Exhaust science

[Tavi]

I figured I would share the information I have gathered over the years.

The most important part of designing a good exhaust system is one that promotes scavenging, has no turns, is free flowing, short, smooth transitions and is equal in length. Obviously there will be many trade offs to accommodate packaging into a vehicle.

Scavenging is when the air velocity of the exhaust creates a vacuum in it's wake as the pulse enters a lower pressure area (a larger pipe). Helping pull out the exhaust pulse from the next cylinder in the cycle. Ultimately pulling the exhaust from the cylinder. Creating a greater vacuum in the cylinder than the piston going down can create. So the exhaust can truly make or break performance potential.

Headers or stock exhaust manifold. Manifolds don't promote scavenging at all. Just dump the exhaust into a common pipe. Headers, long tube, mid length and shortys each provide scavenging in their collectors and depending on primary tube length will perform better at different RPMs. Shorty will (should) bolt right into the stock location. They tend to produce better low end to mid torque and horsepower. Long tubes require you to redesign part of your exhaust and tend to give better performance between mid and high end. Mid length will obviously provide somewhere between the two but still require modification to your current exhaust. Tuned headers, unless the company designed those headers on your engine with a dyno to your preferred RPM range they are not tuned, just marketing words. Equal length is the best you can do before getting into the fancier try-ys type of headers.

X pipe, H pipe and Y pipe true dual. On a V-8 engine just due to design, the firing order puts one cylinder out of order of even fire per bank. One cylinder will directly after another on the same bank. This causes excess pressure in the collector and diminishes scavenging.
-84-6--2 passengers side bank
1--3-57- driver side bank
H pipe will remedy the excess pressure but has almost zero effect on scavenging. Both X pipe and Y pipe create a scavenging effect pulling the pulses from the opposite bank. The X pipe being a 2 into 2 and the Y being a 2 into 1. The closer to the collector, the better the effect will be. But this require the exhaust pipes be close together. True dual will have each bank divorced from the other. This hinders performance, but does give a nice exhaust note due to the odd fire.

Mufflers and catalytic converter. Free flowing as in having no effect on flow. Just due to the nature and design of cats it is nearly impossible to no effect on flow. Mufflers are much easier. If you can look down it and see the other side you can guess it's free flowing. Any turbo or chambered muffler will cause restrictions.

Pipe size, too small and you can choke the engine. Too big will rob you of the scavenging effect. A good guideline I've found is:
150 hp a 2 inch pipe in single or dual.
250-350 hp a 2.5 inch pipe in single or dual.
450 hp a 3 inch single or 2.5 inch dual.
550 hp a 4 inch single or 3 inch dual.
Smaller pipes will promote lower RPM scavenging. The larger in diameter the pipe the deeper the tone. This can be used in part without loosing scavenging if you increase your pipe size after a length.

No two engines, performance and sound desires are the same. The exhaust you want will be dependent on several factors, most of which come down to what you want to do.

I hope this helps at least one person. I am not the is all end all guru. If you have any information contradicting this information or have more details you want to share please and feel free to add it.

 

[Hoplite]

Good write up. Thanks for posting

 

[johnckhall]

I saw an episode of "Engine Masters" (S7 Ep10) that addressed the HP of stock painted vs wrapped vs ceramic coated headers. The old hypothesis is, headers that retained more heat within the interior of the pipe (and absorbed less heat into the wall of the pipe) would get rid of exhaust heat faster and thus increase HP. Hypothetically, this sounds good. However, the results were wrapped vs painted stock headers were essentially the same HP output. Wrapped headers obviously reduced the surface temp of the header (by about 25 degrees) and would be beneficial in racing scenarios where you needed to reduce heat in the engine compartment or near the driver. Ceramic coated headers did have only a slight increase in HP, but nothing very notable at all. They noted the difference was nothing like the claims in HP increase that some builders make (maybe even within some companies). The cost of ceramic coated headers certainly isn't justified by the HP output difference. They do look better with some builds and rusting is reduced, but there's only a minimal increase in performance. Their conclusion was they'd probably still buy ceramic coated headers for the slight benefit and the look/rust prevention. It was a very interesting episode and a very good scientific breakdown of the comparisons.

 

[Ugly Duck]

I have comments:

Scavenging is when the air velocity of the exhaust creates a vacuum in it's wake as the pulse enters a lower pressure area (a larger pipe). Helping pull out the exhaust pulse from the next cylinder in the cycle. Ultimately pulling the exhaust from the cylinder. Creating a greater vacuum in the cylinder than the piston going down can create. So the exhaust can truly make or break performance potential.

This is partly true, but a large part of scavenging is the reflected wave that comes from the high pressure pulse of exhaust hitting the open atmosphere, sending a low pressure wave back up the pipe. The length of tube dictates when this low pressure wave arrives at the exhaust port. If it arrives at the point the exhaust valve is closing, it will suck the last remnants of exhaust out of the cylinder, and if working correctly with overlap it will pull intake charge into the cylinder.

Headers, long tube, mid length and shortys each provide scavenging in their collectors and depending on primary tube length will perform better at different RPMs. Shorty will (should) bolt right into the stock location. They tend to produce better low end to mid torque and horsepower. Long tubes require you to redesign part of your exhaust and tend to give better performance between mid and high end.

Pretty much the opposite is true: The longer the tube, the longer it takes for the reflected wave to arrive at the exhaust port at EVC, so the slower the engine must be turning to achieve the tuning benefit. Long tubes, then, provide a power boost at lower RPM, and shorter tubes provide a power boost at higher RPM.

X pipe, H pipe and Y pipe true dual. On a V-8 engine just due to design, the firing order puts one cylinder out of order of even fire per bank. One cylinder will directly after another on the same bank. This causes excess pressure in the collector and diminishes scavenging.
-84-6--2 passengers side bank
1--3-57- driver side bank

True and false. True that the cylinders fire oddly, not entirely true that it creates an excess of pressure at the collector. What it definitely does is confuse the tuning effect, as the pulse from one cylinder will always affect the others, and in an even-firing header (think a 4 cylinder) this is easily accounted for. But on a SBC the pressure wave from #8 will affect #4 #6 differently (different RPM) than it will affect #2. Under some circumstances the pressure pulse from #8 might not end in time to let #4 create a distinct pulse, so the reflected wave from that cylinder is lost. Etc. However, when the headers are connected to an exhaust, any restriction in the exhaust will create some backpressure and the sequential firing of 8 & 4 will, at some point, create more exhaust volume than the pipe/muffler can pass, therefore more pressure in the collector (as you stated).

H pipe will remedy the excess pressure but has almost zero effect on scavenging.

Not entirely. An H pipe simulates a reduction in pressure (or an opening in volume), so the moment the pressure pulse arrives at the H pipe, a reflected low-pressure wave is returned to the cylinders (even in a system with factory manifolds). It's an old hot rodder's trick to position the H pipe fore or aft in the system to adjust where the torque boost would be. The H pipe is less effective at spreading the pressure across both streams of exhaust than an X pipe is, but it's often more effective at scavenging. At the very least, the X and H pipes scavenge differently, just as they share the pressure pulses across the two exhaust paths differently.

I hope you take my comments well, as I'm not trying to be an ass.

 

[Erik the Awful]

Exhaust velocity is important, but preventing reversion is critical. While it's true that a larger exhaust diameter causes a loss of velocity, air doesn't cram itself into a smaller pipe very well. If your exhaust has to cram itself into a smaller pipe, it creates a pressure wave back up the pipe that slows the exhaust coming out of the exhaust port. Stepping up your exhaust size as it moves out prevents reversion. Megaphone and tri-Y headers were popular for a reason. 4-into-1s really aren't suited to most street engines, but they sell well.

This does beg the question, why do mufflers have the same size outlet as inlet? I'd like to try buying a 3" muffler and welding a 2.5" inlet into it, but I already have too many projects.

 

[0xDEADBEEF]

Pretty much the opposite is true: The longer the tube, the longer it takes for the reflected wave to arrive at the exhaust port at EVC, so the slower the engine must be turning to achieve the tuning benefit. Long tubes, then, provide a power boost at lower RPM, and shorter tubes provide a power boost at higher RPM.

Every before and after dyno I have seen of long tube headers vs shorty shows power everywhere. It's basically the same line but raised up.

You must be registered for see images attach

 

[0xDEADBEEF]

Exhaust velocity is important, but preventing reversion is critical. While it's true that a larger exhaust diameter causes a loss of velocity, air doesn't cram itself into a smaller pipe very well. If your exhaust has to cram itself into a smaller pipe, it creates a pressure wave back up the pipe that slows the exhaust coming out of the exhaust port. Stepping up your exhaust size as it moves out prevents reversion. Megaphone and tri-Y headers were popular for a reason. 4-into-1s really aren't suited to most street engines, but they sell well.

This does beg the question, why do mufflers have the same size outlet as inlet? I'd like to try buying a 3" muffler and welding a 2.5" inlet into it, but I already have too many projects.

It's a balancing act. If you focus on just maximizing one thing, you are giving up too much of something else.

 

[Erik the Awful]

I agree. Shorties are "compromise" headers. If regulation and fabrication aren't obstacles, long tubes are superior in every way.

 

[Ugly Duck]

Every before and after dyno I have seen of long tube headers vs shorty shows power everywhere. It's basically the same line but raised up.

You must be registered for see images attach

Compared to shorties, yeah they'll boost power everywhere. Shorties are WAY too short to make meaningful tuning devices - look at headers meticulously designed for 13-14k RPM engines - they're still very long compared to a shorty.

The LTs probably flow better than shorties & downpipes, so some of this power increase comes from flow, but you can see the bigger boost right around 38-3900 RPM. That's still a pretty "low" RPM, which is what you'd expect to see from the reflected wave. Longer tubes would shift that to the left, shorter to the right. If those LT headers aren't equal length, you'll be getting mini-boosts that will overlap themselves all along the rev range, further boosting power everywhere over a "no tune" shorty (but without a REALLY large peak when they all come into tune).