Since we last exchanged information, I obtained "
Practical gas flow by John Dalton" and "
Supercharged! - Design, testing and installation of supercharger systems" by Corky Bell. I am waiting for some backordered books about exhaust tuning.
One thing starting to emerge from this research is that exhaust requirements for supercharged engines are not the same as for regularly aspirated engines. Most of the info out there does not apply to supercharged engines. The book by Corky Bell provides a lot of formulas, which when I compare them to the results of published dyno tests in some of the magazines, including Y-Block Magazine, appear to work fairly well. In general, the advice about supercharged engines is that bigger is better, but when you start crunching numbers, the objective with the header primaries is to make them big enough that rate of gas flow does not exceed about 250 fps.
Ted Eaton, earlier in this thread, reported good results on a drag racing application with 2-inch primaries. In a detailed conversation with Jerry Christenson, he thought 2-inch primaries would be big enough on a supercharged, 331 cid Y-block with ported heads. Gord McMillan has got good results with 1-5/8 primaries on mildly boosted engines like I propose and did not think I should go bigger than 2-inch on the primaries and 2-1/2 on the collector and tailpipe because I might loose low end torque.
When I worked through all the formulae in the book by Corky Bell, the results indicate that for my application, 2-inch primaries are what I should use for the predicted exhaust gas flow under boost at maximum rpm. As Ted Eaton pointed out, these designs are unique for each engine, so what applies to my engine may not apply to another.
Based on the foregoing considerations, my header design should be simple, right? Well.....maybe not. I am using a centrifugal supercharger which does not even begin to make boost until 3200 rpm and peaks at 6200 rpm. Accordingly, the engine may act more like a normally aspirated engine until boost starts to come up. The question is, how much effect does the header design have at the lower rpm's? Some of the stuff I am reading says none at all, just get rid of all the back pressure, other stuff says low and mid-range torque will be effected by header design.
Here is what I see emerging for my application. First, the cam design will have the most influence over torque and the range of rpm where maximum torque occurs. I am not sure to what extent the historic experience offered about header influence on torque may have been influenced by cam design. Secondly, if I do not have any backpressure or it is less than 0.5 psi in my exhaust, I don't see wave forms and exhaust gas inertia having all that much effect, in other words, if the pressure stays so low as to be hard to measure, that indicates that gas inertia and waves in the exhaust gas are not interfering with the exhaust process. However, thirdly, part of the exhaust gas theory is to try to create a negative pressure at the exhaust port just as the valve opens. Accordly, low measurements on a pressure gage don't mean zip if you are trying to measure a negative pressure, so the pulse and scavenging theories may have some value, particularly at no boost or low boost. Fourth, (and I am waiting for some back ordered info on this), the length of the header primaries effects the rpm range where the torque occurs, assuming the pipe diameter is matched to the gas flow.
In summary, it looks like 2-inch primaries are big enough for my supercharged y-block. The cam design will be a lot more important than anything else. I will still design the header primary pipe lengths to maximize low to mid-range torque, so far as the exhaust goes, recognizing that the supercharger will have more effect than anything at the higher rpm where boost increases. The latter principle will also apply to the cam design, it does not have to be too radical for the supercharger to do its job, so long as I have been paying attention to maximizing airflow through the head.
That is about the extent of my progress on this question to date and I appreciate all the input you folks have provided on this. If I learn anything more, I will add it to the thread.
Mike K