Exhaust Header theory

[deanis]

The shorter the primaries are, the higher the RPM at which they are efficient. And vice versa.

you must be smoking drugs. the LONGER the primaries on an exaust manifold the higher the rpm they are designed for.

perhaps you are confused with intake runner lengths where the shorter the length the higher the rpm?

[fivebob]

The shorter the primaries are, the higher the RPM at which they are efficient. And vice versa.

you must be smoking drugs. the LONGER the primaries on an exaust manifold the higher the rpm they are designed for.

perhaps you are confused with intake runner lengths where the shorter the length the higher the rpm?

Um, perhaps you might care to put that pipe down yourself.

Exhaust theory, as I understand it, is that pipe diameter determines the RPM that peak torque occurs, and length determines the characteristics of the power curve around that point. Shorter primaries improve power above the point of peak torque (at the expense of power below that point) while longer primaries have the opposite effect. Given that exhaust tuning is based on wave theory this make a lot more sense than it being the other way around.

If there is an ideal tuned length that results is the reflected wave arriving at the exhaust valve at the most beneficial time, then, at higher rpm (and therefore higher frequency), the reflected wave has less time to do this. As the wave is travelling at the speed of sound, then, in order to arrive at the right time, it must travel a shorter distance. Therefore the pipes must be shorter to do this.

Perhaps you would care to post an authoritative source that proves your assertions to be true.

[Titties]

lol ginge, boo ern

also just on flow theories...

As far as i know, the intake systems works on sonic pulses caused by the rush of air entering the engine, then hitting a closed valve which pulses back up the runner, off the back of the plenum wall and into the runners again. When the next intake valve opens it's effectively forcing the air into the cylinder.
Thats why short intake runners are good for high rpm...

Now with the exhaust side i'm not 100%.
But i'm pretty sure it doesn't work on waves being reflected.
It works on scavenging, which is pressure differentials.
for a 4-1 system, when an exhaust pulse enters and passes the merge collector, it causes a low pressure in the other 3 runners. If the exhaust valve for the next cylinder opens just as this low pressure is created, it will effectively "suck" (scavenge) the next pulse out, resulting in a much greater flow, and hence more power...
I'm just not sure how Length effects it.
You have to consider that at higher rpm the gas velocity will be greater, and pipe diameter will play a big part in it.
I've read that you should get the air as close to the speed of sound as possible, maybe that was for intakes though... i can't remember.

[deanis]

i dont do theories.

i only look at dyno results. and every header that makes top end power will always have longer primaries than stock.

think about every race header you see, how many have short primaries.

using my car as an example. on a b18c. the primaries are over a meter long and go right down under the engine and under the fire wall.

can you show me a car where they have shortened the primaries and made power?

[Mr Revhead]

deanis. i think you should reread and have a THINK about this:

If there is an ideal tuned length that results is the reflected wave arriving at the exhaust valve at the most beneficial time, then, at higher rpm (and therefore higher frequency), the reflected wave has less time to do this. As the wave is travelling at the speed of sound, then, in order to arrive at the right time, it must travel a shorter distance. Therefore the pipes must be shorter to do this.

[deanis]

i know it sounds good and it makes sence but it doesnt work on a dyno.

[deanis]

unlesssssssssssss

when longer primaries are used they are making the tuned length shorter and then doubling it so there are 2 pulses in the pipe at once.

that explains the theory and the actual readouts on the dyno.

[Mr Revhead]

ok, so its good to have the pulse reversing away from the valve when it opens.
to do so the pulse moves down the pipe
reflects
moves back to the valve
reflects, moves away NOW is the optimum time for the valve to open.

that takes time

at higher rpm the time available is less.
so, to make it happen at say 6000rpm rather than 4000.
the distance traveled needs to be shorter.

correct?

[deanis]

yes i agree with the theory.

a quick look at my headers and a rough measure against factory and other headers that are used for circuit racing (mid range power) and my headers are looking roughly three times the length of the primaries.

pretty much all honda headers that are used for high rpm drag applications use similar lengths to me.

[fivebob]

i dont do theories.

So you don't understand why something makes more power, only that it does. Sounds like an expensive time consuming way to do things

i only look at dyno results. and every header that makes top end power will always have longer primaries than stock.

And you equate that to longer being better for top end power?

That's a classic case of incorrect interpretation of the results arising from failure to understand the theories involved.

I never said short will always produce more power at the top end, it's quite possible that the length is such that the primary frequency they will work at best is outside of the rpm range of the engine, and they work on 1/2 or even 1/4 wavelengths. These are are much weaker because the pressure wave has to travel up and back down the pipe twice (or four times for 1/4 wavelength) each time it does so it get's weaker. The primary reason for stock lengths is more for packaging than performance.

Perhaps I should state it more clearly. When tuned for full wavelength then a longer pipe will produce more power below and a shorter one will produce more power above peak torque rpm point.

think about every race header you see, how many have short primaries.

using my car as an example. on a b18c. the primaries are over a meter long and go right down under the engine and under the fire wall.

can you show me a car where they have shortened the primaries and made power?

Quite a few exist, F1 engines for example. How long are those primaries?

It's not just theory, it's fact, a fact that most exhaust system manufacturers know and use

FYI the tune length of a Header is calculated by the formula Length(mm) = [(21590 x ExhaustDuration) / RPM] - 75. Where Exhaust duration is 180 + the the number of degrees the cam is open before BDC. N.B. This is the tuned length of the header so in the case of a 4-2-1 it's the length of the primaries + secondaries.

Now from this it can be seen that as RPM increases Tuned Length Decreases. As Exhaust duration increases so does Tuned length.

Now doing some real world calculations a Ferrari 308 Engine has an exhaust duration of 234° and peak HP at 7000rpm. Using those calculations that gives a tuned length of approx 650mm, which sounds reasonable, however it's about 1/2 their actual length...

Confused?

Thinking it proves your theory?

Sorry to say it doesn't It's packaging again, only this time they've gone the other way and tuned to 2x wavelength

[AJz]

*scratches head* so lost right now, lol

[deanis]

the theory sounds good.

where is the pipe diameter. this will have an effect on exaust velocity and therefore the length of the pipe.

and why does nobody do it. packaging is a bullshit answer.
Mr engineer saying ill make these headers twice as long and loose power because it looks nicer. if Mr engineer is doing that then he needs a punch in the head.

lots of questions the theory doesnt explain.

[Mr Revhead]

packaging is a bullshit answer.

dude, 99% of everything in a car is to do with packaging....

[deanis]

i dont do theories.

So you don't understand why something makes more power, only that it does. Sounds like an expensive time consuming way to do things

did you ever see the zoom magazine where they had a whole bunch of manifolds that somebody had scored from mazda where they were designing there batman rx7? the development for that was pretty much make a manifold, put it on an engine and chuck it on the dyno and see what it did. these guys have access to all the best flow engineering software, engineers and designers around and they build stuff like me. make something and stick it on a dyno and see what happens.

perhaps time consuming.

but at least you have proved it worked.

[deanis]

packaging is a bullshit answer.

dude, 99% of everything in a car is to do with packaging....

what about on a race car where every last horsepower is accounted for and they still use long headers?

[Mr Revhead]

sorry, how true

the world is wrong
deanis is right!!


zoom magazine....



define "long"
you state making the headers longer is for moving the HP up the RPM range
however all the evidence says not....

UNLESS..... the stock manifold is so short (for pacakging reasons) thats its SHORTER than the desired length.... so THEN making it longer will work.
however....... if the headers are tuned for 4000rpm
making them longer wont tune them for 6000rpm....

[fivebob]

where is the pipe diameter. this will have an effect on exaust velocity and therefore the length of the pipe.

Pipe diameter affects the point at which max torque occurs, length determines the power characteristics above and below that point. Larger diameter pipes move the torque curve up the rev range, a increase of about 3mm will move the peak torque RPM up by about 500rpm.

Formula for calculating the primary pipe ID is 2.1* sqrt(CylinderVolume/(PrimaryLength+75)) where CylinderVolume is the capacity of one cylinder in cc.

and why does nobody do it. packaging is a bullshit answer.
Mr engineer saying ill make these headers twice as long and loose power because it looks nicer. if Mr engineer is doing that then he needs a punch in the head.

Not really anything to do with looking nicer, it's more to do with fitting it in the available space or maybe just ease of assembly.

In the case of the Ferrari it's so the front bank has the same length as the rear bank. Another advantage to making it twice tuned length is that it gives two ramming pulses, one at 3500rpm (the tuned length), and one at 7000rpm (twice the tuned length). If they just made it for 5000rpm as the tuned length then you'd only get one.

lots of questions the theory doesnt explain.

The theory explains why the formulas work, explaining why people do things is not something that can be described with a formula, a psychologist is a more appropriate tool for that.

BTW 4 stroke exhaust theory is a piece of cake compared to 2 stroke expansion chamber design

[sergei]

I look at headers same way as I looked at Organ pipes, the only difference is that the speed of sound is about twice as fast.
Also you could look at them as ported subs, with box being the cylinder and port being the header (Helmholtz).
Having primaries and secondaries complicates the whole model big time as there many resonating frequencies (hence why 4-2-1 is used), although not as strong as 4 in 1 design.

[deanis]

and wtf does cam duration have to do with resonant length. the RPM is the only thing that changes the timing of the exhaust pulses from one cylinder to another. if you have more or less cam duration the exaust valves still only open at 90 degrees after eachother.

[deanis]

you state making the headers longer is for moving the HP up the RPM range
however all the evidence says not....

like your theories or dyno graphs? cause your theory is good but the dyno says otherwise.


the only reason i am doubting any of the theroy is i have seen first hand on a dyno actual results that go in the exact opposite direction of the theory.