Calming The Crankshaft Of Honda’s Six-Cylinder Racer

“This system made crank center into a node, effectively cutting the free vibrating length of the crank in half”

The legend of Honda's RC166 is that with Mike Hailwood at the controls the six-cylinder machine won eight straight 250cc Grand Prix races in 1966, a time in which only a rider's seven best results were scored. Some of you may have heard—live or on a recording—the otherworldly rapid whoop, whoop, whoop of the six being warmed up, a sound so unlike that of any previous internal combustion engine that it was closest to the yelping of hounds.

So light was the six’s crankshaft that when its throttle was closed sound ceased instantly. No audible coasting to a stop. Silence. This lightness was a serious problem for riders, for the slightest inattention stalled the engine (to be revisited 12 years later with the oval-piston NR500). Yet the six’s crankshaft—approximately 14 inches long—was pressed together in classical fashion to make possible the use of one-piece (not split and bolted) connecting rods and main bearings, turning on tiny caged needle bearings.

All crankshafts vibrate in multiple modes. The full-circle flywheels of classic Manx Nortons, Matchless G50s, and Vincent twins vibrate in “elephant ears flapping” mode, which by the later 1950s resulted in premature crankpin breakages (the two massive flywheels in performing this “flapping” were flexing and fatiguing the single crankpin that held them together).

Inline multi-cylinder crankshafts twist slightly at each cylinder’s firing then untwist from their own elasticity, resulting in so-called torsional vibrations in several possible modes. As such a crankshaft is run through its range of rpm the amplitude of torsional disturbance rises and falls as its own natural frequency goes in and out of step with various multiples of firing frequency. It’s a math problem that Mother Nature solves effortlessly.

mike hailwood vintage race action
Mike Hailwood used his toes as pavement feelers and achieved lean angles amazing for the day.Volker Rauch/Cycle World Archives

When the English Napier firm built its first inline-six auto engine in 1903, the subject of crankshaft torsional vibration had not yet been studied, so the loud rattling of this engine’s cam drive from the rapid torsional oscillations of its long crank came as a surprise. When Rolls-Royce built its own first six a year or two later, engineers applied themselves to understanding torsional vibration and finding means of controlling it.

In the years between the world wars, the need to guarantee the reliability of aircraft engines led to the development of “torsiometers,” which could graphically record the amplitude of vibratory twist as a function of rpm. It became understood that as cranks were made lighter (for example, by boring their crankpins hollow and by eliminating all mass not required either structurally or for balance) their natural torsional frequencies increased, possibly to values too high to be excited by the engine’s firing frequencies.

It was against this background that Honda engineer Shoichiro Irimajiri set out to design a four-stroke racing engine able to overcome Yamaha's rapid two-stroke development. The driving force was knowledge that when Yamaha had overcome the RD56 twin's fuel-feed problem at Spa in Belgium, its two bikes, with Japanese riders, had casually motored away to a stunning 1-2 win, leaving Luigi Taveri's Honda four 41 seconds back. This would not do.

mike hailwood dutch grand prix race action
Hailwood at the Dutch Grand Prix aboard the six-cylinder 250.Cycle World Archives

As development of Honda's 250cc inline-four was nearing its limits, greater power could only be had by operating at rpm out of reach of the four-cylinder. At that time, this meant providing more cylinders, smaller in bore and stroke than those of the four and therefore capable of higher revs. Guzzi had in 1955 taken the big step of switching from air- to water-cooling in its 500cc V-8, making it possible to cool the rear bank of cylinders. AJS had shown in its mid-1930s attempt to race an air-cooled V-4 that only water-cooling could prevent rear-bank cylinders from overheating, declining in power, and failing.

If water-cooling was too great a step for the time (Honda would make that step in 1977 with its NR500), all cylinders would have to present themselves equally to cooling air. That meant a transversely mounted inline. While Honda knew how to build reliable four-cylinder racers, Irimajiri knew that overcoming the torsional vibrations of a longer five- or six-cylinder inline crankshaft assembled with press fits would be tough.

Anyone who has rebuilt pressed-together crankshafts like those of the 1960s knows that when a used crank is pressed apart its crankpins and main shafts show the reddish discoloration that indicates slight relative motion between parts—motion resulting from crankshaft vibration (the red color is that of oxidized wear particles). Irimajiri knew that making the crankshaft very light must be the key to success, for the less mass there is to vibrate, the less energy can be stored in the motion. Given tight enough press fits of great enough diameter, slipping at the joints might be prevented.

mike hailwod isle of man race action

Hailwood bends the Honda Six through a corner while spectators bask in a rare sunny day on Man.

Hailwood bends the Honda through a corner on the Isle of Man.Cycle World Archives

In making the crank lighter—giving up the full-circle flywheels of the earlier fours in favor of minimum material, for example—Irimajiri would also be giving up some of the hoop strength that could make pressed joints tight enough to withstand torques from cylinder firings and torsional vibrations.

One important choice had already been made. Honda had dropped the shaft-and-bevel gears cam drive of its first four, the RC160 of 1959, replacing it with a much stiffer central train of spur gears placed between cylinders two and three. To reduce the length of crankshaft able to vibrate torsionally, Honda used a jackshaft to take the drive, not from one end of the crank, but from a gear at crank center to the clutch. This system made crank center into a node, effectively cutting the free vibrating length of the crank in half. Irimajiri adopted this center-drive concept but placed it in Mercedes W196 fashion behind the cylinders rather than between them, allowing a small but useful narrowing of the engine.

To reduce crank mass as much as possible as a means of pushing harmful torsional modes to frequencies too high to be excited by engine firing frequency, Irimajiri reduced the crank cheeks to mere “candy bars.” One end of each bar carried tungsten balancing slugs and the other was either integral with a crankpin or had a bored hole to receive one. Minimum mass.

Because the torques produced by cylinder firing and torsional vibration were least at the free crank ends and greatest at the center, Irimajiri chose to make the crankpins and main shafts in three diameters: smallest for numbers one and six, slightly bigger for numbers two and five, and largest for three and four. In this way, the press fits were strongest where the torques were greatest. Yet where less twist resistance was acceptable (at the ends) the parts were smaller and lighter. The concept worked, resulting in a crank that could complete races. (I can’t be 100 percent sure of my information, but I’ve been told crank life in 1965 was 300 kilometers or 186 miles.)

This engine was developed in extreme haste, design beginning early in 1964, leaving little time for errors and fixes. Introduced at Monza on September 11, 1964, the first six created a sensation but slowed to third in the race, reportedly because of fuel vaporizing in the float bowls. Yamaha won the 250cc championship in ’64, and in ’65 Phil Read and Mike Duff took first and second in the championship, leaving Jim Redman third on the developing six. More work necessary.

Better days were coming for Honda. In 1966, Mike Hailwood and the six proved a fabulous combination—they clinched the 250cc title by the seventh of 12 GPs at Czechoslovakia in July. In 1967, he won again but less easily because of the speed of Yamaha’s new four-cylinder. With rumors of a water-cooled Honda 250 V-8 for 1968 swirling through European paddocks, Honda withdrew from GP racing to concentrate on its developing auto business.