We’ve already had a brief look at airflow, one of the foundations upon which we build horsepower. There are others of course - combustion efficiency and harnessing pressure wave activity being the main ones - but we’ve introduced the way that air behaves and looked at what we need to do to maximise mass flow. Let’s leave flow for the time being and have a look at finite pressure wave activity.
Before we discuss what it is, let’s look at what it can do and why we’d might want to use it. Basically we can only get so much air into (or out of) a cylinder using the pressure differential created by the descending piston. Once we’ve reduced the losses in the intake tract to zero (or as near to zero as practicable) then we might as well hang up the die grinder - there’s nothing left in it. But there are other possibilities for further improvements in cylinder filling and harnessing finite pressure waves is one.
Ever played with a “slinky”, one of those kids toys that are basically just a long steel spring? If you haven’t, go steal one from a kid now because they give a good clear look at wave behaviour.
Grab each end of the slinky and bounce some waves from end to end. The waves that have the coils all bunched together are compression waves, those with the coils spread further apart are called rarefaction waves. Sometimes they’re referred to as positive and negative waves. Now let’s pretend the slinky is the air in the intake tract of an engine and the slinky coils represent the air molecules. Can you see how getting a compression wave to enter the cylinder just before the valve closes can significantly increase the mass of the trapped charge?
Keep in mind that the waves will travel through the medium - whether it’s air or a slinky - regardless of whether it is stationary or moving at some speed. So if the wave is travelling in the same direction as the air flow the absolute speed of the wave will be the sum of the air speed plus the speed of the wave through the medium. If that doesn’t make sense think of this: you’re at the back of a bus travelling at 10kmh. You are capable of running at 15kmh. If you run from the back of the bus to the front your absolute speed while you’re running is 10+15=25kmh.
Of course the opposite is true too; if the wave is travelling in the opposite to the air flow then the absolute speed of the wave will be the air speed minus the wave speed. Back in the bus we’d represent this by running from the front to the back while the bus is travelling forwards. Again, if the bus was doing 10kmh and we ran at 15kmh then our absolute speed would be 10-15=-5kmh. The minus sign indicates that the direction of travel is the opposite of the direction of air flow (or bus travel).
We’ve mentioned the speed of the wave relative to the medium, but what is that exactly? In a lot of ways these finite pressure waves behave very much like sound waves, and you might even hear wave tuning referred to as sonic tuning. Finite pressure waves basically travel at the speed of sound. Or more accurately, the “local” speed of sound.
What’s the local speed of sound? The speed of sound in a gas isn’t a fixed value; it varies according to a number of factors, the main one being temperature. So when we talk about the local speed of sound in an intake or exhaust tract for example we are referring to the speed of sound as it exists in that tract. A simpler way of referencing this is to use the term “Mach number”. If you have a flow velocity of 0.5 mach then it is travelling at 0.5 times the local speed of sound.
Now, one of my pet hates, the word “harmonic”. It’s hard to imagine a more misused or misunderstood term than this one. Racers and engine builders use it to refer to damn near any high frequency vibration, and there’s a whole sticky thread on this forum about harmonics that aren’t harmonics at all. In reality a harmonic is a mathematical multiple of a fundamental frequency.
If you’ve been working with machinery for a while you’ll realize that everything is a spring, or more accurately, everything is a tuning fork. The crankshaft, camshaft, the block itself, the air in the intake tract, even the string of electrons in the ignition coil circuit. All these things have a natural frequency at which they vibrate if they are excited in some way.
If you hang a crankshaft from a string for example, and hit it with a hammer it’ll ring or chime at a certain frequency, let’s call it 600hz for this example. So 600hz then is the natural or fundamental frequency of the crankshaft. If the crank is somehow exposed to forces of around that frequency in the engine it’ll vibrate or oscillate at its natural frequency.
What’s this got to do with harmonics? A harmonic is a multiple or a fraction of that frequency. So for our 600hz crankshaft it would have harmonics at 1200hz, 1800hz, 2400hz and so on. The crank would also vibrate if exposed to forces at these harmonic frequencies.
A harmonic then is nothing more than a multiple of a particular frequency. Remember this because it’s important in regard to pipe tuning.
Now, before we can make use of a pressure wave we first have to create it. With a slinky we just grab one end of it and give it a quick jerk to create a wave. If we pull the end of the slinky we get a rarefaction wave and if we push it we get a compression wave. It’s exactly the same within an engine - if we pull sharply on the end of the intake tract with a descending piston for example, then we create a rarefaction wave that travels out towards the carb bellmouth. And if we crack open an exhaust valve and release gas under high pressure we create a compression wave.
Here’s the interesting bit: if a wave travels along a tube and encounters an open end (like a bellmouth) or even just an increase in cross-sectional-area (like a collector), it’s reflected back in the opposite direction and in the opposite sign. In other words a rarefaction or negative wave would be reflected back from the bellmouth as a compression or positive wave. This is the useful part; if we get everything timed just right then we can trap a nice dense slice of mixture in the cylinder just before the intake valve closes.
The opposite is also true; if a wave travels through a duct and encounters a closed end or a reduction in cross-sectional-area then that wave will be reflected back in the same sign. Ie. a negative wave will be reflected back as a negative, a positive as a positive.
We can also create waves with the exhaust system - when the exhaust valve opens we send a positive wave out that’s reflected back as a negative when it reaches the collector. We can use this negative wave to pull mixture into the chamber. And when this negative wave reaches the bellmouths then it’s reflected back as a positive, and this positive can jam some extra air/fuel into the cylinder as the valve is closing.
Here’s where harmonics come into play: let’s say you’re building an engine and you want to make maximum torque at a particular rpms - we’ll say 4000rpm. So you do the sums and calculate how long your intake tract needs to be to get that positive wave into the cylinder at just the right time. But it works out to be something like a metre or more long. Shit. How are you going to fit that much runner under the bonnet, and how are you going to keep the fuel suspended over such a long distance? Thanks to harmonics you don’t have to.
And here’s how it works: when a wave is jerked into existence and sent out the intake tract it doesn’t just bounce back once and then disappear. It keeps on bouncing from end to end, losing a little energy each time until eventually there’s nothing left of it. If you drop a hard rubber ball on the concrete you don’t have to catch it after the first bounce, you could catch it on the second or third. And this is exactly what we do with our pressure waves.
So if a metre long runner is out of the question we could make it half a metre long and catch that wave on the second bounce, or make it a third of a metre long and catch it on the third bounce. In other words we are harnessing the second or third harmonic.
There’s a catch to this though: with each round trip the waves lose some of their energy; just like the bouncing ball. Which explains why very high output engines like ProStockers use long runners in the tunnel ram manifolds that catch the second harmonic, even though they spin some serious rpms. Ditto for classes that use long mechanical injection stacks. But in other applications where limiting the overall engine package size is more important than peak output the pipes will be tuned to use the third or fourth harmonic.
Which leads to another myth that’s begging to be busted. This one says that for low rpm torque you make your runners longer and for top end you make them shorter. It makes good logical sense but many of us have found out the hard way that as often as not it just doesn’t work like that. Or maybe you’ve wondered why the SU manifolds that give the best top end power seem to have the longest runners. It just doesn’t add up.
It does add up actually. Let’s say your engine currently makes peak torque at 6000rpm and you’d like to increase that rpm number somewhat. Popular wisdom says to cut an inch or two of runner length, so you do that but find that the car is still slower than the competition with longer runners. How does that work?
What probably happened was that shortening the runner did indeed move the tuned length closer to the desired dimension for the third harmonic. But making the runner longer may have worked even better if the new length coincided with the stronger second harmonic. Longer doesn’t always mean low rpms; shorter isn’t always best for high rpms.
I guess the moral of this story is if you want to modify your torque curve by changing the tuned length, you first need to know which harmonic (if any) is being used at present and which harmonic you wish to use after the changes.
Very high output naturally aspirated engines nearly always rely heavily on pressure wave action to produce the big numbers. So it’s very tempting to use these waves in all sorts of applications but before you get too excited you need to know that there’s a definite downside. Unlike port work, which tends to improve the output over a wide rpm range, pressure wave tuning tends to raise peaks in some parts of the torque curve while creating equally big dips at other speeds. For every gain there’s a loss, and sometimes the losses can be very serious. If you don’t have enough gear ratios to keep the engine in its powerband the car can easily end up being slower. Highly tuned N/A engines tend to be very peaky. For a daily driver you might even want to suppress wave activity as much as possible for a nice, flat torque curve.
To illustrate just how high and how peaky the torque curve can be made we’ll take a look at one of the most remarkable racing engines ever: the Norton Manx produced from the late 40’s to the early 60’s. The Manx was a single cylinder 500 with a two-valve hemi head and DOHC; nothing too unusual. What made the engine unusual was its output - it had a BMEP higher than any other engine of the era. BMEP (or brake mean effective pressure) is a calculation of average cylinder pressure based on cylinder capacity and torque output. In other words the higher the torque output per cubic inch the higher the BMEP. Another way to look at BMEP figures is as an indication of cylinder filling or breathing - the better the fill, the more torque it makes and the higher the BMEP.
What really makes the Manx special (apart from all its racing successes) is that not only was its BMEP unmatched back in the day, it’s still unmatched even now. It was only made uncompetitive when the Italian four cylinder bikes were eventually able to outpower the Norton single by virtue of the much higher rpm capability of the short-stroke fours. The Gileras and MVs still couldn’t fill their cylinders as well as the Manx, but they could make up for this by having many more firing strokes in a minute.
Even without pressure wave aids the Manx breathed very well, but at some stage it was discovered that it responded very well to the use of a megaphone exhaust. This diverging cone reflected a negative wave back to the chamber and from there back through the intake duct during the overlap period (these engines used long-overlap, long duration cams). They soon discovered that a short, wide-angle megaphone created a stronger wave and a bigger torque boost than a long, gently tapered cone, so if you wanted to be competitive that’s what you had to run.
It wasn’t all roses though. When “on the pipe” the engine pulled extremely strongly. At other rpms though it ran very poorly indeed. When the rpms were out of phase with the pipe then you could have a positive wave pushing backwards through the engine and out of the carb. When this happened the air received a triple dose of fuel - once on the way in, again when it was blown back out and again on the way back in. The resulting mixture might be so rich it simply wouldn’t fire at all and plug-fouling was a problem. Riders tried all sorts of things like fanning the clutch or even jamming the toe of their boot into the megaphone to try to get the engine through the troublesome rev range. It became known as “megaphonitis” and made the bikes very unpleasant to ride. The trouble was that if you didn’t run a megaphone then you wouldn’t be able to keep up with those that did.
What made the engine breath so well was the combination of these: the megaphone exhaust, the long cam overlap, a very well breathing head with a valve angle that encouraged cross-talk from exhaust to intake and a very clean, tuned length intake tract. They were so concerned with keeping the intake free of anything that might disrupt the flow or wave action that even the metering needle was housed off to one side of the Amal GP carb, so that with the throttle open the entire tract including the carb was nothing more than an open tube.
I hope that this story of the Manx shows two things; one, that pressure wave tuning can make very significant contributions to engine power. And two, that these boosts come at a price - raised torque peaks balanced by deeper troughs. If you have enough gear ratios this mightn’t be a problem at all. Or it could be a show-stopper. There are ways to extend the positive effects of finite pressure waves over a wider rpm range though, and we’ll look at those in a later installment. For now though we’ve had a quick look at what finite pressure waves are and what they can do. Next time we’ll look at the practical side of making them work for us.
View Garage
Any additional info you can tell me in regards to my needs would be a fantastic bonus for me.
