smaller exaust pipe for mpg

[Christ]

I may be incorrect about this, but I believe header primary tube length is measured as the length of the pipe before a sudden change in volume. If you have stepped headers, the "primary" is much shorter than the entire header pipe.

Helmholtz frequency tuning is (again, IIRC) done taking volume changes into account, because each volume change creates a reflection of the sonic frequency in the exhaust.

In intake tuning (which is somewhat close to similar, so I'll mention it here), the intake "runner length" is measured as the distance from the face of the intake valve to the plenum opening, which is a sudden volume change.

Helmholtz tuning doesn't have to be ideal to be effective (I pointed this out earlier), but in a non-ideal situation, efficiency is lost to energy depletion from excess wave travel.

Frequency tuning is done on a system of halves and doubles, on (normally) 4 orders of frequency. If your ideal area is at 10,000 RPM, there are 3 half frequencies and 3 double frequencies, each one with a doubled loss percentage from depleted energy. The halves would be at 5,000, 2,500, and 1,250 RPM, while the doubles would be beyond the reach of most any engine we're interested in.

Notice where the 3rd frequency is? 2,500~... That means that even if you're tuned for the best harmonics at 10,000, you'll still get a gain at 2,500, at a benefit loss ~10-12% or so. 5,000 would be the 3-5% loss that Old Tele Man referred to, and 1,250 RPM would be ~20-25% loss.

Don't focus on the loss, though. It's still "free" energy that would otherwise be wasted.

[gone-ot]

...and, (doing my world-renowned "fancy foot dancing") let me reiterate & summarize: for FE you don't need to use smaller pipes, just don't use pipes that are as big as those typically used for HP (especially the head pipes)! Don't go down in size, just don't go up to the biggest size available!

...also, with headers, it's basically any change in volume, temperature, pressure, etc. that'll cause a reflection.

Originally Posted by Christ

Don't focus on the loss, though. It's still "free" energy that would otherwise be wasted.

...+100! Free is good.

...think of header "tuning" as "drafting" for exhaust valves (wink,wink)!

[nemesis]

Originally Posted by mwebb

cylinder displacement and engine RPM are the two parameters that most affect which pipe diameter to use to achieve the desired mass slug velocity of 300 fps (at HP RPM).

ok
so why then the magic 300fps -could be at lower rpm for efficiency or higher rpm for power
either way ----
does that somehow add or subtract from the
sound waveforms at 1700 fps ?

it can not be the same for an 8 cylinder engine with an optimum of 8 pulses in 720 degrees 90 degrees apart
with exhaust valve open times overlaying them selves even if completely separate dual exhaust s are used

as for a 3 cylinder with 3 pulses in 720 degrees 240 degrees apart
==========

V8 has to be treated as 2 v4 engines, not 2 inline 4's, so the exhaust will have to be treated differently.

[Christ]

No, it won't.

Inline vs V-type doesn't make any difference, the flow can't tell if it's a V or not. It's the pipe convergence angle that makes a difference for scavenging and proper flow velocity, not the profile of the engine.

[mwebb]

Originally Posted by nemesis

V8 has to be treated as 2 v4 engines, not 2 inline 4's, so the exhaust will have to be treated differently.

that does agree with what i have read in the hotrod article as well

but
i have pressure waveforms from the exhaust on V8s
most of it is UNusable hash , there are no distinct pressure / vacuum pulses that can be attributed to any one cylinder

at low rpm or up to about 2k rpm with little to no load

UNLESS you just add ignition timing value + known or measured Exhaust valve open time to measured spark waveform ....
even then , it is just hash , but you know where the hash would change if an event were to happen

to design a "tuned exhaust" for such a mess must involve something
akin to alchemy
===============================
remember
on a V8 there is another exhaust valve opening every 90 degrees or every 180 if you use 2 separate dual exhausts , and Exhaust valve open duration is around 250 degrees ,
4 or 8 events do not fit into 720 degrees without overlapping ....
which creates
hash
=====================================
the only exhaust pressure / vacuum pulses i have captured driving at load are done with a 4 cylinder engine
and you can see where the pulses all line up nice nice
and where they do not

even with the overlap
i am assuming , when the pulses are lined up
nice nice
that is the load and rpm or one load / rpm the exhaust is "tuned" for

and after reading this
i think the test probe in the tail pipe may be introducing something else to be reflected off of
or
another possible source of hash ....

[mwebb]

there are formulas in this thread to calculate pulse size
but
they assume 100% Volumetric Efficiency , which is also the goal
but
for partial load at part throttle VE is considerably less than 100% .

in a perfect world individual pulse volume would be the number of cylinders divided into the displacement
total flow would be that value multiplied by 50% of rpm .

to calculate pulse volume to optimize flow is there a fudge factor multiplier to be considered ?

as an example on an NA engine-
most cars / engines can get to between 90 and 95% calculated load
AT WOT , WideOpenThrottle above 4k rpm
Because there is restriction to flow from the air intake to the tail pipe .
so VE is about 90 to 95% on those engines

you could multiply
displacement X .925 X 50% of rpm to approximate total available flow at WOT above about 4000 rpm and you would be in the ballpark .

is there an online database to study and learn about this stuff ?

[gone-ot]

...you begin with the '1st-order' relationship:

Vg*Ag = Vp*Ap

where:
Vg = velocity of gas slug, fps
Vp = velocity of piston, fps
Ag = area of pipe the gas slug travels through, inches
Ap = area of piston, inches

where: area(A) = (pi/4)*dia^2; so: Ag = (pi/4)*d^2 and Ap = (pi/4)*B^2, where d = pipe diameter in inches and B = piston bore in inches, thus:

Vg / Vp = Ap / Ag = B^2 / d^2 (because the "pi/4" cancel)

...and, the 'mean' piston velocity is: Vp = RPM*S / 360, where S = piston stroke in inches.

...which yields: Vg = (RPM*S / 360)*(B^2 / d^2)

...finally, rearranging and solving for pipe diameter (d) you get:

d = SQRT[ RPM*S*B^2 / X ] ...where: X = 86,400(stock) to 108,00(race)


...if you're curious to know where "X" comes from, ask...

[mwebb]

Originally Posted by Old Tele man

...you begin with the '1st-order' relationship:

Vg*Ag = Vp*Ap

where:
Vg = velocity of gas slug, fps
Vp = velocity of piston, fps
Ag = area of pipe the gas slug travels through, inches
Ap = area of piston, inches

where: area(A) = pi*dia^2; so: Ag = pi*d^2 and Ap = pi*B^2, where d = pipe diameter in inches and B = piston bore in inches, thus:

Vg / Vp = Ap / Ag = B^2 / d^2 (because the "pi's" cancel)

...and, the 'mean' piston velocity is: Vp = RPM*S / 360, where S = piston stroke in inches.

...which yields: Vg = (RPM*S / 360)*(B^2 / d^2)

...finally, rearranging and solving for pipe diameter (d) you get:

d = SQRT[ RPM*S*B^2 / X ] ...where: X = 86,400(stock) to 108,00(race)


...if you're curious to know where "X" comes from, ask...

i am asking
please o please but
while you are at it ...
two sets of variables ..... need to get resolved in one exhaust system -
at a wide range of rpm and load

this is looking like alchemy

..."...with Helmholtz tuning, the goal is to have the "reflected" rarefacation wave arrive back at the just closing exhaust value, so that it can literally "suck" the last vestiges of exhaust gas from the cylinder as the exhaust valve closes, which, if done correctly, will then actually create a slight vacuum within the cylinder that helps it "suck-in" more air when the intake value opens.

...the speed of sound in normal air is about 1,100 fps, but in hot exhaust it's about 1,700 fps (there's an approximation equation available if you're interested).
"...

[mwebb]

photo of the setup with ATS venturi tip ,
exhaust flows thru the venturi tip and the sensor is connected to a tube that bisects , measuring flow NOT pressure
in theory anyway
with laptop and Pico 6 on the passenger's seat


zoomed screen cap , the exhaust pulses are the blue trace
at cruise , part load probably around 45mph

to the right side , blue trace , pulses are lined up well with little to no hash in this system in this car at cruise - at this load at this rpm
i am thinking that that is what is desired in the exhaust pulse train when everything is working at or near optimum efficiency .

the FLS sensor is a pressure differential sensor , it measures change in pressure
it uses a piezoelectric sensor inside -


UNzoomed , different scope and tool using a low pressure transducer
yellow trace is exhaust pulses
synced on secondary spark on cylinder 1
same car same setup

low pressure Dips are long line misfires as exhaust valve opened on vacuum because there was no combustion event

[gone-ot]

1) Velocity of sound thru INTAKE system:

Vs(int) = 49.1*SQRT[ T(air)+460° ] - 50

...where: T is air temperature in degrees F; 49.1 is constant for air.

Vs(int) = 49.1*SQRT[ 100°+460° ] - 50 = 1112 ~ 1100 fps.

2) Velocity of sound thru EXHAUST system:

Vs(exh) = 48.0*SQRT[ T(exh)+460° ] - 50

...where: T is mean temperature (exh.valve to tailpipe) in degrees F; 48.0 is constant for exhaust.

Vs(exh) = 48.0*SQRT[ 1010°+460° ] - 50 = 1790 ~ 1800 fps.


P.S.--header pipe is synonymous with exhaust manifold, but is not the same as exhaust pipe, which is what exits (usually) at the back of the vehicle.