Exhaust scavenging explanation please

[marleyskater420]

What is exhaust scavenging and what is backpressure? How come exhaust scavenging is good for some and not other engines? Same with backpressure.

Do you want more backpressure for more torque?How come?

How would you go about designing a system lets say, around exhaust scavenging..like what kinda of a flow of each do you want for N/A and turbo setups?

And a random question..

Do engine's cylinders fire at the exact same time when in their firing order?

How do you tune the resonance of the intake with exhaust scavenging, dealing with the pulses and whatnot? I probably am making no sense, but thats because I have no idea what Im talking about.

 

[Warp Six]

Imagine that the exhaust in the exhaust system of the car is a slug of moving gas (because it is ). When the exhaust valve is opened, this moving slug of gas effectively helps pull the exhaust out of the cylinder, much like a siphon. This is called exhaust scavenging. The effect can be tuned by the design of the exhaust system. Generally speaking, the smaller the exhaust piping, the faster the exhaust velocity and the more pronounced the scavenging effect, up to the point where the exhaust system becomes a restriction and backpressure overcomes it. The scavenging effect is therefore tuned by the size and length of the header primaries and the exhaust piping. I'm sure there is computer software out there to help with this sort of tuning, but I imagine most engine builders rely more heavily on experience and dyno time.

Since a turbocharger creates a large amount of backpressure, none of this applies for a turbocharged application. The turbo is driven by the high pressure in the exhaust system between the engine and turbo. This is why, generally speaking, you want a cam with a good amount of lobe separation. With too much overlap, the pressure in the exhaust will dilute the incoming fuel/air mixture.

Please note that I'm no professional when it comes to this stuff. I'm quite confident someone will point out any errors I may have made. See my sig for disclaimer.

 

[marleyskater420]

Is lobe seperation overlap?

Thanks for the info.

 

[trbo668bird]

Check out the May 2005 Popular Hot Rodding. It has a very in depth article on exhaust science. I've read it three times and still haven't got it all.

 

[FlaBoy]

Originally posted by marleyskater420
Is lobe seperation overlap?

Thanks for the info.

Well, sorta. Its the same thing, viewed form two different angles. Overlap is how much the cam lobes overlap, like how much time there is when the intake valve has started to open, but hte exh valve hasnt closed all teh way yet. lobe separation is the angular (degreees) distance between the peaks of the int. and exh. lobes on a cam... so the more overlap, the less lobe separation, and vice versa (asssuming a constant duration). make sense?

 

[marleyskater420]

Originally posted by FlaBoy
Well, sorta. Its the same thing, viewed form two different angles. Overlap is how much the cam lobes overlap, like how much time there is when the intake valve has started to open, but hte exh valve hasnt closed all teh way yet. lobe separation is the angular (degreees) distance between the peaks of the int. and exh. lobes on a cam... so the more overlap, the less lobe separation, and vice versa (asssuming a constant duration). make sense?

Less LSA means more overlap, and more LSA means less overlap then, right?

 

[Ormand]

Since a turbocharger creates a large amount of backpressure, none of this applies for a turbocharged application

It is possible to have a tuned exhaust for a turbo, but it is VERY hard to get the pipes to come out even. There have been engines for Formula I, and for Indy/CART that ran tuned exhaust with turbos. The exhaust ends up looking like a "nest of snakes", and the engine makes a lot of power, but only in a narrow band, not good for low end torque! Probably not possible in a street car, but somebody somewhere may have done it.

 

[FlaBoy]

Originally posted by marleyskater420
Less LSA means more overlap, and more LSA means less overlap then, right?

yes, assuming the two cams you're comparing have the same duration and the same ramp speed (how quickyl the cam profile goes from the base of the lobe to peak lift, or how quickly it opens the valve). Essentially if you took one cam, and somehow increased that cams LSA, you would decrease the overlap, and vice versa. But if youre comparing two different cams, it is possible that one may have a narrower LSA, but less overlap, if it has a significantly shorter duration. Alternately, its possible that one cam may have more lsa, AND more overlap than another, assuming its duration is significantly greater... make sense?

 

[bruce]

Originally posted by Ormand
It is possible to have a tuned exhaust for a turbo, but it is VERY hard to get the pipes to come out even.

There have been engines for Formula I, and for Indy/CART that ran tuned exhaust with turbos.
The exhaust ends up looking like a "nest of snakes", and the engine makes a lot of power, but only in a narrow band, not good for low end torque!

Compared to many race cars, a turbo exhuast is a piece of cake.

In the realm of basically unlimited budgets, and where 3 HP, makes the difference of getting on the podium vs, being an also ran, yes, it matters.

You also need to have 6-7 Spd transmissions to keep the engine in it's best operating range.

Once you get past, following the Herd, and investigate Bucci Comments on manifolding for turbos, things get alot better.

The *best* exhaust system for a TR, also means running dual turbos. The you can just the engine'd firing order to best use the impulse of the exhaust gas slugs for spinning the turbo.

 

[marleyskater420]

Originally posted by bruce
Compared to many race cars, a turbo exhuast is a piece of cake.

In the realm of basically unlimited budgets, and where 3 HP, makes the difference of getting on the podium vs, being an also ran, yes, it matters.

You also need to have 6-7 Spd transmissions to keep the engine in it's best operating range.

Once you get past, following the Herd, and investigate Bucci Comments on manifolding for turbos, things get alot better.

The *best* exhaust system for a TR, also means running dual turbos. The you can just the engine'd firing order to best use the impulse of the exhaust gas slugs for spinning the turbo.

Can you explain the impulse exhaust for me once more. I kinda of understood that if you time the intake pulses and the exhaust pulses at the right time, it will spool the turbo up really well. Does this mean it will spool it better than just all the air going to it?

 

[VadersV6]

If the pressure in the exhaust is lower than the pressure in the intake manifold, the fresh air charge will pass right into the intake valve and right out the exhaust valve during the overlap cycle. This is where having some backpressure is beneficial. If the pressure is lower in the intake (too much exhaust backpressure), then the fresh air charge becomes contaminated by the exhaust during overlap. This exists on all engines. At one point in the rpm band, this can be bad news, but you bring it up to a higher rpm, and this overlap actually becomes beneficial. Its all a matter of balance....where you have efficiency, you will also have an area of inefficiency. just depends where in the band you want them to be.
If you want the most out of a forced application, you want exhaust ports that flow 75-80% of the intake. But they have to be ported in a way as to keep velocity as high as possible. You want more exhaust duration as well, although I was surprised when I started seeing all these turbo buick grinds that dont have any more exhaust duration than the intake. These cars continue to surprise me.

 

[Ormand]

If you want the most out of a forced application, you want exhaust ports that flow 75-80% of the intake

That may be true for a SUPERCHARGED application, but I don't think it is for a turbo. On a supercharged application, the exhaust is at atmospheric. There is a large quantity of air/fuel flowing through the motor than would be the case on a normally aspirated engine, so there needs to more exhaust capability relative to intake. Intake is pressurized, exhaust is not.
With a turbo, on the other hand, the exhaust is pressurized more than the intake. So even if there is more air/fuel, the exhaust gas is compressed by the back pressure of the turbine. Higher pressure, less volume, smaller ports are neeed.
Normally aspirated would be "baseline", with intake at slight vacuum, exhaust under slight pressure. For those engines, exhaust flow of about 60-70% of intake flow is usually about right.
Supercharged engines- intake under significant pressure, exhaust under slight pressure. For these engines, the ratio of intake volume to exhaust volume is lower, so exhaust ports need to flow more. Maybe 75-80%, as Vader suggested.
Turbocharged engines- intake under significant pressure, exhaust under even more pressure. For these engines, the compressed exhaust gas has LOWER volume relative to intake. Seems like smaller ports would be in order, and would keep exhaust velocity up for good spooing. 50-60% might work well.
Same logic would seem to work for cam timing, so that turbo engines need LESS exhaust duration than either N/A engines or supercharged engines.

 

[FlaBoy]

Not that i am anywhere close to being an expert on thsi subhject, but I did take 3 years or so of thermodynamics and aeronautics and such at the Air Force academy, so its not total gibberish to me... and so far, the comments Ormand made seem to be the ones that make the most sense to me. Thats jsut based on my basic understanding of the subject, and its very very possible i could be very very wrong, but like i said, when i read what he had to say, it was like "ahhh, that makes sense". Just my .01 (im too poor to afford 2 of em )

 

[VadersV6]

Look at a set of stock exhaust ports. Small and restrictive. They follow the path, up to the turbine. The gases compress and heat up, which spin the exhaust wheel. After all, heat and velocity spin it.
Higher flowing ports, carrying a much greater volume of exhaust gases, reaches the turbine, just like before. But, this greater volume of gases now has to compress a great deal more to squeeze through the same size turbine as before. More exhaust gas, more compression of this gas, more heat, more velocity, more force on the exhaust wheel.
It seems that you would be able to produce the same power at lower boost levels.
You wrote this: "Higher pressure, less volume, smaller ports are neeed."
Higher pressure and low volume in the ports arent what you want. Its higher volume and higher pressure on the exhaust turbine that you want. Of course for this to work, you would need a very good flowing exhaust system after the turbo to keep the pressure differential high enough to spool quickly.
Of course I could be completely full of ****, but this seems to make sense to me.

 

[buickpower]

Imagine that the exhaust in the exhaust system of the car is a slug of moving gas (because it is ). When the exhaust valve is opened, this moving slug of gas effectively helps pull the exhaust out of the cylinder, much like a siphon.

that's called kadenacy theory, and although it works as an intuitive theory, thats not really what happens, and it was disproven ~70yrs ago. For some reason it still gets used alot to explain what is really caused by finite amplitude pressure waves in the exhaust system. Basically they are very powerful pressure waves that bounce back and forth between the exhaust valve seat an the header collector very rapidly. When a wave starts at the opening of the exhaust valve, it hits the collector and gets partially reflected back to the exhaust port. The only thing is, the waves are special in that when they get reflected, they become "negative suction" waves that move mass in the opposite direction that they themselves move. The strength of the wave is inversely proportional to the diameter of the header primary and the design of the collector. If you can tune the length of the header so that the returning negative suction wave hits the exhaust port during the valve overlap period, you not only get some scavenging, but also a mild supercharging effect. Please note that this supercharging effect only happens at a narrow rpm, sometimes at multiple points in the powerband. The rest is just regular scavenging due to the action of the waves that i mentioned previously.

 

[marleyskater420]

Originally posted by buickpower
that's called kadenacy theory, and although it works as an intuitive theory, thats not really what happens, and it was disproven ~70yrs ago. For some reason it still gets used alot to explain what is really caused by finite amplitude pressure waves in the exhaust system. Basically they are very powerful pressure waves that bounce back and forth between the exhaust valve seat an the header collector very rapidly. When a wave starts at the opening of the exhaust valve, it hits the collector and gets partially reflected back to the exhaust port. The only thing is, the waves are special in that when they get reflected, they become "negative suction" waves that move mass in the opposite direction that they themselves move. The strength of the wave is inversely proportional to the diameter of the header primary and the design of the collector. If you can tune the length of the header so that the returning negative suction wave hits the exhaust port during the valve overlap period, you not only get some scavenging, but also a mild supercharging effect. Please note that this supercharging effect only happens at a narrow rpm, sometimes at multiple points in the powerband. The rest is just regular scavenging due to the action of the waves that i mentioned previously.

So let me get this straight.. the exhaust valve opens, and the air waves that come out of it hit the exhaust manifold(header?) and are partially reflected back. They create a "negative suction" that sucks air in,while the wave goes forward. Its also inversally proportional, which means bigger the header,smaller the wave?And higher velocity header, lower velocity wave? So what you would want is to find a perfectly lengthed header so that it will reflect in such a manor that the waves returning create such great suction it brings in air much quicker than it would normally?

Am I getting this straight?

 

[Ormand]

Higher pressure and low volume in the ports arent what you want

Maybe not what you want, but as the worlds greatest rock-and-roll band once said "you can't always get what you want". The turbo requires exhaust gas to be pressurized or it WILL NOT WORK. To drive the turbo, there has to be a higher pressure on the inlet side than on the outlet side. You could use a large turbo, and have lower pressure drop, but it would take forever to spool. Or, you can use a small turbo, which will spool instantly, and cost you power at high rpm by choking the exhaust flow.
Either way, what you GET, is higher pressure in the exhaust system on a turbo engine, relative to a supercharged or a normally aspirated engine. Then, when you have higher pressure, the volume somehow just knows to go down. Funny how the gasses all know to follow the gas laws! It just doesn't matter what you WANT, with a turbo, you are going to have higher pressure/density exhaust gasses. And because they are at higher pressure, and are thus compressed, they don't need as much port size for given velocity, and they don't take up as much room, and they don't need as much exhaust valve duration relative to a normally aspirated engine.

 

[buickpower]

So let me get this straight.. the exhaust valve opens, and the air waves that come out of it hit the exhaust manifold(header?) and are partially reflected back. They create a "negative suction" that sucks air in,while the wave goes forward. Its also inversally proportional, which means bigger the header,smaller the wave?And higher velocity header, lower velocity wave? So what you would want is to find a perfectly lengthed header so that it will reflect in such a manor that the waves returning create such great suction it brings in air much quicker than it would normally?

Yup that's pretty much it, its just that the negative suction waves only help draw in more air during the valve overlap period, the rest of the time they just suck out exhaust because the intake valve is closed (which would leave more room for fresh air the next time around, so I guess technically they do help draw in more air all the time indirectly). Also a higher velocity header (one that has a smaller diameter) will not necessarily mean a low velocity wave. It will however, like you mentioned be stronger than a similar wave in a larger pipe.

I'd also like to point out that this pressure pulse phenomena also exists in the induction system of cars, and designs like a dual plane manifold for a carb and tuned port injection from the 80's were made to exploit these pressure phenomena.

 

[VadersV6]

I explained this theory to marley like a week ago, and how this theory applies to both the exhaust and intake side. Marley- remember resonant tuning?

 

[marleyskater420]

Originally posted by VadersV6
I explained this theory to marley like a week ago, and how this theory applies to both the exhaust and intake side. Marley- remember resonant tuning?

Yes. I remember the intake resonant tuning. About that- do we have an upper and lower plenum? Because I remember seeing our doghouse, which I knew to be called the plenum, sold as a seperate part of the intake manifold..

I just didnt know it applied to exhaust gases.

I learn oh so much from you guys.