How to Test Drive Something That Doesn't Exist

Motor Trend magazine May Issue Features Camaro vs Mustang

ANN ARBOR, Mich., Apr. 4, 2006 — Mechanical Simulation Corporation's CarSim 6 was used to evaulate the current Mustang and the Camaro Concept car in Motor Trend magazine's May 2006 issue.

In February 2006, Motor Trend Magazine contacted Mechanical Simulation with a new and unique proposal: simulate a concept car and compare it against its likely archrival. Thus was born the Chevrolet Camaro Concept vs. the 2006 Ford Mustang. This would prove to be challenging because the data needed for simulating both cars is hard to come by.

Step 1: Obtaining the data.

General Motors has published some data about the Camaro Concept car, including powertrain information, dimensions, and tire sizes. With the help of GM, enough information was gathered about the Camaro Concept car suspension, tires, brakes, powertrain, and performance targets to build a representative simulation in CarSim. This data is in no way fixed and will definitely change should the car go to production. Therefore, these simulations are used for entertainment and educational purposes only and are just one interpretation of the performance level of the Concept Camaro.

Much is already known about the 2006 Ford Mustang GT and there is information aplenty on the internet. While not to be completely trusted, most of the information appeared to be valid and was used in the CarSim simulations. For example, CarSim uses torque curves (Torque vs. RPM at different throttle positions) in order to simulate engine power output. Dyno graphs were found on the internet for the Mustang and put into CarSim, along with transmission gear ratios, and final drive axle ratio. The information found was typically from a chassis dyno; therefore the data had to be scaled to match the published maximum power and torque ratings from Ford to account for drivetrain efficiency losses. Other data, such as spring rates, suspension motion, and damper force vs. velocity curves were obtained from various public sources and applied to the model.

Step 2: Simulating Motor Trend's test courses

Specifying 3D road information is very easy in CarSim. In these particular simulations, all of the roads were assumed to be flat, with no elevation or camber changes. Friction was assumed to be constant as well. However, any road can be input into CarSim with as much detail as needed. CarSim uses tables to describe X-Y coordinates, gross elevation change (hills), and relative elevation changes (camber, curbs, sharp bumps). Friction can be a full map of the road surface, which makes split-Mu braking and patchy ice conditions easy to describe. Motor Trend's Figure 8 course dimensions were translated into X-Y coordinates in CarSim

Step 3: Adjusting the data to match results

Motor Trend Magazine had already published test data on the Ford Mustang GT, so we knew actual performance results (acceleration, braking, lateral accelerations, etc.). And, since it is possible to recreate their test procedures with CarSim, we were able to use the published data as a target for the simulations. Some of the items that were manipulated included the braking torque, tire force plots, and road surface friction.

These adjusted data values were then applied to all of the simulated vehicles in order to eliminate differences between unknown variables. For example, while the tires were different sizes on the Camaro and Mustang, the tire characteristics (Longitudinal force vs. slip ratio, lateral force vs. slip angle, aligning moments, etc.) were assumed to be the same. (Actual tires are carefully selected and "tuned" to a wide range of performance variables for a specific vehicle. It is unlikely that exactly the same tires would be supplied for a Camaro and Mustang. However, by using the same tire data we remove one source of differences in the simulation results.)

Figure 8 Test: Closed loop steering (vehicle centerline offset) and speed (speed vs. position) controllers were manipulated until the Mustang GT closely matched the data from Motor Trend's testing. These controls were then applied to the other Mustang variations and to the Camaro Concept. They were then further manipulated until each of the simulated vehicles could not go any faster.

Acceleration: The acceleration tests were easier to simulate. A clutch-drop start was simulated and launch RPM was adjusted until the acceleration times of the Mustang GT and the simulated Mustang matched from 0-60 mph. Quarter mile times are artificially low and speeds are artificially high because aerodynamic effects were not used in this simulation since that data was not available.

Braking: The cars entered the braking zone at exactly the same speed before the brakes were applied. Both cars are equipped with anti-lock braking systems. CarSim's simplified internal ABS system was instituted and the same characteristics were used for the Camaro and Mustang. Braking output torque applied at the wheel was adjusted until the simulated Mustang GT stopped in approximately the same distance as the real car.

This is another example of the benefits of simulation: driver-related variations can be eliminated when making subtle comparisons.

Slalom: A constant target speed was used in the slalom in conjunction with closed loop steering controls used to determine the offsets necessary at each cone for the slalom. The simulated Mustang was slightly faster than the real life Mustang, but some of this can be accounted for by the differences in the techniques of Motor Trend's professional drivers (we did not have enough time to modify the driver characteristics in CarSim).

Step 4: Building the Cars

Seven different cars were "tested" for this Motor Trend Article: The Mustang GT (baseline car), three concept Camaros (with V6, LS2, and LS7 engines) a V6 Mustang, and two theoretical Mustangs (Boss 330 with 389 hp and Shelby GT500 with 475 hp). Since we had limited information about these cars, several assumptions were made. All of the V8 cars had the same limited slip differential characteristics, and the V6 cars had open differentials. Transmission ratios were taken from other production vehicles (parts bin transmissions) or from published information. Suspension kinematics, spring rates, damper curves, and anti-roll bar rates were entered into CarSim. All Camaros had the same GM suspensions and brakes, and all Mustangs had the same Ford suspensions and brakes. Tire sizes were adjusted, but tire characteristics were the same (as previously explained).

Step 5: Running the Tests

Thanks to CarSim's efficiency, each test run took less than 30 seconds to complete. (Typical simulations run 5 to 10 times faster than real-time.) Once the baseline runs were made with the Mustang GT, CarSim enabled us to make changes very quickly and come up with our "best guess" at the performance of each of the simulated cars. Animations and plots are immediately available for reviewing the runs and determining if the simulation gave proper results.

To see how close our simulations compare to the real Mustang GT, and to see the possible performance of the Camaro Concept vs. the 2006 Mustangs, pick up a copy of Motor Trend's May 2006 issue!

Phil Mather
Technical Support Specialist
Mechanical Simulation Corporation
912 N. Main Street, Suite 210 Ann Arbor, Michigan 49221