The engineering definition of a camshaft is any device that converts rotational motion into translational motion. Cams are used in everything from rifle scopes to most machines from the Industrial Revolution. We had the opportunity to use our high-speed race-engine development tools to improve the cam lobes for the oxygen generator pumps of a fighter aircraft.

These same design techniques are quite useful for direct-injection fuel-pump lobes as well. However, for the scope of this book, I am focused on camshafts for performance and racing, thus, dealing with the camshafts that move the valves of a four-stroke engine.

This direct-injection pump tri-lobe is closer to what is found throughout the rest of the book than the compressor pump lobes, but it may not be quite what you might have in mind. Increasing the stroke of the direct-injection pump while also increasing the dynamic stability at high RPM has given Comp a huge advantage in direct-injection racing applications.

 

 

Image 1-6: This image shows the layout and components regarding how the direct-injection pump fits on the back of an GM LT-1 engine. The spring acts just like a valve spring, so careful profile development here can limit engine speed just as much as it can on the other 16 cam lobes of this engine.

 

Image 1-7: There are a few options regarding engine valves. Poppet valves are dominant over rotary valves and other options because they can open to the high-flow position quickly and seal the chamber effectively. They can do so while requiring minimal intrusion into the combustion chamber. If they did not spend so much time fighting with the piston for real estate in the chamber, they might be considered perfect for internal-combustion engines.

Camshafts open and close poppet valves to allow airflow out of and into a combustion chamber. Then, the valves seal for optimum combustion and power. What first shocked me as I learned about cam design was how few of the modern techniques involve designing the physical camshaft. If you went back to the 1920s, engine designers designed a camshaft lobe geometrically. The standard profiles were initially drawn by hand as a circle for the nose, a circle for the base circle, and two arcs that linked the two (as shown in Image 1-8).

Image 1-8: The first engine cam designs were constructed from three tangent circles. The blue arc is the base circle. The green arc (mirrored on the closing) is called the flank arc. The red arc is the nose. These were typically constructed as a model lobe on a bench grinder that was polished and used to create a master. The brown line tangent to the flank is both the lifter face at 60 degrees and the model grinder setting when this was created.

These three-arc designs were improved by moving from the profile to a tappet motion-centric design approach. Each of the black tangent lines in Image 1-9 represent the flat-tappet face position at every 10 degrees from the nose. In this graphical format, the lobe is static and rotates the tappet bore and engine around the camshaft. This relative motion is a common way to develop the cam surface today and is generally known as the theory of envelopes.

Image 1-9: More detail is shown regarding how that lobe was created. In this view, the lobe is stationary and the tappet rotated around the camshaft. Each gray line represents both the tappet face and the model grinder tool path. That 1:1 relationship between tappet lift and model grinder setting certainly helped the early cam designers. With modern valvetrain systems, these designers would need a lifetime spent without computers to hand-calculate all the conversions.

Image 1-11: I made this graphic with the marketing department for our NHRA trailer with the idea that it could help customers with questions at an event. In the middle, the black shape is a section of a camshaft. This helps, but it shows the endpoint of a cam design and not where we start.

We made Image 1-11 more than a decade ago as a tool to help customers think about camshaft terms, and we never knew it would become such a popular internet meme. It is certainly worth studying. However, when designing camshafts, the lobes surfaces are the final piece to the puzzle but not our focus. This is rather unique to camshafts. When engineers design a connecting rod, they start by drawing a rod shape. When one starts designing a cylinder head, they typically start with a port shape or valve and chamber. Even when designing a crankshaft, the designer starts with the throws and mains and then starts adding and shaping the counterweights.

Unlike these other parts, when I start designing a camshaft, I always start with the motion of the valve that I am trying to control. I probably look at plots like Image 1-12 20 or more times every day and typically place one camshaft valve motion over another to think about how one might perform better than the other in a given application. We will begin by defining the acronyms and terms on this plot and build upon that foundation.

Image 1-12: This format is more useful for looking at a cam design than the lobe surface. We always start with the valve motion desired and work our way backward through the valvetrain to the lobe profile. Sometimes, the system limits what you can do, so the lobe limits can force you to change your design ideas. This was common with 0.842-inch flat-tappet lifters and small journals, but most newer engines have very few design limits.

Basic Terms

There are words that I have used hundreds of times, thinking that I understood them well. Then, I wound up embarrassed when I tried to write a strict definition. This is not just a struggle for our day, but C.S Lewis wrote a wonderful essay titled “The Death of Words” (published in The Spectator on September 22, 1944) that points out how any word is only useful when everyone understands its exclusive meaning. When I describe how long a camshaft lobe keeps a valve open, I often specify its duration at a given tappet lift to give far more information than saying something general like, “It is pretty big,” “3/4 race,” or “Stage 2.”

Slogging through definitions may be a bit burdensome, but a few are rather tricky (lobe separation). Even the simple ones (lift) need qualifiers added (tappet or valve). Going through these terms in detail will increase understanding and hopefully provide useful perspective.

Tappet or Follower

Every cam lobe has a follower (lifter or tappet) that rides along the surface to convert the cam rotation into valve translation. Most people are familiar with flat and roller tappets. Instead of designing around a lobe shape, the actual design is either focused on valve motion (best) or tappet motion, especially in applications that might run any of several various rocker-arm geometries). The lobe shape is simply what is required to achieve that motion.

Image 1-13: Several 0.842-inch flat-lifter designs are shown. All have a 50 to 60-inch radius face (not flat). Most have very small edge chamfers, and the three on the left have diamond-like carbon (DFC)-coated faces. The third lifter from the right has a chilled iron foot insert. When we design a camshaft, we often move quickly from the valve motion to designing how we want this lifter to move up and down the lifter bore relative to the resting position on the base circle.

Image 1-14: Roller lifters (followers or tappets) have a number of advantages over flat-tappet lifters, especially when trying to maximize lift and area. These are Comp Cams Sportsman solid-roller lifters for a big-block Chevy. Just like designing for a flat tappet, we design the lifter motion and not the lobe shape. With a roller, we don’t have to worry about going off the edge, but the flank region can become too concave (inverted) to grind.

-injection racing applications.

 

 

Image 1-15: This shows the valve back though the rocker and pushrod to the lifter and camshaft for a modern LS solid-roller design. Thinking about this as a system is the preferred way to design any camshaft. Note that we are designing how we want the valve to move and convert it back through this system to the required profile.

 

Image 1-16: The wrong approach is to take an earlier design, maybe something from the 1980s that was developed for a small-block Chevy (1.868-inch journal), and grind that same lobe surface just 0.200 inch larger on either this 2.165 LS or even larger for the 2.244-inch Hemi journal sizes. Doing so greatly increases the valve quickness and will most likely result in failure. Some companies may try this approach, but it is the wrong method.

Tappet Lift and Valve Lift

Image 1-17: When we talk about tappet lift, think about the dotted path of the center of the wheel axle around the lobe surface. That is not the same as the wheel radius plus the lobe surface, as the wheel contacts the lobe farther out as velocity increases to create a different path.Lift should be easy, right? Tappet lift is defined as the rise of the tapper (follower or lifter) from its position on the camshaft base circle.

Likewise, valve lift is defined as the distance the valve moves for its resting position on the seat. Maximum lift for any profile will be the highest tappet rise above the base circle achieved at any crank angle.

Image 1-18: Moving from tappet lift to valve lift, it is easiest to put a dial indicator on the retainer. This shows the dial indicator lined up parallel to the valve to measure valve lift accurately. (Photo Courtesy Jeff Smith)

Image 1-19: Chris Brown, the manager who initially hired me into the Comp Cams tech office, made this small-block Chevy inspection stand so that we could practice degreeing camshafts and watch the valves with a dial indicator and see how it opens the ports as they move. It is helpful to have a sectioned engine like this to play with when thinking about camshafts and engines.

TDC

This is the crank angle where the piston stops rising in the bore, changes directions, and begins to drop. The piston stops at the top twice every four strokes: once near the plug-firing event that begins combustion and a second time as the exhaust valve is closing and the intake valve is opening between the exhaust and intake strokes. We will discuss finding top dead center (TDC) in detail in Chapter 2 of High-Performance Cams & Valvetrains: Theory, Technology, and Selection.

Image 1-20: At TDC, the crank throw and rod line up with the cylinder bore to force the piston to the top of the stroke. At TDC overlap, both valves are open and snuggled very close to the piston. About 330 degrees later, the spark plug ignites and combustion begins as the piston approaches TDC firing.

ICL

The intake centerline (ICL) is the angle in crank degrees from the piston reaching TDC (overlap) to the intake valve reaching maximum lift. Looking back at Image 1-12, the intake reaches peak lift 109 crank degrees after TDC, hence a 109 ICL or intake centerline. The engine’s performance and intake valve to piston clearance are highly dependent on the intake centerline.

ECL

The exhaust centerline (ECL) is the angle in crank degrees from the exhaust valve reaching peak lift and the piston reaching TDC. Again, in Image 1-12, the exhaust reaches peak lift 119 crank degrees before TDC, hence a 119 ECL or exhaust centerline. As with the ICL, this ECL is very important to performance and the exhaust valve-to-piston clearance.
Lobe Separation
The term “lobe separation” traces itself back to a time when all intake and exhaust lifters for each bank were always in a line. This is the arrangement on the flathead Ford, small-block Chevy, and continues in most production V-8 engines, including the GM LS, LT, Ford Godzilla, and Chrysler Hemi.

Image 1-21: Even with the drastically different valve angles of the Gen III Hemi, the engineers at Dodge did a great job setting up the rocker system for rather straight pushrod angles with a single lifter bore angle on each cylinder bank. Keeping pushrods straight while reducing machine setup is very important on production engines.

Image 1-22: The GM LS is built around a single lifter-bore angle, but it is much easier with an inline-valve Wedge head. The LS angle is 45 degrees from vertical, which makes the camshaft math really easy.

Image 1-23: The big-block Chevy stock 38.75-degree intake and 45-degree exhaust lifter-bore angles help keep the lifters and pushrods aligned with its canted valve arrangement and the same rocker design on intake and exhaust. However, having splayed lifter-bore angles lead to a turning point in the definition of lobe separation. The engine would not perform properly with the same camshaft-lobe angles on the two banks. To keep things less confusing for the aftermarket, people began using the term lobe separation as the average of the intake and exhaust centerlines and not the camshaft angle.

While inline lifter bores are the most common arrangement, applications like the big-block Chevy, as well as most NASCAR and NHRA Pro Stock engines, have the exhaust lifters rotated out to a different angle than the intake for straighter pushrod angles. Pushrods are very stiff in pure compression but not in bending, so you want a very straight load path. Hence, on applications like Dan Jesel’s Equal 8, where every possible advantage has been investigated, the exhaust lifter bores are rotated considerably compared to the intakes.

Image 1-24: On a modern race engine, such as Dan Jesel’s Equal 8, the lifters can be rotated significantly. Hence, to use the lobe-separation term on something like this to mean what it means on a normal engine, the average centerline definition is extremely helpful.

Looking back at Image 1-23, if the physical angle between the intake and exhaust lobes was 110 degrees on the odd side of a BBC [Author: Should “a BBC” be changed to “BDC”?] the exhaust lift will reach max later (later ECL) by 6.25 cam degrees or 12.5 crank degrees. Hence, it would act like a 97.5 lobe separation in terms of the crank angle between max exhaust and max intake lift. On the even side, this 110-degree cam lobe angle acts like a 122.5-degree lobe separation on straight lifer bores.

Because lobe separation is a key idea for engine builders, the industry changed the definition of this term. Instead of the physical angle between lobes that is implied, industry-wide lobe separation is defined as the average of the intake and exhaust max lift centerlines or (ICL + ECL) ÷ 2. With this definition, we can talk about lobe separation for overhead-cam applications, such as the Toyota 22RE in Image 1-25, even when there are no real lifter bores.

Image 1-25: Using the new definition of lobe separation, we can incorporate this term even with overhead camshafts where there are no lifter bores. We simply design the valve motion and index the lobes so that the max valve lift centerlines are apart by the required average centerline.

Advance

If we went back to the early flathead automotive days, engine builders installed their camshaft with the number-1 piston at TDC and then rotated the camshaft with a level on the exhaust and intake valve faces until they had the same lift and were level. This was called “splitting overlap” and was a common way to install a race camshaft, which typically resulted in a nearly equal ECL and ICL. Going back to the 110 LSA example, this represented a 110 ECL and ICL. However, it did not take long for people to see that if the ICL was moved a few degrees closer to TDC, the engine typically responded well, especially at lower RPM.
The term “advance,” often abbreviated as “ADV,” was coined as the difference between the lobe-separation angle (LSA) and ICL. If you rotate the camshaft 2 cam degrees, which equals 4 crank degrees forward (4-stroke camshafts typically rotate at half crank speed), the ICL is 4 degrees less than the lobe separation.

Looking again at Image 1-12, this camshaft has a 109-degree ICL and 119-degree ECL. The average of these numbers (109 + 119) ÷ 2 is the 112-degree LSA. Calculating the advance is simply the 112 LSA minus the 109 ICL (or 5 degrees). Using lobe separation and advance is a more common way to describe camshafts in the American pushrod V-8 world because there are so many easy ways to change the advance by rotating the camshaft with respect to the crank using adjustable timing sets. If a camshaft was ground for 109 ICL and 119 ECL but installed it with a four-degree retard timing setting, the centerlines will move to 113 ICL and 115 ECL.

Note that we move both the ICL and ECL when we advance or retard a camshaft, but the average of these centerlines or lobe separation will not change. However, in applications with independent intake and exhaust camshafts, it is easier to talk about the ECL and ICL settings, as they are not forced to move together.

Image 1-26: These are the Excel calculation sheets we used based on the timing set pin or keyway angle at TDC, lifter bore angles, lobe separation, and advance to calculate the lobe angles down any camshaft. Some overheads are a little more complicated, as the effective lifter-bore angles can change with lift or base-circle size.

The relationship among LSA, advance, ICL, and ECL is as follows:

• Advance = (exhaust centerline – intake centerline) ÷ 2
• Lobe separation = (exhaust centerline + intake centerline) ÷ 2
• Exhaust centerline = lobe separation + advance
• Intake centerline = lobe separation – advance
• Retard = opposite of advance

Duration

If I have a pet peeve, it is when someone specifies a duration without specifying the lift. If you want to see me squirm, ask me, “What is the best 280-duration camshaft?”

Duration is defined as how many crankshaft degrees a camshaft profile maintains more than a given lift. For most cam-in-block, overhead-valve applications, durations are given with respect to lifter or tappet lift. Duration is only meaningful if the lift at which it is measured is given. Any advertised duration without the specified lift (which is unfortunately very common) is an arbitrary and meaningless number. Almost every profile will be 300 degrees of duration at some tappet lift, but it is also 200 degrees of duration at another lift.
You can ignore any duration number that is given without a corresponding lift value. At Comp Cams, all overhead cams are rated at the valve, so these duration numbers do not seem to correlate well to how those same camshafts check outside the engine with a cam doctor-style gauge.

 

The way many engine builders compare race profiles is to compare the durations at 0.020 inch, 0.050 inch, and 0.200 inch for a given lobe lift (as shown in Image 1-27). The assumption is that for a given 0.050-inch duration, the smaller the 0.020-inch duration and largest 0.200-inch duration, the squarer and better performance a lobe can produce if stable. The problem with this point of view is that air demand and flow are never constant through either the intake or exhaust port. We now know that fast opening/slow closing lobes run differently than slow opening/fast closing lobes, even when all the durations are the same.

Image 1-29: The engine cares a great deal about how the valves move, and these two different tappet-lift profiles were designed to achieve similar valve motion. They are not exactly the same, as the red curve still gets a few thousandths ahead at 120 degrees, but almost no one would look at those tappet lift specs in a lobe catalog and guess how close these are at the valve.

People often confuse tappet durations and valve durations. A 256 at 0.050-inch tappet lift design (such as the Comp part number 23323) with a 2.0:1 ratio is extremely close to the same duration at the valve as a 261 at 0.050-inch tappet lift design (Comp part number 33808) when it is coupled with a 1.8:1 ratio. Note the specs for the 33808 seem more aggressive when looking at the 0.020-, 0.050- and 0.200-inch tappet durations compared to the 23323.

However, with the increased rocker speeding up the motion on the 23323, the resulting valve motion is extremely close to that of the faster 33808 with less rocker ratio. Both these lobe families are extremely successful but only when coupled with the correct ratios. This was only a 0.2 change in rocker ratio, but it corresponded to a 5-degree change in valve duration, which is a nice growth reference to remember. When someone makes an extreme ratio change (for example: 1.6:1 to 2.2:1) to properly select a camshaft, I need to overlay the valve motion plots between the systems. Otherwise, I risk being lost on the new tappet duration.

Written by Billy Godbold and Posted with Permission of CarTechBooks