When R&D Hits a Dead End

[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/)

It’s the 1930s, a time of regular increases in fuel octane number. The first breakthrough came when Charles Kettering ordered Delco researcher Thomas Midgley to search out a medicine against detonation. Using a roomful of knock-test engines, Midgley came up with tetraethyl lead. Yes, it was extremely poisonous, but it would give the engines in Allied warplanes a serious advantage in World War II.

The second biggie came when Dr. Graham Edgar produced exhaustive data concerning the knock behavior of a wide range of pure petroleum hydrocarbons. This led to the octane scale, an extremely useful and practicable means of comparing potential fuel components; Edgar’s Octane Number (ON) scale remains in use to this day.

Having thus become accustomed to a steady rise in pump-fuel ON, builders focused on compression ratio. Unlike varying cam timing or adjusting intake or exhaust lengths, increasing compression boosts engine torque at all rpm.

Increasing Engine Compression

Sport motorcycle engines of the 1930s began at compression ratios in the range of 5:1 or 6:1, and compression increased from there. Because business requires that engineers justify any production changes they want to make, they raised the compression ratios in the cheapest possible way: by giving pistons ever-taller domes, culminating in the 10:1 engines raced from the late 1930s through the 1950s.

By reading Phil Irving’s books Motorcycle Engineering and Tuning for Speed I became aware that, even though the 6:1 two-valve hemi-chamber cylinders of circa 1930 had very open chambers, all of the OHV and OHC engines of the period needed quite early ignition, in the range of 40 degrees BTDC or more, indicating that their combustion wasn’t all that fast.

Hemi Engines and Piston Temperature

Another problem was piston temperature. As compression ratios increased to exploit the improving fuels, peak combustion pressure also increased. As piston domes grew taller, the piston crowns’ heat-gathering surface area increased as well, making pistons run hotter. Fortunately, since up to that time cylinders and heads had been cast in iron, switching to more heat-conductive aluminum cylinders with thin iron liners addressed the piston-temperature problem. Regular growth in total cooling-fin area complemented power increases.

The grand old man of internal combustion, Charles Fayette Taylor (who lived to be nearly 102), had a useful explanation of how turbulence accelerated combustion. He asked us to visualize the many turbulence cells developing as the piston approached TDC as little gears in mesh: If one tooth of one gear is a flame kernel, rotating that gear 180 degrees will ignite the next, and so on across the whole combustion chamber. Accelerating such turbulence (faster rotation of individual turbulence cells) sped up combustion as a whole. With no charge movement at all, flame speed is of the order of one foot per second.

Providing such turbulence as the piston rose to TDC on compression required finding a way to store the intake mixture’s rapid air motion in the cylinder. As the piston came close to the head, that motion would then break up into the desired multitude of turbulence cells.

Axial Intake Swirl

The first successful scheme was axial-charge swirl. That was easy; just re-aim each intake port so that it entered the cylinder on a tangent rather than on a diameter, causing the whole charge to swirl rapidly. Anyone who has used a hose to fill a bucket has played this game, altering how the jet enters the bucket to make the water in it rotate to right or left.

Through the 1930s, swirl successfully reversed the upward trend in ignition timing, pulling it back into the high 30s. After 1950, when Norton engineer Leo Kuzmicki introduced that company to squish, piston domes came down and minimum-for-best-torque ignition timing moved as low as 34 degrees in highly developed two-valve designs. But in less mature designs, best ignition timing might remain very high; witness the 50–55 degrees BTDC required by Harley’s first-try iron XR-750 of 1970. The longer combustion takes, the more of its heat is lost to piston crown and combustion chamber, intensifying cooling problems and substantially reducing power.

Related: Four-Valve Beginnings

Things such as BRM’s trouble with the especially large valves in that organization’s four-cylinder Formula 1 engine of the 1950s impeded the trend toward accommodating larger valves with larger bores and shorter strokes. In veteran designs such as Norton’s single-cylinder Manx roadrace engines, the sheer mass of the valve dictated long valve durations, providing more time for the valve springs to do their job. That in turn narrowed the engine’s range of usable torque.

A New Solution Emerges

Sometimes a chosen R&D direction, which at first unleashes rapid progress, leads eventually to a dead end that requires backing up and trying something new. In the 1960s, progressive designers everywhere backed out of that dead end where the Manx and friends were trapped and started in a new direction: four valves per cylinder. As enunciated by Keith Duckworth in his famous Law, adopting four valves improved everything by roughly the square root of two, which is about 1.41. Individual valves grew lighter, controlled more easily by reasonable springs. Valve durations could be shorter too. Flat piston crowns and narrow valve included angles combined to produce open chambers of minimum surface area, unobstructed by domes. Abandoning air-cooling made fins between the valves unnecessary, so that today both cams live happily under a single cover.

The four-valve solution suggests that we should resist falling too deeply in love with a given set of ideas. That way, when something better comes along we’ll be able to recognize its value. Worst is when we become so deeply committed to the Old Way that we become obstructive old fogeys.

View the full article