Even at Atlanta, it is interesting to note that the throttle was wide open only 58 percent of the time. By way of comparison, 42 percent of the time, the driver was off the throttle. Clearly, Cup cars spend almost as much time off the throttle at Atlanta as they do on the throttle. Since Atlanta is arguably NASCAR's fastest track, you can clearly see from this data why getting on the throttle earlier can be more important than engine power.

In most cases, the shape of the power curve is more important than peak numbers. Over the history of racing, engine speeds have increased. When the engine speed range is reasonably tight, engine speed and gear ratio beat shaft torque, almost every time. You can make more torque at the tire contact patch with gear than is possible with the crankshaft (flywheel torque). To demonstrate how this works, we will compare the power curves of four engines and see how they accelerate a vehicle (Chart A, and Groups B and C).

The comparison shows how four engines accelerate a vehicle (similar to a NNC car) from 165 mph to a distance of 3,300 feet. This would be similar to a straight on a typical 1.5-mile track, comparable to Lowes Motor Speedway.

The first engine is the baseline. The second makes more torque and more power but has the same engine speed range. We will label this second engine "HP & TQ." The third engine makes significantly more torque than the baseline but the same peak power and has the same engine speed range. We will label this engine "TQ." The fourth engine has a small power increase and less peak torque than the baseline, but has greater RPM capability and allows for more gear. We will label this engine "RPM" as shown in Chart A and Graphs B and C.

Actually, these engines are dramatically different. Which engine will accelerate a racecar faster? Well, you may be surprised. A computer simulation program was used to compare how each engine would accelerate a NNC-type car for 3,300 feet, starting at 165 mph.

The data shown in Graph D compares to a car running the baseline engine. The X-axis represents the baseline car and the Y-axis shows by how many feet the other engines would lead or trail the baseline car.

You can see that when all engines were run with a 3.70:1 final-drive gear ratio, the performance was as you would expect. The "HP & TQ" engine, with the +3.3 percent power and the +1.8 percent torque increase, would be fastest. After 12 seconds of acceleration, "HP & TQ" was 13.1 feet ahead of baseline. "TQ" was 2.9 feet ahead of baseline and "RPM" was 4 feet behind baseline. The results are shown in Chart B (next page).

Perhaps the only surprise is that a 2.2 percent increase in peak torque yields only a 2.9-foot lead after 12 seconds of acceleration. Clearly, torque at the expense of power is not always the best solution. The "RPM" engine performed poorly in this test, but the "RPM" engine's extended speed range cannot be favorably used without a change in final-drive ratio Let's install a 4.10:1 and rerun the data. The results may again be surprising.

After 12 seconds of acceleration, the results are dramatically different as shown in Graph E (next page).

The effect of the gear can clearly be seen. Not only does the "RPM" engine with the 4.10 gear jump out to a significant early lead, but it maintains that lead for the entire acceleration run. This is surprising, considering the "RPM" engine is down more than 2.0 percent on peak power and 3.1 percent on peak torque (compared to the "HP & TQ" engine), but the increase in final-drive makes all the difference. Increasing the final drive of the "HP & TQ" engine may help it. Obviously, we can't use as much of an increase as we did with the "RPM" engine because we would hit the rev limit before the end of the run. Now, we'll increase the final-drive ratio to 3.76:1 for the "HP & TQ" engine. The results appear in Graph F (next page).