> Actually if we really wanted to prove this out the right way to do it is
> not with a top speed test but with a coast down test. Top speed tests
> won't be reliable if you don't change the gearing, plus they are
> influenced by how well you came out of the previous corner, or in the
> case of Kyalami, how well the car handles through the kink. A coast
> down test would get rid of these variables. You would need a track with
> a long flat straight, like Monza perhaps. Accelerate up to the
> start/finish line, then cut the throttle and put the clutch in (to
> eliminate the engine braking effect) and let it coast down to zero. To
> be accurate you would have to analyze this in GPLRA or something
> similar, so the person should then start the car up and complete the lap
> so that the replay can be loaded into GPLRA. Then in GPLRA in the speed
> vs. distance graph pick out the time it takes to coast from one speed to
> another. The longer the time, the less the rolling resistance. You
> would have to pick an initial speed that is less than your s/f speed in
> order to be consistent, say 150 mph. Or go even lower to reduce the
> influence of aero and increase the resolution. Also you would not want
> to time it all the way to zero, again for consistency. For example
> compare the time it takes to go from 60 mph down to 10 mph. I don't
> know whether the replays really have enough resolution to be able to see
> the differences that we are talking about, but ideally this is the best
> way to sort out differences in rollling resistance from one setup to
> another. Maybe someone else with a little time on their hands can try
> this out.
> not with a top speed test but with a coast down test. Top speed tests
> won't be reliable if you don't change the gearing, plus they are
> influenced by how well you came out of the previous corner, or in the
> case of Kyalami, how well the car handles through the kink. A coast
> down test would get rid of these variables. You would need a track with
> a long flat straight, like Monza perhaps. Accelerate up to the
> start/finish line, then cut the throttle and put the clutch in (to
> eliminate the engine braking effect) and let it coast down to zero. To
> be accurate you would have to analyze this in GPLRA or something
> similar, so the person should then start the car up and complete the lap
> so that the replay can be loaded into GPLRA. Then in GPLRA in the speed
> vs. distance graph pick out the time it takes to coast from one speed to
> another. The longer the time, the less the rolling resistance. You
> would have to pick an initial speed that is less than your s/f speed in
> order to be consistent, say 150 mph. Or go even lower to reduce the
> influence of aero and increase the resolution. Also you would not want
> to time it all the way to zero, again for consistency. For example
> compare the time it takes to go from 60 mph down to 10 mph. I don't
> know whether the replays really have enough resolution to be able to see
> the differences that we are talking about, but ideally this is the best
> way to sort out differences in rollling resistance from one setup to
> another. Maybe someone else with a little time on their hands can try
> this out.
> > Mmmmm, I don't think Spa is The Test. Going down the hill, well, it's
> > downhill. And going up the hill, well.... Ditto Riverside (at least
> going
> > down the back stretch). I'd try Reims or Le Mans.