Applications - Intake Tract
OK. Up next I'd like to ramble on a bit about intake tracts. By "tract" I mean everything before the cylinder (including the valves, valve seats, ports, carbs, and filters).
As usual, I'll break it into (hopefully) digestible sections and as we're onto applications I'll cover some of the common modifications and how/when they should be undertaken. As always, please feel free to add more into this thread and/or ask questions. It was never my intention for this to be a book, but rather a more participatory kind of thing...
Also, as usual, a little bit of info up front... As you may have noticed by now, I'm addressing these topics in the rough order in which I would apply them to my own build. I'm not saying its the only way, but it is my way. I would not begin making changes to the intake prior to increases in compression or displacement. Nor would I make changes before I've decided upon a camshaft. What you can infer from this statement is that slapping on pod filters before making any other engine changes may not be the best way to go about freeing up some extra ponies.
There are several reasons for completing other engine modifications prior to changing the intake and rather than address them all right now, I'll talk a bit about it as I go through each subsection of this post. The primary reason, however, is that most changes you make to the insides of an engine are going to change the way it breathes and even the way it wants to breathe. If you change your intake first, it's fairly unlikely that you'll get it "right". That said, there are likely to be some changes that can be done as a matter of course, but they're not numerous.
Intake Length - Wave Tuning
The first thing to consider during the modifications of your intake is the total length. This is because the total length of your intake has a performance implication. It's not a big implication (only a couple of percentage points), but it all adds up so you might as well do it right. The reason for deciding upon a length for your intake is because of something called Wave Tuning. This is also known as Pulse Tuning or sometimes, Ram Effect (not to be confused with Ram Air systems).
The theory behind wave tuning has to do with the harmonics of the intake. Like any other "instrument", changing the length changes the tone. Certain tones will correspond to an increase in power at certain RPMs. The ideal length of the intake will depend largely upon the duration of your camshaft and so the total length of the intake is something that should be calculated after you know your intake valve duration. Furthermore, you will need have a decent idea of which RPM range upon which you wish to focus. The ram effect only works best over a fairly narrow RPM range and so you need to choose carefully. You'll want to select a point in your RPM band where you wish to increase your power and common tuning theory insists that you stack as many of your benefits as possible. This usually corresponds to the halfway point between when your cam starts to come on and your redline. This is likely to fall somewhere around 2/3 and 3/4 of your redline. If your redline is at 10,000 RPM, then you're tuning for 6500 RPM or 7500 RPM. Actual mileage may vary.
Now the reason this works has to do with waves or pulses (as the name suggests) that propagate up and down your intake tract. Each time the intake valve closes, it sends a shock wave back upstream along the intake. When that shock wave reaches the beginning of your intake (the air filters, usually) it collides with the relatively slow-moving air of the atmosphere and reflect back again toward the intake port. If this reflection comes back and contacts the intake valve just as it opens, you get a small increase in power due to the shock wave helping to push the air into the cylinder. This shock wave helps to overcome the inertia of the air and quickly speeds it up and into the cylinder. In practice, this shock wave will help even if the valve is already open, or just closing. The greatest benefit, however, will be just after opening when the air in the intake is starting to speed up.
So... how do we calculate it? It's actually not that hard. You do need to know the intake duration of your camshaft, though. For my example calculations, I'll stick with 221° which is the stock duration on a Honda 360 camshaft. Also needed is the desired RPMs of operations and I'll go with 8000 RPM as an arbitrary example.
First up, we need to calculate how long our intake valves remain closed over one complete cycle of the engine. As you probably recall, a complete cycle is two rotations on a four stroke engine and that's the same as 720°. So the time the valves spend open is the 221° (as determined by the camshaft) and so finding the time the valves are closed is just simple subtraction. 720° - 221° = 499°.
Well, that's only part of what we need. In order to get an actual time value from that data we also need to know the speed at which the engine is rotating. We've already decided upon 8000 RPM and so we need to work out how many rotations per second we're seeing. 8000RPM / 60 = 133.33 RPS. Now, we need to convert to degrees per second by multiplying by 360°. This gives us 48,000° of rotation every second. Finally, to get the amount of time the intake valves remain closed, we divide our 499° value by our 48,000° value. The degrees cancel out and we're left with .01034 seconds.
"Great, now what?", I can hear you say. Well... what we've done is to calculate the time that elapses between the intake valve resting into the seat and then coming off it again, or the times it remains closed before opening once more. As I mentioned earlier, our final goal is to find a length and so now that we have a time component for our calculation, we are able to find the distance component once we know the speed.
The speed of the propagating waves is roughly the speed of sound (Remember? Tone? Harmonics?). I usually use 90% of the speed of sound to help account for twists and turns in the intake as well as turbulent air which will also slow down the speed of sound in this environment. The speed of sound is 340.3 meters per second and 90% of that comes in at 306.3m/s. So now that we have speed and time, we can figure out distance. Simply multiply the speed by the time and we arrive at distance covered. 306.3m/s * .01034s = 3.18m. Also, bear in mind that because the wave is reflected at the beginning of the intake, we only need to use half of this value due the wave actually covering the distance of our intake twice. Our final value comes in at 1.59 meters.
Those of you even slightly familiar with the metric system will probably have noticed that the intake length I'm suggesting is pretty damn long. 5.21 feet. This isn't really possible in the real world and so, instead, we take advantage of the fact that the wave, upon hitting a (still) closed intake valve will reflect again back upstream. It will actually continue to reflect back and forth five or six times (in reality, it will reflect many, many, times but as it loses energy during each reflection, it becomes useless for our purposes past the fifth or sixth reflection) and this is something we can make use of.
It is pretty much unheard of to try to capture the first or even second reflection. Not only would it be difficult because the intake would be so long and cumbersome, the added weight and engineering of such a design would likely negate any real benefits. NASCAR intakes are usually tuned for third wave, to give you an idea of what's done in the racing world.
So lets scale it down a bit and shoot for the 4th wave. We can calculate this out simply by dividing our original calculated length by the number of the wave which we are trying to capture. 1.59 meters divided by 4 gives us a value 0.3975 meters, or 397.5mm. A lot more manageable. Especially considering that this length includes the length of the ports, the intake manifold, and the venturi of the carb. The length of our Honda 360's intake port, intake manifold, and stock CV carb venturi comes in at 214mm. This means, we want a further intake length of 183.5mm (7.2 inches) added onto the end of the carbs in order for our intake to be properly tuned.
Changing the duration of the camshaft or the desired wave will obviously affect these calculations. If you were to be running a cam with 251° duration instead of 221°, then the length has just decreased to 160mm because the valves are closed for less time. Likewise, trying to capture 5th (instead of 4th) wave on our original setup gives us a length 104mm. Finally, increasing the RPM target will shorten the intake and decreasing the RPM target will lengthen it.
Should your intake length be fixed, for some reason, you can at least calculate out where your small ram effect gains will occur.
Porting and Polishing
Porting and polishing is a phrase that gets thrown around a lot and it's something I don't think a lot of folks really understand. Many newcomers to performance enhancement believe it is something that every engine undergoes during its transformation from a modest lump of metal into a fire breathing, kitten eater. Unfortunately, this is mostly incorrect.
As mentioned several times previously, decent intake velocity is a key component in getting the most out of your engine. As the ports are opened up, the intake velocity slows down. Remember, the engine can only utilize so much volume of air and opening the ports past a certain size will not increase performance. The reason for this is that the intake stroke actually consists of three separate stages.
The first stage of the intake stroke is overlap. This is when the intake valves first begin to open and the exhaust valves are still open and getting closer to closing. In a well tuned system (at the ideal RPM), the exhaust gases leaving the cylinder will have a low pressure zone behind them that help to pull the fresh intake charge into the cylinder.
The second stage of the intake stroke follows the more generally accepted definition. The piston descends, creating a low pressure area into which the fresh intake charge rushes. Nature abhors a vacuum. Big ports help with this portion.
The third stage acts like a mini compression stroke as the actual compression stroke begins. This is where big ports may actually hurt performance. As the piston begins to ascend, pressure builds within the cylinder because the volume is decreasing. This occurs while the intake valve is still open. With smaller ports, the intake charge moves faster and is more difficult to stop because the air has inertia. If the ports are too large, then the air moves slowly and is stopped more easily. The formula for kinetic energy is E = ½mv². As you can see, increasing the velocity gives you more bang for the buck than increasing the mass. So while big ports are useful for initial filling, they're much less so when it comes time to push that last little bit of air into the cylinder. This problem is even more noticeable as engine revs increase because the time allowed to fill each cylinder during the intake stroke decreases dramatically. As engine revs increase, intake velocity quickly becomes the primary means by which volumetric efficiency is achieved. It is entirely possible to size the ports such that your engine cannot reach the RPM at which that size will be most useful.
Running your head on a flow bench can tell you a lot of information about this, but it won't tell you everything. If you have the opportunity, it's worthwhile to stuff as much clay as you can onto the floors of the intake ports and see how little the CFMs actually change. This is all dead space indicating that ports are oversized. If the addition of clay significantly reduces the CFM, then it may be time to consider enlarging the ports.
For 90% of us, that day is unlikely to come. Most engines will require significant modification before the size of the ports becomes a restriction to making power and most ports come oversized from the factory. And even then, it's likely that a change in port shape may be more beneficial than simply enlarging the ports. Most stock ports are roughly circular, but this is the not the ideal shape for air flow into the cylinder. What you're really after is a 'D' shaped port with the flat portion of the D as the port floor. By allowing the air to spread out evenly along a flat floor, you prevent it from "bunching up" as it turns the radius to go down past the valves and into the cylinder.
What can (and should) be done, however, is to clean up the ports. Casting marks should be removed. Areas around the valve guides and where the head meets the valve seats are likely to have excess metal that disrupts the air flow. The following pic is my own intake port halfway through the clean up process. I've ground off a lot of the casting marks, but you can still see some that are left to go. Also note the casting seam on the port floor (top of the pic, head is upside down) that still needs to come off.
As for a polish? Skip it unless you're running with fuel injection. Polishing your intake ports encourages a boundary layer of air to form along the port walls. This will cause your fuel to more easily drop out of the charge mixture and this fuel will begin to pool. It then gets sucked into the cylinders at uneven intervals making your engine performance less than smooth. After you're done removing the casting marks, finish with a rough grit sand paper. I use 220 though I know some folks will go as high as 400 grit. I don't recommend any higher than that. Definitely stay away from the valve seats during this whole process. If you nick a valve seat with a Dremel or sand it with the paper, it's not likely to seal well and a machinist may need to reseat your valves.
Valve Seats and Intake Valves
There are a few options when it comes to valves, as well. You can often go oversized on valves. Again, enlarging anything will likely result in slower intake velocity. Pursue with caution. Additionally, there are some modifications you can make to the valves themselves. One desirable trait in valves is a narrowing of the valve stem when it begins to meet with the valve head. This promotes smoother (read: better) air flow. Additionally, valves can be backcut to encourage better flow at low lift. This is especially important for getting the air into your cylinder, faster, as the valve begins to open. Titanium valves are nice if you can afford them, but stainless steel is a common performance material as well. These next pic illustrates stem taper (stock valve compared with aftermarket). Taper on a valve is something most machinists won't do because it's pretty darn easy to destroy a valve, even when being careful. You may need to purchase new valves in order to get this feature.
Additionally, backcutting the valves is something that can be done as a matter of course anytime you happen to have the engine open and the valves out. A 30° backcut helps to start flowing air sooner, as the valve begins to open. That can be seen in the following pic as the small metal strip located above the normal 45° valve sealing surface.
The valve seats, themselves, should also undergo treatment. A three angle grind is a standard performance modification and five angle is even better. More angles leads to smoother transitions of moving air and will aid in cylinder filling. Running flat out, good backcutting and a three angle job is usually good for an extra percentage point or two of VE.
Some valve makers also offer a swirl polish that is meant to help more evenly distribute the mixture around the cylinder, but I'm afraid I don't have too much info on this other than its existence. Maybe someone else can offer up their opinions and experience?
For multivalved engines after 1985 (or so), it's likely that your valve design will also include what is known as "tumble". The goal of this is also to ensure more even cylinder filling (not just high VE, but good distribution of fresh mixture). By getting fresh mixture into as many corners of the combustion chamber as possible, we help to keep combustion chamber temperature under control and this permits higher compression and/or timing advance; both of which usually lead to more power. I've not heard of anyone incorporating tumble into a head not already designed for it, but maybe you have a machinist with the knowledge and desire to make it happen?
One of the topics I wanted to talk about here was carburetor selection. It seems all too common for folks to believe that by slapping on a new set of carbs they've somehow created horsepower. I hate to disappoint, but the engineers at your bike's birth place didn't select your carbs on a whim. Possible that the cost control guys came in and told them to choose something else, but I digress...
The point I'm trying to make is that your stock carbs, if working correctly and adjusted correctly, are probably fine for your bike, even after a few engine changes. If anything, they may even be too big. Luckily, there is a way to find out if your carbs are sized correctly.
Before I get to that, there are a couple of other reasons to consider a carb change. Not all carbs are created equal. In addition to metering the fuel, the carbs should restrict the air as little as possible (at WOT, of course), be easy to tune (this is especially important for built engines as you will be messing with the carbs a lot), and do a decent job of atomizing or emulsifying the fuel. This last item is important, but maybe not as much as you'd think. Poor atomization of the fuel by the carbs can often be made up by good chamber design, quench area, tumble/swirl from the valve design, and even increased compression. If the carbs can do a good job of atomizing the fuel, then that's a big plus, but not a deal breaker if they can't.
Anyway... back to sizing. There are two basic ways to determine proper carb sizing. The first method is through calculation and the second method is through experimentation.
I'll cover experimentation first, because the concept is fairly simple. The airflow through carburetors is usually measured at an industry-standard 3" of Mercury for single barrel carbs. Multibarrel carbs and carbs manufactured for automobile use may be rated at 1.5" of Mercury, so pay attention. For those unfamiliar with the concept of inches of Mercury, it's a measurement for determining negative pressures (usually called vacuum, though the terminology of "vacuum" isn't technically correct). This measurement derives from the inches of Mercury that are "pushed" up a low pressure tube by the relatively higher pressure of the atmosphere. The lower the pressure in the tube, the higher the number of inches.
For experimentation purposes, this means that your carbs, when sized well for your engine, should be pulling 3" of Mercury at WOT and redline. The experiment for this is deceptively simple. Hook up a vacuum gauge to your carbs and go for a spin. Take the bike up to redline at WOT and check the vacuum. It should read between 3" and 3.5" of Mercury. If the reading is lower (1.5", for instance) then your carbs are too big. Time to downsize. If the reading is higher, then you need bigger carbs. If you're even a little bit out of this 3"-3.5" range, it would be wise to consider a change. Ensure, however, that the restriction is actually from your carbs. Your air filter may be causing some of this reading. This can be verified by reading the number of inches of Mercury by plugging your vacuum gauge into the intake (before the carbs) and repeating the test.
The calculations for determining carb sizes are also pretty simple, but a bit lengthy. First, you need your engine's displacement in cubic inches. If you already know this value in cubic centimeters then just multiply this value by .061. A stock Honda 360 (356cc) is 21.7 cubic inches. Next, multiply this value by half of your redline value. Assuming a 10,000 RPM redline, we'll use 5,000. So 21.7 * 5000 = 108,500 cubic inches per minute. Conversion to cubic feet gives us 62.8 CFM of air is required by the engine at redline. Next, divide this value by the number of carburetors in use by your engine. Continuing with our example of the Honda 360, we divide by 2 and arrive at 31.4 CFM per carburetor. Finally, we multiply by our estimated volumetric efficiency. For a stock engine, this value is usually around .85 (85%). Built engines push closer to .9 (90%) and full race engines get between .95 and 1.0 (95% and 100%). I'll assume a stock engine and use the .85 value. We now how a final flow rate of 31.4 * .85 = 26.7 CFM.
At this point, your best bet is to find a carb manufacturer that lists the CFM ratings for their carbs and pick the one that most closely matches your requirements. For my own purposes, I usually round down. For instance, lets say we want some new carbs for our 360 in the above example. Ideally, we're looking for something rated at 27 CFM. Supposing we find carbs rated at 26 CFM and 28 CFM, but not 27 CFM. I'd select the set that are rated at 26 CFM. In real world use, your bike isn't going to be seeing WOT very often, even if you ride like a madman. The increased throttle response you get from a slightly smaller carb in the low and midrange will definitely outweigh the potential gains you get from a slightly oversized carb at WOT. It is a common error to select overly large carbs with the idea that it somehow creates horsepower. In theory, perhaps it will free up a few ponies at redline, but in reality you're going to be spending 99% of your riding at part throttle and/or out of redline. Only a full blown race bike (on a relatively straight, high speed course) will see any benefit.
If your carb manufacturer does not provide a CFM, it's time to either bust out some math or hope you have access to a flow bench. I'm assuming the former, so lets get down to it. In order to calculate the area needed to flow 26.7 CFM, we must first know the air speed velocity through the carbs. Unfortunately, this isn't something easy to know. Ideally, this would be measured, but for this model we're going to assume 65 feet per second. As yet another corollary, it's important to realize that the intake system for your engine is constantly decreasing in radius along its track. The carbs will have a much larger radius than the ports and the valve seat area is even smaller (especially with the valves taking up some of that space). It's not uncommon for port velocities to measure between 300 and 500 fps in a well setup engine, but you won't see numbers that high when measuring velocities at the carbs.
Anyway... back to the math. Volumetric flow rate of a fluid (yes, air is still a fluid) through a round pipe uses the formula of v = Q / A where Q is the flow rate, v is the velocity, and A is the cross sectional (inside) area of the tube. Because we know our flow rate (26.7 CFM) and we know our air velocity (65 fps) then all we have to do it convert the units and solve for area.
First, we need to ensure we're using the same units and so we need to convert our air velocity in to feet per minute instead of feet per second. This gets us 3900 feet per minute (65 feet * 60 seconds per minute). And so our equation now reads 3900 = 26.7 / A. Algebra tells us we can swap the locations of the A and 3900 values in order to make this easier to solve (since we're tying to solve for A) and so now the equation reads A = 26.7 / 3900. Completing the division gives us A = 0.00685 square feet. Conversion over to metric (since carb bores are listed in metric) gives is a cross sectional area of 636.39 square millimeters. Divide by PI (3.1415) to get the radius squares and we have 202.58 = r². Next, take the square root of 202.58 and we have a radius of 14.23 millimeters. Double it and we get a diameter of 28.5 millimeters. The ideal sized carbs for our engine will be right around the 28mm mark. VM28s would be perfect.
Finally, it's important to note that the air velocity through the carbs is directly related to the inches of Mercury being pulled. Assuming a WOT condition at max RPMs, increased air speed will lead to decreased pressure. Because it's pressure with which we're concerned, measuring this value is generally of more use than calculating out your carb sizes. Furthermore, multivalved engines can usually maintain a lower vacuum signal across the carbs and still remain responsive. Values as low at 2" of Mercury may be acceptable in those cases but if time and money permit, don't be afraid to experiment.
Intake Filters and Velocity Stacks
OK... we've made it this far. It's time to take a look at the different options available for filtering the air coming into the carbs. First up, there are some carburetors that require some level of restriction in order to work properly. Removing the stock air box from these carbs can create issues that will need to be dealt with.
Mostly this applies to CV carbs where an uneven and turbulent air flow can lead to improper metering of fuel. Usually this will manifest itself in the form of a rich midrange or, assuming you've adjusted that out through jetting changes, a lean top end. There exist several solutions and these include, but are not limited to, lengthening the intake to smooth the incoming air flow, using stronger (or even stretching the stock) diaphragm springs. In many cases it may be more desirable simply to leave the stock system in place.
But if you're building an engine for performance reasons, there are likely to be some gains to be had by freeing up the breathing prior to the carbs. How do you know? Well, it's time to bust out the vacuum gauge and go for a ride. This time, we're going to want to plug the gauge into the intake prior to the carbs, instead of into the vacuum port on the carbs themselves. You may need to tap into the intake system in order to make this happen. Just like when testing vacuum on your carbs, you want to get the bike going and take it to redline at WOT. We're aiming for zero inches of Mercury on this reading. Anything more than that represents a loss in power due, not only, to the reduced air coming into your engine but also the extra work your engine is doing fighting lower pressures (this is an example of pumping losses).
Quite often a simple switch to a new filter or performance filter can resolve this issue with a minimum of headache and tuning. Other times it may be required to remove the stock filtration system and opt for something different. The usual course of action is put on pod filters, but I don't recommend this until after you've tested your intake system. Replacing parts just because everyone else is doing or because it looks good is not often a path to success.
Aside from pod filters, there are also foam filters and other styles available. Largely, they do basically the same thing. They keep the air as clean as possible while restricting the intake as little as possible.
One of the better performance options to investigate is velocity stacks. These don't usually come cheap, but most engines with a well chosen velocity stack will get a slight boost in performance (almost all at the very top end of the power band). A well designed velocity stack can usually allow a carburetor to perform above its rated CFM by a couple of percentage points. A useful feature if you're nearing the top of what your carb can provide but aren't ready to pick up a new set just yet. More often than not, these items get added for looks though, so be aware that there are MANY velocity stacks on the market that will not provide much in the way of performance gains.
The way velocity stacks work is the same way the rest of your intake functions, it just starts the process earlier. A full taper velocity stack (the only kind worth buying, in my opinion) will be wider at the bell mouth than it will be at the carb side. This steady constriction increases velocity of the incoming air and promotes laminar flow prior to the throttle body. By getting the air all moving in the same direction and same speed, the molecules spend less time bumping into each other and more time moving into the engine where they can do some good. This smoothing of the air may also be beneficial, when put into use on CV carbs, where pod filters may not. A velocity stack designed for performance (as aside for just looks) should be polished on the inside and the lip of the bell mouth should make a full 180° turn so that it's edge is pointing back toward the carbs. This have been proven in several tests to promote the maximum air flow.
On the down side, most velocity stacks reduce or eliminate the ability to use an air filter. Even if you don't go riding around in the dirt and mud, it is expected that you will need to re-ring your pistons every 5,000 miles or so. Valve seats and/or valves may last you 10,000 or 15,000, if you're lucky.
As mentioned in the BMEP post a few pages ago, cold air makes more power. If possible, consider heat shielding between your intake and the rest of your engine. Carbs should always be connected to the head via a manifold that does not conduct heat well. Many plastics and rubber compounds fall under this category. If a metal manifold must be used, consider gasket materials that are poor conductors of heat. Keeping heat out of the intake can add significant power to your engine and the best way to do this is ensure you're pulling your air from a cool source and keeping the intake tracts insulated from engine heat as much as possible.
I also wanted to touch briefly on the topic of ram air tubes. Skip these. They do have some benefit at higher speeds, but they're difficult to install and have working correctly (ESPECIALLY on a carbed bike).
Finally, I've seen a few two-into-one or four-into-two intake manifolds floating around. Skip these, too. They actually hurt the performance of your engine. What you may gain in having fewer carbs to mess with, you'll lose in tuning ability. More often than not, these systems are put in place for looks alone and may lead to uneven running of the bike as the runners are rarely balanced.
In my opinion, the intake system of an engine is one of the places where many gains are to be had and perhaps it's because of that reason that so much misinformation exits. It's funny how many people I meet that just swap out their air filters for some pods and think they've added five horse power... Like any other section of engine building, this needs to be approached with a goal in mind, some tests you should expect to undertake, and a methodical process to follow. If you're unsure how to proceed in your tests, find a quiet and straight section of road. Run it at WOT and time yourself. The clock is the great equalizer and it never lies.
You can make additional power in this area and it's not terribly difficult (especially cleaning up your own intake ports or having a machinist backcut your stock valves), but take your time and make sure you understand what you're doing and what you hope to accomplish.