Brake Mean Effective Pressure made easy | Auto Expert John Cadogan | Australia

A proper brainiac calculates BMEP using a formula and established units. If you want to do this, be it entirely on your head.
You can take the torque at maximum power and multiply it by four times pi (in a four-stroke engine; two times pi in a two-stroke) and divide that all by the engine’s displacement volume.
So: Why do we want to know this? It’s a great way to compare engines for how well they’ve been designed, and you can use in when an engine is on the drawing board.
If you’re designing an engine and you know how much torque is required, you can calculate how much displacement you’ll need, given some ballpark knowledge of best-practise BMEP for that kind of engine.
But more interesting for we car nuts who don’t actually design engines, there are two big fudges for making more power on the drawing board.
You can just make an engine bigger. If you go from 2.0 litres to 2.5 you’d expect to get 25 per cent more performance, all other things being equal. It’s an easy win.
Adding a turbo is a bit like that - it artificially pumps up the volume by jamming more air in.
Or, second fudge, even easier, you can just spin the engine faster, right? I call this a ‘mid-life, marketing department upgrade’. This is the cheapest, nastiest way to get a nominal power increase.
Power equals torque times revs. (As long as revs are in radians per second, and not rpm. Because: Applied physics.) So if you can maintain the torque and spin the engine faster, you will get more power.
Let’s say you’ve got some engine. Making the numbers easy: It’s developing a peak of 200 kilowatts at 5500 revs. If you can keep producing the same torque at 6000rpm - that’s about 10 per cent faster - you’ll make 220 kilowatts.
Then, when you refresh the model, you can say: ‘Now with 10 per cent more power.’
Never mind that this extra power is mostly inaccessible because people don’t generally drive north of 5500rpm, and the engine feels exactly the same at all revs below that - the marketing department will be very pleased with that alleged upgrade, and the PR dudes will spruik it endlessly.
(And of course, maintaining the torque 500 revs faster means being able to jam 10 per cent more air and fuel into the engine, and flow 10 per cent more exhaust out - so the existing inlet, fuel and exhaust hardware might need a bit of a tweak.)
It’s still a pretty low-rent and mostly useless alleged power upgrade.
Of course, the third (and best) way to get more out of an engine is with fundamental efficiency gains. By reducing internal friction, by improving the combustion management, by reducing heat rejection into the parts, boosting volumetric efficiency - stuff like that.
So wouldn’t it be nice to have a means of assessing power production that’s independent of revs and also independent of engine capacity? If only we could find one, which is of course what BMEP is. And I guess the mathematics with pi and all that unit conformity is just unfortunate.
But you’re willing to let your grasp on hard science loosen just a bit you can bastardise BMEP and bypass a lot of mathematical complexity. And that’s exactly what you should do if all you want to do is compare engines in terms of their basic design efficiency, independent of revs and swept volume.
All you do is take the peak power, divide it by the volume, and divide it by the number of thousand revs at which the peak power occurs. Simple.
In the example I used yesterday, the big fat Hemi 6.4 V8 was making 350 kilowatts at 6150 revs.
All you do is divide 350 kilowatts by 6.4 litres and divide that by 6.15 for 6150 revs.
You get 8.89 kilowatts per litre per thousand revs. You can do that with other engines and see, essentially, who’s got the best R&D.
You can use cubic inches and horsepower if you want - Retardistan - yesssss! - and you get a different number, but the same proportionality between engines. You have to be consistent - use kilowatts and litres all the time, or the other units, all the time - if you want the numbers to be comparative.
And you are actually calculating a pressure - it’s just not a pressure in units that are meaningful against the standard units we put pressure in. Dimensionally, it’s absolutely a pressure. (Just don’t try to buy a pressure gauge calibrated that way.
Someone will definitely have an intracranial bleed if I lay that out in great detail. Trust me - or confirm it with a physicist or engineer. It’s true. This is the ‘bastardised’ part of the process.
If you go out and compare a whole bunch of different engines - the bigger the number, the better. That’s how this works. More power per litre per thousand revs is good.