By way of review, last month we shared responses from three highly qualified valvetrain authorities: Thomas Griffin and Billy Godbold from Comp Cams and Larry Tores from T&D Machine Products. Their comments were woven together with some personal perspectives on the subject. The following questions and comments are geared toward expanding the Saturday-night racer's understanding of valve motion and how it relates to making power with parts that live.
With some Winston Cup engines now using multiple lobe profiles on the same shaft, how can rocker-arm ratios be selected accordingly? Let's talk about this for a moment. The issue takes us back several months to material that appeared in the Circle Track Sept. 2001 issue dealing with single-cylinder tuning and optimization. The discussion included ways to address how cylinders "talk" to each other in single-plane intake manifolds and "collector-joined" exhaust systems. Various and predictably varying pressure excursions can influence intra-cylinder combustion efficiency and power output, especially in the intake system.
Stated another way, some cylinders receive more air and/or different air/fuel charges than others. Among the reasons for that are inequalities in cylinder-to-cylinder airflow created in a running engine, despite efforts to "calibrate" or equalize flow during air bench tests. As a result, some cylinders make more power than others. While the ability to identify and confront this problem typically requires precision measuring equipment and a fair amount of analysis experience, Engine Cycle Analysis or ECA (also a prior CT subject) is gaining in use and popularity among the more aggressive race teams and engine builders.
Even though some responses to this question may appear a bit off target to its specific content, there's clearly some "between-the-lines" information you can gather from the following comments.
"Most, if not all, of the Winston Cup teams have very extensive dyno testing programs as well as on-track testing," Tores says. "Part of the tuning process for a given racetrack (length, banking, etc.) will be the rockers, as well as valvesprings and valves. Usually, the results of Spintron and dyno testing will dictate which rocker-arm ratios are used; more aggressive for a qualifying engine or configuration and more conservative for the longer race." Obviously, engines for qualifying are not the same for racing.
Griffin responded differently. "Cylinder pressure analysis (ECA) helps to understand the interaction of the 'mixing' of cam profiles on overall performance," Griffin says. "Seat-of-the-pants determines how well any optimization routine works." OK, this goes directly to the prior suggestion that individual cylinder performance follows individual cylinder analysis. Nothing new here. "Reading" spark plugs is age-old technology. However, in-cylinder pressure analysis as a function of crankshaft angle (ECA again) brings the measurement technique inside an operating cylinder to help identify "weak" or "strong" cylinders. Then you can go outside the cylinder to take steps that could include multiple profiles and rocker-arm ratios.
Now, if you'll hold those thoughts for a moment, what Godbold said falls right in line with this reasoning. "By looking at in-cylinder data (perhaps by experimenting with individual cylinder rocker ratios), engine builders sometimes find certain ports are weaker than others," Godbold says. "The weaker the port, the more duration and area (under the lift curve) the cam needs to provide to make power at a given rpm. Every profile Comp designs is intended to work within a range of rocker ratios, hence you cannot just keep adding more rocker ratio to weak ports.
"Typically, you should run the maximum safe rocker ratio on a given profile. However, some engine builders have found increasing the duration on the more restrictive ports can help smooth out the power curve or increase peak power. Most importantly, the larger the difference in port flow from cylinder to cylinder, the greater the potential to find improvements with these techniques." This last sentence should be underlined.
What are the structural reasons some rocker arms break and some don't? According to Griffin, "Rocker arms are loaded (during normal operation) in bending, which is the most horrible way to load a structure. By making these components out of improved materials with stiff cross-sections, high-rpm loading can be less detrimental. Failures also occur from high-stress points in the threads where the adjuster nut is positioned. Since the rocker arm is bending, the surface is being put into tension. As a result, shot-peening has helped increase the fatigue life of aluminum rockers because of the compressive stresses imparted on the surface of the body of the rockers (by this process)."
As you might expect, material selection is critical to parts life. Aluminum is not just aluminum, no more than parts are parts. Tores is particularly clear on this point. "There are a variety of reasons why one rocker arm design works better than another," he says. "Selecting the proper alloy incorporating high strength at elevated temperatures, low notch sensitivity and high levels of fatigue resistance are important factors.
"At T&D, we use SAE 2024 aluminum. This is the most fatigue-resistant of all the hard-alloy aluminums. It also has low notch sensitivity, which is an indicator of how readily a particular material will start to fracture under high stress. In the final design of a rocker, sufficient material should be used in critical areas to ensure a high safety margin.
"Generally, rocker arms do not fail from tensile strength failures. They usually fail from fractures in highly stressed areas such as the radius under the nose or in the rear of the rocker arm or a cut for a valvespring. Failures can result from these radii being too small, not properly distributing the stress evenly to the areas adjacent to the radius. Rocker-arm failures can often come from alloys that do not have good elevated temperature strength. Remember, a race engine may operate at temperatures up to 250 degrees (F), and these temperatures can cut the strength of some aluminum in half."
From a slightly different perspective, Godbold suggests, "Rocker arm failure is typically an alarm for an engine builder to look for problems in other areas. Aluminum work hardens and becomes brittle. Hence, aluminum rockers should be treated with wear limits and replaced at regular intervals. If you're using aluminum rockers that were hand-me-downs or ones with excessive use, a failure should not be a surprise. However, if a rocker fails early in its life, pay careful attention to wear of the retainers, locks and/or valve seats. It is quite likely that valvetrain instability may have contributed to the failure.
"Comp Cams prefers steel rockers to aluminum rockers in any application where steel is available. Steel rockers should only fail if they are used well outside their design application, sufficient care was not taken with design, or if they are improperly modified. In fact, I cannot recall seeing a Comp Cams full-roller steel rocker fail that had not been modified by the customer. Even in steel rockers, bearings should be replaced periodically.
"Some classes of racing require the use of stamped steel rocker arms. The level of competition in many of these classes often results in the cam forcing the pushrod through the pushrod seat of the rocker. This type of failure is very difficult to resolve without choosing a cam profile that makes less power."
How can you equate Spintron testing to actual engine operation, given the fact valvetrain components aren't operating in the changing environment of cylinder pressure? Historically, there are problems associated with relating "simulated" conditions to "operating" conditions, almost independent of the science involved.
It's the nature of "modeling" complex events by whatever method employed, be it via computerized simulation or the use of laboratory devices. In the case of "spinning" valvetrain components and comparing stability and durability performance with that of a running engine, there are issues to be considered.
Countless hours of Spintron work have caused Griffin to draw certain conclusions. "Oddly enough, valvetrain failure rates occur close to the same number of cycles, Spintron to dyno," he says. "Besides the lack of cylinder pressure, engines do not experience the input of combustion heat and its affect on the viscoelastic properties of the valves and seats. As a material's temperature changes, its stiffness changes. Along with that, different materials will have different changes in these properties. It is entirely possible that we never see this effect and it is also possible that the effect is minimal. We just don't know!"
In this same context, Godbold has some additional opinions. "We compare dyno results and other engine data to validate much of our Spintron data," he says. "While some differences appear when making direct comparisons, we've seen almost identical pushrod strain data on the Spintron and a running engine. There's always the possibility that cylinder pressure at exhaust valve opening is substantial. However, the pushrod loads at exhaust opening on the Spintron sometimes exceed 2000 pounds. Hence, forces to overcome the inertia of valvetrain components is most likely much higher than any force due to cylinder pressure.
"On the intake side, pressure differences should be zero at intake closing and small at intake opening. In most applications, we see the same characteristics on the Spintron and a running engine. Other factors, such as valvetrain harmonics and pressure differences across the valve (in a running engine), tend to create data shifts from one measurement environment to the other. Overall, there are significant benefits comparing Spintron information to that obtained from a 'live' engine."
Cycle testing of rocker arms is another viable method of evaluation as companion to Spintron analysis. In this regard, according to Tores, "We test rocker arms on a hydraulic load machine. At loads of about double typical valvespring pressure, we increase the load and cycle frequency until failure is induced. The 'cycle to failure' numbers mimic that of race conditions. The most critical part of testing is the operating temperature, and this can be duplicated on the Spintron. We feel only a small percentage of the load on a rocker arm is from cylinder pressure."
I'm a Saturday-night circle track racer, building my first "serious" engine. What guidelines can I follow in selecting and installing valvetrain components? Caution and exercising judgment is a pretty common thread on this topic. Even the three experts made comments that fall into this line of thinking. Godbold offers, "Every application is so different that very little can be held as true across the board. It's wise to establish a relationship with someone you can trust at a major camshaft company. While you may experiment with other vendors later on, that primary relationship can be extremely helpful toward getting a competitive engine built that will stay together. The second piece of advice is to not let your first few engines become an R&D project for the cam company. Make sure some variation of what you are running has been used successfully in someone else's engine.
"Often someone will use a computer desktop engine program to select the 'perfect' cam specs that has some combination of a big intake, a small exhaust, or really wide lobe separation. Going off the beaten path may be necessary in the long run, but start with something known that can win races. Then go for the experimental parts later. Going the other way around can lead to serious confusion as you try to figure out which unique component of your combination is giving you fits on the track."
Tores doesn't like "tricks," and made that point pretty clear. "Keep it simple," he says. "Save the tricks for later. Allow some room so that other things can be tried, such as different rocker-arm ratios. Don't go out and buy the latest 'trick-of-the-week' combination. Durable parts are more important. If nothing fails after a period of racing, then new components can be tried, following a scientific method of analysis."
Even more to the point, Griffin says, "Listen to the old guys. Go conservative but ask questions about what people have done to really screw things up. If some people don't want to respond to your questions, go ask someone else because the non-responders might not have any real answers anyway." Simple enough.
What is the value (or lack) of lightweight valvetrain components? If there is value, which parts provide the greatest benefit? Multiple pieces of information here, although the perspectives vary slightly. Godbold has this to say: "The closer to the valve, the better your money is spent on lightweight components. Stiffer is almost always more important than lighter as you get to the pushrod and camshaft core. Rockers also typically like stiff just as much if not more than light. Get the valve as light as possible and the pushrod as stiff as possible and work out a good compromise everywhere else.
"Heavy stainless steel valves with steel retainers and standard five-sixteenths-inch pushrods make one of the hardest valvetrain combinations to control at high rpm. The heavy valve system increases pushrod flex (it acts like a spring). When the pushrod springs back, it can surge the valve and valvespring. Then, when it comes back together on the closing side, the pushrod can flex again, resulting in the valve closing well before the tappet gets to the 'soft' part of the closing ramp where rates are controlled. Heavy systems are greatly limited to how hard you can push the components, both in terms of rpm and aggressiveness of useable profiles."
And then, according to Griffin, "Components that will benefit from weight reduction are the valve, retainer, locks and even the spring. The rocker arm benefits from having a lower moment of inertia (where 'lighter' on a weight scale doesn't imply a lower moment of inertia) and a stiffer body. Pushrods benefit typically from having stiffer bending properties, a result of larger diameters, tapers and thicker walls. Another aspect of pushrods that has been examined a small amount is the bending natural frequency which can be increased by making the diameter as large as possible and the wall as thin as possible. On this, little testing has been done to support any theories."
Finally, Tores concludes with these thoughts: "Lightweight valvetrain components are very important. But always remember that you have to be able to finish the race in order to win. So, durability is equally important.
"In most Spintron testing, everything on the valve side of the rocker arm fulcrum (or pivot) is critical. Any weight saved here will usually result in an increase of limit speed with the same valve springs. Weight on the pushrod side of the fulcrum tends to be somewhat less critical because the velocities are so much less (less weight must be started, stopped and re-started into motion). Lightweight retainers and valve locks, valves and rockers are always the best targets for weight reduction as long as they are reliable. A part may be considered reliable if it lasts for one complete race, no matter how long or short. How practical this approach may depend on your racing budget."
By way of review As in the case of most component selections, compromise is almost unavoidable. When one area of concern becomes optimized, another likely is not. Striking that elusive balance between maximum power and optimum durability often boils down to having adequate information upon which to base decisions. Past experience, especially if someone else paid for it and is willing to share, frequently helps avoid repeated mistakes. Seek reliable information sources, ask a rash of questions and draw conclusions based on what you believe best fits your needs. And it's usually a safe bet to assume the manufacturer of a part is a more knowledgeable resource than something you read in a magazine. Well ... unless it was Circle Track, of course!