Learning From Others

[Admin]

Kevin Cameron has been writing about motorcycles for nearly 50 years, first for <em>Cycle magazine</em> and, since 1992, for <em>Cycle World</em>. (Robert Martin/)

In the early 1920s, at its air development center at McCook Field in Dayton, Ohio, the US Army rapidly tested a series of air-cooled overhead valve (OHV) cylinders. Their goal was to find the most reliable design that was capable of sustaining the new and higher power OHV engines were producing.

Meanwhile, at the same time in British motorcycle roadracing, side-valve engines were winning their last events. Motorcycle manufacturer AJS was aggressively pushing its own OHV designs, with its higher airflow capability and power. Unlike the US Army’s OHV test cylinders, which had enclosed valve gear, British OHV prototypes usually featured exposed pushrods and rockers, the aim being to achieve adequate cooling of valves and valve springs by “open air” installation.

The man in charge of the US Army program was S.D. “Sam” Heron, an experienced English engineer who had through World War I served in Britain’s Royal Aircraft Factory, where rapid development of air-cooled engines had been the priority.

Why an English engineer leading a US program? The sad fact was that the US, despite the early achievements of the Wright brothers, lagged far behind Britain, Germany, and France in aircraft engine development. This had meant that US-designed engines played no role in the 1914-1918 air war, despite the US joining that conflict in April 1917. The US Army Air Corps needed his experience.

In the 1921 Senior Isle of Man TT, the OHV power advantage was brilliantly demonstrated when Howard R. Davies won the race on a Junior bike of only 350cc over the side-valve Indian 500s of Freddie Dixon and Bert LeVack.

Related: Survival of the Fittest

Unlike those early OHV motorcycles, the US Army’s aviation program needed to have its valve train enclosed to prevent unacceptable loss of engine oil. In Heron’s eventually successful M-type cylinder, the rocker arms were enclosed in bolted-on rocker boxes, each shaped like a square-section canoe with a bolted-on cover, and equipped at mid-point with a rocker pivot bolt.

The rocker boxes leaked where they bolted to the head, and they cracked. Both were indications that they were flexing. Naturally it was the exhaust rocker box that cracked first; the extra force required to open the exhaust valve against what was left of combustion pressure increased the stress on the magnesium box. Next, Heron substituted rocker boxes made of Duralumin, an early high-strength alloy of aluminum with copper. The exhaust rocker box still cracked, but now only after 120 hours of operation.

Seeing AJS’s success with overhead valves, Eugene and Percy Goodman at Velocette decided to supplement their line of small two-stroke motorcycles with an up-to-date overhead-cam (OHC) four-stroke 350 single. As their design evolved, they too encountered problems caused by the stress imposed on the rocker support structure by the force of valve opening. They wouldn’t be the last.

AJS for a time went back and forth between OHV and OHC, eventually settling upon the latter. But for aviation, Sam Heron’s M-type cylinder would be used on nine-cylinder radial engines. The added complexity of chain or shaft drive to overhead cams on so many cylinders was needless excess.

On the other hand, the extra moving mass of pushrods and rocker arms had made it a rule of thumb in Britain that for equal rpm capability, an OHV engine needed about twice the valve-spring pressure of a side-valve or DOHC design. That would double the stress on whatever structure supported the rocker-arm pivots.

Why not cast or forge the rocker boxes in one piece with the head? Structurally that was a fine idea, but it added complexity to the casting process. The simpler the shape you cast, the lower the scrap rate. When in 1936 Edward Turner at Triumph designed the 500 Speed Twin that was father to the long line of British parallel twins, he gave it bolted-on rocker boxes; but knowing how easy it was for such boxes to flex, leak or crack, he gave them a strong, broad base where they sat atop the cylinder head, protecting them from bending forces. Turner was one of many English engineers who received a substantial part of his education at Britain’s great Brooklands Speedway, 25 miles southwest of London (it had opened in 1907). There, the latest ideas of factory engineers and private tuners alike could be seen up close, as the track was a major development center as well as a popular racing venue. If OHV rocker boxes were cracking, interested parties would see it happen at Brooklands.

That didn’t stop engineers from continuing to design rocker boxes for convenience rather than durability. When in 1940 a Pratt & Whitney team under Luke Hobbs laid out that company’s last major radial piston engine, the four-row 28-cylinder R-4360, he put the ends of its rocker arms where the location of the cam rings dictated: cantilevered out some distance from each cylinder head. Although the head material was high-strength aluminum and the boxes were integral rather than bolted on, the high peak force required to pop open the 1-pound exhaust valves against more than 500 pounds of end-of-stroke cylinder pressure was too much; the rocker boxes cracked in classic fashion. Since by that time the cylinder heads were already tooled, the chosen “fix” was to reduce the initial rate of exhaust-valve lift to something that the structure could survive.

Yet another chapter was to be written by desperate men at Indian as they prepared for the expected flood of pent-up consumer demand as World War II ended. It had been Indian, in the early days of 1916, who realized that something lighter and easier to ride than the traditional 61-inch heavyweights could find a place in the US market. That had been its very successful Scout model, which started life at 37ci and then expanded to 45. From the wild speculation of the post-WWI era, with its flapper girls, rumrunners, jazz, and instant stock-market millionaires, to the sudden shock of the Great Depression In 1929, Indian had repeatedly headed for failure, with annual sales falling to just 1,600 bikes in 1933. Each time it had been the enduring appeal of the Scout that saved them.

While WWII raged on, Indian commissioned a postwar design loosely based on Edward Turner’s 500 Speed Twin, a bike whose lightness gave it winning acceleration and quick maneuvering, and something new to American riders accustomed to the heavy weight and slow steering of traditional big Harleys and Indians. In concept, Indian’s new design was exactly right, for postwar British twins would create a new kind of sporty motorcycling in the US. Indian’s plans would have landed them squarely in the center of this coming trend.

It didn’t happen. At the last minute, an emergency cost-cutting redesign was pushed through. The result was classic rocker-box cracking and a host of other ills that destroyed the product’s reputation before they could be corrected. Indian closed in 1953.

Also toward the end of the war, Wright Aeronautical in New Jersey was planning its high-horsepower postwar engine, a two-row 18-cylinder radial that would ultimately produce 3,700 hp for takeoff. Knowing the problems others had experienced with rocker-box cracking, Wright designed its as box structures interrupted by the smallest possible opening, covered by a domed cap. Through it, the valve-guide bore and valve-spring base could be machined, and the rocker arm itself inserted endwise. This design flew on many great prop-driven aircraft such as the Constellation, Skyraider, and the C-119 Flying Boxcar.

At present, many imagine that future design work will be infallibly carried out by computer. Futurists see engineering experience being built into design algorithms, just as the skills of machinists have been built into the programming of CNC manufacturing equipment. Human review and decision making will, in this view, become unnecessary. The present troubles of Boeing’s 737 Max 8 airplane and its digital flight control, grounded since March 2019, suggest that human supervision remains essential. Although computers are lightning-fast at numerical analysis, they have yet to display much understanding. There are good reasons why engineer/managers at Honda still lead groups of recent graduates on visits to industrial museums. In the great years of US and British engineering it was common practice among leading manufacturing firms to maintain “failure rooms” in which they displayed parts and structures that had failed in testing or service. These were arrayed for their value in instructing incoming junior designers. Attributed to Otto von Bismarck are the words, “Only a fool learns from his own mistakes. The wise person learns from the mistakes of others.”

View the full article