Alas, I found a topic that didn't have a single reference in this forum. I could hardly believe it. So how do you tune these cacophony of waves caused by a step header? I have Larry Muaex's pipe max software, but whats the real story? All lectures, references, and tirades welcome.
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Stepped Header Design
- Thread starter TravisR
- Start date
BrianPetersen
Mechanical
What are you trying to optimize?
A proper engine simulation program is one way to do it. I've heard about but never used Pipemax. Even with engine simulation, it can be tricky to see what the effect of each pulse reflection is - the pressure waves get superimposed so that it's hard to see what's what.
Cut and try is the other option ... first in engine simulation software, then on the engine.
A proper engine simulation program is one way to do it. I've heard about but never used Pipemax. Even with engine simulation, it can be tricky to see what the effect of each pulse reflection is - the pressure waves get superimposed so that it's hard to see what's what.
Cut and try is the other option ... first in engine simulation software, then on the engine.
- Thread starter
- #3
I have engine simulation software, but its really questionable. Its the performance engine analyzer by performance trends. It sometimes acts very oddly, and often states the pumping loss as negative (as in negative negative or positive) So I'm very skeptical to its use.
It got me thinking as to what the mathematical approach to doing something that complicated was. In my mind I know exactly what it looks like, but I have no idea how to solve it even holding rpm fixed.
It seems as if larry's method is to make the total length fo the step header approxiametely the same length as the tuned primary, but add steps so that there are reversion waves every so often which accent other rpm ranges. I have no idea how it arrives at the crossectional steps. Perhaps they are setup just to be sure there is a reflected wave? Then at some point velocity gets very low, so great changes in crossection are needed to create a meaningful pressure wave at the valve. At the same time I would imagine the overall effect of the header is less.
Really I'm just looking for some good literature on it. I have a car to apply it to, its a small displacement 4 cylinder, but I'm more interested in a general approach to any vehicle.
It got me thinking as to what the mathematical approach to doing something that complicated was. In my mind I know exactly what it looks like, but I have no idea how to solve it even holding rpm fixed.
It seems as if larry's method is to make the total length fo the step header approxiametely the same length as the tuned primary, but add steps so that there are reversion waves every so often which accent other rpm ranges. I have no idea how it arrives at the crossectional steps. Perhaps they are setup just to be sure there is a reflected wave? Then at some point velocity gets very low, so great changes in crossection are needed to create a meaningful pressure wave at the valve. At the same time I would imagine the overall effect of the header is less.
Really I'm just looking for some good literature on it. I have a car to apply it to, its a small displacement 4 cylinder, but I'm more interested in a general approach to any vehicle.
hi, if the working principle is the same of 2 stroke expansion chambers (at least the difuser part) this book covers the design process. with some adaptation of the formulas and testing, it may be helpfull.
IceStationZebra
Mechanical
Here is a good forum to check out.
As far as performance goes, most of the "evidence" I have heard of usually shows only small gains over a properly tuned single diameter system of the same design (i.e. tri-Y, 4-into-1, etc.). I believe that many of the tests showing "improvement" can actually be traced to non-optimized merge collectors.
Andy Dunn of Popular Hot Rodding tested 7 sets of headers on one 408 cubic inch engine in 2004. The difference between all 7 was 0.5% in torque and hp over the whole 3000-7000 rpm band. A set of stepped headers was worth only 0.2% compared to a single ID set from the same manufacturer.
ISZ
As far as performance goes, most of the "evidence" I have heard of usually shows only small gains over a properly tuned single diameter system of the same design (i.e. tri-Y, 4-into-1, etc.). I believe that many of the tests showing "improvement" can actually be traced to non-optimized merge collectors.
Andy Dunn of Popular Hot Rodding tested 7 sets of headers on one 408 cubic inch engine in 2004. The difference between all 7 was 0.5% in torque and hp over the whole 3000-7000 rpm band. A set of stepped headers was worth only 0.2% compared to a single ID set from the same manufacturer.
ISZ
Lou Scannon
Automotive
ISZ, I'm with you on the debunking angle. Did the Popular Hot Rodding test happen to include data to indicate whether the differences they documented were statistically significant?
Gordon Blair's, "Design and Simulation of Four Stroke Engines" covers the topic at length including stepped primaries. Going from memory here, but the problem is that the speed of the wave is affected by the pressure, so that waves of higher pressure travel faster down the pipe than lower pressure pulses. Then the speed is affected by superposition with other pulses and local gas velocity. The waves also locally speed up or slow down the gas medium depending on their pressure. A positive pressure wave meeting a negative pressure wave locally increase the gas velocity(possibly to sonic) even though through superposition, their combined pressure might read zero. The stepped header causes the negative wave decrease pressure as it hits progressively smaller cross sections and that's where the benefit comes from the stepped headers. If I remember correctly, Blair derives his equations, then gives coefficients for them and suggests you test the equations on your actual engine. Then the error is supposed to be corrected through the coefficients. Back in my FSAE days, we ended up using the cut and test method with the best success.
Al
Al
IceStationZebra
Mechanical
"Did the Popular Hot Rodding test happen to include data to indicate whether the differences they documented were statistically significant?"
No they didn't. (Do they ever!?) But seeing only a 0.5% difference between 7 different sets of headers over the 3000-7000 rpm range has to make one wonder - though granted they were all picked to match the engine under test. In the end it is all pressure/frequency tuning and a header may hurt you in one spot but help in another.
I was also just thinking that real in-vehicle testing would probably lead to more differentiation as other factors such as gearing, etc. would tend to emphasize a certain portion of the RPM band.
ISZ
No they didn't. (Do they ever!?) But seeing only a 0.5% difference between 7 different sets of headers over the 3000-7000 rpm range has to make one wonder - though granted they were all picked to match the engine under test. In the end it is all pressure/frequency tuning and a header may hurt you in one spot but help in another.
I was also just thinking that real in-vehicle testing would probably lead to more differentiation as other factors such as gearing, etc. would tend to emphasize a certain portion of the RPM band.
ISZ
![[bigsmile]](/data/assets/smilies/bigsmile.gif "[bigsmile] [bigsmile]")
One of the biggest reasons to use stepped primaries has little to do with pulse wave formation. It has more to do with gas density.
As the exhaust gas cools while moving away from the cylinder head it becomes more dense. A larger pipe cross section allows the "slug" of exhaust gas to keep moving freely and maintain the low pressure area behind it.
The over all length of the stepped primary will time the pulse wave return to cylinder accordingly.
As the exhaust gas cools while moving away from the cylinder head it becomes more dense. A larger pipe cross section allows the "slug" of exhaust gas to keep moving freely and maintain the low pressure area behind it.
The over all length of the stepped primary will time the pulse wave return to cylinder accordingly.
..."stepped' headers are surrogates for "constantly-reducing diameter" pipes (almost impossible to manufacture) whose purpose is to maintain constant gas mass VELOCITY down the pipes as the gas temperature (and thus density) reduces with distance and time.
...as I recall, the idea gas mass velocity was about 300 fps, is that still correct? Faster and pipe-wall scrubbing produces more friction/flow resistance; slower and inertial falls off.
...as I recall, the idea gas mass velocity was about 300 fps, is that still correct? Faster and pipe-wall scrubbing produces more friction/flow resistance; slower and inertial falls off.
- Thread starter
- #14
I've heard both theories, in CFD the reflected waves were very small, nearly negligable. If they were significant from the stepping you would imagine the engine dyno readings would exhibit some crazy outputs as the waves moved from constructive to destructive, and back again.
larrylcoyle
Automotive
To the best of my knowledge, testing step headers in the RPM range of 3000 to 7000 is/would be a total waste of time. When this technology first came out it was primarily designed for people running in the Competition Eliminator classes, most of which rarely see the RPM range you are talking about. Even back in the '70's when we ran a lot of de-stroked 283ci cranks to give us a total of 277 cubic inches, we would launch the cars at somewhere between 9500 and 10,500 RPM and the motors would only see 7000rpm if we missed a shift or slowing down at the end of the run.
IMO, testing a set of these headers on a 408ci at the rpm range stated would be nothing but a waste of time and a exercise in futility. Did I read someone state something about constant reducing the diameter of the pipes? Well, that is going the wrong way friend. You start with a small (relatively so) diameter pipe and it is constantly increased to the intended size is reached. The real people to discuss this with is the header manufacturer.
Larry
IMO, testing a set of these headers on a 408ci at the rpm range stated would be nothing but a waste of time and a exercise in futility. Did I read someone state something about constant reducing the diameter of the pipes? Well, that is going the wrong way friend. You start with a small (relatively so) diameter pipe and it is constantly increased to the intended size is reached. The real people to discuss this with is the header manufacturer.
Larry
IceStationZebra
Mechanical
...Well, I guess it doesn't pay to put a new cam in my motor since it doesn't go over 6500rpm like the comp class motors and it's bigger than 300ci. Because obviously there is a different set of fluid dynamics laws at work above that threshold....;-)
ISZ
ISZ
larrylcoyle
Automotive
Hey ISZ.. we talking camshafts or step headers.. lol It all boils down to the application and the combination as you well know.
Larry
Larry
- Status
Similar threads
