HP loss between flywheel and wheels

[cinquecento]

What's the accepted norm when calculating the loss of power in between a horsepower flywheel reading and an "at the wheels" reading.

After some searching, 22% seems a popular figure.. however I understand compression and transmission type can change this.

What an base figure for a Griff 500?

Edited by cinquecento on Saturday 30th October 08:00

[Green Griffith]

If you want the real losses you have to measure them. There are few rolling roads that (really and exactly) can do that, for example the MAHA dynos. You can see e.g. the difference between a cold and a hot wheel bearing on one of those...

[The AJP Griff]

Its a real minefield!Ive now done over 250 runs over the last few years and have used a couple of different brands which calculate the drivetrain losses by measuring the drag on the rundown,but both makes still give significanly different figures.For example,the 'Dyno Dynamics' rollers i've used(2 different ones)have given significantly and more importantly,consistently higher figures than the 2 different 'Sun' branded ones.I used to read that 15% losses were an acceptable figure for most rwd cars,but 20% is often quoted in TVR land! It impressed me that different dynos of the same brand were so close to each other on the results,though.Personally i much prefer the lowest figures,but it matters very little when you think about it.If you compare a rolling road with an engine dyno,who's to say the engine dyno is accurate for example?It's just far more important to use the same dyno for your comparisons,leave the biggest reading ones for the bar room boasters

[Green Griffith]

It really is a minefield, yes. I agree: If it´s only about a nice shootout and some more or less impressive figures for comparison, you´d take every dyno. If it´s about serious diagnostics and developement you´d take the manufacturer´s choice: So if e.g. Mercedes and Porsche uses MAHA dynos, that means something to me. And no, I´m not on their payroll...

[dnb]

How about working it out from first principles?

Find a nice flat straight bit of road, connect up your datalogger (GPS that gives speed will be fine) and drive along as fast as you are allowed (or think you can get away with). At a given mark, drop the clutch and let the car coast to a stop. Time it with the data logger. Do a run the other way too so you can average the results. You might also try in different gears if you find drag losses dominate at high speed too much.

The deceleration, mass of car, frontal area and Cd can be used to compute estimates for drag and transmission losses.

[The AJP Griff]

dnb said:

How about working it out from first principles?

Find a nice flat straight bit of road, connect up your datalogger (GPS that gives speed will be fine) and drive along as fast as you are allowed (or think you can get away with). At a given mark, drop the clutch and let the car coast to a stop. Time it with the data logger. Do a run the other way too so you can average the results. You might also try in different gears if you find drag losses dominate at high speed too much.

The deceleration, mass of car, frontal area and Cd can be used to compute estimates for drag and transmission losses.

Heavy going for mere mortals like myself though Dave.I have a performance meter but still wouldnt be sure how to compensate for the drag of the car
In answer to the op's original question,i reckon 16-17% is about right for drivetrain losses judging by my experiences.

[spend]

I wrote a simple program to work out the stuff you need for performance meters Mike if you need it.

You just enter 60-55 and 15-10 decel in secs along with kerb weight and it gives you the CD and rolling resistance to 'calibrate' the performance meters (at that loading...). Half fueled gives a good enough average for me, just seems sensible to get them in the right ball park rather than obsessed with the variables?

Tubbs is still struggling with the weight measurement though