Another one of my insanely long exhaust tech articles
09-29-2005, 11:15 PM
#1
Registered
Thread Starter
Another one of my insanely long exhaust tech articles
Warning: This is a length record for me when it comes to posting!!! Yes it is all original. I did not cut and paste anything! :D
OK it’s time to get back to doing what I’m good at and originally known for here. Contrary to popular belief it is not being a tight *** as a moderator although I am good at that too but rather it is writing extremely long tech articles that require you to read them over and over again over the course of several sittings or one really long stay on the toilet just to comprehend them fully. Even then you’ll probably be bored to death and die from a brain cramp. How do you think I feel typing these things! I thought I’d deal with exhaust systems in this one even though I’ve dealt with it in the past. I want to go far more in depth on this one though so this one will probably have to be written in a few parts to fit on the forum. There seems to be some misunderstanding as to how certain exhausts are good for one particular application and not another. Hopefully I’ll clear that up here. I also want to deal with exhaust tuning. I know the RX-8 hasn’t been very forthcoming with gains in this area but it will. Give it time. It’s been 2 years and we are only now seeing good forced induction kits and ecu’s hit the market. I’d still like to deal with it anyways as this generally applies to all 4-stroke engines. You’ll find that some of this when it comes to sizing is also directly related to porting as well.
I’d like to start with what a lot of people try to worry about first. That is flow. People like to say that one exhaust must be better because it can flow better than another one. Air however is funny stuff. Just because it looks like it should act one way through a system doesn’t necessarily mean it will act that way. Unless you have tested it, you don’t absolutely know for certain. Just to get to the point, bigger isn’t always better when it comes to exhaust. Many already know this. The main thing people forget though is that the engine can only flow so much air. There is really no point in going any larger in pipe area when the engine can’t flow that much. Or is there? This all really depends on what your goals are surprisingly enough but for the sake of a street driven naturally aspirated car, there isn’t. I really need to deal with this in 3 main parts. These will be naturally aspirated, turbocharged, and supercharged. A lot of people may be surprised that supercharged and naturally aspirated need different systems since there isn’t anything in the way in the exhaust system such as a turbocharger. There is a need difference though.
Let’s start with naturally aspirated. In a perfect world we would know exactly how much air the exhaust ports of the Renesis can flow for any given pressure drop. Too bad we haven’t seen these numbers yet. If anyone has them from their own flow testing kindly let me know the testing procedures and results. I’d love to know. I do have a flow bench but lack an engine to test. Oh well. In the mean time, let’s get back to theory. I know some people hate theory and say that it isn’t accurate. That’s true to a point but theory needs to be tested and generally gets you very close. It is much more accurate than just pure guessing which is the preferred method of many. Even many so called “experts”. Basically on a streetcar, we want a nice broad power band with a flat torque curve. This gives us a very nice driving engine that is very well behaved that can do a broad range of driving conditions very easily. A racecar needs a narrow peaky power band as it will typically only stay in a narrow rpm range and run a close ratio gear box. This of course depends on the type of racing being done. Obviously they have different exhaust requirements from a street car. Have fun guessing which one will work best for each scenario!
The biggest secret to going fast is to remember one very important thing. It is backwards from common thinking. This applies to intakes, exhausts, porting, etc. That secret is that smaller is usually better! High velocities that don’t incur pumping losses is the rule. The whole point of flow testing, porting, etc is not to make something big so it can flow as much as it can. Take a look at the dyno of a mega horsepower Supra sometime. You might find a car that makes 300 hp at 5000 rpm but at 7000 rpm it makes 1000 hp. (for any Supra guys reading it's just an example so get over it!) That just isn't drivable on the street or a road course. It is possible to be good at a drag strip though. These are called dyno queens. These guys like big fat but very short intake runners, big plenum chambers, big turbos, big throttle bodies, big big big everything. This makes no average power and only top end power. I'm not willing to go tell these guys that it is in fact possible to make almost as much peak power but significantly more average power which would be more streetable and spool a turbo faster. They can figure it out on their own. They do have fast cars though so they probably don't care about what I think. OK off of the Supra tangent. On a rotary making a port larger changes port timing and this is also very critical to power band. Why the rotary community predominantly uses this larger logic for porting baffles me. Bigger is not always better. The key is to only go as large as you will need. That’s it. You don’t need to build in room to grow. It’ll only slow you down. I’ll give you an example. If an exhaust port can only flow 124 cfm (I’m making this number up), what is the point in having a nice large exhaust runner that is capable of 224 cfm? You’ll never use it and the added area will only slow the exhaust gasses down. You are losing free energy this way that you could harness to gain power. Why make a port rally large when the runner behind it can’t flow as much as the port window? Just to confuse you guys I’m going to go back later and show you how even this simple rule doesn’t always apply. You’ll find there are no absolutes here. There are generalizations within a certain application that should be adhered to but they may not necessarily apply in other situations. You can see how guessing which one is best is really difficult to get right. Especially when the people doing the guessing don’t even know about the generalization differences.
Back to street exhausts since this is what most here will be concerned with. On the market so far we have many different systems. None have been shown to do a whole lot in the way of power gains over stock although there is some improvement. There are systems anywhere from 2 ½” to 3” in total size. If we added up the total area of the exhaust runners leaving the engine, we’d actually find that when combined they have a total area equal to a single 3 1/8” pipe. Why then do we want a 2 ½” or 3” exhaust? First of all remember that air is in pulses through the exhaust. Just because the port is open does not mean that it is always flowing the same amount of air. In fact the exhaust pulses when moving down the exhaust runner are shaped more like a teardrop in the sense that most of their intensity is at the port opening and then tapers off as the exhaust phases gets farther and farther along. Since we have 2 rotors, we have alternating pulses. Yes we do have pulse overlap but every engine does. It also doesn’t matter in the cat back section if we have 6 pulses from 2 rotors or 6 pulses from 6 pistons (another example don’t worry about where the number came from). It all takes up space and is energy in the pipe. Since all of this energy is not present at any one spot at any one moment in time, we don’t need to total area of all of the exhaust ports combined which is why 2 ½” to 3” works just fine. Personally I’d lean towards the 2 ½” side for naturally aspirated use.
Another reason why we don’t need the total area is that this is the total area for the exhaust runners and not the ports themselves. The runners actually have more area than the actual ports do and the port runners should flow air potentially than the ports can. This means that the total area of each runner isn’t being used to it’s fullest and therefore we don’t need that much total area downstream. Are you starting to see that because of this we might actually pick up some power by making the exhaust runners smaller?! Why though? Air has energy. Faster moving air has more energy as it is more air over a distance per the same amount of time. Why not utilize it? This is what we are trying to do by not getting carried away and going with a large exhaust pipe that can flow the needed air requirements for an entire city but rather optimize it to flow what is needed for the engine.
How does backpressure come into play in all of this though? There are many unfortunately wrong people that think that more backpressure means more low end power and that you need backpressure to have it. Um, no. Wrong. Terribly wrong. First of all backpressure is nothing more than positive pressure in the exhaust. It is entirely possible to have moments of negative pressure in the exhaust. In we always have backpressure at some point in the exhaust. You never have all of only one as there are pulses at work. It all goes back to flowing exactly as much as you need to and nothing more. Remember that if we design an exhaust system that flows exactly what the engine needs to make best power at 9000 rpm, this means that we have compromised every rpm below that as we only need that flow at the max possible flow level. This means that at 3000 rpm’s, we have a pipe that is too large with velocity that is too slow. What’s the cure that a lot of people use? Add a restriction in the system to give it more backpressure. Of course power on the low end picks up but power on the top end falls off. We can no longer flow as much as necessary at peak rpm. Why did this happen if it wasn’t because of backpressure? Backpressure is a rather abstract subject. To understand what pressure is we must first have a frame of reference. On a flow bench, we have to establish a base pressure in the system before we can take a cfm reading on airflow. This is because at different pressure levels, different amounts of air molecules are present for the same area of space. Knowing how much something flows in cfm is a worthless number if you don’t know what pressure it was measured at. An example I like to use is my air compressor out in the garage. I use this to illustrate turbo sizing too. My compressor tank is sitting idle. It has 0 cfm of airflow. However it is sitting at 120 psi of pressure. What if I blow through a drinking straw? I have a measurable amount of airflow but far less pressure. I surely couldn’t blow through a drinking straw at 120 psi of pressure (insert explicit joke here). So basically we need to know pressure and flow. Here’s an interesting fact, too much backpressure can actually cause the earlier onset of detonation! Why? Not enough of the exhausted air gets out of the chamber. It therefore gets carried back into the next combustion cycle, which lowers the oxygen content and raises the temperature. Yes this can happen with no port overlap too!
OK it’s time to get back to doing what I’m good at and originally known for here. Contrary to popular belief it is not being a tight *** as a moderator although I am good at that too but rather it is writing extremely long tech articles that require you to read them over and over again over the course of several sittings or one really long stay on the toilet just to comprehend them fully. Even then you’ll probably be bored to death and die from a brain cramp. How do you think I feel typing these things! I thought I’d deal with exhaust systems in this one even though I’ve dealt with it in the past. I want to go far more in depth on this one though so this one will probably have to be written in a few parts to fit on the forum. There seems to be some misunderstanding as to how certain exhausts are good for one particular application and not another. Hopefully I’ll clear that up here. I also want to deal with exhaust tuning. I know the RX-8 hasn’t been very forthcoming with gains in this area but it will. Give it time. It’s been 2 years and we are only now seeing good forced induction kits and ecu’s hit the market. I’d still like to deal with it anyways as this generally applies to all 4-stroke engines. You’ll find that some of this when it comes to sizing is also directly related to porting as well.
I’d like to start with what a lot of people try to worry about first. That is flow. People like to say that one exhaust must be better because it can flow better than another one. Air however is funny stuff. Just because it looks like it should act one way through a system doesn’t necessarily mean it will act that way. Unless you have tested it, you don’t absolutely know for certain. Just to get to the point, bigger isn’t always better when it comes to exhaust. Many already know this. The main thing people forget though is that the engine can only flow so much air. There is really no point in going any larger in pipe area when the engine can’t flow that much. Or is there? This all really depends on what your goals are surprisingly enough but for the sake of a street driven naturally aspirated car, there isn’t. I really need to deal with this in 3 main parts. These will be naturally aspirated, turbocharged, and supercharged. A lot of people may be surprised that supercharged and naturally aspirated need different systems since there isn’t anything in the way in the exhaust system such as a turbocharger. There is a need difference though.
Let’s start with naturally aspirated. In a perfect world we would know exactly how much air the exhaust ports of the Renesis can flow for any given pressure drop. Too bad we haven’t seen these numbers yet. If anyone has them from their own flow testing kindly let me know the testing procedures and results. I’d love to know. I do have a flow bench but lack an engine to test. Oh well. In the mean time, let’s get back to theory. I know some people hate theory and say that it isn’t accurate. That’s true to a point but theory needs to be tested and generally gets you very close. It is much more accurate than just pure guessing which is the preferred method of many. Even many so called “experts”. Basically on a streetcar, we want a nice broad power band with a flat torque curve. This gives us a very nice driving engine that is very well behaved that can do a broad range of driving conditions very easily. A racecar needs a narrow peaky power band as it will typically only stay in a narrow rpm range and run a close ratio gear box. This of course depends on the type of racing being done. Obviously they have different exhaust requirements from a street car. Have fun guessing which one will work best for each scenario!
The biggest secret to going fast is to remember one very important thing. It is backwards from common thinking. This applies to intakes, exhausts, porting, etc. That secret is that smaller is usually better! High velocities that don’t incur pumping losses is the rule. The whole point of flow testing, porting, etc is not to make something big so it can flow as much as it can. Take a look at the dyno of a mega horsepower Supra sometime. You might find a car that makes 300 hp at 5000 rpm but at 7000 rpm it makes 1000 hp. (for any Supra guys reading it's just an example so get over it!) That just isn't drivable on the street or a road course. It is possible to be good at a drag strip though. These are called dyno queens. These guys like big fat but very short intake runners, big plenum chambers, big turbos, big throttle bodies, big big big everything. This makes no average power and only top end power. I'm not willing to go tell these guys that it is in fact possible to make almost as much peak power but significantly more average power which would be more streetable and spool a turbo faster. They can figure it out on their own. They do have fast cars though so they probably don't care about what I think. OK off of the Supra tangent. On a rotary making a port larger changes port timing and this is also very critical to power band. Why the rotary community predominantly uses this larger logic for porting baffles me. Bigger is not always better. The key is to only go as large as you will need. That’s it. You don’t need to build in room to grow. It’ll only slow you down. I’ll give you an example. If an exhaust port can only flow 124 cfm (I’m making this number up), what is the point in having a nice large exhaust runner that is capable of 224 cfm? You’ll never use it and the added area will only slow the exhaust gasses down. You are losing free energy this way that you could harness to gain power. Why make a port rally large when the runner behind it can’t flow as much as the port window? Just to confuse you guys I’m going to go back later and show you how even this simple rule doesn’t always apply. You’ll find there are no absolutes here. There are generalizations within a certain application that should be adhered to but they may not necessarily apply in other situations. You can see how guessing which one is best is really difficult to get right. Especially when the people doing the guessing don’t even know about the generalization differences.
Back to street exhausts since this is what most here will be concerned with. On the market so far we have many different systems. None have been shown to do a whole lot in the way of power gains over stock although there is some improvement. There are systems anywhere from 2 ½” to 3” in total size. If we added up the total area of the exhaust runners leaving the engine, we’d actually find that when combined they have a total area equal to a single 3 1/8” pipe. Why then do we want a 2 ½” or 3” exhaust? First of all remember that air is in pulses through the exhaust. Just because the port is open does not mean that it is always flowing the same amount of air. In fact the exhaust pulses when moving down the exhaust runner are shaped more like a teardrop in the sense that most of their intensity is at the port opening and then tapers off as the exhaust phases gets farther and farther along. Since we have 2 rotors, we have alternating pulses. Yes we do have pulse overlap but every engine does. It also doesn’t matter in the cat back section if we have 6 pulses from 2 rotors or 6 pulses from 6 pistons (another example don’t worry about where the number came from). It all takes up space and is energy in the pipe. Since all of this energy is not present at any one spot at any one moment in time, we don’t need to total area of all of the exhaust ports combined which is why 2 ½” to 3” works just fine. Personally I’d lean towards the 2 ½” side for naturally aspirated use.
Another reason why we don’t need the total area is that this is the total area for the exhaust runners and not the ports themselves. The runners actually have more area than the actual ports do and the port runners should flow air potentially than the ports can. This means that the total area of each runner isn’t being used to it’s fullest and therefore we don’t need that much total area downstream. Are you starting to see that because of this we might actually pick up some power by making the exhaust runners smaller?! Why though? Air has energy. Faster moving air has more energy as it is more air over a distance per the same amount of time. Why not utilize it? This is what we are trying to do by not getting carried away and going with a large exhaust pipe that can flow the needed air requirements for an entire city but rather optimize it to flow what is needed for the engine.
How does backpressure come into play in all of this though? There are many unfortunately wrong people that think that more backpressure means more low end power and that you need backpressure to have it. Um, no. Wrong. Terribly wrong. First of all backpressure is nothing more than positive pressure in the exhaust. It is entirely possible to have moments of negative pressure in the exhaust. In we always have backpressure at some point in the exhaust. You never have all of only one as there are pulses at work. It all goes back to flowing exactly as much as you need to and nothing more. Remember that if we design an exhaust system that flows exactly what the engine needs to make best power at 9000 rpm, this means that we have compromised every rpm below that as we only need that flow at the max possible flow level. This means that at 3000 rpm’s, we have a pipe that is too large with velocity that is too slow. What’s the cure that a lot of people use? Add a restriction in the system to give it more backpressure. Of course power on the low end picks up but power on the top end falls off. We can no longer flow as much as necessary at peak rpm. Why did this happen if it wasn’t because of backpressure? Backpressure is a rather abstract subject. To understand what pressure is we must first have a frame of reference. On a flow bench, we have to establish a base pressure in the system before we can take a cfm reading on airflow. This is because at different pressure levels, different amounts of air molecules are present for the same area of space. Knowing how much something flows in cfm is a worthless number if you don’t know what pressure it was measured at. An example I like to use is my air compressor out in the garage. I use this to illustrate turbo sizing too. My compressor tank is sitting idle. It has 0 cfm of airflow. However it is sitting at 120 psi of pressure. What if I blow through a drinking straw? I have a measurable amount of airflow but far less pressure. I surely couldn’t blow through a drinking straw at 120 psi of pressure (insert explicit joke here). So basically we need to know pressure and flow. Here’s an interesting fact, too much backpressure can actually cause the earlier onset of detonation! Why? Not enough of the exhausted air gets out of the chamber. It therefore gets carried back into the next combustion cycle, which lowers the oxygen content and raises the temperature. Yes this can happen with no port overlap too!
Last edited by rotarygod; 09-30-2005 at 01:00 AM.
09-29-2005, 11:15 PM
#2
Registered
Thread Starter
Cont...
Moving this all back to backpressure at lower rpm’s, many people will say that you need backpressure at lower rpm’s to make good torque but none up top to make good power. First of all torque and horsepower are mathematically related so you can’t have more or less of one without the other. Next this thought process is inherently flawed, as a law of physics is not going to dictate that 2 things that are opposite apply to the same thing. The real issue is that what we want is only as much flow as we need at any particular rpm. By adding so called backpressure into an exhaust system, we decrease the total flow capability of the system and make the max flow capabilities equal to that required by a lower rpm. Of course anything above this will suffer. Too bad that many will still use a larger exhaust pipe but then cork the end of it with an insert in a muffler or some other archaic device. All this does it to weaken the exhaust pulses within the entire length of the pipe and then speed it back up to max velocity right at this one spot. Why not just use a smaller pipe with this much total airflow ability and allow the pulses to stay strong the entire length? The rest of the world will never know but this forum will! Pat yourselves on the back for suddenly knowing more than the average rice tuner type with their restrictive plug in mufflers. It’s funny that they would probably make more power with the stock exhaust! Remember it’s not backpressure that makes low end power. It’s flowing what is needed for that rpm. In a perfect world, our exhaust pipes would expand in area and length as rpm and load increases. We can simulate some of this but for a total system like this, we’ll have to wait until Startrek technology becomes a reality.
So now how do we tune an exhaust around a certain power band on the RX-8. This is very difficult as most of our exhaust tuning is done by the time it gets to the cat back. The stock exhaust manifold collects the pulses into a certain area where they interact and then move on towards the rear. Obviously anything that we do to the exhaust after this point is only going to have a relatively small impact on what our power band does. We could use a smaller pipe to speed up total airflow at a lower rpm (or for the simple people, add “backpressure”) to gain a little bit of power down low with a sacrifice higher in the power band. We can use a very large pipe to keep max power up high at the slight expense of the bottom end. Don’t expect miracles though.
What really needs to be dealt with is the exhaust manifold area. It has been shown up to now that a header doesn’t really do much more than give us more than 4-5 hp at the very top of the powerband. This isn’t terribly functional and certainly not feelable on the street so most are shying away from this option. The fact that these run anywhere from $500 to well over a thousand dollars just makes it not worth it for what you get. Why have there been issues though? Is it because the engine can’t flow much more air and that we are limited based on this? Is it due to the siamesed center exhaust port that combines the pulse of the front and rear rotors before we ever get a chance to tune it on our own? Is it due to a lack of port overlap? Is it a combination? Am I asking too many questions? Racing Beat has done some homework on this subject and is still continuing to do so. A year ago at the RB open house after Sevenstock, RB showed us 2 different variations on headers that they had tried. One used 3 equal length runners into a collector. The other separated the center ports so the had 4 runners. I’d have though that one would do better than the other but apparently they both provide about the same power gains. Strange. So it’s apparently not anything to do with the Siamesed center port. What about the lack of port overlap? Could that be it? In looking at my pictures that I took of those headers, I noticed that it appears that the pipes they used are way too large. It could be the pictures but I’m pretty sure they used the size based off of the flange and not the actual runner area. The flange is completely off in size. The whole stock exhaust manifold on the car is just plain weird. If those pipes are too large, by virtue of the gas velocity taking longer to reach max velocity speed of 240-260 ft/sec it only makes sense that both headers only show an increase at the top of the power band only. (I’ll explain this number in a minute) This makes perfect sense. Why not try a smaller pipe? Not sure they haven’t but I hope they will.
OK what’s this 240-260 ft/sec max exhaust velocity number I just mentioned? When you size an exhaust pipe, you size it according to a certain amount of airflow at a certain rpm. More specifically you are aiming for a certain velocity at a certain rpm which will determine where you peak torque will be. Note that this is only in relation to where peak torque occurs according to how the exhaust has an effect on it. Peak torque is also affected by the displacement of the engine, intake tuning, etc. This is just one aspect so don’t chastise me thinking this is how you change where your engine makes it’s peak. It does contribute though. As you can see from this number, a larger pipe will need a higher rpm to make peak torque. Even if we keep the same sized exhaust runners out of the engine, we can still affect where our torque peak will be by using a smaller or even larger pipe. It is all based on velocity. The larger the pipe gets, the higher up the peak will be or more specifically the higher in the power band the tuning affect will be at. The larger the pipe, the less effect down low and greater up top. The smaller, the greater down low and less up top. Remember how I said RB used larger header pipes and only got power up high? This is why. They actually need a total exhaust runner pipe area that is even smaller than the area of the exhaust runners at the flange as they flow more than the port windows do. This will give a nicer power gain lower in the power band where it is more usable. How much more? I don’t know. Go try it and let me know but it according to what we know it should work. Please remember that nothing is fact until it is proven but theory is more accurate and closer to the truth than guessing. Therefore this has a greater likelihood of actually working as theory supports it and guessing wouldn’t have even though of it. If it works and you get rich off of it, remember where you heard it. If it fails, it was all your idea!
Let’s take this one step farther. What if we could replace the exhaust sleeves in the housings so that they flowed exactly what the port windows flowed? This would mean higher velocity. Remember higher velocity is a greater pulse that interacts with the opposite rotors pulses and aids in scavenging. How can we scavenge when we have no port overlap? Well, what if we have so much velocity from our wonderfully designed exhaust that the exhaust pulse actually pulls more air from the chamber than just the exhaust gasses? This would mean that it pulled the chamber into the negative pressure zone below ambient pressure. This is definitely not backpressure! What happens if something is at a lower pressure and you open a doorway into it, like say, an intake port? Air speeds up very fast and rushes in easily. Suddenly we just got help on the intake side by pulling more air into the chamber. If we get more air than ambient, we have boost! This is very possible to do and is how the engine can exceed 100% volumetric efficiency at spots. This would just make it better. If we got more air in the engine because the exhaust pulled more air out, that is called scavenging. However we still accomplished it without overlap. It has always been though that you couldn’t scavenge without overlap. Not true. Scavenging during overlap is actually using the exhaust velocity to pull intake charge into the engine while the intake and exhaust are open. Yes this also speeds up airflow into the engine at the beginning of the intake stroke but it also allows some of our fresh intake charge to go right through and out the exhaust. That’s a waste. It also works best within a fairly narrow power band and only at higher rpm’s. The more overlap, the higher in the power band it begins to be an advantage and the worse the performance and emissions at lower rpm’s. If we can get this effect with no overlap, we can do it at all rpm’s, and not suffer from the low speed drivability issues of a large overlap engine. Why aren’t people doing this yet? I’ll tell you why. It seems to go against common way of thinking. I’m basically saying that we can get more power from this engine by going smaller on the entire exhaust system assuming of course we start at the engine. Typical thinking is to optimize what we have rather than make it worse. Remember, if we already flow more than we need, we don’t really need to flow more. We are only getting slightly more power from current exhausts because of the limited effects of the total speed of sound through the system which I’ve already mentioned. It can’t do miracles but it can do something.
How exactly can the exhaust after the collector actually influence the power band though? Many people think that the critical spot in an exhaust is before the collector and that anything after it is unimportant. This is one area where acoustics come into play. I have been talking about the effects of the actual air pulses in the exhaust system up to this point. We also have acoustic waves as well. Sound waves travel through the air with peaks and troughs just like any other wave does. They do not travel like ocean waves though. They are more like the visible waves in a spring. They travel back and forth and not up and down. This is also similar to how the exhaust gasses pulse through the pipes. The difference is that the exhaust pulses travel at various speeds depending on rpm. The acoustic pulses always travel at one speed, the speed of sound. When we tune an exhaust around length, we are tuning it around a resonant acoustic frequency. When we collect header pipes at a certain point, we are tuning around acoustics. We are using the actual pulse movement through the pipes by adjusting the total area but we are utilizing the acoustic waves by varying the length and collection points. Each wave has a positive and negative pressure zone. The key is to time a negative wave to return up the system and arrive at the exhaust port just as the next chamber is opening. This will help it pull gasses out faster. However if we have a spot that it helps at, we also have spots that it can hurt and this is when a positive wave arrives back at the port opening as it opens. This somewhat impedes the flow of the exhaust. Like I said earlier, the perfect exhaust would change in area and length continuously to always allow perfect tuning. Now you know how length and area relate to each type of pressure wave in the system.
When these acoustic waves come to a collector in the system some neat things start to happen. First off this is the point where the waves will decide to reverse direction and travel back up each pipe that collects here. This is the phenomenon that determines collector point and it’s effect I just described. Another neat thing happens though. As we all know noise still gets past this point and out of the exhaust. At the collector, this acoustic pulse completely reverses its phase 180 degrees. This means When a positive (or negative) wave hits the collector, that wave will reverse direction and head back towards the engine while the opposite acoustic pulse will travel through the collector and out towards the tailpipe. The reversing pulses keep bouncing back and forth over and over again in the system and each time loses some intensity. This means that the same pulse will affect more than one spot in the power band and it is usually in exactly half or twice multiples of the main tuning frequency. Whoosh, right over some people’s heads with that one! Trust me though it works that way. A neat thing about this phase reversal in the collector is that when a high pressure wave gets to the collector, it changes into a negative wave and travels outwards down the pipe pulling some charge with it. Of course there are negative effects too that would allow the collector to be just as big of a hindrance to tuning. How can we make a collector work better going outward rather than inward?
Megaphone! Install an expanding collector that starts at one size and tapers to a larger one farther downstream. This is done all the time on racecars. This is also the first half of a 2-stroke expansion chamber that should be more appropriately called and expansion/convergence chamber. What this expansion does is exactly like the horn of a trumpet. It increases the amplitude and strength of the outgoing sound waves. This increases scavenging. However it only works well in one direction. Noise doesn’t like to go backwards into a horn so the bad opposite waves are drastically decreased and their negative effect on the power band effectively lowered. That’s a good thing. This effectively makes the area before and after the collector 2 independent parts that work together in harmony. Otherwise you’d get the negative effects from after the collector working on the area before. On a 2-stroke engine the convergence cone is designed to push some of the pressure back towards the engine and push some of the exhaust back in. I’m not going to get into why but lets just say that is a bad idea for a 4 stroke engine, even the Renesis.
Even if you don’t use a megaphone style merge collector, you can still tune for power band. The primary header tubes typically tune lower the longer they get. As usual there is nothing for free and you hurt power somewhere else. The other way to tune lower is to use smaller diameter pipes to approach the critical max torque exhaust velocity of 240-260 ft./sec. So you have 2 ways to do it, change the pipe size, or change the primary pipe length. This only is in regards to velocity though and you will still have to account for the acoustic aspect in all of this. You can see how carefully balancing length, and diameter is critical for best tuning. After the collector you can also have the same effect happen. Remember that multiple pulses come together in a collector and that their combined pulses equal more energy per area. The same 240-260 ft./sec rule takes effect here too. A larger pipe after the collector will raise the power band regardless of primary runner length and a smaller one will lower it. This too is ignoring the acoustic benefits of the megaphone style collector. Did I mention that the collector area before the megaphone has a part in this too? How can someone guess at all of this? There’s too much to take into consideration. Use theory and math to get close and then tweak it on the road or a dyno. That’s the only way to get it right. Guessing and building will not do it. Don’t be discouraged from trying though.
What about stepped headers? Well they have a similar effect to a megaphone expansion in theory in that they aid in scavenging but making it harder for waves to back up. The total expansion from one pipe size to another isn’t very much though and you do lose some velocity before the collector. It should help the acoustics though. The use of steps in headers has been the subject of much debate however and there is no definitive proof that a stepped header is necessarily any better than one that isn’t. There are just as many stepped headers out there that work and don’t work good as there are conventional ones that do and don’t. I’m not going to get into any math formulas for figuring out header lengths on rotaries vs piston engines.
So how does all of this relate to supercharged engines, turbocharged engines and naturally aspirated engines? By now you should have an idea of what to look for in a total exhaust design for a naturally aspirated vehicle. You should also have a reasonable idea of how to tune them around an intended power band. I’m going to deal with supercharging next as many people think the same exhaust setup applies to them as to naturally aspirated engines. Oh how wrong can people be!
Remember a supercharger is adding air to an engine. Remember also that the critical velocity number in the exhaust is between 240/260 ft./sec. This is determined not only by what rpm the engine is at but also by how much air went into the engine. If you added more air in the form of boost, more air is going to also come out and take up space. This means that a supercharged cars exhaust will reach this critical velocity in the before the naturally aspirated car will even though they both possess the same exhaust and both have no exhaust restriction such as a turbo. The supercharged car will need a larger diameter exhaust for sure. Depending on where you want the power band to be, length may also vary in the header. Like I said, don’t worry about headers for the RX-8, we’ll see these worked on yet and we’ll see improvements from them when everything is all said and done. There is a tradeoff with the larger exhaust as we’ve already learned. That is that low-end power suffers. When you are under part throttle at lower rpm’s and the supercharger isn’t working, you are also not getting the greater scavenging benefits of the smaller exhaust. The good news is that a supercharger is always moving air towards the engine, even if it isn’t boosting it so this will be a fairly negligible effect. Just remember a bypass valve installed in some superchargers is only a pressure equalization port. There is actually very little airflow through it which sounds weird and this results in air always moving towards the engine.
Since most people here are concerned with street use, I’ll deal with how I personally would design an exhaust around the different supercharger options. The exhaust can play a crucial role in the boost curve of a supercharger! For a roots style supercharger, I’d design it to tune for a higher rpm rather than low. This is because a roots is more efficient down low and needs all the help it can get up high. By utilizing the exhaust to optimize performance in the blower’s weak spots, we can provide for a flatter torque curve. Yes it’s true you could potentially get more power down low but at the expense of good top end, which is already a weakness of a roots.
For a centrifugal supercharger, which makes most of its power up high and none down low, I’d do the opposite. I’d tune the exhaust for max performance at a lower rpm and let the blower attempt to compensate on the top end. Yes this will decrease ultimate top end performance but we are talking about a street system here where average power is king and makes for a more drivable car.
For a supercharger such as a twin-screw unit, they seem to do pretty well across the power band. I’d still be inclined to design a system that makes peak power exactly where the car did stock so it just feels like an overall larger engine with the same albeit stronger power band.
The Axial Flow supercharger is supposed to be a sort of cross in the boost curve of a centrifugal and a positive displacement. It should be a linear increase in boost as rpm’s rise. The actual design of the unit can be tweaked to provide greater efficiency at certain spots though so this may not necessarily be true. Not quite sure how I’d do an exhaust for one of these but I’d have to guess that I’d aim for somewhere in the middle when it comes to exhaust tuning. Not too sure yet as we haven’t seen the power band of one yet.
Moving this all back to backpressure at lower rpm’s, many people will say that you need backpressure at lower rpm’s to make good torque but none up top to make good power. First of all torque and horsepower are mathematically related so you can’t have more or less of one without the other. Next this thought process is inherently flawed, as a law of physics is not going to dictate that 2 things that are opposite apply to the same thing. The real issue is that what we want is only as much flow as we need at any particular rpm. By adding so called backpressure into an exhaust system, we decrease the total flow capability of the system and make the max flow capabilities equal to that required by a lower rpm. Of course anything above this will suffer. Too bad that many will still use a larger exhaust pipe but then cork the end of it with an insert in a muffler or some other archaic device. All this does it to weaken the exhaust pulses within the entire length of the pipe and then speed it back up to max velocity right at this one spot. Why not just use a smaller pipe with this much total airflow ability and allow the pulses to stay strong the entire length? The rest of the world will never know but this forum will! Pat yourselves on the back for suddenly knowing more than the average rice tuner type with their restrictive plug in mufflers. It’s funny that they would probably make more power with the stock exhaust! Remember it’s not backpressure that makes low end power. It’s flowing what is needed for that rpm. In a perfect world, our exhaust pipes would expand in area and length as rpm and load increases. We can simulate some of this but for a total system like this, we’ll have to wait until Startrek technology becomes a reality.
So now how do we tune an exhaust around a certain power band on the RX-8. This is very difficult as most of our exhaust tuning is done by the time it gets to the cat back. The stock exhaust manifold collects the pulses into a certain area where they interact and then move on towards the rear. Obviously anything that we do to the exhaust after this point is only going to have a relatively small impact on what our power band does. We could use a smaller pipe to speed up total airflow at a lower rpm (or for the simple people, add “backpressure”) to gain a little bit of power down low with a sacrifice higher in the power band. We can use a very large pipe to keep max power up high at the slight expense of the bottom end. Don’t expect miracles though.
What really needs to be dealt with is the exhaust manifold area. It has been shown up to now that a header doesn’t really do much more than give us more than 4-5 hp at the very top of the powerband. This isn’t terribly functional and certainly not feelable on the street so most are shying away from this option. The fact that these run anywhere from $500 to well over a thousand dollars just makes it not worth it for what you get. Why have there been issues though? Is it because the engine can’t flow much more air and that we are limited based on this? Is it due to the siamesed center exhaust port that combines the pulse of the front and rear rotors before we ever get a chance to tune it on our own? Is it due to a lack of port overlap? Is it a combination? Am I asking too many questions? Racing Beat has done some homework on this subject and is still continuing to do so. A year ago at the RB open house after Sevenstock, RB showed us 2 different variations on headers that they had tried. One used 3 equal length runners into a collector. The other separated the center ports so the had 4 runners. I’d have though that one would do better than the other but apparently they both provide about the same power gains. Strange. So it’s apparently not anything to do with the Siamesed center port. What about the lack of port overlap? Could that be it? In looking at my pictures that I took of those headers, I noticed that it appears that the pipes they used are way too large. It could be the pictures but I’m pretty sure they used the size based off of the flange and not the actual runner area. The flange is completely off in size. The whole stock exhaust manifold on the car is just plain weird. If those pipes are too large, by virtue of the gas velocity taking longer to reach max velocity speed of 240-260 ft/sec it only makes sense that both headers only show an increase at the top of the power band only. (I’ll explain this number in a minute) This makes perfect sense. Why not try a smaller pipe? Not sure they haven’t but I hope they will.
OK what’s this 240-260 ft/sec max exhaust velocity number I just mentioned? When you size an exhaust pipe, you size it according to a certain amount of airflow at a certain rpm. More specifically you are aiming for a certain velocity at a certain rpm which will determine where you peak torque will be. Note that this is only in relation to where peak torque occurs according to how the exhaust has an effect on it. Peak torque is also affected by the displacement of the engine, intake tuning, etc. This is just one aspect so don’t chastise me thinking this is how you change where your engine makes it’s peak. It does contribute though. As you can see from this number, a larger pipe will need a higher rpm to make peak torque. Even if we keep the same sized exhaust runners out of the engine, we can still affect where our torque peak will be by using a smaller or even larger pipe. It is all based on velocity. The larger the pipe gets, the higher up the peak will be or more specifically the higher in the power band the tuning affect will be at. The larger the pipe, the less effect down low and greater up top. The smaller, the greater down low and less up top. Remember how I said RB used larger header pipes and only got power up high? This is why. They actually need a total exhaust runner pipe area that is even smaller than the area of the exhaust runners at the flange as they flow more than the port windows do. This will give a nicer power gain lower in the power band where it is more usable. How much more? I don’t know. Go try it and let me know but it according to what we know it should work. Please remember that nothing is fact until it is proven but theory is more accurate and closer to the truth than guessing. Therefore this has a greater likelihood of actually working as theory supports it and guessing wouldn’t have even though of it. If it works and you get rich off of it, remember where you heard it. If it fails, it was all your idea!
Let’s take this one step farther. What if we could replace the exhaust sleeves in the housings so that they flowed exactly what the port windows flowed? This would mean higher velocity. Remember higher velocity is a greater pulse that interacts with the opposite rotors pulses and aids in scavenging. How can we scavenge when we have no port overlap? Well, what if we have so much velocity from our wonderfully designed exhaust that the exhaust pulse actually pulls more air from the chamber than just the exhaust gasses? This would mean that it pulled the chamber into the negative pressure zone below ambient pressure. This is definitely not backpressure! What happens if something is at a lower pressure and you open a doorway into it, like say, an intake port? Air speeds up very fast and rushes in easily. Suddenly we just got help on the intake side by pulling more air into the chamber. If we get more air than ambient, we have boost! This is very possible to do and is how the engine can exceed 100% volumetric efficiency at spots. This would just make it better. If we got more air in the engine because the exhaust pulled more air out, that is called scavenging. However we still accomplished it without overlap. It has always been though that you couldn’t scavenge without overlap. Not true. Scavenging during overlap is actually using the exhaust velocity to pull intake charge into the engine while the intake and exhaust are open. Yes this also speeds up airflow into the engine at the beginning of the intake stroke but it also allows some of our fresh intake charge to go right through and out the exhaust. That’s a waste. It also works best within a fairly narrow power band and only at higher rpm’s. The more overlap, the higher in the power band it begins to be an advantage and the worse the performance and emissions at lower rpm’s. If we can get this effect with no overlap, we can do it at all rpm’s, and not suffer from the low speed drivability issues of a large overlap engine. Why aren’t people doing this yet? I’ll tell you why. It seems to go against common way of thinking. I’m basically saying that we can get more power from this engine by going smaller on the entire exhaust system assuming of course we start at the engine. Typical thinking is to optimize what we have rather than make it worse. Remember, if we already flow more than we need, we don’t really need to flow more. We are only getting slightly more power from current exhausts because of the limited effects of the total speed of sound through the system which I’ve already mentioned. It can’t do miracles but it can do something.
How exactly can the exhaust after the collector actually influence the power band though? Many people think that the critical spot in an exhaust is before the collector and that anything after it is unimportant. This is one area where acoustics come into play. I have been talking about the effects of the actual air pulses in the exhaust system up to this point. We also have acoustic waves as well. Sound waves travel through the air with peaks and troughs just like any other wave does. They do not travel like ocean waves though. They are more like the visible waves in a spring. They travel back and forth and not up and down. This is also similar to how the exhaust gasses pulse through the pipes. The difference is that the exhaust pulses travel at various speeds depending on rpm. The acoustic pulses always travel at one speed, the speed of sound. When we tune an exhaust around length, we are tuning it around a resonant acoustic frequency. When we collect header pipes at a certain point, we are tuning around acoustics. We are using the actual pulse movement through the pipes by adjusting the total area but we are utilizing the acoustic waves by varying the length and collection points. Each wave has a positive and negative pressure zone. The key is to time a negative wave to return up the system and arrive at the exhaust port just as the next chamber is opening. This will help it pull gasses out faster. However if we have a spot that it helps at, we also have spots that it can hurt and this is when a positive wave arrives back at the port opening as it opens. This somewhat impedes the flow of the exhaust. Like I said earlier, the perfect exhaust would change in area and length continuously to always allow perfect tuning. Now you know how length and area relate to each type of pressure wave in the system.
When these acoustic waves come to a collector in the system some neat things start to happen. First off this is the point where the waves will decide to reverse direction and travel back up each pipe that collects here. This is the phenomenon that determines collector point and it’s effect I just described. Another neat thing happens though. As we all know noise still gets past this point and out of the exhaust. At the collector, this acoustic pulse completely reverses its phase 180 degrees. This means When a positive (or negative) wave hits the collector, that wave will reverse direction and head back towards the engine while the opposite acoustic pulse will travel through the collector and out towards the tailpipe. The reversing pulses keep bouncing back and forth over and over again in the system and each time loses some intensity. This means that the same pulse will affect more than one spot in the power band and it is usually in exactly half or twice multiples of the main tuning frequency. Whoosh, right over some people’s heads with that one! Trust me though it works that way. A neat thing about this phase reversal in the collector is that when a high pressure wave gets to the collector, it changes into a negative wave and travels outwards down the pipe pulling some charge with it. Of course there are negative effects too that would allow the collector to be just as big of a hindrance to tuning. How can we make a collector work better going outward rather than inward?
Megaphone! Install an expanding collector that starts at one size and tapers to a larger one farther downstream. This is done all the time on racecars. This is also the first half of a 2-stroke expansion chamber that should be more appropriately called and expansion/convergence chamber. What this expansion does is exactly like the horn of a trumpet. It increases the amplitude and strength of the outgoing sound waves. This increases scavenging. However it only works well in one direction. Noise doesn’t like to go backwards into a horn so the bad opposite waves are drastically decreased and their negative effect on the power band effectively lowered. That’s a good thing. This effectively makes the area before and after the collector 2 independent parts that work together in harmony. Otherwise you’d get the negative effects from after the collector working on the area before. On a 2-stroke engine the convergence cone is designed to push some of the pressure back towards the engine and push some of the exhaust back in. I’m not going to get into why but lets just say that is a bad idea for a 4 stroke engine, even the Renesis.
Even if you don’t use a megaphone style merge collector, you can still tune for power band. The primary header tubes typically tune lower the longer they get. As usual there is nothing for free and you hurt power somewhere else. The other way to tune lower is to use smaller diameter pipes to approach the critical max torque exhaust velocity of 240-260 ft./sec. So you have 2 ways to do it, change the pipe size, or change the primary pipe length. This only is in regards to velocity though and you will still have to account for the acoustic aspect in all of this. You can see how carefully balancing length, and diameter is critical for best tuning. After the collector you can also have the same effect happen. Remember that multiple pulses come together in a collector and that their combined pulses equal more energy per area. The same 240-260 ft./sec rule takes effect here too. A larger pipe after the collector will raise the power band regardless of primary runner length and a smaller one will lower it. This too is ignoring the acoustic benefits of the megaphone style collector. Did I mention that the collector area before the megaphone has a part in this too? How can someone guess at all of this? There’s too much to take into consideration. Use theory and math to get close and then tweak it on the road or a dyno. That’s the only way to get it right. Guessing and building will not do it. Don’t be discouraged from trying though.
What about stepped headers? Well they have a similar effect to a megaphone expansion in theory in that they aid in scavenging but making it harder for waves to back up. The total expansion from one pipe size to another isn’t very much though and you do lose some velocity before the collector. It should help the acoustics though. The use of steps in headers has been the subject of much debate however and there is no definitive proof that a stepped header is necessarily any better than one that isn’t. There are just as many stepped headers out there that work and don’t work good as there are conventional ones that do and don’t. I’m not going to get into any math formulas for figuring out header lengths on rotaries vs piston engines.
So how does all of this relate to supercharged engines, turbocharged engines and naturally aspirated engines? By now you should have an idea of what to look for in a total exhaust design for a naturally aspirated vehicle. You should also have a reasonable idea of how to tune them around an intended power band. I’m going to deal with supercharging next as many people think the same exhaust setup applies to them as to naturally aspirated engines. Oh how wrong can people be!
Remember a supercharger is adding air to an engine. Remember also that the critical velocity number in the exhaust is between 240/260 ft./sec. This is determined not only by what rpm the engine is at but also by how much air went into the engine. If you added more air in the form of boost, more air is going to also come out and take up space. This means that a supercharged cars exhaust will reach this critical velocity in the before the naturally aspirated car will even though they both possess the same exhaust and both have no exhaust restriction such as a turbo. The supercharged car will need a larger diameter exhaust for sure. Depending on where you want the power band to be, length may also vary in the header. Like I said, don’t worry about headers for the RX-8, we’ll see these worked on yet and we’ll see improvements from them when everything is all said and done. There is a tradeoff with the larger exhaust as we’ve already learned. That is that low-end power suffers. When you are under part throttle at lower rpm’s and the supercharger isn’t working, you are also not getting the greater scavenging benefits of the smaller exhaust. The good news is that a supercharger is always moving air towards the engine, even if it isn’t boosting it so this will be a fairly negligible effect. Just remember a bypass valve installed in some superchargers is only a pressure equalization port. There is actually very little airflow through it which sounds weird and this results in air always moving towards the engine.
Since most people here are concerned with street use, I’ll deal with how I personally would design an exhaust around the different supercharger options. The exhaust can play a crucial role in the boost curve of a supercharger! For a roots style supercharger, I’d design it to tune for a higher rpm rather than low. This is because a roots is more efficient down low and needs all the help it can get up high. By utilizing the exhaust to optimize performance in the blower’s weak spots, we can provide for a flatter torque curve. Yes it’s true you could potentially get more power down low but at the expense of good top end, which is already a weakness of a roots.
For a centrifugal supercharger, which makes most of its power up high and none down low, I’d do the opposite. I’d tune the exhaust for max performance at a lower rpm and let the blower attempt to compensate on the top end. Yes this will decrease ultimate top end performance but we are talking about a street system here where average power is king and makes for a more drivable car.
For a supercharger such as a twin-screw unit, they seem to do pretty well across the power band. I’d still be inclined to design a system that makes peak power exactly where the car did stock so it just feels like an overall larger engine with the same albeit stronger power band.
The Axial Flow supercharger is supposed to be a sort of cross in the boost curve of a centrifugal and a positive displacement. It should be a linear increase in boost as rpm’s rise. The actual design of the unit can be tweaked to provide greater efficiency at certain spots though so this may not necessarily be true. Not quite sure how I’d do an exhaust for one of these but I’d have to guess that I’d aim for somewhere in the middle when it comes to exhaust tuning. Not too sure yet as we haven’t seen the power band of one yet.
09-29-2005, 11:16 PM
#3
Registered
Thread Starter
Cont...
You can see that exhaust isn’t just a generic bolt on for every situation type of product. This may be true in the video games but not in reality. There is so much to think about.
What about exhausts on turbocharged cars? The best exhaust after a turbo is no turbo at all. Any backpressure (positive pressure) in the system after the turbo will slow down its spool rate and create lag. There have been people that have claimed that they had more low-end power and faster spool with a smaller turbo exhaust but what was really happening they failed to understand. The problem was the their engine’s were best tuned for use with the slightly restrictive exhaust. When they put a better flowing one on or none at all, their air/fuel ratios were all messed up and the engine didn’t respond very well. Makes sense to me. This is one reason why I can’t stand people asking for different maps for their ecu’s when a map based system is not tolerant of a different exhaust, intake, turbo, porting, etc. They are very sensitive and you are in danger of blowing your engine up if it isn’t the exact same setup as the car the map is coming from. People never learn. Some have reported bad results with the Racing Beat cat back exhaust when used with the Greddy turbo kit. This is because it was designed around the flow requirements of the stock engine. It isn’t a hindrance to the stock engine but to a turbo, it doesn’t flow enough as a turbo is sending more air through the system. RB did a great job of designing a system that is perfect for it’s intended use and nothing more. In this case, people need to use a larger exhaust that flows better. They also need a corresponding retuning of the ecu to compensate as they risk detonation and blowing their engines up. They map the kit comes with is designed for the stock exhaust system and is not appropriate for any other setup. Hell, it isn’t terribly appropriate for its intended job!
Hopefully I’ve covered all of the bases. If anything I probably have just spawned a whole round of new ones and possibly some disagreements. You should be able to decipher which parts are my own ideas and which ones are proven theory. It’s not too hard to tell. At least I didn’t get into the ram air exhaust collector! I’ll write that one up some other time. I may have forgotten some things so I may add on to this yet!
Thanks for putting up with my ramblings this long.
You can see that exhaust isn’t just a generic bolt on for every situation type of product. This may be true in the video games but not in reality. There is so much to think about.
What about exhausts on turbocharged cars? The best exhaust after a turbo is no turbo at all. Any backpressure (positive pressure) in the system after the turbo will slow down its spool rate and create lag. There have been people that have claimed that they had more low-end power and faster spool with a smaller turbo exhaust but what was really happening they failed to understand. The problem was the their engine’s were best tuned for use with the slightly restrictive exhaust. When they put a better flowing one on or none at all, their air/fuel ratios were all messed up and the engine didn’t respond very well. Makes sense to me. This is one reason why I can’t stand people asking for different maps for their ecu’s when a map based system is not tolerant of a different exhaust, intake, turbo, porting, etc. They are very sensitive and you are in danger of blowing your engine up if it isn’t the exact same setup as the car the map is coming from. People never learn. Some have reported bad results with the Racing Beat cat back exhaust when used with the Greddy turbo kit. This is because it was designed around the flow requirements of the stock engine. It isn’t a hindrance to the stock engine but to a turbo, it doesn’t flow enough as a turbo is sending more air through the system. RB did a great job of designing a system that is perfect for it’s intended use and nothing more. In this case, people need to use a larger exhaust that flows better. They also need a corresponding retuning of the ecu to compensate as they risk detonation and blowing their engines up. They map the kit comes with is designed for the stock exhaust system and is not appropriate for any other setup. Hell, it isn’t terribly appropriate for its intended job!
Hopefully I’ve covered all of the bases. If anything I probably have just spawned a whole round of new ones and possibly some disagreements. You should be able to decipher which parts are my own ideas and which ones are proven theory. It’s not too hard to tell. At least I didn’t get into the ram air exhaust collector! I’ll write that one up some other time. I may have forgotten some things so I may add on to this yet!
Thanks for putting up with my ramblings this long.
Last edited by rotarygod; 09-29-2005 at 11:18 PM.
09-29-2005, 11:51 PM
#4
Kaiten Kenbu Rokuren
Wow, that was a really good read. I can say that I learned at least a few things. What about porting though Fred? It seems it looked very promising about 6-8 months ago and all of a sudden a lot of people lost interest. I thought you guys said that the ports are one of the biggest bottlenecks on the renesis and if we want to really tune the exhaust that is what we need to fix. I remember seeing pictures and the angles of the ports, at least from my understanding are far to extreme and make for a poor flowing exhaust port. Maybe you could elaborate a bit there? Thanks, and thanks for the writeup, good to have the RotaryGod back 
09-30-2005, 12:12 AM
#5
Registered
Thread Starter
The reason I didn't get into porting is because we haven't seen anyone properly tune a ported Renesis. I personally think that we shouldn't worry about making ports any larger but obviously that is an area of differing opinions. At least not on the intake side. Just clean up the intake ports and manifolds. The exhaust side needs alot of attention though. It's not so much the exhaust port size but rather the shape inside the runner and exhaust sleeves that need attention. I don't feel making the exhaust ports much bigger is a good idea and I certainly don't condone cutting into the water jacket to do so. Larger isn't always better but as we have seen there is differing opinion on this. When we make ports bigger we are also changing the port timing. This affects the power band. Why do we want to do that? Why do we want less low end power just to get a few up top? That's not a nice tradeoff on this car. If we could make new exhaust sleeves that kept the flow velocity high, port turbulence low, but port area very close to stock, we would not only have our stock power band but also more power everywhere else in the powerband with nothing suffering anywhere else. That should be the goal. Optimize what you have before you attempt to make it larger. I think we can benefit from internal engine work but it isn't going to be in the same way the older rotaries did it and by that I mean it isn't going to be by just porting it. It'll be more involved than that. Just porting will probably give some gain but I truly believe more power can be had by making certain things smaller, some maybe ever so slightly larger and just doing them overall better. This isn't the same engine we've known for decades with the same port styles. We can't treat it as such. How much can we ultimately get from a naturally aspirated Renesis? Who knows. Someone do it properly and tell us. I don't own an RX-8. If I did I'd try it. I also don't think trying to bridgeport this engine is a good idea either. Alot of potential of this engine lies in the fact that it has no overlap. We obviously don't have a total airflow issue as it already breathes well enough to rev as high as it does. This engine is far and away more powerful than the older 13B and it has done it with no overlap and less emissions. I feel that this engine is another prime example of how going against common thinking is going to work out very well in the end.
Last edited by rotarygod; 09-30-2005 at 12:15 AM.
09-30-2005, 01:23 AM
#7
The Professor
You talk about resonant oscillation in the collecting chambers dealing with both exhaust pules and "sound waves", but I dont really see what sound in the sense that you describe it would have to do with effective tuning of an exhaust system.
I'm sure it would be satisfactory to use a certain frequency, characteristic of what kind of pulse cfm airflow you expect from an engine as its ports expel air, to begin tuning at.
You are right, air is a funny thing... all sound is is pressure waves. So "sound" would simply be equivalent to the pressure waves pulsing out of the engine in this case.
For example. You play a C note on a violin at 50 decibels. The reason it sounds like a C, is because it is only one frequency being played. Now imagine you played an A and a C note both at 50 dB, the human ear could differentiate these two because of different frequency sensitive parts of the Organ of Corti in the Cochlea, but thats getting into physiology... The fact is two frequences are being played and we can distinguish both of them effectively.
Now imagine you played every frequency from 0 to 30,000 Hz at 50 dB, what would it sound like? A bang( a square wave of magnitude 50dB) ! This is because all the frequencies of sound are rushing at you and your brain cannot make sense of all of them, even though they are all there, all the greatest symphonies' notes, all the most beautiful bird songs, are all there, but their mixed in with all the other notes and are indistinguishable to us. Thats what loud chaotic noises are, just heaps of sound(pressure waves) being produced.
So, in short, there is only one type of tuning required for an engine exhaust, that is turning around engine pluses, because if it moves air in any way, it is a form of sound.
It is a good assumption to say that harmonics build in the exhaust header, but in reality, so much and so many different waves are coming out of the engine that any specific frequences that do decide to collect and linger in standing wave form are most likely going to be instantly cancelled by a wave of equal and opposite magnitue coming out of the engine at the same time.
I'm sure it would be satisfactory to use a certain frequency, characteristic of what kind of pulse cfm airflow you expect from an engine as its ports expel air, to begin tuning at.
You are right, air is a funny thing... all sound is is pressure waves. So "sound" would simply be equivalent to the pressure waves pulsing out of the engine in this case.
For example. You play a C note on a violin at 50 decibels. The reason it sounds like a C, is because it is only one frequency being played. Now imagine you played an A and a C note both at 50 dB, the human ear could differentiate these two because of different frequency sensitive parts of the Organ of Corti in the Cochlea, but thats getting into physiology... The fact is two frequences are being played and we can distinguish both of them effectively.
Now imagine you played every frequency from 0 to 30,000 Hz at 50 dB, what would it sound like? A bang( a square wave of magnitude 50dB) ! This is because all the frequencies of sound are rushing at you and your brain cannot make sense of all of them, even though they are all there, all the greatest symphonies' notes, all the most beautiful bird songs, are all there, but their mixed in with all the other notes and are indistinguishable to us. Thats what loud chaotic noises are, just heaps of sound(pressure waves) being produced.
So, in short, there is only one type of tuning required for an engine exhaust, that is turning around engine pluses, because if it moves air in any way, it is a form of sound.
It is a good assumption to say that harmonics build in the exhaust header, but in reality, so much and so many different waves are coming out of the engine that any specific frequences that do decide to collect and linger in standing wave form are most likely going to be instantly cancelled by a wave of equal and opposite magnitue coming out of the engine at the same time.
09-30-2005, 01:45 AM
#8
Registered
Thread Starter
The acoustics aspect is not my own idea. This part is known fact and every car today integrates this into it's engine design in one way or another.
An engine only makes one frequency for any one rpm. Increasing load at this rpm doesn't change the frequency but it does change the amplitude. Ever seen the video of the F1 engine revving on a dyno that was programmed to play a "how dry I am"? A certain length/diameter exhaust pipe will only resonate at one particular frequency as well. Ok it will also resonate at the harmonics. Each harmonic decreases in magnitude as it increases in order. Obviously these sound waves also get shorter in length as rpm rises. This means that at certain spot we will have positive interaction as well as negative interaction. On exhaust systems that do not collect, we are only relying on acoustic tuning principles as the pipes can't collect to scavenge each other. This is called the organ pipe resonance principle. Every different system has it's drawbacks and advantages though. If you have an engine that has many different sounds coming out of it at once, you have an engine that is only going to spin around a few more times before never spinning again! It is true that there will be spots that you will see cancellations and negative effects from acoustic interaction but that's what proper tuning is getting around. The goal is to place the negative aspects in an area of the powerband that you don't care about and the positive effects where you do.
The sound waves in an exhaust system travel at one speed and one speed only. The speed of sound. This never changes regardless of rpm. However the actual exhaust pulses out of the engine are not always moving this fast. All sound waves move at the same speed in the same medium. Their intensity and usefulness changes with rpm though. At low rpm's they are moving much slower than at high rpm's. The key for peak torque at the peak rpm in this respect is between 240-260 ft./sec of actual gas velocity. Sound is still moving much faster than this. You can plainly see that there are in fact 2 different waves in the system. An intake system is actually designed around acoustic tuning and not actual airflow scavenge tuning. This is the principle of the RX-8 intake system. Acoustics to provide a gain. By utilizing this acoustic ramcharging, Mazda even got up to a 2 psi advantage over not using it on the 89-91 RX-7's. Acoustics are a larger gain on the intake side while direct exhaust pulses are a stronger effect on the exhaust but they are both important and very useful none the less. It is more accurate to say that no system can effect where peak torque will occur but rather say they will effect how strong the powerband is before or after the peak torque spot. The usefulness of acoustic tuning is over a narrow rpm range. The usefulness of scavenge tuning is also over a very narrow rpm range. If we design a system that tunes each one at different spots about a thousand or so rpms away from each other, we can noticably increase our average power. Fortunately for us this is already being done on many cars.
An engine only makes one frequency for any one rpm. Increasing load at this rpm doesn't change the frequency but it does change the amplitude. Ever seen the video of the F1 engine revving on a dyno that was programmed to play a "how dry I am"? A certain length/diameter exhaust pipe will only resonate at one particular frequency as well. Ok it will also resonate at the harmonics. Each harmonic decreases in magnitude as it increases in order. Obviously these sound waves also get shorter in length as rpm rises. This means that at certain spot we will have positive interaction as well as negative interaction. On exhaust systems that do not collect, we are only relying on acoustic tuning principles as the pipes can't collect to scavenge each other. This is called the organ pipe resonance principle. Every different system has it's drawbacks and advantages though. If you have an engine that has many different sounds coming out of it at once, you have an engine that is only going to spin around a few more times before never spinning again! It is true that there will be spots that you will see cancellations and negative effects from acoustic interaction but that's what proper tuning is getting around. The goal is to place the negative aspects in an area of the powerband that you don't care about and the positive effects where you do.
The sound waves in an exhaust system travel at one speed and one speed only. The speed of sound. This never changes regardless of rpm. However the actual exhaust pulses out of the engine are not always moving this fast. All sound waves move at the same speed in the same medium. Their intensity and usefulness changes with rpm though. At low rpm's they are moving much slower than at high rpm's. The key for peak torque at the peak rpm in this respect is between 240-260 ft./sec of actual gas velocity. Sound is still moving much faster than this. You can plainly see that there are in fact 2 different waves in the system. An intake system is actually designed around acoustic tuning and not actual airflow scavenge tuning. This is the principle of the RX-8 intake system. Acoustics to provide a gain. By utilizing this acoustic ramcharging, Mazda even got up to a 2 psi advantage over not using it on the 89-91 RX-7's. Acoustics are a larger gain on the intake side while direct exhaust pulses are a stronger effect on the exhaust but they are both important and very useful none the less. It is more accurate to say that no system can effect where peak torque will occur but rather say they will effect how strong the powerband is before or after the peak torque spot. The usefulness of acoustic tuning is over a narrow rpm range. The usefulness of scavenge tuning is also over a very narrow rpm range. If we design a system that tunes each one at different spots about a thousand or so rpms away from each other, we can noticably increase our average power. Fortunately for us this is already being done on many cars.
Last edited by rotarygod; 09-30-2005 at 01:47 AM.
09-30-2005, 10:14 AM
#10
Kaiten Kenbu Rokuren
Originally Posted by rotarygod
The reason I didn't get into porting is because we haven't seen anyone properly tune a ported Renesis. I personally think that we shouldn't worry about making ports any larger but obviously that is an area of differing opinions. At least not on the intake side. Just clean up the intake ports and manifolds. The exhaust side needs alot of attention though. It's not so much the exhaust port size but rather the shape inside the runner and exhaust sleeves that need attention. I don't feel making the exhaust ports much bigger is a good idea and I certainly don't condone cutting into the water jacket to do so. Larger isn't always better but as we have seen there is differing opinion on this. When we make ports bigger we are also changing the port timing. This affects the power band. Why do we want to do that? Why do we want less low end power just to get a few up top? That's not a nice tradeoff on this car. If we could make new exhaust sleeves that kept the flow velocity high, port turbulence low, but port area very close to stock, we would not only have our stock power band but also more power everywhere else in the powerband with nothing suffering anywhere else. That should be the goal. Optimize what you have before you attempt to make it larger. I think we can benefit from internal engine work but it isn't going to be in the same way the older rotaries did it and by that I mean it isn't going to be by just porting it. It'll be more involved than that. Just porting will probably give some gain but I truly believe more power can be had by making certain things smaller, some maybe ever so slightly larger and just doing them overall better. This isn't the same engine we've known for decades with the same port styles. We can't treat it as such. How much can we ultimately get from a naturally aspirated Renesis? Who knows. Someone do it properly and tell us. I don't own an RX-8. If I did I'd try it. I also don't think trying to bridgeport this engine is a good idea either. Alot of potential of this engine lies in the fact that it has no overlap. We obviously don't have a total airflow issue as it already breathes well enough to rev as high as it does. This engine is far and away more powerful than the older 13B and it has done it with no overlap and less emissions. I feel that this engine is another prime example of how going against common thinking is going to work out very well in the end.
That's kinda along the lines of my question, I'm not saying we should make the ports bigger necessarily, just more open. I'm not sure I'm describing it right. From the old pictures, I remember that the exhaust ports had very extreme angles that the gases had to pass through, a big constriction of flow if you will. I also remember you and Ajax and Zoom mentioning you saw them in person and had the same ideas, that they needed to be cleaned up badly. I hope someone goes ahead and does this soon...or just my wishes of you getting your hands on a renesis sometimes soon :p
09-30-2005, 10:56 AM
#11
Go Texas Longhorns!
a couple of points of clarification for several questions I'm sure your asking yourself about this article:
1. Yes RG (Fred) has a girl friend and a job, how I'm not totally sure but I've met her and I have a business card.
2. Fred actually builds models of exahust systems out of plastic buckets and saranwrap to test his theories
3. Fred was asked to speak at 7stock on the finer points of flow testing but had to decline as he was chairing a 12 step help group for gearheads
4. Yes we are working on getting Fred his own blog to save memory on the forum. Several moderators were concerned we might need an extra server soon to hold his posts.
That is all :D
1. Yes RG (Fred) has a girl friend and a job, how I'm not totally sure but I've met her and I have a business card.
2. Fred actually builds models of exahust systems out of plastic buckets and saranwrap to test his theories
3. Fred was asked to speak at 7stock on the finer points of flow testing but had to decline as he was chairing a 12 step help group for gearheads
4. Yes we are working on getting Fred his own blog to save memory on the forum. Several moderators were concerned we might need an extra server soon to hold his posts.
That is all :D
09-30-2005, 12:10 PM
#13
I WAS BEES
Join Date: Jun 2004
Location: San Diego
Posts: 279
Likes: 0
Received 0 Likes
on
0 Posts
Good read. It also got me thinking, which is dangerous. Seeing as how the last thing I read last night was about the Axial Flow Super charger and oil distribution. Things were running through my head all night about how to work it. *shake* anyway back on topic.
My idea was, well feel free to slap me or whatever because of lack of knowledge.
What if someone were to put a belt driven vacuum on the exhaust (I was going to say supercharger, but we wouldn't want to compress anything further at that point would we, maybe a supercharger put on backwards?.. *shrug*), to suck the air out and create the negative pressure that you mentioned. Which would be a good thing. Right?
Also instead of a star trek dynamic expanding pipe technology, what if there was an iris valve that opened wider when the rpm's went up, or dynamically based on whatever matches your setup? I guess they do something similar on fighter jets don't they? I've seen the exhausts that expand and contact in the navy videos.
Anyway I'm rambling, I don't even know if that all makes any sense.
My idea was, well feel free to slap me or whatever because of lack of knowledge.
What if someone were to put a belt driven vacuum on the exhaust (I was going to say supercharger, but we wouldn't want to compress anything further at that point would we, maybe a supercharger put on backwards?.. *shrug*), to suck the air out and create the negative pressure that you mentioned. Which would be a good thing. Right?
Also instead of a star trek dynamic expanding pipe technology, what if there was an iris valve that opened wider when the rpm's went up, or dynamically based on whatever matches your setup? I guess they do something similar on fighter jets don't they? I've seen the exhausts that expand and contact in the navy videos.
Anyway I'm rambling, I don't even know if that all makes any sense.
09-30-2005, 01:17 PM
#14
Registered
Thread Starter
You are referring to what would be known as negative supercharging. This takes place anytime you have a lower pressure in the exhaust than you do ambient. Look up negative supercharging online. There have been some guys in Australia I believe that have made this work in a more extreme way than I have described but they also suffer from tradeoffs such as not a terribly high rpm range. They are using exhaust velocity to do this but they are also drastically using a smaller exhaust pipe than the area of the exhaust ports. While I believe the idea is sound, I personally don't want to go to this extreme as I don't want to suffer in other areas. I just want to maximize what we have and hopefully reach a compromise that will give us good low end and top end power. getting too greedy on either end will cause a loss in the other. Using some of the principles is a good idea though and should work well. The biggest issue with putting an actual supercharger on the exhaust is heat. Getting it to survive could be an issue. Not saying it is impossible. It "should" work. Also curious if the parasitic losses of the supercharger would cancel out the benefits.
You could do a butterfly valve in the exhaust to in essence simulate a different sized pipe. This is already done on some Ferrari's and even the new Z-06 Corvette except those all use a butterfly valve to open an additional exhaust tip. The concept does work to an extent. I wouldn't use an iris style valve. Not only due to the turbulence through it but because of tolerances. There was a person who tried one on a car (forget where) and he found that the air pressure put just enough suction on the iris that it would bind when trying to open and close. An iris looks neat but like a slide valve is really only advantageous when fully open.
Fighter jet exhaust cones do change size to change and focus the exhaust velocity. Actually the outlet to the cone changes size. The inlet doesn't. That would be a neat idea but I'm not quite sure how much effort it would be to design and build one. It doesn't look easy.
You could do a butterfly valve in the exhaust to in essence simulate a different sized pipe. This is already done on some Ferrari's and even the new Z-06 Corvette except those all use a butterfly valve to open an additional exhaust tip. The concept does work to an extent. I wouldn't use an iris style valve. Not only due to the turbulence through it but because of tolerances. There was a person who tried one on a car (forget where) and he found that the air pressure put just enough suction on the iris that it would bind when trying to open and close. An iris looks neat but like a slide valve is really only advantageous when fully open.
Fighter jet exhaust cones do change size to change and focus the exhaust velocity. Actually the outlet to the cone changes size. The inlet doesn't. That would be a neat idea but I'm not quite sure how much effort it would be to design and build one. It doesn't look easy.
D:D
09-30-2005, 03:11 PM
#16
Registered User
Join Date: Jun 2002
Location: Misinformation Director - Evolv Chicago
Posts: 3,086
Likes: 0
Received 0 Likes
on
0 Posts
Holy crap...it just kept going and going and going and going...as you said, RG, haven't seen you post one of those in a long time!!!
09-30-2005, 03:46 PM
#17
Registered
Thread Starter
Originally Posted by brillo
2. Fred actually builds models of exahust systems out of plastic buckets and saranwrap to test his theories
09-30-2005, 07:39 PM
#19
The Professor
Originally Posted by rotarygod
The acoustics aspect is not my own idea. This part is known fact and every car today integrates this into it's engine design in one way or another.
An engine only makes one frequency for any one rpm. Increasing load at this rpm doesn't change the frequency but it does change the amplitude. Ever seen the video of the F1 engine revving on a dyno that was programmed to play a "how dry I am"? A certain length/diameter exhaust pipe will only resonate at one particular frequency as well. Ok it will also resonate at the harmonics. Each harmonic decreases in magnitude as it increases in order. Obviously these sound waves also get shorter in length as rpm rises. This means that at certain spot we will have positive interaction as well as negative interaction. On exhaust systems that do not collect, we are only relying on acoustic tuning principles as the pipes can't collect to scavenge each other. This is called the organ pipe resonance principle. Every different system has it's drawbacks and advantages though. If you have an engine that has many different sounds coming out of it at once, you have an engine that is only going to spin around a few more times before never spinning again! It is true that there will be spots that you will see cancellations and negative effects from acoustic interaction but that's what proper tuning is getting around. The goal is to place the negative aspects in an area of the powerband that you don't care about and the positive effects where you do.
The sound waves in an exhaust system travel at one speed and one speed only. The speed of sound. This never changes regardless of rpm. However the actual exhaust pulses out of the engine are not always moving this fast. All sound waves move at the same speed in the same medium. Their intensity and usefulness changes with rpm though. At low rpm's they are moving much slower than at high rpm's. The key for peak torque at the peak rpm in this respect is between 240-260 ft./sec of actual gas velocity. Sound is still moving much faster than this. You can plainly see that there are in fact 2 different waves in the system. An intake system is actually designed around acoustic tuning and not actual airflow scavenge tuning. This is the principle of the RX-8 intake system. Acoustics to provide a gain. By utilizing this acoustic ramcharging, Mazda even got up to a 2 psi advantage over not using it on the 89-91 RX-7's. Acoustics are a larger gain on the intake side while direct exhaust pulses are a stronger effect on the exhaust but they are both important and very useful none the less. It is more accurate to say that no system can effect where peak torque will occur but rather say they will effect how strong the powerband is before or after the peak torque spot. The usefulness of acoustic tuning is over a narrow rpm range. The usefulness of scavenge tuning is also over a very narrow rpm range. If we design a system that tunes each one at different spots about a thousand or so rpms away from each other, we can noticably increase our average power. Fortunately for us this is already being done on many cars.
An engine only makes one frequency for any one rpm. Increasing load at this rpm doesn't change the frequency but it does change the amplitude. Ever seen the video of the F1 engine revving on a dyno that was programmed to play a "how dry I am"? A certain length/diameter exhaust pipe will only resonate at one particular frequency as well. Ok it will also resonate at the harmonics. Each harmonic decreases in magnitude as it increases in order. Obviously these sound waves also get shorter in length as rpm rises. This means that at certain spot we will have positive interaction as well as negative interaction. On exhaust systems that do not collect, we are only relying on acoustic tuning principles as the pipes can't collect to scavenge each other. This is called the organ pipe resonance principle. Every different system has it's drawbacks and advantages though. If you have an engine that has many different sounds coming out of it at once, you have an engine that is only going to spin around a few more times before never spinning again! It is true that there will be spots that you will see cancellations and negative effects from acoustic interaction but that's what proper tuning is getting around. The goal is to place the negative aspects in an area of the powerband that you don't care about and the positive effects where you do.
The sound waves in an exhaust system travel at one speed and one speed only. The speed of sound. This never changes regardless of rpm. However the actual exhaust pulses out of the engine are not always moving this fast. All sound waves move at the same speed in the same medium. Their intensity and usefulness changes with rpm though. At low rpm's they are moving much slower than at high rpm's. The key for peak torque at the peak rpm in this respect is between 240-260 ft./sec of actual gas velocity. Sound is still moving much faster than this. You can plainly see that there are in fact 2 different waves in the system. An intake system is actually designed around acoustic tuning and not actual airflow scavenge tuning. This is the principle of the RX-8 intake system. Acoustics to provide a gain. By utilizing this acoustic ramcharging, Mazda even got up to a 2 psi advantage over not using it on the 89-91 RX-7's. Acoustics are a larger gain on the intake side while direct exhaust pulses are a stronger effect on the exhaust but they are both important and very useful none the less. It is more accurate to say that no system can effect where peak torque will occur but rather say they will effect how strong the powerband is before or after the peak torque spot. The usefulness of acoustic tuning is over a narrow rpm range. The usefulness of scavenge tuning is also over a very narrow rpm range. If we design a system that tunes each one at different spots about a thousand or so rpms away from each other, we can noticably increase our average power. Fortunately for us this is already being done on many cars.
It wouldn't be easy though. :D :D
My last post was a bit rash as I was tired and I had incorrectly assesed something that was highly difficult as something that was impossible. Apologies for my arrogant tone.
Last edited by staticlag; 09-30-2005 at 07:44 PM.
10-03-2005, 09:14 AM
#20
Registered
Join Date: Jan 2005
Posts: 615
Likes: 0
Received 0 Likes
on
0 Posts
A lot of this information can be found in the old-school book: http://www.amazon.com/exec/obidos/tg...books&n=507846
I don't think its any secret that smaller ports can increase flow velocity, leading to bigger power, and/or better power curves. The guy who writes Mototune has been saying that for a while: http://www.mototuneusa.com/
I would think the best way to develop an exhaust would be to use pressure sensors at various places in the exhaust, and datalog the pressure versus degrees of rotation of the engine. You could do the same with the intake. This should show reversion and the exhaust pulses.
The real question is: why can't we see any gains from headers? And I think the answer is the lack of overlap. Sure, you can achieve some form of scavenging thru harmonics, but I think the problem is that without overlap, the exhaust flow stalls quite close to the exhaust ports. If there was overlap, there would be more velocity to get the flow farther down the header before the pulse encounters reversion.
So the idea of a narrow-diameter header makes sense. The crucial variable is determining at what point in the travel of the exhaust, at a particular rpm, the reversion is occuring. The idea would be to either step the header to a larger diameter at that point, or place the collector at that point, creating a lower pressure area that should help minimize the force of reversion.
I still think there is something to the cross-flow of the exhaust in the rotor. The force of the exhaust going out two headers leads to contention in flow inside the rotor. Therefore, I think that the center runners should NOT be the same length as the outside runners. Even if the center runners were achieving equal flow and velocity as the outside runners (which I doubt), there would still be contention as to which way the air will go. Basically its a tug-of-war. I'd be interested to see if there are benefits to blocking off the center runners.
I don't think its any secret that smaller ports can increase flow velocity, leading to bigger power, and/or better power curves. The guy who writes Mototune has been saying that for a while: http://www.mototuneusa.com/
I would think the best way to develop an exhaust would be to use pressure sensors at various places in the exhaust, and datalog the pressure versus degrees of rotation of the engine. You could do the same with the intake. This should show reversion and the exhaust pulses.
The real question is: why can't we see any gains from headers? And I think the answer is the lack of overlap. Sure, you can achieve some form of scavenging thru harmonics, but I think the problem is that without overlap, the exhaust flow stalls quite close to the exhaust ports. If there was overlap, there would be more velocity to get the flow farther down the header before the pulse encounters reversion.
So the idea of a narrow-diameter header makes sense. The crucial variable is determining at what point in the travel of the exhaust, at a particular rpm, the reversion is occuring. The idea would be to either step the header to a larger diameter at that point, or place the collector at that point, creating a lower pressure area that should help minimize the force of reversion.
I still think there is something to the cross-flow of the exhaust in the rotor. The force of the exhaust going out two headers leads to contention in flow inside the rotor. Therefore, I think that the center runners should NOT be the same length as the outside runners. Even if the center runners were achieving equal flow and velocity as the outside runners (which I doubt), there would still be contention as to which way the air will go. Basically its a tug-of-war. I'd be interested to see if there are benefits to blocking off the center runners.
10-03-2005, 02:36 PM
#21
Registered
Join Date: Aug 2004
Posts: 664
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by rg
When you size an exhaust pipe, you size it according to a certain amount of airflow at a certain rpm.
Since the Renesis has a relatively linear torque curve, you can assume that the BMEP is constant over a wide rpm range. This means that the pressure in the expansion chamber is about the same when it opens the exhaust port regardless of the engine speed.
The rotor speed is relatively low compared to the velocity that the gases reach when the exhaust port opens (at a pressure of approximately 60 psi), so the changing rotor speed shouldn't greatly affect exhaust velocity. Therefore different rpms shouldn't really affect the airflow per pulse just the total airflow.
As the rpm and with it pulse frequency goes up I see why you ideally would adapt the length of the exhaust pipe (with increasing rpm), but why would you want to change the diameter as well, considering the fact that the airflow per pulse stays approximately the same?
10-03-2005, 02:46 PM
#22
Registered
Join Date: Jan 2005
Posts: 615
Likes: 0
Received 0 Likes
on
0 Posts
Airflow per pulse is not the same per RPM's (at least I'm pretty sure its not). Think about it: if the airflow was the same, it would mean that the airflow going into the engine is also the same per combustion event. If that was true, there would be no point to the additional ports that open at higher rpms.
The vacuum created in the intake gets stronger at higher rpms, thus creating a charging effect that pressurizes the chamber to more than ambient. At higher rpms, the velocity is greater, thus increasing this effect. Everything I've read supports the fact that airflow per pulse is quite dependent on rpm.
The vacuum created in the intake gets stronger at higher rpms, thus creating a charging effect that pressurizes the chamber to more than ambient. At higher rpms, the velocity is greater, thus increasing this effect. Everything I've read supports the fact that airflow per pulse is quite dependent on rpm.
10-03-2005, 03:08 PM
#23
Registered
Join Date: Aug 2004
Posts: 664
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by tuy
Think about it: if the airflow was the same, it would mean that the airflow going into the engine is also the same per combustion event. If that was true, there would be no point to the additional ports that open at higher rpms.
The reason why I believe the airflow per pulse is the same is because the torque curve is pretty linear. If the torque curve is linear that means BMEP stays roughly the same and if BMEP stays roughly the same the amount of air that is being combusted is roughly the same and therefore the pressure before the exhaust port opens should roughly stay the same as well.
Originally Posted by tuj
The vacuum created in the intake gets stronger at higher rpms, thus creating a charging effect that pressurizes the chamber to more than ambient. At higher rpms, the velocity is greater, thus increasing this effect. Everything I've read supports the fact that airflow per pulse is quite dependent on rpm.
Last edited by globi; 10-03-2005 at 03:22 PM.
10-03-2005, 03:45 PM
#24
Registered
Join Date: Jan 2005
Posts: 615
Likes: 0
Received 0 Likes
on
0 Posts
Sorry, but I don't agree. You have claimed you believe the airflow per pulse to be the same, regardless of rpm. However, I claim that more air enters the engine per combustion event (pulse) at higher rpms. If more air is entering the engine, and there is zero overlap, more air must be exiting the engine, meaning that the airflow per pulse must increase as rpms increase.
The only way the airflow per pulse would be constant across the powerband is if the amount of air entering the engine (per pulse) is constant. I don't think that's the case, although I'd be interested if you can show it is.
The only way the airflow per pulse would be constant across the powerband is if the amount of air entering the engine (per pulse) is constant. I don't think that's the case, although I'd be interested if you can show it is.
10-03-2005, 05:15 PM
#25
Registered User
Join Date: Oct 2005
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
Globi, I understand what you are saying, I think the diameter might want to change because of the factor -- time. As rpms increase more pulses per second are getting sent down the tube. Even if the pulses are the same size, there is a lot more of them now at high rpms, which is going to test the limits of the tube's size, and probably pressurize it or something 
If this wasn't the case I think we'd all be running some smallll tubes if they could handle high rpms.
If this wasn't the case I think we'd all be running some smallll tubes if they could handle high rpms.