heat release out of the same amount of fuel by adding a little more oxidizer. Note however that the heat released spreads out to heat more product molecules, therefore the adiabatic flame temperature will actually go down.
  25. Dec 27, 2005 #24
    Hi there:
    Check out the following link to help you out with the combustion calculations etc.: http://members.aol.com/engware [Broken].
    Last edited by a moderator: May 2, 2017
  26. Jun 27, 2006 #25


    User Avatar
    Science Advisor
    Gold Member

    Just remembered about this thread, and learnt something on a combustion course last week that might help out if you're still interested!

    In a gasoline engine, mixtures weaker than stoichiometric will increase cylinder temperatures for two reasons:

    The primary reason is that lean burning in a homogeneous charge is slower, and allows the heat to soak into the cylinder components (piston, bore, and flame face). The peak temperature is actually lower than with a stoichiometric mixture, but at stoichiometric the burning is so rapid that the engine components don't get chance to heat up as much. It's like swiping your finger slowly through a yellow bunsen burner flame, or quickly through a blue flame; your finger will get burnt if you move it slowly through even a cool flame, but is quite happy to dart through a blue flame.

    The secondary reason I think I already stated; fuel has a higher specific heat capacity than air and serves to suck some heat out of the cylinder.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook