Assumptions
We started by gathering a full set of actual test data on a real 300-horsepower Mustang GT and rounding up all the dimensions and suspension parameters we could find on the production Ford and the concept Camaro. There's a remarkable amount of data available on the Internet if you know where to look (none of which is 100-percent trustworthy), but many variables on both cars still had to be guesstimated from scratch. In cases where information was missing for both cars, an approximation was tailored to make the Mustang simulation match the real-world test data and then applied equally to the Camaro as well.
The biggest single unknown variable is the tire performance. CarSim needs some esoteric data, like longitudinal and lateral tire forces versus slip angle, camber thrust, and tire spring rates. That's all well beyond what TireRack or the manufacturers publish, so CarSim guru Phil Mather rounded up figures for similar performance tires, tweaked them on the Mustang GT model, and applied the same properties to all versions of the Camaro and Mustang, varying the tire sizes to fit. Drag coefficient and frontal area are the next most difficult parameters to guess. Ford doesn't share its aero stats, and the Camaro has never been near a wind tunnel, so rather than guess at factors for two cars that might be wildly different, we ran both cars "in a vacuum," with no wind resistance at all. All simulated times to speeds above 60 mph will therefore be artificially low.
Complete engine performance maps for the Mustang engines, the Corvette LS7, and the 3.9-liter GM V-6 were scaled from manufacturer power and torque curves measured at wide-open throttle; GM supplied engine map data for the show car's 540-horse LS2 engine, which was scaled down to match the expected production 6.2-liter's power and torque peaks. A limited-slip differential with a 75 pound-foot preload was assumed for all V-8s; the V-6s run open diffs. Spring and damper rates were provided for the concept Camaro and scrounged for the Mustang GT, and then applied to all versions of both cars, absent any better information. Ditto the brake systems. We know the ZL1 and GT500 will have stiffer suspensions and bigger brakes, but we can't accurately guess by how much, so expect better performance from the real versions of those cars.
To illustrate how a tiny change in any of the parameters can have unforeseen effects on performance, senior development engineer David Hall ran a CarSim animation showing two Porsche 911 Turbos, identical except for slightly different antisquat geometry. The more level car's aerodynamic advantage helped it win a drag race decisively. This also helps illustrate that CarSim is an engineering tool, not an entertainment device. Video games like GT4 may do a reasonable job of mimicking a car's measured behavior, but CarSim accurately predicts it. Still, with all the guesses in play here, we can guarantee "your results will vary."
All that said, our simulated Mustang GT managed to match the vehicle dynamics of the real one relatively closely. The lateral grip is spot on at 0.87 g, and the digital driver trailed Neil Chirico by just 0.2 second around our figure-eight course. Digit-man managed to snake through the slalom 3.8 mph faster, but remember that there was no wind in his face, and some of our chassis assumptions might not hold up as well with so many quick and radical side-to-side weight transfers. The acceleration data looks less impressive. Launching a manual-transmission car is a tricky business in the digital realm, and Neil's hole-shot was almost a half-second quicker; but by 60 mph, the cars are nearly even (5.3 seconds for real, 5.5 virtual), and above that speed, the car with no pesky atmosphere to fight pulls ahead, hitting the quarter mile 0.3 second and 10 mph ahead. For a much fairer fight, let's head out to the all-digital races.
...
>>next page