Crosspost: Exhaust System 101
05-13-2001, 02:19 PM
#3
Elder Member
Join Date: Mar 2000
Posts: 39,901
Likes: 0
Received 0 Likes
on
0 Posts
how does restriction reduce noise
The explanations of absorption (glass packs) and reflection do a good job of explaining how they reduce noise emitted from the tailpipe.
The restriction section doesn't do anything to explain how restriction limits noise. Now, to some people that may be fine, but I'd like to know how restrictive muffling works.
Can anyone elaborate?
The restriction section doesn't do anything to explain how restriction limits noise. Now, to some people that may be fine, but I'd like to know how restrictive muffling works.
Can anyone elaborate?
05-14-2001, 10:12 AM
#4
AudiWorld Senior Member
Join Date: Jun 2000
Posts: 1,284
Likes: 0
Received 0 Likes
on
0 Posts
Critique of the article
To answer Steve's question, the noise of the motor is caused by the hot exhaust gasses leaving the port and hitting cooler ambient air with some energy.
By introducing a restriction into the exhaust system it robs the gas of some of its energy, therefore it won't turn that energy into noise at the tailpipe.
I think it is false to say that turbo motors require less muffling because the turbo is a restriction. Turbos do indeed introduce additional backpressure, but the amount of noise reducion purely from that restriction is only part of the equation. The turbo absorbs a LOT of energy from the flowing exhaust gas to drive itself, which is the whole point of the exercise. The EGTs drop several hundred degrees in the transfer through the turbo housing, and that alone is responsible for a large reduction in noise. Additionally, the turbo tends to absorb and stretch out the individual exhaust pulses from each cylinder opening.
The character of an exhaust pulse has as much to do with the noise level as the overall energy it contains. Look at two stroke motors versus four strokes. In a four stroke, the exhaust valve is opened by the cam and takes 20-50 degrees of crankshaft rotation to come fully open to maximum lift, and then is closed again. There are limits to how quickly a cam can push a valve. In contrast, when the piston clears the port wall of an exhaust port two cylinder, the port opens to full width immediately, and to full height within a few (possibly 20) crank degrees, and remains open for scavenging while the piston dwells at bottom dead center and then rises. This abrupt opening rather than the smoother opening of a poppet valve is what gives a two stroke its "ring ding" character and why they are notoriously hard to silence. This is also true of radial port rotary engines like the 13B Mazda. Noisy, not because of the total flow, but the suddenness of the pulse.
Because the pulse of a turbo motor must go into the turbo housing, wrap almost 300 degrees around the scroll, go thru the turbine blades, and exit out at 90 degrees to its input, there is a lot of lengthening of the pulse and reduction in its sharpness. 14 liter diesel over the road rigs routinely run with no mufflers and make less exhaust noise than a car with a 5.7 liter v-8 and glass packs right next to it (Except when the jake brake is on, but that is another story, there the compressed air from the compression stroke is bled off thru the exhaust suddenly, making all the noise. This helps slow the truck by making the engine waste energy on compressing the air, but robbing it of the chance to use that air as a spring on the following (power) stroke. The result is that you can get 600 hp of braking from a Cummins 600 hp 14 liter.)
Basically to get a good exhaust system you have to reduce the energy of the exhaust gas while not restricting its flow. Reflection and expansion are your friends. The expansion room of the muffler is what slows the exhaust gas greatly, again cooling it and robbing it of energy. As the gas expands, it creates a new train of pulses with their own frequency, generally lower than those of the cylinders. Standing waves from reflection can cancel out or greatly reduce the incoming pressure pulses, but there is a bell curve of the effective engine speed at which they work because the pulses come faster at higher speeds. You hope the new pulse created by the gas hitting the expansion chamber of the muffler is stronger than the pulses from each cylinder, then you tune the resonance of the muffler to that pulse length, not the pulse length of the cylinders.
In the worst case, this resonance can actually AMPLIFY noise. I had a drag race truck with a 460 motor in it with a crossover pipe between the header collectors. at Idle, the motor was LOUDER than at 1500 rpm because of a resonance between the cylinder banks when the odd-man-out cylinder fired. It was a profound bass throb that would do a low rider proud.
Too large a pipe is bad because the gas can cool too quickly, so it doesn't make it to the muffler with enough energy to set up resonance pulses, which is why Civics with 6" exhausts sound like a bumble bee in a barrel. You actually hear the cylinder pulses and not the muffler resonances. I see no need for 4" exhausts. A 300 HP 1.8T can probably use the 3" exhausts available. If you want to check, take out the o2 sensor, attach a 6" length of 1/8" pipe, and hook a vacuum/pressure auge (like a boost gauge) to it. see if you get much restriction. Bell in his book sas on the order of 2-3 pis restriction is about as good as it gets, so if you are in that range at peak power, you are fine.
Engines don't NEED backpressure, but having almost none on a motor with a large amount of valve overlap means that you will lose torque down low because you will scavenge the cylinders too efficiently. No, you won't warp your exhaust valves w/o an exhaust "because they cool down too quickly". That myth is laughable to anyone who has ever worked (like I have) around aircraft engines. The exhaust pipe on our 9-cylinder Wrights was about 6" long. No warped valves there...
Anyway, I am just a self-profefssed expert, so what do I know.
By introducing a restriction into the exhaust system it robs the gas of some of its energy, therefore it won't turn that energy into noise at the tailpipe.
I think it is false to say that turbo motors require less muffling because the turbo is a restriction. Turbos do indeed introduce additional backpressure, but the amount of noise reducion purely from that restriction is only part of the equation. The turbo absorbs a LOT of energy from the flowing exhaust gas to drive itself, which is the whole point of the exercise. The EGTs drop several hundred degrees in the transfer through the turbo housing, and that alone is responsible for a large reduction in noise. Additionally, the turbo tends to absorb and stretch out the individual exhaust pulses from each cylinder opening.
The character of an exhaust pulse has as much to do with the noise level as the overall energy it contains. Look at two stroke motors versus four strokes. In a four stroke, the exhaust valve is opened by the cam and takes 20-50 degrees of crankshaft rotation to come fully open to maximum lift, and then is closed again. There are limits to how quickly a cam can push a valve. In contrast, when the piston clears the port wall of an exhaust port two cylinder, the port opens to full width immediately, and to full height within a few (possibly 20) crank degrees, and remains open for scavenging while the piston dwells at bottom dead center and then rises. This abrupt opening rather than the smoother opening of a poppet valve is what gives a two stroke its "ring ding" character and why they are notoriously hard to silence. This is also true of radial port rotary engines like the 13B Mazda. Noisy, not because of the total flow, but the suddenness of the pulse.
Because the pulse of a turbo motor must go into the turbo housing, wrap almost 300 degrees around the scroll, go thru the turbine blades, and exit out at 90 degrees to its input, there is a lot of lengthening of the pulse and reduction in its sharpness. 14 liter diesel over the road rigs routinely run with no mufflers and make less exhaust noise than a car with a 5.7 liter v-8 and glass packs right next to it (Except when the jake brake is on, but that is another story, there the compressed air from the compression stroke is bled off thru the exhaust suddenly, making all the noise. This helps slow the truck by making the engine waste energy on compressing the air, but robbing it of the chance to use that air as a spring on the following (power) stroke. The result is that you can get 600 hp of braking from a Cummins 600 hp 14 liter.)
Basically to get a good exhaust system you have to reduce the energy of the exhaust gas while not restricting its flow. Reflection and expansion are your friends. The expansion room of the muffler is what slows the exhaust gas greatly, again cooling it and robbing it of energy. As the gas expands, it creates a new train of pulses with their own frequency, generally lower than those of the cylinders. Standing waves from reflection can cancel out or greatly reduce the incoming pressure pulses, but there is a bell curve of the effective engine speed at which they work because the pulses come faster at higher speeds. You hope the new pulse created by the gas hitting the expansion chamber of the muffler is stronger than the pulses from each cylinder, then you tune the resonance of the muffler to that pulse length, not the pulse length of the cylinders.
In the worst case, this resonance can actually AMPLIFY noise. I had a drag race truck with a 460 motor in it with a crossover pipe between the header collectors. at Idle, the motor was LOUDER than at 1500 rpm because of a resonance between the cylinder banks when the odd-man-out cylinder fired. It was a profound bass throb that would do a low rider proud.
Too large a pipe is bad because the gas can cool too quickly, so it doesn't make it to the muffler with enough energy to set up resonance pulses, which is why Civics with 6" exhausts sound like a bumble bee in a barrel. You actually hear the cylinder pulses and not the muffler resonances. I see no need for 4" exhausts. A 300 HP 1.8T can probably use the 3" exhausts available. If you want to check, take out the o2 sensor, attach a 6" length of 1/8" pipe, and hook a vacuum/pressure auge (like a boost gauge) to it. see if you get much restriction. Bell in his book sas on the order of 2-3 pis restriction is about as good as it gets, so if you are in that range at peak power, you are fine.
Engines don't NEED backpressure, but having almost none on a motor with a large amount of valve overlap means that you will lose torque down low because you will scavenge the cylinders too efficiently. No, you won't warp your exhaust valves w/o an exhaust "because they cool down too quickly". That myth is laughable to anyone who has ever worked (like I have) around aircraft engines. The exhaust pipe on our 9-cylinder Wrights was about 6" long. No warped valves there...
Anyway, I am just a self-profefssed expert, so what do I know.
05-14-2001, 11:41 AM
#5
New Member
Join Date: Aug 2000
Posts: 236
Likes: 0
Received 0 Likes
on
0 Posts
Re: Critique of the article
This article was filled with silliness, starting from the statement about restrictive exhaust manifolds cutting power because the pistons have to push against the pressure in the manifold. While the basic statement has some truth, it's not the primary power killer. The reason for most of the lost power is because the residual exhaust gases in the cylinder (which don't burn) mean that less fuel and fresh air is ingested on the next intake stroke.
There are lots of examples, including the light treatment of what happens to gases that actually allow for resonance tuning, light treatment of how exhaust velocity aids scavenging, etc.
Perhaps I'm just being cranky, though. The piece didn't really address a question that I'd hoped it would:
What I'd really like to know is why exhaust back pressure aids torque production at low and medium engine speeds. I've read that it does, and I've been told it does by people whose opinions I respect. I've even seen it work on a number of occasions, on a dyno, on several different cars - only one of which had enough cam to qualify for statements regarding a lot of overlap. There's even an article in the current Roundel magazine about putting a little restriction into a stock M car that added torque.
But nobody's been able to tell me the why and how.
RangeR BoB?
There are lots of examples, including the light treatment of what happens to gases that actually allow for resonance tuning, light treatment of how exhaust velocity aids scavenging, etc.
Perhaps I'm just being cranky, though. The piece didn't really address a question that I'd hoped it would:
What I'd really like to know is why exhaust back pressure aids torque production at low and medium engine speeds. I've read that it does, and I've been told it does by people whose opinions I respect. I've even seen it work on a number of occasions, on a dyno, on several different cars - only one of which had enough cam to qualify for statements regarding a lot of overlap. There's even an article in the current Roundel magazine about putting a little restriction into a stock M car that added torque.
But nobody's been able to tell me the why and how.
RangeR BoB?
05-14-2001, 12:27 PM
#6
Elder Member
Join Date: Mar 2000
Posts: 39,901
Likes: 0
Received 0 Likes
on
0 Posts
bumble bee in a barrel! LOL
That's a good one. I call them "chainsaw farting noises".
I'm totally used to those noises now with all the ricers around here. It's starting to sound normal to me. I was going to whine about how on some cars (the Civics in question) you do sound like you are hearing the exhaust pulses straight from the cylinders, but you explained it perfectly and how it is different from a more normal exhaust.
I wonder if the supposed "S2000*2" new NSX will sound like hornet in a barrel or like a real V8?
I'll have to go back and reread the article. Although it seems it had a little b***s*** in it, I thought it was pretty good all in all. I was skeptical about the "turbo" (including "turbo" mufflers) section, but otherwise pretty impressed. I'll go back again and see if I was just fooled.
I'm totally used to those noises now with all the ricers around here. It's starting to sound normal to me. I was going to whine about how on some cars (the Civics in question) you do sound like you are hearing the exhaust pulses straight from the cylinders, but you explained it perfectly and how it is different from a more normal exhaust.
I wonder if the supposed "S2000*2" new NSX will sound like hornet in a barrel or like a real V8?
I'll have to go back and reread the article. Although it seems it had a little b***s*** in it, I thought it was pretty good all in all. I was skeptical about the "turbo" (including "turbo" mufflers) section, but otherwise pretty impressed. I'll go back again and see if I was just fooled.
05-14-2001, 02:32 PM
#7
AudiWorld Senior Member
Join Date: Jun 2000
Posts: 1,284
Likes: 0
Received 0 Likes
on
0 Posts
Cylinder filling is a function of many things...
You can enhance optimum cylinder filling in several ways: the intake tract, the cam and the exhaust tract. Peak torque always occurs where the engine has maximum volumentric efficiency.
You can tune the intake tract to resonate at a specific frequency, and that will cause a "ram" function of the air going into the cylinder. The air column moving down the runner moves into the cylinder, drawing air out of the plenum. The rate at which it does this before the intake valve closes sets up a pulse much like an exhaust pulse in reverse, and by tuning the length and diameter of the pipe (length sets RPM, size sets the strength of the signal felt in the plenum). You can actually get Volumetrc Efficiency > 100% doing this, over a VERY narrow rev range, tho it still breathes adequately (>85% is GOOD for a normally aspirated motor) at other rev ranges.
This resonance is like a header in reverse, so it works best at a single frequency, which corresponds to a single rev range. This can be spread out by using multiple intake tracts and valves to switch (Audi v8, Porsche, Mercury Cougar 2.5, many others) the length to two or more values to widen the torque peak.
The Cam is the master component and is far and away the determinant factor in cylinder filling. When and how much the valves open is important.
On to your question: Backpressure helps low speed cylinder filling because the bulk volume of air that you are moving is less at lower revs. If the exhaust is too open then you get a weak signal from the tuned exhaust pulses of the header/exhaust combo coming back to the cylinder to aid in scavenging. This reduces scavenging, increases reversion and decreases V.E.
As the exhaust gas leaves the combustion chamber, it first causes a pressure wave to leave from the exhaust valve and travel down to the, at which point it expands and actually creates a suction on the other cylinder's pipes. This vacuum signal (like blowing across a coke bottle creates a vacuum in the bottle) helps reduce pressure in the adjoining cylinder's header pipes, so when the exhaust valve opens there is an even larger pressure differential and flow starts off smartly.
As a cylinder's charge is exhausted into the collector, the expansion from a, say, 1.5" primary into a 3" collector causes a vacuum signal to be felt in the header pipe, which helps scavenge the cylinder efficiently. The moving exhaust gas has inertia, and is "pulled along" by the expansion into a larger volume. Think of a line for a ride at Disneyworld. The queue is backed up for the duration of the ride, then suddenly they raise the bar, and a bunch of people flow forward and disperse to get on the ride (like exhaust in the collector), causing a surge (albeit slower) in the queue. Everyone in the queue gets dragged along. This reduced pressure in the header pipe (its not really a vacuum, but the pressure in the pipe is much less than it was, and gas flow is like sex in Little Rock, its all relative) aids residual gas extraction from the cylinder.
Remember all engines have a bit of overlap between the intakes and exhausts, with the intake opening as the piston is still coming up on the exhaust stroke. As the intakes open, the reduced pressure in the combustion chamber (from header scavenging) means that intake flow will start even before the piston reaches TDC on the exhaust stroke and starts its descent on the intake stroke.
This is optimal, but with a big pipe on a street motor, if you get too free a flow, you don't get as strong a pulsing of signals throughout the exhaust tract back to the cylinder, so initial flow thru the intake will be less. In the worst case you get a bit of reversion of exhaust into the intake tract as the valve opens, and reverse flow as the piston clears TDC and EXHAUST gas flows BACK into the cylinder from the header for the first few degrees of crank rotation until the valves open enough to really fill the cylinder.
So a little back pressure helps make a stronger pressure signal which helps keep too much reversion from happening, which helps cylinder filling, which makes power. The down side is less total flow at high revs, but that is one of the tradeoffs. Lots of power over a narrow band, or adequate power everywhere.
The ultimate expression of resonance tuning is in 2-stroke bike motors. There the pull of the expansion chamber is what completely drives the tuning of the motor. On the pipe, you can make 60 hp from a 250 cc motor. 500 rpm off the pipe, you are down to 30 hp and no go! Extreme, but with 6 gears right on top of each other on a dirt bike or 500 cc GP bike (180 hp from a naturally aspirated, crank scavenged, 500 cc V4 is not too shabby) and they are tolerable. Miss a corner entry and hit the brakes, and its down 3 gears to get going again.
Does that help?
You can tune the intake tract to resonate at a specific frequency, and that will cause a "ram" function of the air going into the cylinder. The air column moving down the runner moves into the cylinder, drawing air out of the plenum. The rate at which it does this before the intake valve closes sets up a pulse much like an exhaust pulse in reverse, and by tuning the length and diameter of the pipe (length sets RPM, size sets the strength of the signal felt in the plenum). You can actually get Volumetrc Efficiency > 100% doing this, over a VERY narrow rev range, tho it still breathes adequately (>85% is GOOD for a normally aspirated motor) at other rev ranges.
This resonance is like a header in reverse, so it works best at a single frequency, which corresponds to a single rev range. This can be spread out by using multiple intake tracts and valves to switch (Audi v8, Porsche, Mercury Cougar 2.5, many others) the length to two or more values to widen the torque peak.
The Cam is the master component and is far and away the determinant factor in cylinder filling. When and how much the valves open is important.
On to your question: Backpressure helps low speed cylinder filling because the bulk volume of air that you are moving is less at lower revs. If the exhaust is too open then you get a weak signal from the tuned exhaust pulses of the header/exhaust combo coming back to the cylinder to aid in scavenging. This reduces scavenging, increases reversion and decreases V.E.
As the exhaust gas leaves the combustion chamber, it first causes a pressure wave to leave from the exhaust valve and travel down to the, at which point it expands and actually creates a suction on the other cylinder's pipes. This vacuum signal (like blowing across a coke bottle creates a vacuum in the bottle) helps reduce pressure in the adjoining cylinder's header pipes, so when the exhaust valve opens there is an even larger pressure differential and flow starts off smartly.
As a cylinder's charge is exhausted into the collector, the expansion from a, say, 1.5" primary into a 3" collector causes a vacuum signal to be felt in the header pipe, which helps scavenge the cylinder efficiently. The moving exhaust gas has inertia, and is "pulled along" by the expansion into a larger volume. Think of a line for a ride at Disneyworld. The queue is backed up for the duration of the ride, then suddenly they raise the bar, and a bunch of people flow forward and disperse to get on the ride (like exhaust in the collector), causing a surge (albeit slower) in the queue. Everyone in the queue gets dragged along. This reduced pressure in the header pipe (its not really a vacuum, but the pressure in the pipe is much less than it was, and gas flow is like sex in Little Rock, its all relative) aids residual gas extraction from the cylinder.
Remember all engines have a bit of overlap between the intakes and exhausts, with the intake opening as the piston is still coming up on the exhaust stroke. As the intakes open, the reduced pressure in the combustion chamber (from header scavenging) means that intake flow will start even before the piston reaches TDC on the exhaust stroke and starts its descent on the intake stroke.
This is optimal, but with a big pipe on a street motor, if you get too free a flow, you don't get as strong a pulsing of signals throughout the exhaust tract back to the cylinder, so initial flow thru the intake will be less. In the worst case you get a bit of reversion of exhaust into the intake tract as the valve opens, and reverse flow as the piston clears TDC and EXHAUST gas flows BACK into the cylinder from the header for the first few degrees of crank rotation until the valves open enough to really fill the cylinder.
So a little back pressure helps make a stronger pressure signal which helps keep too much reversion from happening, which helps cylinder filling, which makes power. The down side is less total flow at high revs, but that is one of the tradeoffs. Lots of power over a narrow band, or adequate power everywhere.
The ultimate expression of resonance tuning is in 2-stroke bike motors. There the pull of the expansion chamber is what completely drives the tuning of the motor. On the pipe, you can make 60 hp from a 250 cc motor. 500 rpm off the pipe, you are down to 30 hp and no go! Extreme, but with 6 gears right on top of each other on a dirt bike or 500 cc GP bike (180 hp from a naturally aspirated, crank scavenged, 500 cc V4 is not too shabby) and they are tolerable. Miss a corner entry and hit the brakes, and its down 3 gears to get going again.
Does that help?
Trending Topics
05-14-2001, 02:37 PM
#8
AudiWorld Senior Member
Join Date: Jun 2000
Posts: 1,284
Likes: 0
Received 0 Likes
on
0 Posts
Unless the V-8 Honda has a flat crank, or a bundle of snakes header, it will sound like a V8
That is just a function of firing order, and is inescapable in a 90 degree crank V8
I had a post here somewhere about it, relating to the W8 from VW having a flat crank.<ul><li><a href="https://forums.audiworld.com/performance/msgs/29781.phtml">What the heck is a flat crank, and why do V-8s rumble?</a></li></ul>
I had a post here somewhere about it, relating to the W8 from VW having a flat crank.<ul><li><a href="https://forums.audiworld.com/performance/msgs/29781.phtml">What the heck is a flat crank, and why do V-8s rumble?</a></li></ul>
05-14-2001, 05:24 PM
#9
Junior Member
Join Date: Mar 2000
Posts: 1,004
Likes: 0
Received 0 Likes
on
0 Posts
JASON, ANYONE, put RangeR BoB's responses below in a FAQ!!
I think I learned more about exhausts and things like scavenging in the 15 minutes it took me to read and digest his responses than I've learnt the whole time I've been on Audiworld (3 yrs)!!!
Better yet, RangeR BoB, write a new FAQ please!!!! I want to be able to send other people over to Audiworld to read stuff like what was in your replies!
Ash
(feeling like I should go back to school again!)
Better yet, RangeR BoB, write a new FAQ please!!!! I want to be able to send other people over to Audiworld to read stuff like what was in your replies!
Ash
(feeling like I should go back to school again!)
05-15-2001, 01:04 PM
#10
AudiWorld Senior Member
Join Date: Jun 2000
Posts: 1,284
Likes: 0
Received 0 Likes
on
0 Posts
Thanks! I'd be happy to make a better FAQ on exhausts if folks want me to.
Not to diss the other FAQ, which was indeed a 101 level course, but how about a 201 course with my content in it?