When the Japanese motorcycle makers decided to make their names in international Grand Prix roadracing, they had a tremendous amount to learn but had useful European examples before them. On a 1950s trip to Europe, Soichiro Honda sent home racebikes he'd bought from NSU in Germany and Mondial in Italy. Suzuki and Yamaha were able to study such programs as that of DKW with its three-cylinder 350 and MZ's rise from obscurity to winning the occasional 125cc or 250cc GP.

Although Kawasaki didn’t contest European GPs until later, a scrapbook I found in an office (where I was told to wait during my visit in 1972) was filled with photos of 1950s and ’60s GP bikes, every photo covered with marks and lines indicating attempts to measure important dimensions (I knew because I had done the very same starting in 1964). I assume that Suzuki and Yamaha were at least as active in this sort of data gathering.

Yamaha, upon return of its wartime factory in August 1954, showed a two-stroke single clearly derived from the DKW RT 125. Japanese rivalries were fought out in events such as the Mount Asama “volcano race,” using production-based racers. By 1957, Yamaha had a twin in competition as well as a single. For 1961, engineer Noriyuke Hata was asked to design a 125cc single for European GPs.

Suzuki’s Colleda ST 125cc single of 1955 also grew from the RT 125 seed. Participation in domestic races ran Suzuki straight into all the hard problems of higher-power two-stroke engines: piston cooling, seizure, mixture control, accurately timed ignition, short big-end bearing life, exhaust-pipe design. And most of all, detonation. In that work, improvement in one area lets you advance only a short way before being stopped cold by the next limitation and the next. And the next. A 10-man racing department was organized in 1956 and in 1959 plans were made to contest the 1960 Isle of Man 125 TT.

Both companies began with the synthesis created at MZ by engineer Walter Kaaden: air-cooled, Adolf Schnürle-scavenge, rotary-disc intake, two or three transfer ports, and that crucial “sonic exhaust valve,” the counter-cone pipe or “expansion” chamber.

Both companies also vigorously attacked the problem of piston cooling for the obvious reason that thermal expansion of hot pistons often produced seizure. Designers of aircraft piston engines had suffered this problem and had responded with a combination of closely spaced cooling fins with tight baffling to force all air taken into the engine nacelle to pass through fin space. But racing motorcycles at their lower speeds could not generate enough ram pressure to push cooling air through closely pitched fins (Suzuki tried such fins on at least one race engine and thereafter returned to the 0.25- to 0.35-inch fin pitch usual in motorcycling). And why not just make the fins bigger? The farther a fin projects from the hot object it’s intended to cool, the lower its temperature falls, yielding diminishing returns. Air-cooling thus had limits, albeit not reached until the middle 1960s.

Another method of improving piston cooling is to make pistons smaller. This was why the first two-stroke world championship came in the 50cc class. The bore of Suzuki’s winning RM62 was just 40mm. One crude measure of piston distress is horsepower per square inch of piston area. In the case of RM62, this was 4.1 hp per square inch.

How could the heat expansion of aluminum pistons be controlled? Ultimately, alloying aluminum with silicon provided relief, at the same time increasing the hot strength of the material.

Both Yamaha and Suzuki were manufacturing production bikes whose engines had durable but hot-running cylinders of cast iron. To improve cooling, both did what European makers had done after 1930: switch to cylinders of higher-heat-conductivity aluminum with wear-resisting liners of cast-iron. Even this was not enough. One Suzuki engineer at the time commented that, “Our pistons swelled like cakes.”

By drilling holes into test cylinders to various depths and placing thermocouples in them, it was possible to record a radial temperature profile—highest at the bore surface. This revealed that the lower conductivity of the iron liner was acting as a “blanket,” keeping the piston hotter than need be. Okay, the liner had to go, but aluminum was too soft to endure the friction of piston and gas-sealing rings.

Yamaha first tried hard anodizing, but results were inconsistent. Suzuki had success with hard chrome plating on the aluminum cylinder of its RM62. Yamaha adopted this solution as well. How could the heat expansion of aluminum pistons be controlled? Ultimately, alloying aluminum with silicon provided relief, at the same time increasing the hot strength of the material.

Tempting also as an emergency measure is to run a rich fuel-air mixture, as this reduces flame temperature. Yet rich operation was also risky, for if on a given circuit there were hot and cool zones, in cooler air the engine would lean out, make more power and heat, and perhaps seize without warning.

Step by painful step, problems were ground down. Once operational variables came under some control, two-strokes in larger sizes became successful. In 1963, Suzuki took the 125 championship with its 25.5-hp RT63 twin, whose power per piston area was now 5.7. In that same year, Yamaha’s RD56 air-cooled 250cc twin dominated the super-fast Belgian GP at Spa. Once fuel delivery problems caused by its remote-float carburetors were solved, the two Japanese riders Fumio Ito and Yoshikazu Sunako just motored away from the previously dominant Honda fours (forcing that company to begin design of a 250cc six). Power per square inch of piston area was now 5.9.

Further development of the RD56 boosted it to 54 hp, which Phil Read used to defeat Honda in the 250cc class in 1964 and ’65. Piston power loading rose to 7.1 hp/square inch. In this process, it was discovered there are limits to the power that can be sustained with air-cooling. As with four-strokes as well, compression ratio (and therefore torque) is limited by the high temperature of piston crown and cylinder head, which encourages detonation. Yet conversion to water-cooling was neither easy nor straightforward, with much development required.

In cylinder porting, only minor advances were being made. Yamaha had only advanced from two transfer ports to three in 1962, its sabbatical year for further development.

The combination of the simple three-transfer-port scavenge system and the common belief in the value of a high crankcase compression ratio conspired to make two-stroke GP bike powerbands very narrow.

Let’s consider yet another figure of merit in engine performance: brake mean effective pressure (BMEP), which is stroke-averaged net combustion pressure (net means with friction and pumping loss subtracted). It is a measure of how well an engine fills its cylinder(s) with fresh mixture, and how efficiently it burns that charge. In the case of Suzuki’s 125cc twin, this was 112 psi, and for Yamaha’s RD56 of 1964–’65, 130 psi. Compare this with good four-stroke peak BMEP numbers, which for years have been (and today remain) in the range of 200 psi. Bear in mind that a two-stroke produces this effect every time a piston comes to TDC, while in a four-stroke it is only every other time.

The combination of the simple three-transfer-port scavenge system and the common belief in the value of a high crankcase compression ratio conspired to make two-stroke GP bike powerbands very narrow. In response, Suzuki gave its race engines ever-increasing numbers of transmission speeds—seven in the 1962 50cc, eight in the 1963 125cc twin. By 1967, the last year of this classic era, its 50cc water-cooled twin would have 14 speeds. Although fascinating as a radical extreme, this was not a progressive development. Listen to the audio recordings made of racing in those years, gear changing is constant: “bee-bee-bee-bee…”

No one in GP racing challenged the orthodoxy of three-transfer scavenging, even though it seemed to limit BMEP to about 135 psi. That put them in the same boat with four-stroke developers, who, faced with BMEP limited to 180–200 psi, were forced to seek increased power by boosting rpm.

Honda replaced its 125cc twin with a higher-revving four and then with a 125cc five-cylinder that was safe to 21,500 rpm. Two-strokes, because they had to fill their cylinders in roughly one-third of a revolution (rather than the three-quarters of a revolution available in four-strokes), were less able to rev. Add to that the fact a four-stroke’s valve area is limited by bore (because the valves are located in the head), while a two-stroke loses port area as bore is increased and stroke is shortened (because its ports are in its cylinder wall).

The result was that Japanese two-stroke designers accepted the limited BMEP of three-transfer scavenging and focused development on increasing power by using larger numbers of higher-revving cylinders. The result was very complex V-4 engines. By the end of 1967, Suzuki had achieved 42 hp at 16,500 rpm from its RS67 four-cylinder 125. That corresponded to a piston power loading of 6.8 hp/square inch of piston area and a BMEP of 135 psi. For Yamaha, the 250cc four RD05 ended the era at 73 hp at 14,000 rpm for 7.7-hp/square inch of piston area with BMEP equaling 138 psi.

As we know that the last 125cc two-stroke singles in GP racing made 55 hp—a piston power loading of 15.5 hp/square inch and a BMEP of 224 psi—there was clearly a lot of development yet to come.