One for the boffins - Intake Pulses

addo

addo

Top Poster
VIP
REDLINE 10K
OK, an induction pulse has velocity, wavelength and frequency, right?

What is the basic theory for tying all this together - ignoring the interrupting effect of throttle position/movements?

I am assuming that this all occurs irrespective of turbulent flow - that is; rather like the fuzzy sound of distortion on an electric guitar there is still an underlying dominant waveform (and harmonics).

If you have equal valving and cam timing and strokes on adjacent cylinders, each one generates the same "suck" or intake velocity. But intake runners may dictate different wavelengths, so frequency must alter to honour the velocity...

Am I right so far? Don't want to charge too hard into a line of thought, in case not!

Thanks, Adam.
 
Get the book (The Scientific Design of Intake and Exhaust Systems) and don't plan anything for about two weeks so you can read it a few times. It will answer your questions.
 
In a nutshell, the intake pulse hits the end of the runner at the valve and part of it is reflected back up the runner where it encounters another end of track, and is reflected back again towards the valve. When the reflected pulse is exactly in time with the next incoming pulse, you get the harmonic you are describing and it has the desired effect of increasing the VE and therefore the output of the engine.

Changes in inlet tract length change the RPM at which this effect occurs. That's the theory behind the 2 different lengths of air horns seen on the Lucas injectors on old Can-Am engines and the variable length tracts seen on current street and race engines.

The basic fromula is:

L = T x V / RPM

Where

L = inlet track length in inches
V = speed of sound, approx 1100 FPS
T = time for the pulse to travel to the end and back, expressed in degrees of crankshaft rotation. Use 90° for this value, if you want to know why you're going to have to read the book.
 
Idea # 1
While at Crane when I looked at my stock 88 GT's intake runners, I envisioned a intake system with long runners that were connected to the next runner in the firing order. It would also need a reed valve at the begining of each runner to contain the velocity. My thought was that the velocity was energy and the pulse from the valve shuting on this cylinder would help the push more air down the next runner.
When I asked the engineering guys there they just kind of look at me odd.
I realize this thing would look like a ball-O-snakes

Idea #2 How about makeing a long itake runner out of a flexable material.
At high vacum (no load and low airflow) the runner might be .750 dia.
At low vacum (high load and high air flow) the runner opens up to 1.5 dia.
Just think of how much better the throttle response would be.
 
Idea #2 has a problem. The energy used to reduce the diameter of the tube would be removed from the incoming charge, reducing its velocity. You need to power the changes in inlet track length/diameter from the output of the engine either through hydraulics off the PS pump of electrics off the alternator.
 
Does this explain why intake ports are sometimes carboned up?

With enough boost it shouldn't really matter what the harmonics are will it? :lol:
Joe
 
I am surprised V=speed of sound in free air.

Also, what SR describes as a "harmonic" is maxed-out sympathetic resonance of the gas body inside the tract, right? And if that is effective (defined) enough, then it's how VE can exceed 100%.

I could see T=90° making sense in terms of the pulse being ¼ wavelength.
 
The basic calculation also assumes a dry-flow environment; there are density and resonance corrections for a wet-flow environment. At Roush, we tested everything wet. It has obviously paid off for him in recent years...

Adam- your are mostly correct given a static throttle setting & valve event schedule, which are rarely seen in road racing. From what I've seen, throttle events are logged for standard opening and closing angles over a series of runs (say, 37% to 100%), then that data is used to tune- vice shooting for oen target setting (ie WOT). This is, of course, for N/A engines.

The NASCAR / speedway types do multiple harmonic testing to get more than one harmonic to add to the intake wave- hence some of the very high VEs in that series. They also use the intake valve & spring 'bounce' effect to assist this wave. I know that the F1 pneumatic stuff does this, and has been used as a traction control by BMW before (bleeding cyl pressure off).

I was at a builders conference a couple years ago where Waddell Wilson talked about being able to time the pulse right as the valve was opening- over about a 800-1000 rpm range. He said it took the right combo of port shape, cam ramps, cam timing, and intake shaping. Notice he didn't say much about runner length- he has been known to alter cast intakes to the point of having it paper-thin on one side, and a half-inch thick on the other just to re-tune the runner length.

anyway- sea story over. I really love this subject... I just wish I had some of the lab stuff I had access to before.
 
Wish I could find the pics of a Mazda GT Prototype race engine.

A two rotor wankel with four intake runners, each was made like the slide on a trombone, little electric motor drive a rack extending/ retracting in time with rpm changes controlled by the computer to keep intake in resonace and inertial ram effect at different rpm points.

The one thing to remember along with all the formulas, there is an inertial effect as the air (gas) pulse slams against the close valve compressing the pulse. At some given rpm the pressure is greatest as the valve open giving the highest ram effect. Before that rpm the pressure as not reached max, after pressure is starting to backflow and decrease. Sort of like a wave slamming against a seawall an the inertial mometum (pressure) drive the wave up the wall

The same type inertial effect happen on the exhuast side except reverse, the gas pull a vaccum behind it. Max cylinder scavenging is that rpm where the vaccum is highest when the exhaust valve opens.

Possible to tune length where for max torgue, intake and exhaust occur at the same rpm. Or different rpm less torque but wider peak.

IIRC the different length on the CANAM BBC engine was to compensate for the two different port lengths cast in the original BBC racing heads.
 
Newer BMW engines have individual sliding intake "trumpets" inside a common plenum. IIRC, the cars don't even need a throttle body, because the trumpets are so efficient at optimizing the intake length.

Neat, but not cheap or easy...
 
I once tried the "slide trombone" thing on a Briggs & Stratton engine with both the intake and exhaust tracts. One at a time, and then, together at the same time. It was amazing how little difference in length affected power at a given RPM. The intake was far more sensitive to change in length which was contrary to what I was expecting. It became immediately apparent that a computer would be needed to control this "system" of varying tract lengths because not only was RPM a factor but load/throttle position were variables that were in constant change and impossible to manually keep tuned. It was a fun experiment and I learned just how important these tract lengths are to volumetric efficiency. The technology is available but my prediction is that we will not see much of it be put to use for one simple reason. The internal combustion engine has embarked on its road to retirement which will be spent in hybrid vehicles operating at a steady RPM/loads which completely negates the need for variable tract lengths. :?
 
You must log in or register to reply here.
Back
Top