Exhaust gas flow theory?

Dr.Jeff

True Classic
This is specifically regarding the exhaust manifold (header) to cylinder head interface.

I've seen two theories about the flow of exhaust gas from the head to the exhaust system. One is basically to have the smoothest possible transition between the two components with the ports perfectly matched. The other is to have an increase in diameter in the manifold's ports, or "step" transition, to create a reversionary pulse effect. Some tubular exhaust headers have more than one step increase as the gases transition along the tubes. But with a cast manifold it would be just at the junction to the head.

Frankly I do not know if there is any clear advantage to either design. I've seen a few tests that claim to show each is better than the other. So that really does not help. But I'm also not too current on this subject. Anyone have more info about this? And for the 'stepped' concept are there any basic criteria for how it should be designed?
 
The step should be at the lower half of the port, where the air speed is lower. Step should be approx 1 to 1.5mm. The upper half of the port should be a smooth match with no step. The lower step will prevent flow reversion in the slower moving air along the port floor. This isn't theory it's proven in practice.

The stepped header design is for a slightly different reason, this is much more theoretical, IMO is more for ease of manufacture than outright performance.

If you're intending to turbocharge the engine this really doesn't apply, you have a large restriction in the system called a turbine blade that's trying to recover energy from the hot exhaust gasses and creating a fair amount of back pressure compared to a regular N/Aspirated engine... so until you get to the stage where you have a tuned length exhaust manifold this isn't going to make a difference.

SteveC
 
Last edited:
Stepped or not, be sure to check the gasket. I find the stock manifold gaskets can have smaller openings than the port in the manifold or the openings in the exhaust header/manifold. Same holds true for the intake.
 
be sure to check the gasket. I find the stock manifold gaskets can have smaller openings
I also found that the stock manifolds gaskets that I have do not match all of the bolt holes very well. If I place the gasket over the studs, some of the port openings are slightly off. If I place the gasket so all of the port openings line up well, then the studs do not fit. Only required a little enlargement of some bolt holes. But also the middle point where the two gaskets meet needed some trimming to prevent them from interfering with one another.
 
The step should be at the lower half of the port, where the air speed is lower. Step should be approx 1 to 1.5mm. The upper half of the port should be a smooth match with no step.
Thanks Steve. I remembered you said something about this in our old "header design" thread a few years ago, but wasn't sure if anything different has been discovered since.

For other engines, the conventional "step" design has the exhaust manifold opening larger all the way around the head port (circumferentially). But that is a generic design, not one specifically developed for one particular head's charactorestics.
 
Thanks Kevin. That excellent article is mainly focused on the overall length of the exhaust system, mergers, and the pressure waves created by these "downstream" elements. In the case of the "step" design manifold I think it is a little different. As I understand, it has to do more with the velocity of flow as it exits the head's exhaust valve/port. The increased diameter creates a slight pressure change that accelerates the velocity locally for just that cylinder. I suppose the theory may be somewhat the same at a different level, but perhaps the application is different? However I am not that well versed on the details or any recent research so I'm certainly open to more info.
 
Thanks Kevin. That excellent article is mainly focused on the overall length of the exhaust system, mergers, and the pressure waves created by these "downstream" elements. In the case of the "step" design manifold I think it is a little different. As I understand, it has to do more with the velocity of flow as it exits the head's exhaust valve/port. The increased diameter creates a slight pressure change that accelerates the velocity locally for just that cylinder. I suppose the theory may be somewhat the same at a different level, but perhaps the application is different? However I am not that well versed on the details or any recent research so I'm certainly open to more info.

I've always believed that an increase in diameter/area of the exhaust caused a decrease in velocity. But, I am not exactly sure where the increase is. Just after
the valve or 6" downstream in the header?

Edit:
Ok, I read Steve's description again and my question is whether the step reduces the port diameter or increases it.
At one time tuners were adding "anti-reversion" cones just inside the header entry, these reduced the area and were an attempt to stop backflow into the cylinder.
 
Last edited:
The way I understand it, the velocity will decrease after the diameter gets bigger, as the gasses expand to fill the increased area. But that causes the velocity to increase just before the diameter gets bigger, as the gases rush to fill the lower pressure of the expanded area. That is what I believe Steve has achieved by increasing the manifold diameter at the lower half of the port.
2017-06-20_17-18-38.jpg



I might be remembering it wrong, but aren't the anti-reversion cones actually bigger than the port and taper back down to the same size as they went into the header tubes?
Anti_ReversionChamber.jpg



Tubes-3.jpg
 
The way I understand it, the velocity will decrease after the diameter gets bigger, as the gasses expand to fill the increased area. But that causes the velocity to increase just before the diameter gets bigger, as the gases rush to fill the lower pressure of the expanded area. That is what I believe Steve has achieved by increasing the manifold diameter at the lower half of the port.
View attachment 22707


I might be remembering it wrong, but aren't the anti-reversion cones actually bigger than the port and taper back down to the same size as they went into the header tubes?
View attachment 22708


View attachment 22709

Not sure where/when I read the article but the one I saw used truncated cones welded into the entry of the header. I never persued it further, could have been early experiments that didn't work out.
What you show with the expanded section is probably later than I was looking at.
When you take temp readings on the outside of several header tubes the readings seem rather inconsistent. Hot/cold areas never seem totally predictable even on similar looking tubes. Is higher/lower temps an indicator of higher pressure locations?
 
Gene, I've also seen some cone "inserts" available but don't know their purpose or theory. Could be the same idea but executed differently?

Good question about temps. I imagine there are several factors that relate to temp differences. Pressure would be one of them. I know from physics that as the temperature of a gas in a closed space goes up the pressure does too. But what the exact cause and effect is here isn't so clear.

In the last picture I posted (complete exhaust header for a motorcycle), the pipes increase diameter but do not reduce again. This is what they refer to as a "stepped" design, rather than a "anti-reversion" design. It may have the same end effect but I believe the principle and theory is a bit different. Some have multiple "steps" at certain points along the header tubes.
 
Gene, I've also seen some cone "inserts" available but don't know their purpose or theory. Could be the same idea but executed differently?

Good question about temps. I imagine there are several factors that relate to temp differences. Pressure would be one of them. I know from physics that as the temperature of a gas in a closed space goes up the pressure does too. But what the exact cause and effect is here isn't so clear.

In the last picture I posted (complete exhaust header for a motorcycle), the pipes increase diameter but do not reduce again. This is what they refer to as a "stepped" design, rather than a "anti-reversion" design. It may have the same end effect but I believe the principle and theory is a bit different. Some have multiple "steps" at certain points along the header tubes.

Normally when you think about header design (or maybe operation for our purpose), you assume the flow is all smooth and equally distributed in the pipe. That may not be the case. Sometimes flow runs in a "ribbon" that wanders through the pipe. I think that's what causes the hot(ter) sections in the header. Not sure what that does or tells you about the pressure at that particular area.
As for the MC header, it might be exactly what we are talking about for reversion. Another thought is that at 12 to 14K RPM there is still exhaust expansion happening that far away from the valve and more area is required to keep the increased volume moving.
Edit:
Do we know that the larger area is not just double wall pipe and not increased tube size?
 
Last edited:
Do we know that the larger area is not just double wall pipe and not increased tube size?
For the particular example I posted the title said it is a stepped design. So I'm sure it is not just a double layer. I've also seen race car headers that look like that, but even do it more than once - with bigger and bigger pipes as it progresses. Somewhere I read something about how they determine the increase the diameter with the step but I don't recall any details.
 
This discussion and Kevin's link to the Burn's article on total exhaust length reminds me of the very first hardback book I ever bought for myself as a kid. I must have been around seven years old when we went to a real bookstore. Living in a rural area we did not have such stores near us, so this was part of a trip out of town. And it was the first time I had ever seen so many books other than at a public library. Having discovered the automotive section I began looking at the titles. One stood out at me; "The sports car engine. Its tuning and modification" by Colin Campbell. I had always been a huge fan of European sports cars, and in particular ones modified for performance and racing. This book discussed how to do that in detail, with the engineering science and theory to support things, and it really spoke to my interests. So I bought it despite being what seemed like a lot of money at that age. One of the concepts the book explained was how the total length of the exhaust system influenced the engine's performance characteristics. Something similar to the Burns article but in much more detail, including the mathematics involved. I still have the book and much of it is still just as relevant today.
 
The way I understand it, the velocity will decrease after the diameter gets bigger, as the gasses expand to fill the increased area.

That is correct, Boyles gas laws dictate that if the volume increases, the pressure and temperature will decrease.

You are sort of talking about two different things here... a "Step" is an abrupt change of cross sectional area, in our case characterised by a sharp right angle at the point where the cross section increases.

and yes this picture you found Jeff is exactly what I was trying to describe
2017-06-20_17-18-38.jpg

with the upper part of the port "matched" and the lower portion with a sharp / abrupt "step" which will help prevent exhaust reversion back into the cylinder on overlap.

But that causes the velocity to increase just before the diameter gets bigger, as the gases rush to fill the lower pressure of the expanded area. That is what I believe Steve has achieved by increasing the manifold diameter at the lower half of the port.
View attachment 22707

This is the "second" thing... not to do with "anti reversion" but what is known in gasflow terms as "Pressure Recovery" and requires a completely different gasflow scenario. This is what you try to achieve as the air rushes past the valve and into the cylinder / combustion chamber. If the form and angles involved in the valve seat and the angles at which it merges with the combustion chamber / head face are "JUST RIGHT"... then you see an increase in port velocity as the flow exits the valve seat throat and passes around the lip of the valve.

Not the same sort of thing as what's happening in our exhaust port / exhaust runner with a sharp angled change of cross section, so there is no localised increase in air speed involved, it's simply an anti reversion step to prevent exhaust gasses finding their way back past the valve while the inlet valve is open and the piston has started to move down the bore again during the overlap phase.

SteveC
 
Back
Top