Are Equal Length Headers All Wrong?
Thread Starter
Racer
Joined: Sep 2016
Posts: 274
Likes: 45
From: South Florida
Are Equal Length Headers All Wrong?
Are Equal Length Headers Wrong? This is what David Vizard is saying. The manufactures are doing us a disfavor, and charging a premium penny (especially the Corvette tax) for poor engineering.
Note: I copied most of this text from someone else asking the same question, reiterated below.
In David Vizard's most recent Big Block Chevy book there is a section about header primary lengths. He says in the book (and fleshes it out further in a 30min YouTube presentation, link below
that headers with equal length primaries are effectively wrong in length, especially for cross-plane crank V8's. The big thing being exhaust pulses entering the collector at unequal intervals.
Firing order being 1-8-4-3-6-5-7-2
-meaning that on the odd bank you've got cylinders 1 and 3 separated 270 degrees, then 180 between 3 and 5, then 90 between 5 and 7, then 180 between 7 and 1.
And the same issue on the even bank, in a different order of course. All this meaning that exhaust scavenging is being held back by unequal pressure waves in the exhaust system.
The "fix" is unequal primary lengths, picking an ideal length for one cylinder, then modifying the other 3 lengths to time the exhaust pulses equal between each cylinder equally.
Looking around on this forum, I see effectively no mention of this concept anywhere. Has anyone explored this concept?
Note: I copied most of this text from someone else asking the same question, reiterated below.
In David Vizard's most recent Big Block Chevy book there is a section about header primary lengths. He says in the book (and fleshes it out further in a 30min YouTube presentation, link below
that headers with equal length primaries are effectively wrong in length, especially for cross-plane crank V8's. The big thing being exhaust pulses entering the collector at unequal intervals.Firing order being 1-8-4-3-6-5-7-2
-meaning that on the odd bank you've got cylinders 1 and 3 separated 270 degrees, then 180 between 3 and 5, then 90 between 5 and 7, then 180 between 7 and 1.
And the same issue on the even bank, in a different order of course. All this meaning that exhaust scavenging is being held back by unequal pressure waves in the exhaust system.
The "fix" is unequal primary lengths, picking an ideal length for one cylinder, then modifying the other 3 lengths to time the exhaust pulses equal between each cylinder equally.
Looking around on this forum, I see effectively no mention of this concept anywhere. Has anyone explored this concept?
Pro
Joined: Aug 2010
Posts: 622
Likes: 169
From: Montgomery Texas
I doubt that any other five people combined have done as many dyno runs as DV. He is 80+ years old and he is still going strong! To my mind, what he is saying makes logical sense but, it's merely a Band Aid for a compromise crankshaft / firing order design. The actual fix would be a flat crankshaft. With a flat crank, equal length primaries make sense, not so much otherwise. Due to the speed and timing of exhaust pulses and their reflections. It would be relatively easy to design a header for one specific RPM, but what works for one RPM may be totally abysmal for a different RPM. So, unless you are building for a boat, or a dedicated drag car that spends the bulk of its time at, or around one specific RPM, every system, regardless of design, will be a compromise ... which includes an engine with a flat crank and equal length primary pipes.
Consider modern composite intake manifolds. They employ multiple length tuned intake runners in order to create torque peaks at multiple different RPMs. IIRC, the C8 has three discrete intake runner length combinations. I expect that the next logical step will be constantly variable intake manifolds, eh? Thus far, I'm not aware of anyone who has designed a practical analogue for the exhaust side of the engine.
Another fly in the ointment is that unless you are racing, or otherwise dealing with a high performance sports car (high RPMs), the aforementioned is largely irrelevant because 4 into 1 systems only come into their own above somewhere around 6k to 7k RPMs. Below that, you would generally want a tri-y system for best performance. DV, himself, acknowledges this. When I worked on Toyotas two lifetimes ago, all of those cars came from the factory with a tri-y exhaust manifold to eek out the most performance possible from those cute little engines. For a competition car, budget and class rules permitting, you would naturally hire a professional exhaust specialist to design and build a system for your specific application.
There are a lot more variables, so if you are interested, I'd suggest that you track down a copy of "The Scientific Design of Intake and Exhaust Systems," by Smith & Morrison. It's a fascinating read, but at the end of the day, it all depends on the specific car and its use case. Once those are identified, the book gives all sorts of ideas to try on the dyno, because there is no such thing as one size fits all. Every possible design includes a collection of compromises. The trick is to find which specific group of compromises will work best for your particular situation.
Consider modern composite intake manifolds. They employ multiple length tuned intake runners in order to create torque peaks at multiple different RPMs. IIRC, the C8 has three discrete intake runner length combinations. I expect that the next logical step will be constantly variable intake manifolds, eh? Thus far, I'm not aware of anyone who has designed a practical analogue for the exhaust side of the engine.
Another fly in the ointment is that unless you are racing, or otherwise dealing with a high performance sports car (high RPMs), the aforementioned is largely irrelevant because 4 into 1 systems only come into their own above somewhere around 6k to 7k RPMs. Below that, you would generally want a tri-y system for best performance. DV, himself, acknowledges this. When I worked on Toyotas two lifetimes ago, all of those cars came from the factory with a tri-y exhaust manifold to eek out the most performance possible from those cute little engines. For a competition car, budget and class rules permitting, you would naturally hire a professional exhaust specialist to design and build a system for your specific application.
There are a lot more variables, so if you are interested, I'd suggest that you track down a copy of "The Scientific Design of Intake and Exhaust Systems," by Smith & Morrison. It's a fascinating read, but at the end of the day, it all depends on the specific car and its use case. Once those are identified, the book gives all sorts of ideas to try on the dyno, because there is no such thing as one size fits all. Every possible design includes a collection of compromises. The trick is to find which specific group of compromises will work best for your particular situation.
Melting Slicks
Joined: Jul 2015
Posts: 2,264
Likes: 222
From: Round Rock Texas
Okay think of the biggest HP monster out there top fuel dragster. Review picture of headers They all look the same to me. So if they are not using it to squeeze out more HP it does nothing
Drifting
Joined: Jun 2020
Posts: 1,677
Likes: 721
From: Stay dangerous my friends
I think Dave is right, but only if you also look at "for what"... There's a top Winston Cup engine guy who said the same.
But if you aim for top peak dyno numbers, or a very narrow power band, optimized for that may be equal length.
Same goes for matching parts, if you tune the intake lengths and cam lobe timing and exhaust primary all for maximizing the exact same RPM, you may end up with something very peaky that is a dog until it hits the powerband and then blows off the tires.
But if you aim for top peak dyno numbers, or a very narrow power band, optimized for that may be equal length.
Same goes for matching parts, if you tune the intake lengths and cam lobe timing and exhaust primary all for maximizing the exact same RPM, you may end up with something very peaky that is a dog until it hits the powerband and then blows off the tires.
