Over time, in this column, we have touched upon several aspects of not only how to achieve higher levels of volumetric efficiency (as it relates to torque) but the importance of this feature in a racing engine, particularly for circle track applications. In the course of those discussions, we've attempted to establish an understanding of how induction and exhaust system design and dimensions play into this subject. But before we try to develop topics specific to yet another perspective of how torque curves can be linked with these two major systems, a brief review of a few previously-mentioned points is intended to provide some background information.

For example, we know that in both an intake and exhaust system, we're dealing with pulsating, unsteady flow. We also know that as a particular function of piston displacement and rpm, each system can pass through what we'll call a "resonant" point often associated with a peak torque value. At this particular rpm, both systems are generally the most efficient and can be associated with a certain "mean flow velocity" found at such peaks, regardless of engine combination. We can also label this flow rate as the "critical" velocity.

As each system approaches this flow velocity, there is not sufficient rpm to achieve the rate and beyond peak torque rpm it will exceed that value, thus the shape of a typical torque curve. We also know, from a design standpoint, that each can be treated as a separate system. That is to say each will effectively produce its own torque curve. In reality, they combine to create a "net" or resulting curve shape just described.

It turns out that a third variable influencing at what rpm an engine of specific piston displacement will reach its critical intake or exhaust flow velocity is the cross-section area of the flow path. For the sake of simplicity, let's say these paths are of constant cross-section, neither tapered nor "stepped" as in practice. Actually, for purposes of our discussion, this aspect of the subject is irrelevant. What's important is understanding that it's possible to "tune" intake and exhaust systems separately. Let's restate that suggestion.

On a given engine, it's possible to select passage section areas to achieve peak torque (volumetric efficiency) points at specific rpm where such boosts are desired. Now, let's consider the possible value in having this measure of influence over an engine's torque characteristics.

Suppose we decided to have an intake and exhaust system broaden (or make flatter) a net torque curve. One approach would be to adjust intake and exhaust passages sufficiently different to spread their respective torque peaks farther apart in the rpm range. The net effect of this would be to create somewhat of a "depression" in the curve's shape between the two peaks and, actually, remove some of the net peak torque.

However, the effect could benefit off-the-corner acceleration during the lower engine speeds while having some torque in reserve (higher rpm peak or boost) getting past the flag stand. We're not dealing with "maybe so" issues here, there have been numerous instances where, when reduced to practice, the concept works. There are patents verifying the results.

Now let's take this approach one step further. Building on the same concept that flow passage size, piston displacement, and rpm are closely tied with where in an engine's speed range torque boosts occur. Here's another thought; suppose we configured a set of headers with two sizes of primary pipes for a V-8 engine, choosing to order the pipes in a way that paired every other cylinder with the same size pipe. As an example, if we assume a firing order of 1-8-4-3-6-5-7-2, every other cylinder would be served by the same size primary pipe. Functionally, we will have created a header system that treats the engine like two V-4s, each of which will contribute to the overall torque curve at separate and different rpm, but do so with a flatter "exhaust system torque curve," if you will.