I did some coast down testing today. Changes from the pictures above are:
1. A bigger grill block made of dark gray foam, it's harder to see and covers more of the grill.
2. More of the foam stuffed in around the radiator area to direct airflow through the radiator as best possible (the combination of the two left me with no overheating problems during testing)
3. An Uglyback/Kammback that still allows my trunk to open and close.
It was nice out today, mid 70's, light breeze every now and again, SUNNY.
My test track is flat with a bend about half way through. When I say flat, there is less than 1 meter of elevation change through the course. (Thanks go to bwilson4web for the tip on the location)
Coast down started at either 55 or 60 mph (depending on how fast I could get going before having to coast), and with no traffic encountered on the road, I was able to coast down to 15 or even 10 mph (depending on the start speed). I did 5 bi-directional runs, and the car was warmed up when I got to the test site. Total time (including getting to and from the site) was approximately 2 hours... almost as much time as making sure I had my data set up correctly!
Results (from excel) of the testing:
Cd = 0.277939368
Crr = 0.011853646
Realistically:
We can call that Cd,Crr 0.278,0.0119... and that is probably putting a lot of confidence in my timing abilities!
Confidence:
I also did some "smoothing of numbers" to correct for inaccuracies in my data collection, and the numbers were very close 0.286,0.0115... I guess this is why ASME recommends 10 bi-directional runs. If I had some good data collection software, I think that would help too. I guess you should be able to do this with a Garmin, since you can export the speed and time data, or a video camera and a digital or analog readout, but I had my iPod touch and a pad of paper.
I don't know what the individual benefits of each mod are, but collectively this is a 0.04 drop in Cd (given a stock Cd = 0.32). Not bad, but not finished either.