There's much more to header collectors than just shoving four pipes together. Using a leng
Header Size
Consider this: It is the downward motion of a piston that creates cylinder pressure less than atmospheric. Intake flow velocity then becomes a function of piston displacement, engine speed, and the cross-section area of the inlet path. On the exhaust side, a similar set of conditions exists. In this case, exhaust-flow velocity depends on piston displacement, engine speed, the cross-sectional area of the exhaust path, and cylinder pressure during the exhaust cycle.
Of the similarities between the intake and exhaust process, piston displacement, engine speed, and flow-path cross section are common. Therefore, there must be a functional relationship among rpm, piston displacement, and flow-path section area, and there is (see the section on calculating pipe sizes).
Note how this shorty header minimizes the collector length. Generally, this is done to mak
This suggests the possibility of sizing primary-pipe diameter to produce torque boosts (as contributed by the exhaust system) to an engine's net torque curve. The previously mentioned mean flow velocity (240-260 fps) found in primary pipes around peak torque rpm is a function of pipe diameter. So, selecting diameters that correspond with the rpm at which torque boosts are desired is one method of header selection or sizing.
Matching Headers to Objectives
If we know any two of the three previously mentioned variables (piston displacement, rpm, or primary-pipe diameter), we can apply some simple math to solve for the other. Here's how that works.
1. Peak torque rpm = Primary pipe area x 88,200 / displacement of one cylinder. Given this relationship, we can perform some transposition to solve for the primary-pipe cross-section area.
2. Primary pipe area = peak-torque rpm / 88,200 x displacement of one cylinder. We can also determine the required displacement of one cylinder (multiplied by the number of cylinders for total engine size) by:
Out of all the variables to consider, one of the most important is that the headers fit th
3. Displacement of one cylinder = Primary pipe area x 88,200 / peak-torque rpm.
Equations 1 and 2 provide a method for determining peak-torque rpm (as contributed by the primary pipes) if you have already selected a set of headers and know the engine size. In equation 3, primary-pipe area can be determined if the desired peak-torque rpm and engine size are already known. It will also calculate engine size based on a known set of headers and rpm at which peak torque is desired.
Here's an example of how this approach can work. Suppose you have a 350ci small-block (43.75 cubic inches per cylinder). A primary-pipe torque boost around 4,000 rpm is your target engine speed. The choices for pipe size are 15⁄8 inches, 13⁄4 inches, and 17⁄8 inches. If we assume a tubing wall thickness of 0.040 inch, each of these od dimensions requires subtracting 0.080 inch when computing cross-section areas.
Using the formula, Area = (3.1416) x (id radius) x (id radius), we obtain the following cross sections: 15⁄8 inches = 2.07 square inches; 13⁄4 inches = 2.19 square inches; 17⁄8 inches = 2.53 square inches.
Remember that headers are just one part of the power equation. When trying to improve powe
Plugging each of these values into equation 1, we find the selection of peak torque becomes (in the same order of pipe sizes), 4,173, 4,415 and 5,100 rpm. Based on an intention to provide a torque boost around 4,000 rpm, 15⁄8-inch-diameter primaries appears to work. In accord with our previous comments about primary-pipe length, extending these primaries will increase torque below 4,000 rpm at the expense of torque above this point, which is an additional tool to manipulate a torque curve about its peak (see "Torque Peaks").
While this method will not predict header-pipe area as precisely as some contemporary computer-modeling programs, it can be a valuable quick-and-dirty tool when making decisions about header choice or application of sets already on hand.
Conclusion
There is much more to the science of exhaust-system tuning and headers that space does not allow us to include. It's worth noting once again that the final combination of parts must take into account all the components as a system, rather than looking at the headers as a separate entity. Any engine will make its best overall power when treated as a complete system.