The theory of exhaust systems ?

**GotF2TinGD** · 3-10-04, 8:11

Now i dont know anything for certain but this is what i have come up with in my head

The whole idea with the faster moving exhaust (vs. slower moving but larger cross-sectional area) is that the momentum of the gases will create a vacuum behind it as it's leaving between exhaust pulses. This literally "sucks" out the gas from the next cylinder when it's exhaust port opens. That's why a well designed header will give you more power, the exhaust pulses reach the collector just as the other cylinder opens, effectively sucking the exhaust gases out of the other cylinder. This reduces the work the piston has to do on the exhaust stroke, and you get more power

If it's moving slower, true, you should get the same mass flow rate, assuming no compressibility effects. But, although small, they are present, and can give you a nice boost if you use them properly.

On the other hand, you make the exhaust too small, yes, it will flow fast, but you will get alot of energy loss due to friction, and higher pressure will make it harder for the pistons to pump the exhaust out.

Am i on the right track ?

**White90MX6** · 3-10-04, 18:42

yea, thats how I would consider it... I could be wrong though.

I think also you could add that gas going from a reasonably small volume (in the engine) to a larger volume (in the exhaust) the gas (which is hot, and high in kenetic energy) will expand quickly (due to the good old p1v1 = p2v2 eq) and therefore 'suck' more air through the engine.

Also, the lack of back pressure etc in larger exhausts due to there not being enough exhaust gas (from the engine) to create enough pressure within the exhaust to effectively suck air from the engine (especially at low RPM). Like having a 3" exhaust on a 1.5L non-turbo pulsar.... hey, it happens.

Tripharn

**2Bad** · 3-10-04, 19:30

heheh I had the same problem with power vs. torque, which one is better. Its all very well to get theoretical (power is always better than torque because you can use gearing to multiply torque), but in real world situations it often another story (low end torque is what really matters for day-to-day drivability, who cares how high it revs?).

My point is how does this theory affect the kinds of exhaust designs we should actually put on our cars? Does your thoery mean that the good ol' 3" mandrel from the turbo back with hi-flow cat and straigh-through muffler is not the best exhaust for our cars (1gen)?

I'm not critisising you or anything I'm just wondering what real world implications are there for this theory? I think the most important factor for exhausts is top end gain vs. low end loss (see the power vs. torque thing is actually relevant to exhausts!)

What do you guys reckon?

**White90MX6** · 3-10-04, 20:22

If you are talking to me, then no, I think that in a turbo car that, that exhaust system would be a good upgrade, especially since it is in the high RPM that you get the majority of power from our cars. It is the larger volume of the exhaust that allows the gas to expand, therefore greater a larger pressure difference on each side of the turbo... thereby pulling more air through the turbo, and allowing the turbo to spool faster... which in turn is more power.

That is why turbo cars have much bigger exhausts than NA cars.

But, in a non-turbo car, then I personally believe that you would be doing more bad than good.

Tripharn

**rolli** · 3-10-04, 21:52

You ever noticed when you replaced you stock exhaust with a new high flow 3", you lost a bit of power down low, but this was greatly out weighed by the gain once the rev rose above 3000rpm???

This is due to exactly what you said. The extra scavaging from the exhaust helps to pull the gasses out of the cylender, however once the turbo starts moving, the spent gasses have enough charge to force itself out of the cylender faster than any negative pressure in the exhaust could try to pull it out, thus it is better to have a low speed/high volume exhaust than use gas speed to create the volume (smaller piping), as the negitive pressure will actually slow the high(er) pressure gass trying to get out.

NB: perfect examples of this is in NA cars, where once your get over about 2.5", no gains or even losses may result from increase exhaust size.

**MissaT** · 3-11-04, 2:13

I would totaly agree with this theory of exhaust gases.

In regards to the Turbo Vs NA issue, my understanding of why you can get away with a larger exhaust on a turbo than on an NA is because the turbo unit itself provides that degree of restriction to help make the scavenging effect work to its full potential (and it also has the effect of quieting the exhaust note).

The headers on NA cars are tuned (ie the length of the tubes) so that each 'pulse' of exhaust gas from the cyclinder reaches the collector exactly the same time apart from each other to help provide the smoothest and most effective scavenging. In addition, I havent really been able to explain it but a rule of thumb with NA headers is the 4-2-1 type provide a tradeoff between torque and high end power where as the 4-1 units are just made for pure high end power at the expense of lower end torque. I suppose the next logical thought would be towards optimising a turbo manifold in the same way as a set of headers to providethe most efficient, smoothest stream of gas. Apparently for engines below 400hp? (ie anything that we are likely to encounter) it makes bugger all difference.

**mx6mat** · 3-11-04, 11:46

These theories capture the basics.

The fact is there is an optimum diameter that is primarily dependent on the velocity of air in the pipe, the friction in the pipe, the density of the air. For fluid flow in any pipe there is a value obtained from charts using this information called the darcy friction factor. The optimal condition is where the velocity in the pipe is the fastest but still producing layered steady flow(laminar flow). The pipe will give the least resistance and consequently the losses in the pipe will be minimal.

The difference in pressure between the engine and the exhaust is negligible when using different size exhausts. The volumetric flow rate through the exhaust doesnt change that is why when you decrese the diameter the air flow has to be faster. In the perfect system the gases exiting the engine will arrive at the collector at the same time. There shouldnt be any lag other wise there will be an imbalance in the fluid flow at this point in the system and the gases will be traveling faster in one pipe than the other to reach the collector at the same time or you may get air flow in the opposite direction back into the engine in the worst case. The optimum case has to be when The air arrives at the collector allowing the maximum amount of air to flow through the exhaust.

Fluid flow can be described in 2 possible ways. Laminar or Turbulent or chaotic. In turbulent flow the friction losses in the exhaust are greater there fore this will result in a drop in pressure of the lenght of the exhaust and hence decrease the velocity of exhaust air flowing and hence the volumetric flow rate. ( ie less air exiting exhaust). At the point just before the flow becomes turbulent the flow is perfectly laminar (layered) and this gives you least pressure loss through the exhaust. Therefore greatest flow rate.

The next step would be to minimise the pressure losses due to friction in the pipe by lowering the darcy friction factor. This can be done by choosing a smoother material or polishing the inside of the exhaust )if possible. The only other way is to make the exhaust as short and as straight as possible.

Welcome to your first lesson in engineering fluid flow.

As for having less torque. I suggest this is due to the fact that because you have reduce the restriction in the exhaust, the gas will suck out of the cylinder faster and the total force on the piston at the begining of the stroke will be less untill the engine starts revving fast enough that the escaping air actually aids the piston on the up stroke and gives a higher negective pressure is the inlet opens to suck more fuel in.

The collecter also provides the smoothest possible transition into the main exhaust pipe. AT there would be the highest losses if not done properly.

Of course this description has been essential for a NA engine but the principles for the exhaust of a turbo engine are essential the same except the turbo becomes huge restriction in the system. Hence the efficiency of your turbo becomes another variable in the system. (turbine size, fiction losses in the turbo[ie ball bearing turbo] ) Allowing the gases to escape faster allows the turbo to work at its optimum.

A better exhaust hence increases efficiency as you would expect.

This is all off the top of my head from theoretical point of vue. Some of it my not be 100% so feel free to correct me if you know better but Im sure it is pretty close.

**crusty** · 3-11-04, 16:20

[Ralph Wigham's voice]
I glued my elbow to my forehead...............
[/Ralph Wigham's voice]

FRINK · 3-11-04, 17:20

Take away the chasis, wheels and gearbox. just leave the motor. what does it do? it becomes a real big air pump. the faster the exhaust fumes get out, the more cleanly it runs.

**GotF2TinGD** · 3-12-04, 2:40

Quote:

Originally posted by FRINK
Take away the chasis, wheels and gearbox. just leave the motor. what does it do? it becomes a real big air pump. the faster the exhaust fumes get out, the more cleanly it runs.

debatable.

**2Bad** · 3-12-04, 3:37

Yeah that's fine if you only want to pump air or just get a big kw figure, but a car DOES have a chassis, trans and wheels etc. and the point of the engine is to turn those wheels, and its how fast it can turn those wheels that we are concerned about, otherwise we'd just leave the stock exhaust on. So if the exhaust is so big that the engine don't turn the wheels fast at low rpm, then that's bad. mmk.

Its fine on the drag strip to produce high end power but for the street we have to consider low end torque. I'm big on this, because everyone (including car manufactors) can forget it in pursuit of big numbers. Just remember bigger don't always mean better, even with turbo cars.

**mx6mat** · 3-12-04, 8:38

with just the motor you take away the effect of cold air induction. Without moving cool air feeding in to the engine how can you expect altering the exhaust to make any difference. The chassis, wheels and gearbox are separate issues. The topic was exhaust.

It is true that high end power isnt everything. How do you think F1 engineers make a living and earn so much money, and why do they need so many.

A commercial every day car is a compromise between noise, power, torque, fuel effieciency and everyday drivability. A performance car is a compromise between power and torque race drivability. How gives a sh@# how much noise it makes if we want it to be the fastest machine. Every situation changes the compromise point. But as EPA regulation become more strict it will become harder and harder for manufacturers to increase power output with F1 technology. There is only so far that they can go. Notice smaller car engine capacities are slowly increasing.

If you like to plant your foot and feel the power increase at higher revs its lots of fun. But people are paid millions of dollar to maximise everyday drivability and that happens at low revs.

**MissaT** · 3-13-04, 8:33

Quote:

As for having less torque. I suggest this is due to the fact that because you have reduce the restriction in the exhaust, the gas will suck out of the cylinder faster and the total force on the piston at the begining of the stroke will be less untill the engine starts revving fast enough that the escaping air actually aids the piston on the up stroke and gives a higher negective pressure is the inlet opens to suck more fuel in.

This sits well with me because if you consider that the definition of peak torque is when the rev speed is at the optimal time for filling the cylinders with air. ie. It below peak, it has too much time to fill the cylinders, above peak, it doesn't have enough time to fully fill the cylinders.
A less restricive exhaust would allow quicker exiting of the gas from the cylinders, allowing more time fill the cylinders with air thus making anything below peak EVEN LESS efficient...... Holding that thought, would that mean that the peak torque would be moved up the rev range?

The scavenging effect would throw another spanner into the already complex theoretical base as well.

Question through, does exhaust gas eddie like water does? Ie if you look at a river flowing, you will see that the water around the banks is moving slower than that in the middle, this is because the bank, not being perfectly smooth causes turbulance thus causing some water to stagnate. This is actualy a good thing for flow becuase this stagnant water basicaly acts as the 'the pipe' for the water in the middle becuase what is going to cause less surface resistance, the bank or the stangant water...interesting effect. Anyway, does this occur with exhaust gasses? I believe it does on the intake side of things. I read about some research by NASA that found that a slightly rough pipe flowed better than a perfectly polished and smooth one because of the 'eddieing' (A point to remember when you take your head to get ported and polished).

**GotF2TinGD** · 3-13-04, 9:04

Quote:

Originally posted by MissaT
I read about some research by NASA that found that a slightly rough pipe flowed better than a perfectly polished and smooth one because of the 'eddieing' (A point to remember when you take your head to get ported and polished).

yes is it is totally polished the air/fuel doesnt atomise (spelling?) correctly

**mx6mat** · 3-13-04, 13:00

I think you will find that the torque curve becomes broader. I would expect it to reach peak torque quicker and hold it for longer.

That affect that you are referring to doesnt involve eddies. Eddies are a form turbulence. There is an effect like that in pipes all though its not as dramatic as that in the rivers as the bank provides a lot more friction then Pipe while and the density of water is a lot greater.. Because the gas is traveling under pressure velocity diminishes very quickly at the wall. If it is not possible to get a smooth wall then there may be optimum flow condition with a more rough pipe. This however effectively reduces the cross sectional area of the pipe resulting in the gas having to flow faster to achieve the same volumetric flow rate. Because this boundry layer of air on the wall of the pipe effectly removes friction. You will probably achieve laminar(layered) flow at higher velocities. This effect results in a pressure drop over the length of the pipe. This effect can be calculated.

Without consulting the text books i believe the worst case scenario is calculating the flow rate assuming no frictional losses against the wall of the pipe. To accurately modell this you would have to do some testing. Assume a worst case scenario or optimum operating condition possible then you know you are
below this and the exhaust is optimised for all conditions.

3-10-04, 18:42	#2 (permalink)
White90MX6 Join Date: Jan 2003 Location: iTrader: (0)	yea, thats how I would consider it... I could be wrong though. I think also you could add that gas going from a reasonably small volume (in the engine) to a larger volume (in the exhaust) the gas (which is hot, and high in kenetic energy) will expand quickly (due to the good old p1v1 = p2v2 eq) and therefore 'suck' more air through the engine. Also, the lack of back pressure etc in larger exhausts due to there not being enough exhaust gas (from the engine) to create enough pressure within the exhaust to effectively suck air from the engine (especially at low RPM). Like having a 3" exhaust on a 1.5L non-turbo pulsar.... hey, it happens. Tripharn
	150kW ATW... 12psi.

3-10-04, 20:22	#4 (permalink)
White90MX6 Join Date: Jan 2003 Location: iTrader: (0)	If you are talking to me, then no, I think that in a turbo car that, that exhaust system would be a good upgrade, especially since it is in the high RPM that you get the majority of power from our cars. It is the larger volume of the exhaust that allows the gas to expand, therefore greater a larger pressure difference on each side of the turbo... thereby pulling more air through the turbo, and allowing the turbo to spool faster... which in turn is more power. That is why turbo cars have much bigger exhausts than NA cars. But, in a non-turbo car, then I personally believe that you would be doing more bad than good. Tripharn
	150kW ATW... 12psi.

3-11-04, 2:13	#6 (permalink)
MissaT Join Date: Nov 2003 Location: Melbourne, Australia iTrader: (0)	I would totaly agree with this theory of exhaust gases. In regards to the Turbo Vs NA issue, my understanding of why you can get away with a larger exhaust on a turbo than on an NA is because the turbo unit itself provides that degree of restriction to help make the scavenging effect work to its full potential (and it also has the effect of quieting the exhaust note). The headers on NA cars are tuned (ie the length of the tubes) so that each 'pulse' of exhaust gas from the cyclinder reaches the collector exactly the same time apart from each other to help provide the smoothest and most effective scavenging. In addition, I havent really been able to explain it but a rule of thumb with NA headers is the 4-2-1 type provide a tradeoff between torque and high end power where as the 4-1 units are just made for pure high end power at the expense of lower end torque. I suppose the next logical thought would be towards optimising a turbo manifold in the same way as a set of headers to providethe most efficient, smoothest stream of gas. Apparently for engines below 400hp? (ie anything that we are likely to encounter) it makes bugger all difference.
	Common sense is what tells us the world is flatapuss FE-DOHC : Go baby, Go!!!!!! Sir Charles Lyndon: Are you done with my lady? Redmond Barry: I beg your pardon, sir? Sir Charles Lyndon: Come come now Mr. Barry, I'm a man who would rather be known as a cuckold than a fool. *1991 Black TX5 Ghia*

3-11-04, 11:46	#7 (permalink)
mx6mat Join Date: Jan 2004 Location: Bunbury Western Australia, Australia Age: 30 iTrader: (0)	These theories capture the basics. The fact is there is an optimum diameter that is primarily dependent on the velocity of air in the pipe, the friction in the pipe, the density of the air. For fluid flow in any pipe there is a value obtained from charts using this information called the darcy friction factor. The optimal condition is where the velocity in the pipe is the fastest but still producing layered steady flow(laminar flow). The pipe will give the least resistance and consequently the losses in the pipe will be minimal. The difference in pressure between the engine and the exhaust is negligible when using different size exhausts. The volumetric flow rate through the exhaust doesnt change that is why when you decrese the diameter the air flow has to be faster. In the perfect system the gases exiting the engine will arrive at the collector at the same time. There shouldnt be any lag other wise there will be an imbalance in the fluid flow at this point in the system and the gases will be traveling faster in one pipe than the other to reach the collector at the same time or you may get air flow in the opposite direction back into the engine in the worst case. The optimum case has to be when The air arrives at the collector allowing the maximum amount of air to flow through the exhaust. Fluid flow can be described in 2 possible ways. Laminar or Turbulent or chaotic. In turbulent flow the friction losses in the exhaust are greater there fore this will result in a drop in pressure of the lenght of the exhaust and hence decrease the velocity of exhaust air flowing and hence the volumetric flow rate. ( ie less air exiting exhaust). At the point just before the flow becomes turbulent the flow is perfectly laminar (layered) and this gives you least pressure loss through the exhaust. Therefore greatest flow rate. The next step would be to minimise the pressure losses due to friction in the pipe by lowering the darcy friction factor. This can be done by choosing a smoother material or polishing the inside of the exhaust )if possible. The only other way is to make the exhaust as short and as straight as possible. Welcome to your first lesson in engineering fluid flow. As for having less torque. I suggest this is due to the fact that because you have reduce the restriction in the exhaust, the gas will suck out of the cylinder faster and the total force on the piston at the begining of the stroke will be less untill the engine starts revving fast enough that the escaping air actually aids the piston on the up stroke and gives a higher negective pressure is the inlet opens to suck more fuel in. The collecter also provides the smoothest possible transition into the main exhaust pipe. AT there would be the highest losses if not done properly. Of course this description has been essential for a NA engine but the principles for the exhaust of a turbo engine are essential the same except the turbo becomes huge restriction in the system. Hence the efficiency of your turbo becomes another variable in the system. (turbine size, fiction losses in the turbo[ie ball bearing turbo] ) Allowing the gases to escape faster allows the turbo to work at its optimum. A better exhaust hence increases efficiency as you would expect. This is all off the top of my head from theoretical point of vue. Some of it my not be 100% so feel free to correct me if you know better but Im sure it is pretty close. Last edited by mx6mat : 3-11-04 at 11:51.

3-11-04, 16:20	#8 (permalink)
crusty Join Date: Mar 2003 Location: Boostville, Queensland, Oz, population....Me. Age: 37 iTrader: (0)	[Ralph Wigham's voice] I glued my elbow to my forehead............... [/Ralph Wigham's voice]
	When they outlaw freedom, only outlaws are free

3-10-04, 19:30	#3 (permalink)
2Bad Join Date: Jul 2002 Location: Melbourne iTrader: (0)	heheh I had the same problem with power vs. torque, which one is better. Its all very well to get theoretical (power is always better than torque because you can use gearing to multiply torque), but in real world situations it often another story (low end torque is what really matters for day-to-day drivability, who cares how high it revs?). My point is how does this theory affect the kinds of exhaust designs we should actually put on our cars? Does your thoery mean that the good ol' 3" mandrel from the turbo back with hi-flow cat and straigh-through muffler is not the best exhaust for our cars (1gen)? I'm not critisising you or anything I'm just wondering what real world implications are there for this theory? I think the most important factor for exhausts is top end gain vs. low end loss (see the power vs. torque thing is actually relevant to exhausts!) What do you guys reckon?

3-10-04, 21:52	#5 (permalink)
rolli Join Date: Dec 2002 Location: Perth, West Australia, Australia Age: 27 iTrader: (0)	You ever noticed when you replaced you stock exhaust with a new high flow 3", you lost a bit of power down low, but this was greatly out weighed by the gain once the rev rose above 3000rpm??? This is due to exactly what you said. The extra scavaging from the exhaust helps to pull the gasses out of the cylender, however once the turbo starts moving, the spent gasses have enough charge to force itself out of the cylender faster than any negative pressure in the exhaust could try to pull it out, thus it is better to have a low speed/high volume exhaust than use gas speed to create the volume (smaller piping), as the negitive pressure will actually slow the high(er) pressure gass trying to get out. NB: perfect examples of this is in NA cars, where once your get over about 2.5", no gains or even losses may result from increase exhaust size.

3-11-04, 17:20	#9 (permalink)
FRINK Join Date: Feb 2004 Location: Canberra, Australia Age: 25 iTrader: (0)	Take away the chasis, wheels and gearbox. just leave the motor. what does it do? it becomes a real big air pump. the faster the exhaust fumes get out, the more cleanly it runs.
	Boost......its a beautiful thing........

3-12-04, 3:37	#11 (permalink)
2Bad Join Date: Jul 2002 Location: Melbourne iTrader: (0)	Yeah that's fine if you only want to pump air or just get a big kw figure, but a car DOES have a chassis, trans and wheels etc. and the point of the engine is to turn those wheels, and its how fast it can turn those wheels that we are concerned about, otherwise we'd just leave the stock exhaust on. So if the exhaust is so big that the engine don't turn the wheels fast at low rpm, then that's bad. mmk. Its fine on the drag strip to produce high end power but for the street we have to consider low end torque. I'm big on this, because everyone (including car manufactors) can forget it in pursuit of big numbers. Just remember bigger don't always mean better, even with turbo cars.

3-12-04, 8:38	#12 (permalink)
mx6mat Join Date: Jan 2004 Location: Bunbury Western Australia, Australia Age: 30 iTrader: (0)	with just the motor you take away the effect of cold air induction. Without moving cool air feeding in to the engine how can you expect altering the exhaust to make any difference. The chassis, wheels and gearbox are separate issues. The topic was exhaust. It is true that high end power isnt everything. How do you think F1 engineers make a living and earn so much money, and why do they need so many. A commercial every day car is a compromise between noise, power, torque, fuel effieciency and everyday drivability. A performance car is a compromise between power and torque race drivability. How gives a sh@# how much noise it makes if we want it to be the fastest machine. Every situation changes the compromise point. But as EPA regulation become more strict it will become harder and harder for manufacturers to increase power output with F1 technology. There is only so far that they can go. Notice smaller car engine capacities are slowly increasing. If you like to plant your foot and feel the power increase at higher revs its lots of fun. But people are paid millions of dollar to maximise everyday drivability and that happens at low revs.

3-13-04, 13:00	#15 (permalink)
mx6mat Join Date: Jan 2004 Location: Bunbury Western Australia, Australia Age: 30 iTrader: (0)	I think you will find that the torque curve becomes broader. I would expect it to reach peak torque quicker and hold it for longer. That affect that you are referring to doesnt involve eddies. Eddies are a form turbulence. There is an effect like that in pipes all though its not as dramatic as that in the rivers as the bank provides a lot more friction then Pipe while and the density of water is a lot greater.. Because the gas is traveling under pressure velocity diminishes very quickly at the wall. If it is not possible to get a smooth wall then there may be optimum flow condition with a more rough pipe. This however effectively reduces the cross sectional area of the pipe resulting in the gas having to flow faster to achieve the same volumetric flow rate. Because this boundry layer of air on the wall of the pipe effectly removes friction. You will probably achieve laminar(layered) flow at higher velocities. This effect results in a pressure drop over the length of the pipe. This effect can be calculated. Without consulting the text books i believe the worst case scenario is calculating the flow rate assuming no frictional losses against the wall of the pipe. To accurately modell this you would have to do some testing. Assume a worst case scenario or optimum operating condition possible then you know you are below this and the exhaust is optimised for all conditions.