I think the MPG data should be presented as a series of 4-curves, with MPG on the Y-axis and MPH on the X-axis. In descending order, the curves would be:
1) MPG vs MPH with driver only, no A/C
2) MPG vs MPH with driver only, A/C operating
3) MPG vs MPH with driver, max. passenger(s)/load, no A/C
4) MPG vs MPH with driver, max. passenger(s)/load, A/C operating.
Initially, the data would come from actual, on-the-road vehicle tests, but as soon as the preliminary data was collected and analyzed, computer-simulations would be sufficiently accurate enough to eliminate further road-tests...until something aerodynamically or mechanically on the subject vehicle was changed.
|