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

My internal combustion engine is more efficent than 30%

  1. Sep 27, 2010 #1
    I have an spark ignited 4 cylinder 2000 cc engine in a wagon and I think its 50% thermally efficient at idle due to the mods I have done on the intake manifold.

    My engine uses (mesured on video) roughly 9.807 cc/min at approx 750 rpm and the only reference I have come accross is a similar engine uses 5kw of fuel just to idle which works out at 29.55 cc/min of fuel.

    Is 5kw for idle a good guess?

    I have the video of the idle test on youtube, if anybody is intrested I will post the links.
    so I did the percentage 29.55 - 9.807= 19.743 then (19.743/29.55)*100= 66.8% then assuming an engine is 30% efficent 30*.668= 20.04 then 20.04% +30% = 50.04% is it ok to add the thermal efficency on to the original number?

    If not how would I work it out the thermal efficiency at idle?

    Another thing my city consumption is 58% better about 7l/100km and my hwy consumption is approx 34% better at approx 7l/100km my engine seems to work better under light loads.
    Its very strange to have hwy and city nearly the same that is pushing a law of physics.

    Another thing my engines idle increases by 9.33% when the temp goes from 160 deg f to about 200 deg f the above tests where done with a 160 deg thermostat, I now have a 195 deg F thermostat and I will fit it soon and retest my engine to see if I get a 9.33% increase.

    My first real test of my engine in 2005 was 340km to approx 14 L of fuel at 80km/h with a 190 or 195 deg F thermostat this is nearly 100% more efficent but it was running very close to knocking most of the way, since then some one has sliped silicon in my engine and damaged it so its not as good as it used to be.

    << link to commercial whereis website removed by berkeman >>
    Last edited by a moderator: Sep 27, 2010
  2. jcsd
  3. Sep 27, 2010 #2


    User Avatar
    Science Advisor
    Gold Member

    An idling engine is 0% efficient. With no load there is no conversion of heat to mechanical energy.

    If you wish to measure efficiency accurately you'll need to hook the engine up to a controlled load so you can measure the work done per fuel energy used.

    At idle all the mechanical work is lost to engine friction. You can't extrapolate from similar engines because the friction could vary significantly. Little things like oil type, temperature, rings, whether you broke in the engine properly or not, and simple variations in manufacturing tolerances may dramatically change the engine friction.

    I suggest you link your engine to a generator and use that to power something like a heating element or bank of lights to produce a controlled load. You can then either measure the electrical power output or measure the torque and rotation speed to calculate power (work per unit time). Then work out(load)/energy in (fuel) is the efficiency.
  4. Sep 27, 2010 #3
    The stock 3.4 litre engine in my minivan uses 16.7 cc per minute. That works out to 4.9 cc per litre of engine displacement, which is exactly the same as yours. Sorry.
  5. Sep 27, 2010 #4
    Ok so how did you measure it I am intrested?
    Is it an EFI or carby?
    My city cycle is approx 58% better and I have a tacho graph see below.
  6. Sep 27, 2010 #5
    EFI with a fuel consumption readout. It takes six minutes of idling to use .1 litre (100 ml).

    The reason that city fuel economy is usually worse than highway is that it involves traffic. Most cars will get very good fuel economy at a steady 60 kph, often better than at 100 kph.

    What mods did you make? And 58% better than what?
  7. Sep 28, 2010 #6
    I can gaurantee you that you've not increased your thermal efficiency to over 50% purely by arsing around with the inlet geometry.

    The fact is you've openly guessed at the vast majority of your numbers. The sheer amount of error involved with 'guesstimating' everything can lead to so very funny results/conclusions.

    You'll need to accurately measure a heck of a lot more to get a better idea as to what you are using, and therefore the real efficiency.

    Even using your numbers you get a very odd answer, shoing that they are most likely wrong.

    Thermal efficiency = total energy available from fuel / total energy output.

    Output: 5kW for 1 minute = 300kJ energy.
    Input: Assuming standard petrol contains 34.8 MJ/l
    You used about 10cc of fuel in a minute. 0.01 * 34.8 = .348 MJ energy = 348 kJ.

    So your indicated thermal efficiency is 300/348 = 86%. Which is clearly wrong.
    Last edited: Sep 28, 2010
  8. Sep 28, 2010 #7

    Ranger Mike

    User Avatar
    Science Advisor
    Gold Member

    Metrology is the science of measurement
    Measurement is the language of Science
    learn how to measure
  9. Sep 28, 2010 #8
    I suspect the mod has more to do with inlet temperature.
  10. Sep 28, 2010 #9
    ok so the .1 L is what the computer is saying. What is the resolution on the computer?

    The mods I made are like a golf ball surface that simple.

    58% better than some one who has owned one from yahoo answers see http://answers.yahoo.com/question/index?qid=20100220205012AAUkG6f"

    The city test was done driving arround the block 12 times and at speeds of 50 60 70 km/h see post with graph.
    Some rough stats from the data based on engine rpm, crusing 82.61%, idle 7.41% rpm above 2440 rpm 9.94%.
    I also calculated when the acceleration rate exceeded a threshold eg when in lower gears accelerating the figure is 19.49% of the time the engine was running during the test.
    Last edited by a moderator: Apr 25, 2017
  11. Sep 28, 2010 #10
    Yes for guessing I am using what wally said on a forum that a 30% efficient engine similar to mine uses 5kw of fuel just to idle he calulated it he wrote

    fuel = 34.656 Mj/litre
    specific weight majority of cases = 0.72 (can go as high as 0.79 and as low as 0.71)
    engine efficiency 30% = 69mg/kW (0.09858cc/kw) = 5.91cc/kw/min
    0.868 l/s air /kW at stoich 14.7

    Thus 5kw=29.55 cc/min at idle.
    I do not know how accurate the 5 kw is I have no idear.

    Increased thermal efficency, then why is my city and hwy consumptions very similar?
    see tacho graph on other post.

    I know I need to accurately measure all sorts of things but I am on a pension and do not have any money left to spent on it.

    Another thing is my idle speed increases by 9.33% when going from 160 deg f to above 200 deg f, How normal is that?

    Another thing I could do is hire a video camera and record my city test curcuit then host on youtube.

    And I could just add that this is my second prototype my first one used to knock when the timing was fired 4-5 deg before top dead center above 3000 rpm with no load, I ended up destroying it before I could fix the timing problem later on I threw it out.
    Last edited by a moderator: Apr 25, 2017
  12. Sep 28, 2010 #11


    User Avatar

    Staff: Mentor

    Fuel economy isn't easy to measure consistently due to the amount of driving you have to do to get accurate measurements. In any case, 7l/100km is good but not terribly exciting. I drive a 2.4L Mazda 6i and get about that under good conditions.
    That doesn't sound much like city driving to me - how often did you come to a complete stop?
    Last edited: Sep 28, 2010
  13. Sep 28, 2010 #12
    But you have to compair eggs with eggs.
    I have a 1984 nissan bluebird wagon with a carby with a 2 point vacumme advance. I can only speculate that if I had an modern EFI engine I might me another 20% to 30% better off if the load map was correct and having fuel injectors.

    For the city test that I did, I drove arround the block 12 times, 6 laps turning right and 6 laps turning left so I would cop the stop signs on 1/2 the journey and doing 67 km in one go.

    Acording to my tacho graph my engine was idleing for 7.41% of the time I think this is a bit low but I can not find a standard anywhere to test my car against eg the amount of idle time I think it should be neer 15%, thats easy to derate my figures
    the test ran for 91 min and I am short 7.59% so say 7 min at 10 cc/min I used 4.76 L in 91min +70cc any way about 7.184 l/100km
    Last edited: Sep 28, 2010
  14. Sep 29, 2010 #13
    There is no way it should have been knocking with that little timing at 3000 rpm and no load.
  15. Sep 29, 2010 #14
    It was knocking and the spark plugs would only last 3-4 days
    see http://www.youtube.com/watch?v=XnwJuNbvULY" for a look at my plugs back in 2002.

    This is the timing of my first engine http://ampair.tripod.com/Graph/My_first_engine.html" [Broken]
    The 1st and 2nd graphs would stop my engine knocking but the power level was low.
    The 3rd and 4th are when I lowered the compression ratio to 8:1.

    My second engine timing plots at http://ampair.tripod.com/Graph/My_second_engine.html" [Broken]
    note notice how there is no bend at 3000 rpm.
    Over time my timing base line seems to go lower.
    I think its due to carbon build up.
    Just for the record my current timing base line is as follows
    750 rpm 1.18 deg BTDC
    2000 5.74
    4000 13.03
    6000 30.32
    10000 35

    Also I have a normal laptop computer running Microsoft's Quick Basic v4.5 controling my spark timing eg running in realtime.
    Last edited by a moderator: May 5, 2017
  16. Oct 2, 2010 #15


    User Avatar
    Science Advisor
    Gold Member

    Dear me, this is pretty shocking. What is engine speed supposed to show you anyway?

    To compare fuel consumption with different engine setups, you really have to do tests on a dynamometer to hold all possible test conditions constant. I really don't believe that meaningful conclusions can be drawn with the test methodology currently in use.
  17. Oct 2, 2010 #16
    The big problem is I do not have any money to spend on testing it, so I can only do what I can and report my results.
    When my engine was undamaged in 2005 we did a trip of 340 km on a flat road in the country crusing very carefully at 80 km/h in overdrive with the engine slightly knocking under acceleration and used approx 14 L of fuel thats 4.11 L/100km, but since then I have had silicon put in my engine and damaged the cam and crank also I have a very low temp thermostat as well and I do not push my engine into the knocking zone anymore.
    I used the low temp thermostat because the timing drifts greatly with temp, soon I will be fitting a high temp thermostat again and run some more tests.
  18. Oct 3, 2010 #17


    User Avatar
    Science Advisor
    Gold Member

    I fully understand why you're not doing controlled condition tests, but let me put it like this:

    My job is all about developing engines, and fuel consumption is an increasingly important parameter. I'd never even dream of wasting money trying to do this in a vehicle, because the variability in SFC which can be attributed to uncontrollable factors (humidity, operator mood, wear, traffic, ambient temperature to name a few) far outweighs any change in fuel consumption that I would reasonably foresee in making the kinds of modifications I would make. That's all assuming the engine is a constant; you're talking about knock, damage, and several years between tests. Your methodology simply cannot yield valid results, I'm sorry.

    So, I understand (ish!) what you're trying to do and why you're trying to do it, but I hold no faith in your test methodology. You said it yourself - you have to compare apples with apples and it's impossible to do this by road testing.

    More crucially, I'm not sure what your 'idle' experiments are trying to show; the fuel used at idle balances the FMEP and pumping losses. Volumetric efficiency (which is presumably what you're trying to improve) makes bugger all difference at idle, because the air flow is so low; and even less in a spark ignition engine because your flow is intentionally throttled anyway!

    We're not here to piss on your chips, we're here to help, but what you're doing is flawed and we'd rather see you learn a bit about conducting a proper experiment than waste time and fuel on your current methodology.
  19. Mar 21, 2011 #18
    Ok 1 problem its 5 HP not kw so I redid the numbers, If you want the links to my evidence please let me know and I will post them.

    Output: 5 HP = 3.7284kw for 1 min = 223 kj energy
    Input: My best is 10cc of fuel/min 0.01 * 34.8 =.348 MJ = 348 kj

    So my indicated thermal efficiency is 223 / 348 = 64%
    I have it on video on youtube

    I did a 40 min test and measure in 2008 and the average was 13.66 cc/min

    Output: 5 HP = 3.7284kw for 1 min = 223 kj energy
    Input: my average 13.66 cc of fuel/min 0.01366 * 34.8 =.475 MJ = 475 kj

    So my indicated thermal efficiency is 223 / 475 = 49%

    I did a more recent test and used approx 15 cc/min

    Output: 5 HP = 3.7284kw for 1 min = 223 kj energy
    Input: my best is 15 cc of fuel/min 0.015 * 34.8 =.522 MJ = 522 kj

    So my indicated thermal efficiency is 223 / 522 = 42%

    so its dropped off a bit when I have some money to spare I will rip it apart and look at whats gone on

    Note: I used 5 HP not 5kw because that's what is quoted around the traps for a 2L 4 cylinder engine.
    I have even been quoted 30cc/min just to idle as well for a normal 4 cylinder but my program puts that down to 20 cc/min for 5 HP.

    Thanks for your time Chriss
  20. Mar 21, 2011 #19


    User Avatar

    Staff: Mentor

    A guess is still just a guess.
  21. Mar 22, 2011 #20
    I came to a complete stop roughly 28 times the data log is a bit coarse
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook