To round off the articles on power and torque, here is a real example of how the coastdown losses from an actual car were measured on a rolling road dyno. Some time ago I asked a colleague to run a series of tests for me on their rollers. What we needed was a car with a reasonably well quantifiable flywheel bhp and one that we could run in any gear without getting wheelspin on the rollers. This ruled out modified cars that had not been on an engine dyno and anything with too much power. Some time later a completely standard cvh engined Fiesta XR2i came into their workshop and this seemed as good a choice as any. The engine was in good condition and absolutely unmodified according to the owner - a good chance therefore of it producing close to the quoted horsepower.
The aim of the test was to see how wheel bhp and coastdown losses change depending on which gear you run the test in. The rolling road in question is a Bosch flywheel system, which means it has a heavy flywheel attached to the rollers and the system works out power according to how quickly the car can accelerate this large mass. It can't take "steady state" power figures which can be a hindrance when setting up fuel and ignition systems but on the other hand there is nothing for the operator to tinker with and distort the readings - you just sit in the car and floor the throttle and wait for the run to reach maximum rpm. At this point you can put the car in neutral while it "coasts back down" and the system measures these coastdown losses. Some dyno systems then add these losses back to the wheel bhp and call the result "flywheel horsepower". Proponents of this method claim that the "flywheel horsepower" figures so produced are more consistent and repeatable than wheel bhp figures. Hopefully this article will show the pitfalls in relying on coastdown losses by means of this real example - anyway on with the plot.
Ford quote 110PS (i.e. about 108.5 bhp) as the standard flywheel power for the car in question. Obviously every individual engine will differ slightly and this quoted figure can only be a guide to the spread of power outputs that a selection of engines would produce. To restate my own rules for estimating wheel bhp from flywheel bhp - about 15% transmission losses for front wheel drive cars and 17% for rear wheel drive is a rule of thumb. This tends to overstate the losses for high powered engines and understate them for smaller ones. A more sophisticated guide is to deduct 10% of the flywheel power plus another 10 bhp for FWD and 12% plus another 10bhp for RWD cars. The XR2i is FWD of course so if we apply those two rules to 108.5 flywheel bhp we get either 92 or 88 bhp at the wheels respectively. So that's the sort of level of wheel bhp that one would be expecting if the quoted flywheel bhp is correct.
To run the test, the car was warmed up and given a couple of runs on the rollers to stabilize the temperature of the tyres, gearbox oil and engine. A power run and a coastdown were then done in each of 3rd, 4th and 5th gear with a few minutes for the car to cool down between each run to keep the figures consistent. So first let's look at the how the wheel bhp changed in each gear. The figures are as follows:
3rd gear - 95 bhp at the wheels
4th gear - 92 bhp at the wheels
5th gear - 88 bhp at the wheels
So why do the figures show a drop in power as a higher gear is used? The engine of course is producing exactly the same flywheel power regardless of which gear the car is in - what is changing here is the real transmission and tyre losses. A higher gear means that the tyre speed on the rollers goes up too - this leads to more power being absorbed as heat and friction - the measured wheel bhp therefore goes down a bit. There are other factors at work here too but it is not the aim of this particular article to go into all of these in depth. The key thing is that the figures show a reasonable and predictable trend and are in the estimated bhp range calculated above.
It makes the point though that there is no such thing as just one true wheel bhp for a given car on a given set of rollers - it depends on tyre pressure, gear ratio and a host of other things that have already been covered in previous articles. Now devotees of the "coastdown loss" system would say that it should compensate for this - it should reflect the larger losses in higher gears by showing a larger coastdown loss which when added back to the wheel bhp ought to give a flywheel bhp that stays the same in each gear. So let's now look at the coastdown losses that were measured on each of those runs and see if they actually do what is claimed. The coastdown losses were as follows:
3rd gear - 17 bhp coastdown loss
4th gear - 27 bhp coastdown loss
5th gear - 44 bhp coastdown loss
We can add those losses back to the wheel bhp to get the estimated flywheel bhp that so many rolling roads these days quote you.
3rd gear - 95 + 17 = 112 bhp
4th gear - 92 + 27 = 119 bhp
5th gear - 88 + 44 = 132 bhp
Well clearly something isn't working here. The coastdown losses (whatever it is that they are actually measuring) are rising much more in a higher gear than the actual transmission losses are, leading to larger "flywheel" bhp figures in the higher gears. The engine is producing the same power all the time and although we can never know for certain exactly how much power this particular engine had, we can be fairly certain it isn't far away from the factory quoted power. Even the 3rd gear "flywheel" figure is a tad on the high side but it is within the realms of possibility - the figures in the other two gears are obviously not.
The wheel bhp data show a consistent and understandable pattern. Adding back the coastdown losses leads to power figures which vary much more and make less sense. The point to remember is this - if the coastdown losses really were an accurate measurement of the true transmission losses then we would expect to end up with the same estimated flywheel bhp in all 3 gears. The fact that this does not happen means by definition that the coastdown losses are measuring something other than true transmission losses - in turn this means that adding them back to wheel bhp cannot result in true flywheel bhp. The fact that they result in horsepower numbers much larger than the 108.5 bhp claimed for this engine only go to reinforce the message.
So the moral, for the last time hopefully, is to look at the wheel bhp as well as (or preferably instead of) the estimated flywheel bhp. It won't be a figure you can take for gospel and it will change from day to day and from rolling road to rolling road. With a modicum of common sense in keeping the test conditions the same and applying reasonable amounts for transmission losses it will get you "in the ball park" of what the true flywheel figure might be. The flywheel figure generated from coastdown losses though, can vary from the sublime to the ridiculous. Every now and then it might come up with a realistic bhp number but it might equally well be a country mile out.
To estimate true flywheel power from wheel power just apply the rules given at the start of this articles in reverse.
FWD cars - add 10 to the wheel bhp and then divide the result by 0.9
RWD cars - add 10 to the wheel bhp and then divide the result by 0.88
Remember these are estimates. The only way of knowing true flywheel bhp for a particular engine is to run that engine on an engine dyno.
Back to Main Menu page