The engineering department at Edison2 greatly appreciates the informed comments and gentle prodding we get on the Blog page. We have and will continue to do our best to put actual information up here. We hope everyone will realize that we also have our jobs to do so sometimes it might be a while to answer even really good questions.
We didn’t have stopwatch on it but it seemed to take only about 5 minutes before Kevin pointed out that the C term in our last blog post didn’t make a lot of sense. He’s absolutely right, it doesn’t, and the reason for that is also the reason that the car industry needs coastdown testing and does not just rely on wind tunnel numbers.
The SAE standard that defines coastdown testing is J2263. It’s available for download from multiple places online but because it’s copyright material you have to have to pay for it. It’s an interesting read if you’re into this kind of thing (and a great cure for insomnia if you’re not) and it delves into some of the complexities that surround this. For example, how do you account for rotating weight? A wheel not only has linear inertia because it’s travelling but it also has rotating inertia because it rotates. J2263 discusses this at length and in detail.
Whether or not stuff like rotating inertia is significant depends on the numbers you’re trying to find. If all you really want to know is how much drag there is at any given speed, the coastdown method is great. Plot speed against acceleration (and, since mass is constant, acceleration gives you force) from the coastdown and you have the drag profile. It happens that car drag profiles very reliably fit the A + Bx + Cx^2 three term general form.
Here’s a parallel: if you want to know downforce, do you take a bunch of pressure taps and attempt to integrate the pressure contours you generate over the car’s planform area? You could but it’s kind of messy. Far better to measure the downforce directly because that intrinsically integrates the air pressure over the whole body.
So what happens is, the coastdown people do a curve fit on the speed/drag plot and that gives the A, B and C numbers and all the complexities are handled right there. While the C number, being squared, sort of corresponds to aero, it doesn’t necessarily do so exactly. It’s just like saying we don’t care very much what our pressure distributions are because we know accurately how much downforce we have.
That said, we at Edison2 like to be sure of our ground as much as the next person so we asked the coastdown engineers to dive into the sea of numbers and calculate our Cd given our 1.702 m^2 (18.3 ft^2) frontal area. These guys are good at this and proved it when the number they delivered, 0.157, agrees within 2% the number we saw in the wind tunnel.
It’s an interesting factoid that the Very Light Car rolled about 8100ft (over 1½ miles) while coasting down from about 71 to 10 mph (the standard is actually from 115 to 15 km/h). If we were to assume linear speed decay, the force to decelerate would be about 20.4lb. This matches the ABC coastdown numbers at a little over 40mph, right in the middle of the speed range. Overall, we’re happy that we’re dealing with facts.
And if you really want to see how good the VLC is, take your own road car up to a bit over 70 on a level road, knock it into neutral and see how far it goes. Even though it weighs some multiple of the VLC, bet it’s less than 1.5 miles. Please use common sense and do this only where it’s safe and legal.
Edison2’s engineering department goes to a lot of trouble to model performance, which is why we were able to make some very good primary decisions about our cars’ layout and characteristics. But we’ve also learned a hard lesson over the years: there comes a point where you just have to go out and try it. When what you observe in reality doesn’t line up with the model, the reasons why are worth study.