EPA Ignores Aerodynamics?

[gascort]

Originally Posted by Frank Lee

I always thought the last gen EPA ratings were pretty accurate. The new ones "legitimize" the asinine driving of today.

x3 - their "highway" and "high speed" tests aren't what I call highway. If I drove 50 "highway" miles, I would assume 49 miles at 55+ mph and 1 mile of side roads, onramps, and offramps.
From the epa (fueleconomy.gov) website:
Highway test:

High Speed (modern test): (this reminds me of drag racing in my mustang, lol)

[Randy]

The aerodynamics are from a roll-down test done by the manufacturer. They do a curve-fit, ending up with 3 parameters (A,B,C) that they send to the EPA. That plus actual weight is all the dyno needs. I think the 'set' parameters are to compensate for the roll-down on the dyno itself.

You can see these on the detailed test data at this EPA page. Obviously this doesn't take into account hills, but I think that's why the tests speed up and slow down so much... it's like going over hills.

The only question is: what exactly are A, B, and C? They're never really defined by the EPA. Edit: see below for an improved guess at the formula.

[azraelswrd]

Cool! Found my car's data...and now I don't know what to do with all these numbers staring back at me.

[Randy]

I think I finally figured it out from an example in this paper. The target parameters are in pounds, pounds per mph, and pounds per mph squared. So F=A+Bv+Cv^2, where F is in pounds-force and v is in mph.

So power at a particular speed would be P = (Av + Bv^2 + Cv^3)/375, where P is horsepower and v is mph. This seems reasonable... about 40 hp at 60mph for the worst SUV's, about 15 for the best cars.