I was reading an interesting article in the May issue of Popular Hotrodding on “Exhaust Science Demystified”. Now this is not my area of expertise so I will outline the article as best as I can and then pose a question or two which those in the know can chime in. Keep in mind this is not a crossover system.
The author (David Visor) built a 400lb-ft, 404hp engine and found that no matter which muffler was put on, he lost 20lb-ft and 25hp. His objective was to build a no loss system that was not too noisy. His work along with an acoustics engineer, resulted in the now Walker/Dynomax muffler. In a shootout of mufflers, their 2.25” muffler beat all the 2.5” models. He gave an educational section on how race engines with large cams (290 in duration) and tuned length exhaust system generate negative pressure waves to scavenge the cylinder during valve overlap. Thus making exhaust flow more important than intake as it (intake) benefits from the exhaust more than it generates on its own. Well-tuned engines can generate partial vacuum as high as 6-7 psi at the exhaust valve and can translate that into 4-5 psi in the intake during overlap. This is pulling the air about 500% more than the piston can by itself. He referred to this engine as a five-cycle unit with 2 consecutive induction events. With this information and proper tuning, one should be able to build a system that is quiet and is able to develop within 1% of its open exhaust power.
Primary tube sizing will often times result in lost torque by choosing a primary tube that is too large. The goal is to size the pipe for optimal output over a range of RPM. The primary tube’s velocity is influenced by the flow capability at peak valve lift. This has been pretty much standardized through much testing and resulted with the following graph.
So if you know your flow, choosing the primary size can be made a little easier rather than trial and error. For most street driven engines, it is a little better to under size the primaries.
In primary lengths, he states that the “old” idea of equal length primaries doesn’t hold and is a waste of time. (Remember, he’s not talking about crossover design primaries) Most want maximum scavenging over a bandwidth of up to 4000 RPM and racers of 3000 or less. With today designs, the tuning of one cylinder to another may vary by 1000 RPM and if all else is equal, the lengths could be different by as much as 9”without affecting performance very much. One advantage to the different lengths is that higher flowing radiuses and convenient routing in a crowded engine bay can be used. The old inline 4 cylinders were primary sensitive and the modern V-8 is really 2 V-4s not 2 inline 4s. Test with primaries varying between 24 and 36” in 3” increments had little variation.
On secondary pipes (collectors), the diameter was just as critical as the primaries and that given a 4 into 1 collector should have a diameter of 1.75 times the primaries. He found that the length was a sensitive factor. If the engine was going to peak in the 8500-RPM range, it benefited from a longer (aprox. 20”) collector, whereas one peaking in the 5000RPM range benefited from a shorter (10”) collector. You can see the benefits in the two pics of added collector extension.
This is all a bit technical and it brings me to the first question. Given that most of us use the 302 and 351, what is the difference in what he has said so far with regards to the 180-degree crossover systems we use on the GT40? Should we add length to the collectors if we increase the HP? If so, how much, or better put, what is the optimum length and how do we go about finding that out without spending a fortune on dyno test?
The next part of the article I found much easier to handle and more relevant to everyone here. It was about mufflers. How to pick one out, and how to build a system that wouldn’t cost HP and remain reasonably quiet? Most people pick out a muffler based on the size of the collector pipe, and according to him that’s all wrong. If you want to have the same HP of an open exhaust system (or close to it), you have to understand what the system” sees” with a muffler on it. A mufflers inlet pipe should only be for fitment. The engine doesn’t care what size the muffler pipe size is, it only cares what it flows, and that is what selection should be based on. His example shows why:
With a length of pipe equal to the length of the muffler and rated the same test pressure as the carb, the flow equals 115cfm per square inch. With this in mind a 2.5” pipe will flow about 560 cfm. If you chose a 2.5” muffler that flows say 400 cfm that correlates to a pipe of 2.1” in diameter. All mufflers (just about) restrict flow between the entry and exit of the system. To get a system that equals the diameter of the collector would require a larger muffler. His example stated a 4” muffler. The problem here is that bigger may be better, but at what cost (money and sound).
They did a lot of testing and came up with the following graph.
What they found out was that with a cam of 290 degrees of seat duration (advertised), the peak HP numbers grew less than 1% above 2.2 cfm per HP. So,,, If you can make a reasonable guess as to open exhaust power potential, multiply that by 2.2 for the total flow needed to prevent back pressure. If you have 500 HP then flow requirements are 500 x 2.2= 1100 cfm. Two 550 cfm mufflers should do the trick and HP loss will be less than 5HP.
Now the other shoe will drop. Remember pressure wave tuning? Well adding a muffler is like adding a length of pipe to the collector, and there goes your just found HP. The trick is to add a muffler that doesn’t alter the tuned length of the collector. If the muffler has a sizable increase in cross-sectional area (open chamber design) like Flowwmaster for example, the pressure wave sees no change in length and the reflection wave occurs. The flow through glass pack types acts as a pipe extension to the collector and will drop the power.
In order to get the same result with a glass pack as the flow through type, a pressure wave termination box (resonator box) has to be added to the system. If it is of sufficient size, it makes everything downstream look invisible to the systems primary and secondary-tuned lengths. If the muffler has the 2.2 cfm requirement, it too is invisible to the system from a flow standpoint and should give the same power as an open exhaust. His suggestion was a resonator box with a minimum volume of 8 x CID of 1 cylinder.
So I guess the last question is whether or not the collector length makes a difference in our exhaust systems. If it does and we want all the HP of the system, then the engine bay is going to get awfully crowded for some of us.
What do you think guys???
Bill
The author (David Visor) built a 400lb-ft, 404hp engine and found that no matter which muffler was put on, he lost 20lb-ft and 25hp. His objective was to build a no loss system that was not too noisy. His work along with an acoustics engineer, resulted in the now Walker/Dynomax muffler. In a shootout of mufflers, their 2.25” muffler beat all the 2.5” models. He gave an educational section on how race engines with large cams (290 in duration) and tuned length exhaust system generate negative pressure waves to scavenge the cylinder during valve overlap. Thus making exhaust flow more important than intake as it (intake) benefits from the exhaust more than it generates on its own. Well-tuned engines can generate partial vacuum as high as 6-7 psi at the exhaust valve and can translate that into 4-5 psi in the intake during overlap. This is pulling the air about 500% more than the piston can by itself. He referred to this engine as a five-cycle unit with 2 consecutive induction events. With this information and proper tuning, one should be able to build a system that is quiet and is able to develop within 1% of its open exhaust power.

Primary tube sizing will often times result in lost torque by choosing a primary tube that is too large. The goal is to size the pipe for optimal output over a range of RPM. The primary tube’s velocity is influenced by the flow capability at peak valve lift. This has been pretty much standardized through much testing and resulted with the following graph.

So if you know your flow, choosing the primary size can be made a little easier rather than trial and error. For most street driven engines, it is a little better to under size the primaries.
In primary lengths, he states that the “old” idea of equal length primaries doesn’t hold and is a waste of time. (Remember, he’s not talking about crossover design primaries) Most want maximum scavenging over a bandwidth of up to 4000 RPM and racers of 3000 or less. With today designs, the tuning of one cylinder to another may vary by 1000 RPM and if all else is equal, the lengths could be different by as much as 9”without affecting performance very much. One advantage to the different lengths is that higher flowing radiuses and convenient routing in a crowded engine bay can be used. The old inline 4 cylinders were primary sensitive and the modern V-8 is really 2 V-4s not 2 inline 4s. Test with primaries varying between 24 and 36” in 3” increments had little variation.
On secondary pipes (collectors), the diameter was just as critical as the primaries and that given a 4 into 1 collector should have a diameter of 1.75 times the primaries. He found that the length was a sensitive factor. If the engine was going to peak in the 8500-RPM range, it benefited from a longer (aprox. 20”) collector, whereas one peaking in the 5000RPM range benefited from a shorter (10”) collector. You can see the benefits in the two pics of added collector extension.


This is all a bit technical and it brings me to the first question. Given that most of us use the 302 and 351, what is the difference in what he has said so far with regards to the 180-degree crossover systems we use on the GT40? Should we add length to the collectors if we increase the HP? If so, how much, or better put, what is the optimum length and how do we go about finding that out without spending a fortune on dyno test?
The next part of the article I found much easier to handle and more relevant to everyone here. It was about mufflers. How to pick one out, and how to build a system that wouldn’t cost HP and remain reasonably quiet? Most people pick out a muffler based on the size of the collector pipe, and according to him that’s all wrong. If you want to have the same HP of an open exhaust system (or close to it), you have to understand what the system” sees” with a muffler on it. A mufflers inlet pipe should only be for fitment. The engine doesn’t care what size the muffler pipe size is, it only cares what it flows, and that is what selection should be based on. His example shows why:
With a length of pipe equal to the length of the muffler and rated the same test pressure as the carb, the flow equals 115cfm per square inch. With this in mind a 2.5” pipe will flow about 560 cfm. If you chose a 2.5” muffler that flows say 400 cfm that correlates to a pipe of 2.1” in diameter. All mufflers (just about) restrict flow between the entry and exit of the system. To get a system that equals the diameter of the collector would require a larger muffler. His example stated a 4” muffler. The problem here is that bigger may be better, but at what cost (money and sound).
They did a lot of testing and came up with the following graph.

What they found out was that with a cam of 290 degrees of seat duration (advertised), the peak HP numbers grew less than 1% above 2.2 cfm per HP. So,,, If you can make a reasonable guess as to open exhaust power potential, multiply that by 2.2 for the total flow needed to prevent back pressure. If you have 500 HP then flow requirements are 500 x 2.2= 1100 cfm. Two 550 cfm mufflers should do the trick and HP loss will be less than 5HP.
Now the other shoe will drop. Remember pressure wave tuning? Well adding a muffler is like adding a length of pipe to the collector, and there goes your just found HP. The trick is to add a muffler that doesn’t alter the tuned length of the collector. If the muffler has a sizable increase in cross-sectional area (open chamber design) like Flowwmaster for example, the pressure wave sees no change in length and the reflection wave occurs. The flow through glass pack types acts as a pipe extension to the collector and will drop the power.

In order to get the same result with a glass pack as the flow through type, a pressure wave termination box (resonator box) has to be added to the system. If it is of sufficient size, it makes everything downstream look invisible to the systems primary and secondary-tuned lengths. If the muffler has the 2.2 cfm requirement, it too is invisible to the system from a flow standpoint and should give the same power as an open exhaust. His suggestion was a resonator box with a minimum volume of 8 x CID of 1 cylinder.

So I guess the last question is whether or not the collector length makes a difference in our exhaust systems. If it does and we want all the HP of the system, then the engine bay is going to get awfully crowded for some of us.
What do you think guys???
Bill