According to Comp's VP of New Product Development, Brian Reese, "Our objective with the design was to shift the torque curve 'up' over stock, across the entire torque band. This is no small task, as typically it is easy to move a torque peak within the rpm band, but such tuning typically comes at the expense of a lower torque average at another point in the rpm band.

"For instance, lengthening the runners will shift peak torque to a lower rpm, but at the expense of erasing higher rpm torque. We found this an unacceptable solution for 95 percent of the market. To make our manifold commercially viable and applicable to 95 percent of the market, we set a design requirement to increase torque 'all over.' To accomplish this goal, the tuning is absolutely critical.

"Our CFD capabilities today are a tremendous tool, as they correlate to a flow bench within 1 percent, which negates the need for extensive prototyping and physical flow bench work. So if the CFD doesn't look good, it isn't usually worth pursuing any further."

It's particularly noteworthy that Comp had sufficient belief and confidence in the concept of optionally-sized intake passages that fit into the same basic manifold to have incorporated this feature as an available feature. From experience, I will share that enhancing the performance of individual cylinders (or group of cylinders in a given firing order) through specific intake and exhaust dimensions and then linking these with the appropriate valve events can have a material influence on where torque is boosted in the rpm range.

During development of the Victor 4+4, a street performance engine was modified using the following approach. Given a firing order of 1-8-4-3-7-6-5-2, cylinders 1, 4, 7, and 5 were sized for a torque boost at an rpm lower than for cylinders 8, 3, 6, and 2. Correspondingly, the primary header pipes for these two sets of cylinders were dimensioned in a similar way, relative to the overall rpm range. This was capped off by installing a camshaft with intake and exhaust lobes (and displacement angles) associated with a "low" and "high" rpm torque output.

Since dimensional variations were deliberately applied to alternating cylinders, the engine not only ran smoothly but produced a very broad torque range. We also learned that cylinder-to-cylinder air/fuel charge distribution corrections could be made by changing runner and/or plenum configurations.

With regard to Comp's manifold, according to Reese, "We experimented with different runner configurations in an effort to equalize cylinder-to-cylinder air/fuel ratios. Air doesn't like to 'bend' and even less so once it gets some velocity and momentum. So every cylinder fills differently, particularly when the throttle is central, not individual. This problem is not exclusive to EFI either. Carbureted engines are actually worse off, as they lack the individual cylinder fuel control option. There is power to be had by balancing individual cylinders and it is best to do it by way of airflow, not by way of fuel trimming, whenever possible."

Fortunately, contemporary computational means and methods far beyond what was available 25 years ago have moved induction system design and development considerably past prior techniques. It's comforting to know that such technology evolution has accelerated the process of scrutinizing a variety of internal combustion engine components and functions in ways that clearly benefit the racing community. You can be assured such tools often find homes in the more progressive specialty automotive parts manufacturers and race engine shops very quickly, sometimes before just about any place else.