Before we dig into this subject matter, we’d like to preface what will follow with some background comments. We’ve had hands-on experience designing intake manifolds and exhaust headers using the techniques we’ll share with you. And, as a matter of fact, there have been U.S. Patents issued on some of what will be discussed. So these will not be theoretical approaches. Rather, they have stood the test of time and proven. We just hope you’ll find them as helpful as has been our experience.
To begin, we know that both intake and exhaust systems operate with un-steady, pulsing flow. On the intake side, by comparison, this is probably more important to consider since (particularly when using carburetors) we’re dealing with the problem of keeping atomized fuel in suspension with air. Especially because of differences in mass between air and fuel and particularly during flow directional changes, the two can become separated. That’s worst case. Best case, atomized fuel particle size can change during changes in direction but still be combustible. We’ll get into the subject of burn rate and fuel droplet size at another time, but suffice to say air flow quality on the intake side is important because it is the conveyor of fuel into the combustion space. The unsteady flow conditions on the exhaust side are less critical.
We also know that an engine’s volumetric efficiency capabilities are directly linked to how much torque it can produce. Although there are other influencing factors, both the intake and exhaust systems on an engine have a material effect on v.e. As such, both systems can play a significant role in determining where in an engine’s torque range a peak will occur. In fact, peak v.e. and peak torque will or should occur at the same or approximately the same rpm. So, since v.e. is particularly influenced by flow rate, we come to the point of discussing mean flow velocities in the intake and exhaust systems.
Sponsored Video: Varsity Math: Rational Water
There is substantial data supporting the notion that at peak torque (v.e.), the mean flow velocity is around 240 feet/second, wherever in the rpm range this occurs. Some will argue it’s slightly higher or lower than this, but for the purposes we’ll describe, that flow rate is sufficient.
Now, all else being equal, flow passage cross-section area determines flow rate. Stated another way, and by running an engine at a constant rpm, if we were to increase or decrease cross-section area at that rpm, the flow rate at that rpm would vary accordingly. So, since we are targeting 240 feet/second as our mean flow velocity (at peak torque, don’t forget) we can manipulate the intake manifold’s contribution to the overall torque curved by deciding on a specific cross-section area for the manifold’s runners.
This brings us to a potential point of contention. Some will say that a constant cross-section is the best approach. Others hold that intake manifold runners should exhibit a measure of taper, and there’s data in support of both. However, once again, for our purposes here, it will likely have little or no influence either way. If a runner has taper, you can compute an “average” cross-section area by adding the entry area to the exit area and dividing the results by two.
OK, let’s re-visit the paragraph just before the last one. Given the fact that runner cross-section area is a major factor in how the intake manifold influences torque, why not find a mathematical equation that allows us to quantify this influence, based on piston displacement and rpm? Well, look no more. Here it is:
torque peak (rpm) = (88,200 x cross-section area) / displacement of one cylinder
where cross-section area has units of square-inches and piston displacement has units of cubic-inches.
With just a dab of algebraic manipulation, this little equation can provide you with multiple bits of information. For example, you may already have an intake manifold and want to find out at what rpm it makes a contribution to peak torque. If this is so, just plug the numbers into the equation as shown above. Let’s say you’re switching to an engine of larger or smaller piston displacement and will be using the same manifold. If that’s the case, the equation as shown still works.
But let’s say you want a certain rpm at which the intake is making a significant contribution to torque. If that’s the case, you can algebraically re-arrange the equation to the following:
cross-section area = (displacement of one cylinder x torque peak) / 88200
If you are getting the impression that we can “tune” an intake manifold to make its contribution to the overall torque curve, you’re correct. In fact, the same approach can be taken when designing or evaluating headers. However, on the exhaust side you’re not dealing with conveying suspended fuel along with the inlet air. Plus, with the possible exception of “stepped” headers, you are working with a constant cross-section.
It may be that you’ll find this mathematical approach to building/modifying/comparing intake and exhaust systems more valuable on the exhaust side than intake, simply because of the measure of control you experience. Either way, it can be a revealing time-saver.
However, there’re some additional perspectives we can take on this tuning issue. From experience and for practical purposes, we know it’s possible that the intake and exhaust systems can materially influence where (rpm) torque boosts are provided. In fact, we also know that you can “tune” them to peak at different rpm points. For the most part, sizing the cross-section areas is the controlling feature. Taking this yet another step, it’s possible to broaden an overall torque curve by tuning the intake and exhaust systems (the rpm at which each peaks) farther apart, thus tending to broaden the net torque curve and reducing the peak. Years ago, while working with a prominent circle track engine builder, he used this technique to boost torque in the lower rpm ranges (off the corner torque) with a comparable boost at higher rpm (heading past the flag stand).
Before we button this up, there’s one final thought we’d like to share. If, as it turns out, cross-section area of intake runners and exhaust pipes have the influence on where torque occurs that they do, why not use more than one size runner or pipe on the same engine? But before we try to expand on this, here’s something else to keep in mind. Cross-section area determines the point at which the 240 feet/second mean flow velocity previously discussed occurs. That fixes the peak torque rpm. Lengthening or shortening the passage without a change in cross-section area simply rocks the curve about that rpm point. Lengthen the passage and torque is taken from above the peak and placed beneath it. Shortening the passage removes torque from below the peak and adds it above. Without a change in cross-section area (intake or exhaust), at least insofar as either of these influences peak torque, the peak will remain where these two areas determined its rpm.
Now, about different cross-sections in a given intake manifold or header set. It works, to the extent of obtaining similar results in terms of broadening an overall torque curve. In fact, it’s possible to construct an engine package with an intake manifold with two different runner section areas, a header set of two different section areas and a camshaft with different intake and exhaust valve timing…matching the two-sized intake and two-sized headers. But that’s a whole new subject to consider. That one also worked as well.