My engine theories are based on a heavy research effort since the late 80's. I have studied engineering texts, selected magazine and internet articles, as well as talking with engine builders, racers, and cam specialists. The people I have had the priviledge to exchange ideas with, as well as learn a few really cool things from, have proven to be invaluable. I am a fan of most all engine combinations that are built within the structured hardware limitations of class racing. As a mechanical engineer, I do not subscribe to the idea that a motor works because 'it just does' or from mystic voodoo black magic than only a few understand. An engine is a mathematically quantifiable system and should be treated and/or modified in this light. However, the problem that will always exist with math calculations is: garbage in equals garbage out. An accurate model is required to give accurate results.
I am an engine enthusiast to say the least. Cubic dollars have prevented me from furthering my involvement in race engines other than my own bracket car and engines in some of my friends’ cars. My ideas come from my studies as a mechanical engineer and reading technical data of any shape or form pertaining to engine theory.
In 1988, I began formulating a simple calculation that would enable me to approximate the power level of a particular engine size and family, based on intake flow and rpm. At that time, I was using maximum cylinder head flow in my calculations and assigning compression ratios to a particular cam duration. Since then, I have quantified the role of port volume/area as well as cylinder pressure in the engine parameters. My engine calculations are continually evolving and are getting fairly detailed, however most of the last few years (since '95) have been spent formulating valve timing calculations. After several years of relying on curve-fit equations for rpm versus duration, I came to the realization that this was not reality. The equations provided a very convenient baseline; but at some point, valve timing must become a collection of opening and closing points, not an imposed duration value.
I am including general ideas on cylinder heads, cams, and to a lesser degree (but no less important) intake manifolds and headers because these components are among the most common bolt-on parts, and therefore the most commonly mismatched parts.
Cylinder Head Requirements:
Maximum runner flow is somewhat meaningless without the total flow curve throughout the complete lift range of the cam. Flow occurs from the time the valve is opened and until it is closed. If the flow values do not match the needs of the engine throughout the lift range, then the ability to make good power is gone everywhere but at that those specific matching valve lifts. The end result is poor cylinder filling because of flow inadequacies. If you apply this power loss to every cylinder over the whole rpm range, it becomes quickly apparent that the power output has taken a significant hit. In short, the cylinder is not adequately being filled and the cylinder pressure is suffering. Inadequate cylinder pressure adds up to reduced torque.
A useful relationship to keep in mind is the runner's flow per cross-sectional area. If the flow area is calculated over the entire lift range and divided by port's cross-sectional area, the resulting proportion is a great benchmark for comparing runner effectiveness. Using a method like this helps to average the total flow capability, over the complete valve lift range, and yields a more realistic look at the power output based on intake flow capability.
The intake manifold’s ability to affect the cylinder head’s flow capability cannot be overlooked. It is beneficial to have flow data with the intake manifold passage attached to the head, as well as the intake manifold runner volumes and cross-section areas. The combination of the two port runners make up the engine’s intake port, not just the 3-6 inch length cast into the cylinder head.
This is an area where my theories may deviate from what I classify as normal thinking. Correct flow quality must be maintained within the cam’s lift range, just as on the intake port. I think of the exhaust port as an extension of the valve seat/bowl geometry and the header tube as the actual exhaust runner. Because of this relationship, the header tube should be installed when flow data is taken on the exhaust, or at least a tube stub simulating the header diameter and transition with enough length to introduce flow losses like the actual header. All modifications to the port cross-section should be in the interest of maintaining an effective transition into the header tube, as well as increasing flow. The exhaust runner will behave differently than the intake runner because it is a tube with constant cross-section (usually). Short of using inertial wave effects to improve the overall port effectiveness, optimizing the seat geometry, valve bowl area, and effectively transitioning the flow to the header tube is about all than can be done because the maximum flow capability is dictated by the port's cross-sectional area. The header primary tube is the critical dimension in the exhaust path and therefore controls maximum flow capability. This total flow is even effected by the exhaust system as a whole.
Runner Volume and Cross-Sectional Area:
Runner Volume, like maximum runner flow, is meaningless without the complete picture. The port's cross-sectional area is the real parameter to characterize. Compare the small block Chevrolet intake port volume to the Ford's. A 200cc runner in the Chevrolet head can be considerably different than the same volume in a Ford small block head. The stock Chevrolet runner is about and inch longer, which for a given port volume will yield a smaller port cross section. If runner volume is the only quantity known, then the port length has to be measured to calculate an average cross-sectional area or the port must be disected to find the most representative cross-sectional area. An engine’s torque peak is directly related to the intake and exhaust runner’s cross-sectional area. Large cross-sections, in proportion to the cylinder volume will produce higher rpm torque peaks.
Along with port cross-sectional area, is the requirement of the port to produce flow velocities indicative of an efficient port design. Therefore a minimum flow rate exists for a particular cross-sectional area to maintain this velocity. It is important to realize that forced induction and high volumetric efficiency motors have breathing capabilities beyond their cubic inch limitations and will follow these trends a little differently. In these special cases, the engine components must reflect this increased breathing capability. In short, if an engine has substantial intake flow capability, then the exhaust side must be equally sized, regardless of the engine's cubic inch displacement or horsepower rating.
The camshaft is the brain of the motor. It controls everything that happens. Because of this, I have basically abandoned all cam catalog applications and go directly to the lobe specification sections. An off-the-shelf cam is a compromise, at best, to get optimum valve timing into a wide range of engine set-ups. Because of this marketing scheme, you see information such as minimum compression ratio, generic intake and exhaust requirements, and rpm ranges for particular families or series of grinds.
Look at two engines, each with the same displacement and geometry, one of which has very large cylinder head runners with very high flow capability and the other with virtually stock heads. Each engine is outfitted with identical intake manifolds and headers/exhaust. The two engines will exhibit very different requirements for valve timing for similar performance and rpm capability. This is the sort of information that is not addressed in most catalog offerings. Equally different, are two engines outfitted identically, but with differing cubic inch displacements. In these examples, assigning camshaft specifications based on broad generalized ideas used by most off-the-shelf marketing techniques can fall short of realistic engine requirements. A camshaft selected on rpm alone is not going to match the engine's performance needs unless the camshaft was designed around a specific set of engine parameters, and your set-up matches it identically.
The cam should always be the last thing chosen on a motor set-up. The ramp profiles can crutch low-lift flow issues, intake and/or carburetor restrictions, and exhaust inadequacies. Overlap quantities are selected to increase cylinder scavenging, aid in low or high flow velocity situations, or to meet special idle or driveability requirements. Intake closing points tailor rpm requirements. Exhaust opening points will reflect the exhaust system capability. All of these characteristics paired up with the vehicle's gear ratios will define the overall package's effectiveness.
: Specific durations and lobe separations have not been mentioned. Engines do not read cam catalogs or magazine articles. An engine makes power from cylinder pressure, port velocity, and rpm. Valve openings and closings are the real numbers to be concerned with. We always speak in terms of duration and lobe separation; but keep in mind, those values are byproducts of the valve openings and closings.
One of my early theories was to apply a minimum static compression ratio to a particular camshaft duration. This was only a part of the puzzle. Compression ratio helps, but what about the low compression motors that make big time horsepower? Cylinder pressure and port velocity are the key players in determining power output. The camshaft is the controlling factor for both of these parameters. By optimizing the timing events within the crank cycle for the required rpm band, the cylinder pressure can be tailored to compliment the engine geometry and components. Typically, the goal is to maximize (or atleast optimize) the cylinder pressure within the engine's intended rpm range.
The media globally labels the Lobe Separation Angle (LSA) as a major selection criteria for choosing a camshaft. Narrower LSA cams are usually touted as having poor idle and driveability characteristics. Part of the reason they have such limited usage guidelines is because of their supposed 'peaky' powerband characteristics. If the cam design is allowed to compliment high port velocity and cylinder pressure for the intended rpm range, then idle and driveability will be no worse than a wider LSA cam used in its correct application, for the same rpm range. In reality, the LSA is a relative value and does not pertain to the intended usage of the engine combination. Poor idle and driveability characteristics come from mismatched components and from the purchase of components with design requirements way out of line from the engine’s needs.
No text on cam design can be complete without discussing overlap. Overlap has been a 'buzzword' around camshaft discussions since hotrodding began. Technically, overlap is the point in the crank cycle at which there is charge exchange during the intake opening and exhaust closing at TDC. It can be an extremely misunderstood process. Like any engine subject: more is better, less is better, it depends on whoever's musings you read or listen to. Which is true, why is there so much contradiction? The media has applied guidelines, like the lobe separation theme, that are unconfounded. The ideas may not be wrong, it's just that the complete story is not told.
Overlap is required to aid in the air exchange, increase the overall breathing capacity of an engine, and tailor rpm band width. Two terms generally associated with overlap are 'scavenging' and 'reversion'. The situation exists that if there is too much pressure in the exhaust, due to flow restriction or inertia wave pulses, then it is possible to reverse the flow in the intake port (reversion). Likewise if the exhaust pressure is too low, creating a vacuum, it is possible for the incoming charge to be sucked past the chamber and go out the exhaust (over-scavenging). The goal is for the pressure differential and charge momentum to be just enough so that the intake charge is helped into the chamber, but the exhaust valve closes just in time to prevent excessive intake charge from being carried out with the exhaust. The key players in determining an engine's overlap requirements are the port cross-sectional areas, flow capability, rod/stroke ratio, static compression, and rpm.
Overlap is a quantity of flow per degree of crank rotation; not so much an angular value based on the engine's intended operating environment. When you look at overlap as a quantity of air exchange per crank rotation or piston displacement, it becomes apparent why very high flow capability motors use less overlap and therefore wider LSA's. The angular measurement is less, but the actual air exchange is still occurring.
I did not get into any of the Inertial Wave Tuning effects theory. I am not sure I have an engineering grasp on the topic or the assumptions that can be made to account for the temperature and pressure variations. The topics I have spoken about are in the interest of maximizing power output per a given engine set-up. These topics along with the inertial wave tuning are vitally important and play a major role in the engine’s power making ability. But, I believe that by applying the relationships of the port size, flow, and rpm, you can at least be in the game even if you are not hitting home runs. Based on these ideas, it appears that there may be power limitations for an engine set-up by omitting inertial effects. I believe this is true. However, a good baseline set-up is going to provide a much better starting point. Designing an engine set-up around inertial effects without careful consideration of the flow parameters (cross-sections, flow, velocity, valve events, cylinder volumes, etc) is pointless.
I think the ‘Class’ racers/builders have the best knowledge of engine set-ups, as opposed to the builders that have the ‘eternal quest’ for maximum power in an unlimited environment of hardware and components. I am referring to the ‘class’ engine builders in drag racing, stock car, and sports cars that are forced to work within a particular hardware ‘envelope’. Once you get knowledge of how to make a set-up work within an ‘envelope’, then all buildups will benefit because of the understanding of what various components do and what the limitations are.
Overlap and Compression- A very common idea, although for the most part incorrect, is that overlap bleeds off compression. Overlap, by itself, does not bleed off compression. Overlap is the angle between the exhaust closing and intake opening and is used to tune the exhaust's ability draw in additional intake charge as well as tuning idle vacuum and controlling power band width. Cylinder pressure is generated during the compression cycle, after the intake valve has closed and before the exhaust opens. Within practical limits, an early intake closing and late exhaust opening will maintain the highest cylinder pressure. By narrowing the Lobe Seperation Angle 'LSA' for a given lobe duration, the overlap increases, but the cylinder pressure can be increased as well. Thus cylinder pressure/compression can actually increase in this scenario, by the earlier intake closing and later exhaust opening. By increasing duration for a given LSA, the overlap will increase, the intake closing will be delayed, and the exhaust opening will occur earlier. This will decrease cylinder pressure, but the decrease/bleed-off of compression is not due to the overlap, it is due to the intake closing and exhaust opening events.
Adjusting Lash on Mechanical/Solid Cams- If valve lash changes significantly over time, then something is wrong. Cam wear is very slight, along the order of .002 or less. If the lash setting changes more than .005 then there has been a component failure (loosened hardware or actual mechanical failure). Lash settings should be taken/adjusted at the same temperature and same order as the previous or original setting. This is the only way to rule out expansion/contraction of the components from temperature changes. This temperature delta is usually the culprit of most valve lash dilemmas. At initial start-up and break-in of a new set-up: cam, lifters, rockers, pushrods, valve job, etc., the lash may move around during the break-in procedure and for a short time after. This is because all the parts are seating into their new wear patterns. Once this occurs, the lash setting should stay steady.
Hydraulic Lifter PreLoad- Hydraulic lifters are intended to make up for valvetrain dimensional differences as well as providing a self-adjusting method of maintaining valve lash, or rather the lack of. By setting the valvetrain so the lifter plunger is depressed slightly, the lifter is able to compensate for these differences, making a convenient hassle-free valvetrain set-up. For performance applications, lifter preload is not needed or wanted. As rpm's increase, the lifter has a tendency to bounce over the back of the lobe as it comes back down from the maximum lift point. The pressurized oil fills the lifter body to account for this bouncing. Eventually, after several engine revolutions, the oil can completely fill the lifter body and the plunger will be pushed up to its full travel (pump-up). Higher oil pressures can amplify this problem. With the lifter pre-loaded, this can cause a valve to run off it's seat and can cause piston clearance issues if and when pump-up occurs. By setting the valvetrain at 'zero' preload, lifter pump up is eliminated and in most cases, the cam will rev higher. Ford tech articles in late 60's actually urged 'stock' class racers to run .001-.003 lash on hydraulic cams.
Piston To Valve Clearance- Piston clearance is a function of lobe geometry and phasing to the piston. Cam lift should not be a deciding a factor in clearance issues. Valves will hit the piston in the overlap period, while exhaust is closing and intake is opening. Exhaust clearance problems will typically occur just before TDC and intake just after TDC, not at max lift. Some cylinder head venders and other component manufacturers advertise a max duration or lift before clearance issues arise. This is very misleading. Maximum safe duration is a totally bogus value, and is completely worthless without knowing anything about the ramp rates or actual timing/phasing events of the installation. At least with maximum safe lift, the vendor can a apply a rediculously fast ramp at a very early opening/closing and arrive at a somewhat meaningful measurement, but without knowing the design specifics the information is still next to useless.
Custom Ground Camshafts- When the performance of a particular engine combination is wanted to be optimized, the camshaft design parameters are calculated from the engine and vehicle specifications to perform within specific conditions. Let me emphasize that last statement, 'within specific conditions!'. In no way was total maximum power for the engine implied. The intent is to maximize performance within the intended design parameters. If that means taking a pro-stock motor and wanting to run it from 2000-5000 rpm, then so be it.
The camshaft's seat timing events, ramp rate, and lift are directly related to the intake and exhaust flow capabilities, crankshaft geometry, static compression, rpm range, as well as other criteria. A camshaft selected in this manner, becomes personalized to that particular engine combination. Usually a custom grind is selected as an intake lobe and exhaust lobe with a particular phasing to each other (lobe separation angle, LSA) and sometimes a specified amount of advance or retard is built in. Although, it could easily end up having completely reengineered lobe characteristics, requiring new lobe masters with specialized ramp requirements. It is possible for an off-the-shelf camshaft to be a classified as a 'custom'. If the cam design is calculated for a particular combination and an off-the-shelf part number fits the bill, then for all practical purposes that part number is a 'custom' cam (but only for that particular set-up).
Typically, cam catalogs do not specifically list custom ground camshafts, because the possibilities are endless. They stick to particular series or families of camshafts. The superstock grinds come closest to an off-the-shelf grind that is truly optimized for a combination. There will be small differences due to header sizes and engine builder's 'secrets, but usually the catalogs are pretty close to a good baseline. Likewise, brand to brand, the grinds will be very similar because of the 'class' dictated combinations and the flow characteristics being so well documented
Degreeing Camshafts- There is no special magic involved for degreeing a camshaft during installation, but this is not the same thing as random advancing, retarding, or installing the gears 'lined up'. Degreeing a camshaft involves definite known values for valve events. Typically this is specified as an Intake Centerline or as opening/closing events at specific lobe lifts. This is done to insure the cam is installed per specific requirements, such as a recommendation from an engine builder or the vendor's data sheet for that camshaft grind. Manufacturing tolerances and shop practices do not guarantee that the cam matches the data sheet, when installed at crank gear 'zero'. The cam will usually need to be advanced or retarded to the correct location. If it is correct, at crank gear 'zero', then the cam has still been degreed. It just did not require any additional tweaking to meet the requirements. This is what degreeing a cam is all about; the verification of the installation. A common mis-used term is the 'straight-up' installation. Typically this is described as installing the cam at crank gear 'zero'. This is 100% wrong. Straight-up refers to the Intake and Exhaust Centerlines being the same. In other words the cam will have no advance or retard at the installation, regardless of the amount of advance/retard ground in by the vendor. In reality, the cam may have to be advanced or retarded (from crank gear 'zero') significantly to arrive at a straight-up installation.
Exhaust System Diameter and Engine Horsepower- A popular idea is to select/size the exhaust system components to the engine's horsepower output. This idea typically attributes a header diameter or an exhaust system diameter to a particular horsepower level. To resolve this, look at how an engine operates and consider one cylinder. The cylinder will move a volume of air based on its crankshaft geometry, rpm, and sealing capability. The amount of air that can enter the cylinder is dependant on the intake flow capability, crank geometry, rpm, and valve timing as a minimum consideration. Likewise, the amount of air that exits the cylinder is dependent on the same characteristics.
An engine's output is usually thought of in terms of horsepower. Actually, an engine produces torque, and the horsepower is calculated through a units conversion. The amount of torque an engine can produce is directly related to the amount of cylinder pressure generated. This is all affected by the same previous characteristics (intake and exhaust capability, crank geometry, rpm, valvetiming, etc). So basically an engine's power output is about air exchange capability. Using this line of thinking, look at the exhaust path again. The exhaust system is more reflective of the engine's ability to move air, as opposed to horsepower numbers. Engine output does not address the breathing aspects of the engine and is probably not a good rule to use for exhaust sizing.
There is a very good reason that tuners/engineers/specialist have attempted to assign exhaust to intake relationships around 70-80% for a typical natural aspirated set-up. In non-detailed terms, it is a range that offers a good balance for power capability. Other relationships, such as 1:1, are used and they work very well, but these methods have to be applied and tuned for very specific circumstances. This relationship does not stop on the flow bench, it goes all the way from the intake path opening to the exhaust system termination. In short, try to maintain exhaust sizes that are inline with the intake capability. Also, do not stop your analysis at the intake and exhaust paths. If the engine already has the camshaft, look at the valve events. If the specs favor a restricted exhaust (indicated by early and wider exhaust openings with wider lobe separation angles), then size it accordingly by using exhaust components with smaller cross-sections. If the valve timing specs favor the intake, then the engine needs some serious exhaust flow capability which is only possible with larger cross-sections.
This section was written with natural aspirated combinations in mind. However, by using the 'air exchange' rationale, it becomes apparent why forced induction engines typically benefit from increased exhaust flow capability. Also, look at the nitrous combinations. The intake system remains virtually unchanged, yet with the major increases in cylinder pressure it acts like a substantially larger engine on the exhaust side, requiring earlier exhaust openings and/or higher exhaust flow capability.
Pushrod Length- Incorrect pushrod length can be detrimental to valve guide wear. Most sources say that centering the rocker contact patch on the valve stem centerline at mid valve lift is the correct method for determining the optimum pushrod length. This method is wrong and can actually cause more harm than good. The method only applies when the valvetrain geometry is correct. This means that the rocker arm lengths and stud placement and valve tip heights are all perfect. This is rarely the case. To illustrate this, think of the valve angle and the rocker stud angle. They are usually not the same. If a longer or shorter valve is installed, then the relationship of the valve tip to the rocker stud centerline has changed. Heads that have had multiple valve jobs can also see this relationship change. Note, the rocker length (pivot to tip) remains unchanged, so the rocker contact patch will have to move off the valve centerline some particular distance for optimum geometry to be maintained.
The optimum length, for component longevity, is the length that will give the least rocker arm contact area on the valve stem. In other words the narrowest wear pattern. This assures that the relationship is optimized and the rocker is positioned at the correct angle. This means that the optimum rocker tip contact point does not necessarily coincide with the valve stem centerline, and probably will not. What is the acceptable limit for being offset from the valve stem centerline? That will depend on the set-up. A safe margin to strive for is about +/-.080" of the centerline of an 11/32 diameter valve stem. This means that no part of the wear pattern should be outside of this .160" wide envelope. As the pushrod length is changed, the pattern will change noticeably. As the geometry becomes closer to optimum, the pattern will get narrowest. If the narrowest pattern is too far from the valvestem centerline, then the valve to rocker relationship has to be changed. In this case, valve stem length will need to change.