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

Automotive Heat/energy balance for a high power petrol engine

  1. Oct 18, 2012 #1
    Hi, first post newbie here, so go easy please:smile:

    I'm trying to get a grip on the understanding of a car cooling system for a friends race car. It seems the more I look into it the more questions I'm left with. For this post I would like to clarify the coolant side, leaving the air side for another post.

    Having trawled the net I see most opinion says the power output from the engine is around 33% of the fuel energy available, with the rest rejected by the engine to the cooling system & exhaust, i.e. roughly equal distribution of fuel energy between the 3, power = cooling = exhaust.
    SAE J1349 (Jun 90) originally suggested 85% of power produced is delivered (after friction & ancillaries etc) to leave a nett available flywheel power of 28.3% which again is a figure I've found in my trawls.

    Sorry if this is a long & rambling post, but the following questions leave me uncertain.
    If it would be best to break these into seperate posts let me know.


    What I can't find is how this % distribution changes if a turbocharger is fitted. While some exhaust heat is used to drive the turbo to pump more air & hence produce more power, some is rejected into the cooling system by the water & oil used by the turbo.
    I would guess its a relatively small change if indeed there is any, but if the turbo is extracting heat from the exhaust it must alter the distribution?

    If the engine is running non stoichometric (λ <1 typically) mixtures then surely not all the potential fuel energy is being released as heat?
    The engine I'm looking at runs λ typically around 0.85 (AFR ~ 12.5:1) when under steady state full load, and AFR of ~11.1:1 during acceleration enrichment.
    This being the case, how much does this alter(reduce) the heat?
    Engines I have knowledge of all run 'rich' mixtures under full load. Has this been accounted for in the basic 'rule of thumb'?
    if say for (an extreme) example the engine power produced was only 20% of the available fuel energy would the heat rejection (to cooling) still be equal to the produced power? or does the heat reduce in a non-linear fashion?

    Bear in mind as this is a race engine it has a seperate large oil cooler;
    If you use the 'std' 33% to determine the heat rejection into the cooling system, how would this typically be distributed amongst the various elements conduction, radiation, Water, Oil?

    If I use the basic rule of thumb values above to rough out some figures for a 700bhp turbocharged petrol engine then I get:
    available power: 700bhp = 522kW's
    total mech power: 522/0.85 = 614kW's
    target coolant temp = 90 deg C
    Radiator dT = 15 deg C
    (values for mass & Cp found from here:
    if this was rejected solely into the water cooling it would require a cooling system flow rate of approx 600 litres/min (134 gpm). This is a high rate so clearly some heat is dissipated by the other 3 elements?

    I found this SAE paper on the Ford GT cooling design:
    In this they seem to be designing for a heat rejection to the cooling system of only around half the available output power of the engine rather than the nominal equal value total mechanical power produced. This is for a supercharged engine, but the figures they use are significantly different to the 'rule of thumb' figures, even more so as the oil cooler is an oil to water (rather than air) exchanger.

    I found some 'finger in the wind' suggestions of 1litre/min/bhp which suggests this engine requires 700 litres/min!
    Elsewhere I've seen figures for similar power engines using 100 gpm (~450 l/m) but also an oldish BMW F1 engine of circa 850 bhp using 450 l/m. These higher power (esp the F1) engines also use a much lower value of dT than I have in the approximation above(q3)

    I realise the engine doesn't spend it life under full load condition, but any help in understanding these greatfully received.

    Thanks if you got this far!

  2. jcsd
  3. Oct 20, 2012 #2

    jack action

    User Avatar
    Science Advisor
    Gold Member

    I don't know if this will help but Caterpillar offers some nice data for their engines. On page 20 of their application and installation guide for cooling systems, they talk about Cooling System Sizing and give some definitions for their technical data. The technical data for their engines gives the heat loss to exhaust, surroundings and cooling system. For example, the 3412C TA diesel generator (800 ekW output) have the following values:

    • Heat rejection to coolant (total): 511 kW
    • Heat rejection to exhaust (total): 860 kW
    • Heat rejection to atmosphere from engine: 132 kW
    • Heat rejection to atmosphere from generator: 36.8 kW
    So there is an 836.8 kW output for the engine (36%), 860 kW for the exhaust (37%) and 643 kW through cooling (27%). But only 22% goes to the water jackets.

    In the guide, with the example for the coolant flow calculation for that engine, CAT finds 198 GPM for their engine (with 508 kW heat rejection, similar to your engine's calculation).

    This site talks about a «typical Winston Cup» engine (similar to yours) with a 100 GPM coolant flow.

    So you must not be far with your estimation.
  4. Oct 22, 2012 #3
    Thanks for those links Jack.
    The one discussing the cup car seems to be taken from the Stewart web site
    http://www.stewartcomponents.com/tech_tips/Tech_Tips_6.htm [Broken]

    I presume is based on the generator o/p @ 95% efficiency?

    The Cat site certainly gives more info than Cummins.

    Given the race cars will usually have seperate oil & inter(after) coolers then this obviously lightens the load on the water system.:redface:

    Page 86 is very illuminating & sums things up nicely in combination with the worked example they give on page 87 certainly helps me better understand things.

    The info provided certainly seems to tie in better with the flow rates I've been finding in my searches, in particular a suggestion by someone that bhp/3 = litres/minute.

    My friend is on to the dyno to get the full logs from the runs they did. With this we should be able to finally understand the situation.

    thanks again

    Last edited by a moderator: May 6, 2017
  5. Oct 22, 2012 #4

    jack action

    User Avatar
    Science Advisor
    Gold Member

    The electrical output is 800 ekW and the heat rejection from the generator is 36.8 kW; I just assumed that there is no other loss, so the generator's input (hence, the engine's output) must be 836.8 kW.
  6. Nov 29, 2012 #5
    What is the intended purpose of the race car?
  7. Nov 29, 2012 #6
    The car has already completed a seasons racing in classic thunder championship.
    This is 2 races per meeting each approx 35 miles duration for which I was helping a friend during the seasons racing. Found it to be a good championship, just lacking some suitable competition in his class.
    While it completed the season, it was always running much hotter than the owner believed and I was trying to convince him of this, hence the background info I asked for above.
    Certain parts of the car were specced & built by someone else & I always felt the cooling system was far from capable for the power level.
    The owner has now decided to have the cooling system redesigned...
  8. Nov 29, 2012 #7
    What model car and engine and where on the engine is the temperature sampled?
  9. Nov 29, 2012 #8
    Nissan Skyline with 3l single turbo engine measured just short of 800 bhp at the hubs.
    Where the temp(s) were sampled was part of the trouble I had convincing him on.
    He finally accepted the water temp was measured at the intake to the engine, & regularly saw temps of 95+ deg C. Oil temp was also being measured at intake to engine & this got to 110+ deg...
    I had previously measured another (almost same spec, but 140bhp less power) and found water temps across the rad of 25 deg & more.

    I know F1 engines & the like run this hot, but I've not seen an inline 6 survive these. Surprised it did the season, althought the engine has just been stripped & it would not have lasted another meeting apparently.
  10. Nov 30, 2012 #9

    Ranger Mike

    User Avatar
    Science Advisor
    Gold Member

    Our Formula Car typically races at 95 tp 102 C. with Oil temp 20 to 30 degrees hotter..Our super late model big block V8 cars ran 115C after a good race and never had a problem..those big engines like it hot to make a lot of heat. You can not boil over though. Keep in mind, you make HP with the heat cylce engine and detonation is THE factor to avoid. Hot is good as long as you control it. Valve springs are the biggest casualty of heat and piston rings. But, if you run a dry sump and proper rated oil cooler that is mounted where air stream can properly cool it along with proper size radiator ( same air stream rules) you are good to go.
    Some things to think about- reverse cooling, electric water pump, multiple radiators, Water Wetter,,never antifreeze
    Last edited: Nov 30, 2012
  11. Nov 30, 2012 #10
    From what I've experienced, 300F is the magic number for oil temperature.

    Many a Z car story from improved touring 240z to IMSA GTO Z32 usually contains "I watched the oil temperature climb and first heard the rods at 300 (degrees F)........a couple of turns later, it was done".

    The L engines have an issue with the engine management system sampling from near cylinder 5 while the gauge samples from the thermostat housing but I believe the RB series has the temp sensors clustered on the intake manifold.

    To make 800 from a RB requires substantial revving and you might be running into water pump cavitation.

    It's also possible you may need to run external coolant lines to the trouble spots.

  12. Dec 1, 2012 #11
    Bear in mind the figures I quoted are intake temps, not the usual outlet temps.
    I agree hotter is better & in this case no knock triggers have been seen on the datalogs.
    But it was found the deck needed 20 thou & head face 6 thou to true them up again so it suggests these engines do not like such high temps (esp combined with the large water dT)

    It was decided (when the car(s) were being built) that they would try rear mounted water rads & the lack of water flow combined with lack of air flow to the rad led to the engine(s) running hotter than intended.
    1 car runs a mech pump & due to lack of funds has only been out for testing this year, but also runs cooler. It was this car that we took several measurements from (air + water in & out temps, air speed through the ducting) while the other (that did the full season) ran electric pumps.
    These electric pumps have long been under suspicion as being part of the problem. 2 Davies Craig pumps were used on the recommendation of 'someone who should know'. If you look at the flow/pressure graphs on the company's web site in my opinion they are nowhere near suitable for these kind of power levels
    All these probs are now being addressed (translated = the experiment didn't work...:rolleyes:).

    Given the temps I mentioned, the outlet temps were approaching this.

    Not really, both cars produce peak torque around 4000 & peak hp around 7000. Both have the limiters set to 8200 (soft limiter @ 8k)
    IIRC the production R32 was redlined at 8000
    There is yet another 'team member' that has run a redline of 8800 for the last 4 years racing with a std mech pump & front mounted rad without any cooling issues at all.

    Really comes down to the reality that the rear mounting of the rad was not designed correctly in both air & water flow needs.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook