I'm willing to have a

serious discussion with you, because you know what you are talking about...

It first started with discussions of Buick's engines vs. Toyota's, when some people mentioned low-end torque. Then ifidonlike, in a post to "teach you guys something", quoted an article to support his claim that only peak hp matters in terms of vehicle performance.

I object to this claim, and further, I claim that even the entire hp curve is not really relevant in determining performance. Here, performance is understood to mean all types of acceleration times.

The simplest thing to consider is one car only, driving at 25 mph in 2nd gear on a highway on-ramp, accelerating under wide-open-throttle up to 65 mph. I want to measure the time needed to go from 35->45, 45->55, and 55->65, respectively. Looking at the power curve only, can you tell me which of the three times is the longest, and which is the shortest? The power curve is mostly increasing in the rpm ranges used in this acceleration run and I bet you cannot answer my question. On the other hand, a simple glance at the torque curve gives you the answer. And the answer may vary from car to car, depending on the exact shape of that torque curve.

More generally, the question is what determines total acceleration time for any car, from one speed to another. Here, gear ratios obviously matter, and after I did all the calculations, I find that the answer is given by the torque curve and the gear ratios. Even if you started with the power curve and used energy conservation to derive it, you have to divide the power curve by rpm in the final formula, reducing it to the torque curve.

My point is therefore: If the power curve is less useful than the torque curve in determining even single-vehicle performance data, then the peak-hp figure is even more irrelevant to comparing performance of two vehicles.

Finally, now I want to correct a mistake I made in a previous post. I had conceded that peak-hp could be used to determine maximum track speed. Now I contend that even this hypothesis is false.

If a car is at maximum speed, its power must be equal to drag-force times that speed, or in equation form:

P(V_max) = F_drag(V_max) * V_max.

But F_drag(V_max) = 0.5 * (Drag_constant) * V_max^2,

so putting all these together,

P(V_max) = Some_function_1(V_max).

On the other hand, there is one gear ratio (usually top gear), which converts V into rpm:

P(V_max) = Some_function_2(V_max).

It's the equality

Some_function_1(V_max) = Some_function_2(V_max)

that determines V_max.

Can V_max solved by this equality correspond to the rpm where power is maximized? Most certainly no, because Some_function_2 is determined by the drivetrain (engine and transmission) is built, whereas Some_function_1 is determined by the body (drag coefficient and frontal area) is built. The two are independent from each other and need not fit together to give you V_max at peak power.

Now that your car's maximum speed is not achieved at peak power, what that speed is doesn't depend on peak power either. Sure, it's still somewhere "near the peak", but can you now tell me how near, and whether the power curve's shape there is flat or peaky?

Once more, I'll say that I'll apply Ockam's Razor to cut the concept of power, especially peak power, from the theory altogether.