In Chapter 3, we looked at what was needed in the way of mixture quality and air/fuel ratios. Now our focus is the design of the intake manifold. Here, I discuss what it takes to minimize flow restrictions and optimize intake pressure waves.
This Tech Tip is From the Full Book, DAVID VIZARD’S HOW TO BUILD HORSEPOWER. For a comprehensive guide on this entire subject you can visit this link:
SHARE THIS ARTICLE: Please feel free to share this article on Facebook, in Forums, or with any Clubs you participate in. You can copy and paste this link to share: http://www.musclecardiy.com/performance/horsepower-secrets-intake-manifolds/
Many of you are reading this book to get some idea of how much extra power you are likely to see from a manifold change. It is reasonable to expect that an aftermarket performance intake manifold actually can deliver a cost-effective amount of extra power. Unfortunately, this is not always the case. I have tested intake manifolds that gave really mediocre results. In the mid 1980s, I did a big intake manifold test for an article in Hot Rod magazine. From 15 different intake manifolds, only about four of them produced what I call worthwhile results. Two or three produced barely more than the stock intake, and one actually reduced the 325-hp test engine by a whopping 56 hp.
To make sure it was not a problem just with my test engine, I sent it off for another race shop to test. Same result! The moral here is: You would think that if you have two manifolds that appear the same at critical points (such as runner junctions to a plenum, runner lengths, etc.), they should produce similar results, but that is not always the case. Why? Because not enough consideration was taken to understand what is needed.
Photo 4-1 shows four intakes for a 2-liter non-VTec Honda. One might expect the intake that had the highest average flow was best. In this instance, it was the second intake from the top. However, it showed a minimal gain in top end (about 3 hp) while costing twice that much in the 3,000- to 4,500- rpm range, which is used so much more often. To make any plenumstyle intake work, we have to consider the size and efficiency of whatever feeds the plenum in the first place. Next, the entry into the runner has to be optimized by suitably reshaping it. After that, the diameter and length of the runner itself has to be determined for best results. Fail at any one of these factors and the results suffer.
Still using this Honda as an example, let’s start with how the plenum is fed. The object of the exercise here is to supply the plenum with as much air as needed with the minimum restriction. The obvious solution is to use a throttle body with a really big bore but, in fact, that is the wrong approach. The throttle body runner is, in effect, a tube supplying a plenum and can be an effective Helmholtz resonator. If the throttle body runner is of the right diameter and length, Helmholtz tuning can be used to great effect. What this means is flow derived from an efficient but smaller throttle body runner is far more effective than a bigger less-efficient runner, even if that bigger runner does flow more air. Let us not lose sight of the fact that air is much heavier than often realized. When a cylinder draws air, the velocity of the air and its mass within the throttle body runner can be a significant source of kinetic energy. We can either dissipate it as heat in an inefficient runner or use it to boost pressure in the plenum. Obviously, the second option is our number-one choice.
Basically, a Helmholtz plenum works like a high-velocity piston in a tube, compressing air in a closed chamber in front of it. When all of the piston’s kinetic energy has been used up compressing the air in front of it, the piston reverses back down the tube. When another induction pulse occurs, the piston (which in reality is a slug of air) is sucked back in, builds speed in the tube, and the whole process starts again. Added to this is the natural resonant frequency between the plenum and the tube.
By adjusting the three variables that dictate the frequency, we can boost the pressure in the plenum at a time when the runners are out of phase and doing the reverse of ramming the cylinders. This is usually a low-speed deal and, if such a system is tuned right, it can be very effective at pumping up low-speed torque.
Although it works well with four cylinders attached to the plenum,the best is to have three cylinders so attached. All eight cylinders of a V-8 reduce the effect to minimal proportions, but if one bank of a twoplaned crank V-8 operates from one plenum and the other bank from another, the Helmholtz effect here can be very pronounced.
To produce a system like this, the starting place is to calculate the intake diameter required. Although not yet totally conclusive, my dyno testing to date has indicated that 170 to 180 feet per second air speed in the intake tube at peak power is about as high as needed for optimal results. This leaves us with the following formula:
If you are working in metric, then use:
Where: D is Displacement in CC or inches
V is Velocity in ft/sec or meters/sec
VE is Volumetric Efficiency percent
Length and Volume
So far, we have arrived at a working formula for a subject known for its mathematical complexity. Unfortunately there are, at this point, some serious obstacles to achieving mathematical simplification of the plenum design yet still achieving near-optimal results.
To be of the greatest benefit, a Helmholtz resonator needs to boost power just below the RPM at which the intake runner length comes into tune. Although it can be made to work with all eight cylinders attached to one plenum, it does so at about 2,000 rpm and with a much smaller intake than predicted here. For best operation, there must be a welldefined stop/start flow to the intake. This means the plenum needs to be attached to a maximum of four cylinders, but three is close to optimal. A V-8 with two induction pulses just 90 degrees apart looks (to the plenum) like a three-cylinder engine with one of the cylinders bigger than the other two. So for this system to work on a V-6 or a V-8, two plenums are needed. The intake runner must then be sized for half the engine’s displacement.
In practice, a Helmholtz resonator does not act exactly as such on an engine because the engine keeps on drawing air from the system. Under these circumstances, the formula for determining what is optimal is extremely complex, so a more cut-and-shut method is called for. Let’s first deal with plenum volumes.
This selection can be tricky; the intake runners need to be able to draw air out of the plenum efficiently. This often changes what can be effectively made. But as a first approximation, a plenum having a volume of about 20 percent more than the combined displacement of the cylinders being fed is a good starting point.
Now for the length required. For an engine turning 10,000 rpm, a plenum intake runner length of about 7 to 8 inches is required. For every 1,000 rpm below that, increase the length by 13 ⁄4 to 2 inches.
Intake Runner Lengths and Diameters
We now get onto the tricky subject of runner dimensions. To appreciate why these two dimensions are so important, you need to understand a few basic facts. First is that air is heavier than you may think (an average school gym contains about 40 tons of air!) A suitably high port velocity helps ram the air into the cylinder at the end of the induction stroke. As the valve closes, the air piles up, creating a positive pressure that, during the last few degrees the valve is open, helps push the last few CC of air into the cylinder. On a well-tuned system, the pressure just before and at valve closure can reach 7 psi above atmospheric pressure.
Also (if the exhaust is tuned right for the RPM involved), a very strong negative-pressure wave can arrive at the exhaust valve during the overlap period. This low-pressure wave is communicated to the intake through the combustion chamber. This lowpressure wave can also be very strong. A well-tuned exhaust can pull 4 psi of vacuum here, and an optimally tuned race system, as much as 7 psi. This, on an all-out pushrod V-8, can result in the intake charge moving into the cylinder at speeds up to 80 mph before the piston has even started on its way down the bore. To make all this happen at the desired RPM, the intake needs to be a certain length.
The following formula gives the required length:
L = Intake Length (from the intake valve to the open end of the intake runner)
ECD = Effective Cam Duration
V = Pressure Wave Velocity (about 1,300 ft/sec)
RV = Reflective Value (usually 2 but, for a tuned length for lower RPM,
can predict an impractically long intake length. In this instance, a less-effective but more-convenient RV of 3 or even 4 can be used.) D = Diameter of Intake (at the end of the intake tract just before any entry flare)
ECD is an assessment of when the valve is open sufficiently for some useful activity to be occurring. For a typical engine, subtract 15 degrees from the 0.020 tappet-lift duration to arrive at a good approximation. (By the way, cam manufacturers’ catalogs have all these). As an example, let us assume we want to tune the intake length to 8,000 rpm for a fuel-injected four-cylinder engine.
ECD at 0.020 = 285 degrees
V = 1,100 ft/sec RPM = 8,000
RV = 2 D = 2.25 inches
Inserting these numbers into our equation, we have:
This equals 15.47 inches.
From this, we subtract half the inlet diameter, which is is 2.25 inches in our example.
So, 15.47 – 1.125 gives us the final length of 14.34 inches.
V-8 Intake Manifolds
So far, we have looked at the relatively simple Helmholtz resonator plenum-style intakes and independent runner intakes required for even-firing inline engines of 2 to 6 cylinders. However, the most popular type of engine to modify is still the two-plane crank V-8. By having two rods on each journal and four journals, each spaced 90 degrees apart, the two-plane crank makes a V-8 function as two V-4s—not two inline 4s. This layout produces crank firing angles of 270, 180, 90, 180, and back to 270 to repeat again. It is these angular differences that give a two-plane crank V-8 its distinctive throbbing exhaust note, leaving the impression of fewer RPM than is actually the case.
This layout has given rise to two principal types of intake manifolds: the two-plane, and the single-plane. A two-plane intake divides the engine so that the runners joined to either half of a 4-barrel carb draw 180 degrees apart. For this reason, this type of intake is also known as a 180-degree intake. Its principle advantage is that there is no overlap of induction cycles; the interaction of one induction stroke on the next, and its negative impact on idle and cruise vacuum, is negated. The down side is the runners must follow a more tortuous route from the carb to the intake port and the effective carb flow seen by any cylinder is halved.
By contrast, a single-plane intake seeks to produce the best results at WOT and at higher RPM. With this type of intake, all eight cylinders draw from a common plenum. This not only means a more direct routing of the runner, but also each cylinder in effect “sees” all four barrels of the carb. An extension of this type of intake is the tunnel ram (see page 45).
Producing an effective two-plane intake is no easy task. Although no factory V-8 has been produced with this type of intake since the mid 1980s, it continues to be popular because, conceptually, it can be very effective for a high-performance, street-driven V-8. Since the early 1990s, considerable developmental effort has led to some truly effective two-plane intakes that deserve credit for their functionality. These designs owe their success to countless hours on the flow bench and in recent years to the use of computational fluid dynamics.
Because these modern intakes have so much more airflow capability than earlier designs, it has become necessary to reevaluate just how much carburetion they need for optimal results. In essence, we see that these intakes thrive on much greater carb CFM than is traditionally accepted. If we combine that with the fact that each cylinder only sees half the carb’s CFM, then it becomes clear that these high-efficiency two-plane intakes should theoretically require much more carb CFM.
This, in fact, has been borne out by my own dyno testing. By adopting this “bigger than generally accepted” philosophy with high-efficiency two-plane intakes, I have been able to produce well-mannered street motors with very reasonable vacuum for brakes, idle, etc., and outstanding top-end numbers for those fast times on the drag strip. Passing the 500-hp mark with a two-plane has historically been a difficult target to reach. However, using a carb with more than 1,000-cfm capacity, I have had street-drivable Chevys of a little more than 400 inches displacement produce in excess of 550 hp.
Without a doubt single-plane intakes are the favored style for allout performance. Using one generally improves the power curve from about 4,200 rpm on up. With the ability to have a superior routing from the plenum to the port runner, a single-plane has a far better chance at producing big top-end numbers, but success is by no means guaranteed. Factors, such as runner area, length, taper, plenum volume and their effect on mixture quality, and fuel distribution, all must be dealt with. If any significant design error exists in these areas, the effect on power can be relatively big to catastrophic.
An example serves to make the point here. While doing some intake-manifold-to-cam compatibility testing for a well-known intake manifold manufacturer, I found that one of the single-plane intakes being tested was sensitive to a small change in the plenum to the tune of a 30-percent drop in output! That said, often the changes, with experience, can for the most part, be somewhat intuitive. The factor that can often be the most problematical is fuel distribution but even that, with experience, can, even if time consuming, be second nature to establish a fix.
As mixture equality between cylinders is approached, the final steps to achieve it can be made by “stagger jetting.” This procedure,without a wideband O2 unit in each exhaust, is hard to do without a keen eye for plug reading. It is a skill that, without such instrumentation, may take you years to perfect, but it is a skill worth developing if your field of endeavor is likely to be centered around single-plane, four-barrel, V-8 induction systems.
Plenum volume is a significant factor toward making optimal output in the desired RPM range. Fortunately, this is easily adjusted by means of a carb spacer. These can come in a variety of forms. The simplest is the open spacer, which simply acts as a means to extend the plenum volume by raising the carb by 1 or 2 inches. The next most common is the four-hole spacer. These holes can be of a parallel-wall design or they may be of a contoured-form design, intended to improve the airflow as it exits the carb. Contoured spacers can add 10 to 15 cfm to the carb’s airflow capacity. For the record, big-block Chevys using most of the commonly available Dominator intakes respond well to the use of a spacer of as much as 2 inches, even when an already large plenum exists.
Tunnel Ram Intakes
A tunnel ram intake typically utilizes two 4-barrel or four 2-barrel carbs that align each carb barrel directly over an intake runner. These runners feed from a suitably sized plenum so that, at WOT, the plenum appears (to each intake runner) to be open air that has been pre-mixed with fuel.
The fact that all eight barrels of carburetion, each of relatively large CFM, can be seen by any individual cylinder means the plenum runs at barely below atmospheric pressure. The absence of a butterfly within the runner also means that the runner is uncompromised both in terms of airflow and pressure-wave reflection. This makes the tunnel ram the number-one power producer when it comes to manifolds for normally aspirated engines.
It is often said that a tunnel ram is a race-only setup but, in practice, this is not the case. Assuming that the budget and hood lines stretch to requirements, a tunnel ram installation produces a low-speed output that is midway between a single 4-barrel-carbed two-plane and a single-plane intake. Assuming a stout but nonetheless streetable cam, the tunnel ram starts to outpace the twoplane intake at about 4,200 rpm, and the single-plane intake at about 5,000 rpm. After that, it really is no contest. If a motor makes 550 hp on a 4-barrel-equipped, single-plane intake, it should make at least another 25 hp on a tunnel ram and turn an additional (and useful) 300 to 400 rpm. It may not be the cheapest of intakes to build, but the results are worthwhile.
Tunnel ram installations intended to use carburetors and a pair of 4-barrel injector throttle bodies can be used to equally good effect.
Although cast tunnel rams are common, the casting process does make the whole manifold assembly somewhat heavier than might be desired for an all-out race vehicle. This has brought about the popularly used sheetmetal intake so commonly seen on Pro Stock engines and the like. Indeed, the sophistication of specialist-built, custom tunnel ram intakes has taken the state of art in this field to levels hardly conceived of—even in Formula 1—just 15 years ago.
Written by David Vizard and Posted with Permission of CarTechBooks