There are a few things that I feel are important about headers and clearing some misconceptions.
Pipe Diameter
Pipe diameter is based on the cylinder volume and RPM. Basically it comes down to the volume of exhaust gasses that are being evacuated from the cylinder. The general rule of thumb is to start with a primary INSIDE DIAMETER that is within 1/8" of the exhaust valve diameter. Another is that changes to pipe diameter rock torque output about peak torque. Larger diameter pipes will aid torque production above peak torque RPM. Pipe diameter can be calculated to optimize "tune" for a specific displacement and RPM to get you a starting point.
If you look at the exhaust (header) system as a whole, from combustion chamber to atmosphere, the greatest restriction is the valve seat geometry. At low valve lift, the restriction is the seat geometry and at high lift, the restriction is the valve seat throat (the transition between the machined seat and the cast port). This area forms a venturi shape, enhancing velocity, and is so critical to port flow that I once lost about 30CFM in an SBC Vortec head by increasing the throat diameter, while learning head porting. I say to emphasize how important velocity is to flow and therefore performance.
In thinking about my application, I am pondering the pros and cons of a "tuned" set of headers. I currently have a set of through transom Bassetts. I don't think any thought of "tuning" went into the design of these headers. I think they are intended to look big and impressive and chrome on an open engine compartment boat. If any gains in performance occurred over the manifolds they replaced, bonus. The first thing that jumps out at me is the primary pipe diameter. Comparing the huge mismatch between these headers and the exhaust ports, this becomes an area of gaseous expansion and a corresponding area of decreased velocity. On a side note, I think this contributes to reversion.
In a supercharged application, the gaseous expansion is less of a concern than in a naturally aspirated application. The exhaust performs less of a duty of scavenging, creating a point of low pressure in the cylinder to decrease restriction of atmospheric pressure pushing fresh air fuel mixture into the cylinder. At our level of supercharged applications, the exhaust performs more of a "evacuating exhaust gas from the passenger compartment with minimal restriction" duty rather than aiding performance through "tuning".
I do think there is performance to be had with tuned headers and the most benefit would occur at max RPM. This speaking purely full-throttle performance rather than cruise economy or pulling skiers. I would bias pipe diameter for high RPM. I would also step the pipe diameter to maintain the highest level of velocity.
Lastly, I think that biasing the torque curve to favor max RPM, therefore killing off a little bottom end, will help the pump hook and load better. Supercharged applications tend to shock the pump harder, causing the RPM to tag the rev-limiter instead of launching harder. Although, with a centrifugal supercharger, you could soften the hit with a blow off valve by dumping boost until the pump loads.
Pipe Length
Organ pipe theory describes wave tuning and basically states that each pipe length will resonate at its own frequency. When the engine "comes on the pipe," a low pressure area is created in the cylinder, aiding volumetric efficiency. I think it is important to note here that at all other RPM, the wave tuning may or may not be detrimental to VE, bit it is for sure not aiding VE.
I have spent a lot of discussing header theory with two basic 'school of thought' people. One, and the most common is: all the pipes need to be the exact same length, or at least within 2" of each other. The other is: each pipe needs to flow exactly the same, regardless of length. Pipes with more bends or sharper bends will be more restrictive than straight pipe. Therefore a pipe with more bends will be necessarily shorter than one with less bends. Both schools of thought completely disregard the other. The equal flow guys disregard wave tuning because it occurs at only one RPM.
Both conditions exist in a running engine and both have merit. In a car, an untuned length header can actually help produce a broader, flatter torque curve. In a jet boat, I don't see that as being beneficial. One successful jet boat racer is of the philosophy that the engine should produce the most peak horsepower and should carry it for the last 1,000 RPM of the RPM range before falling off. This aids in maintaining consistent performance throughout the day of changing weather conditions.
A single RPM engine of a jet boat lends itself to tuned length headers. At the same time, fewer gradual bends are less restrictive and cosmetically appealing. Flowbench tuning of headers is unrealistic at our level. One area of great importance is the transition from the ports into the primary pipes. A short piece of straight tubing, about 2" long, before the first bend is important. This tube should not be perpendicular to the header gasket surface, but should closely replicate the angle of the port roof and have a smooth transition. A mismatch is acceptable and preferred at the floor. If reversion will occur, it will be along the floor and short-turn radius of the port.
All headers will "tune" somewhere. I don't really see performance being that different between the off-the-shelf one-size-fits-all production headers when pipe/collector sizes and lengths are so similar. There might be a difference based on port alignment. I think the Borla or MSF would have an edge, especially if the MSF are tuned to your application.
I hope this gives you some insight on features to consider when shopping for headers.
Cheers,
Joe
