[aerohead]

The coastdown must be performed on a 'level' couse,with no more than a fraction of a degree grade discrepancy over the entire length of the test course.Muliple runs are run in both directions to cancel-out any difference in grade.The car must be completely warmed-up,all fluids must be at their' respective equilibrium temps for ambient conditions.Tires must be fully warmed-up.Temperature,barometric pressure,air density must be constantly monitored during the test.A tape recorder can be used to capture velocities,as the vehicle deaccelerates,then later,a stopwatch can be used to ascertain the time intervals between velocities,for which to construct the velocity/time curve.An accurate frontal area must be known ( often published in CAR and DRIVER Magazine,or can be estimated).At no time can wind speeds exceed 7-mph during testing.Fuel should be topped-up between runs to maintain constant mass.No wet roads.Your curve from 20-mph and below will help you isolate your car's rolling resistance,everything above 20-mph is a mix of aero and rolling resistance.It's a rigorous procedure,I won't do it anymore.The wind hasn't stopped blowing in Texas for years now,making any efforts futile.A back door into Cds,is to track your mpg ( scangauge can speed this up!).At a constant 55-mph,any 2-percent drag reduction will yield a 1-percent improvement in fuel economy.At 70-mph,a 10% reduction in drag will yield a 6% improvement in economy.Dr.Glen Scharf (SP?),Director,Aerodynamics Lab,General Motors Corporation,is comfortable with these relationships,and since you should be monitoring your mpg anyway,if you modify your car,and you see an improvement at the pump,the drag coefficient will fall out of your data.And you skip the hassle of the coastdown headaches!