Engine Architecture: Twins

When motorcycle and aircraft pioneer Glen Curtiss needed more power than his home-built single-cylinder engines could provide, the next step was clear: He altered the original crankcase to accept a second cylinder and linked that cylinder’s connecting rod to the single’s existing rod. The result was a V-twin.

Because of Harley-Davidson’s longtime association with the layout, we may think of V-twins as an American innovation. But lots of V-twins were built in many places before US makers Curtiss, Indian, and Harley-Davidson built theirs. The first-ever such engine, a 20-degree V-twin built by Daimler in 1889, can be seen in the Deutsches Museum in Munich. It was the least expensive way of adding more displacement, not a contest to be first. Curtiss twins had a 50-degree cylinder angle, Indian favored 42 degrees for years, and Harley chose, and for over 100 years stuck with, 45 degrees.

The method of adding the second connecting rod varied. Harley and Indian placed the two rods in the same plane by attaching them to the single crankpin, fork-and-blade style. Curtiss added a boss to the big-end of the first or “master” rod and attached the link or “slave” rod to it with a pin. Vincent, and others, placed the two rods side-by-side on the single crankpin.

Why all the different V angles? With a single crankpin, the strongest construction, the big problem was piston skirts hitting each other near bottom dead center. Because early engines had bores much smaller than strokes (the smaller the piston, the cooler it ran), very small V angles were workable. In the 1889 Daimler, with its 60mm x 100mm bore and stroke, the very narrow 20-degree V was made practicable by giving it quite long connecting rods.

Late-20th-century V engines aimed at high performance were naturally given bores bigger than their strokes, to allow using larger valves; this required a greater V angle to achieve piston-to-piston clearance.

Aprilia wanted to avoid the extra engine length of a 90-degree V (for years this was a problem for Ducati, who only gradually addressed it), so it decided to close it 1,000cc RSV Mille’s cylinders to a more compact 60 degrees and handle the resultant primary shaking with dual balance shafts. Harley, too, chose 60 degrees for its VR1000 Superbike racer project. In the case of Erik Buell’s Rotax-built twins, a 72-degree V angle was chosen to allow more room for its intake system. Yet when Indian subsidiary Swissauto designed the FTR dirt-track race engine, the absolute need for a quick-responding short wheelbase pushed the choice to a narrower 53 degrees. Different priorities, different choices.

Long before this, Triumph’s Edward Turner, a pioneer who was more product psychologist than engineer, showed the way to build standout parallel twins that were almost as cheap as singles. The resulting bikes, beginning with the Triumph Speed Twin of 1937, struck people as different in a good way, an alternative to the thudding singles they had known for years. The Triumph twins had a fresh sound, and because they fired every revolution of the crank instead of every other revolution, they propelled their riders more smoothly.

Other British manufacturers, such as BSA, Norton, Royal Enfield, and AMC, understood the appeal and built their own. All of those designs relied upon the light weight of their small pistons (the engines began as 500s) to keep vibration moderate at the usual 6,500-rpm peak of those times.

Why did the pistons in these British twins move together, giving a 360-degree firing interval? Why not put the two crankpins at 180 degrees, to provide some self-cancellation of piston shaking force? The even firing interval made single-carburetor operation easy, and 180-degree crankpins would still have left a “rocking couple” vibration.

After World War II, Sunbeam tested the oft-made assertion that a bike made more like a car would achieve vast sales. The S7 was a parallel twin with 360-degree crankpins, but with its crank axis fore-and-aft rather than transverse. Alas, vast sales eluded the S7.

Today the parallel twin is back, and for some of the same reasons that motivated Edward Turner. Compactness was important to Turner because his Speed Twin had to replace a single in an essentially unchanged frame. Compactness is important now, when the twins’ leanness, light weight, and reduced parts count are so often replacing the “beaminess” of the inline-fours that have been with us since Honda’s 1969 CB750.

These new-generation twins offer something else: 270-degree crankpin spacing, something which Vincent engineer Phil Irving (author of the classic text Motorcycle Engineering) recommended back in the 1950s. Why? With crankpins at either 360 or 180, both pistons are stopped at top and bottom center, imposing an “inertia torque” on the crankshaft. Having to start and stop both pistons simultaneously imposes a fluctuating load on the crank. This motivated Yamaha engineer Masao Furusawa, now retired, to give Valentino Rossi an inline-four MotoGP engine whose crankpins were spaced at 90 degrees to each other rather than the traditional 180. That fluctuating load from piston starting/stopping interfered with traction on the track; with Furusawa’s and Irving’s 90-degree crankpin spacing half the pistons are stopped while the other half are near their top speed. Thus they exchange energy with each other, and crankshaft instantaneous speed varies but little.

But why a crossplane crank for production parallel twins? The real reason is probably engine sound—the 90-degree crankpin spacing transforms the normally droning even firing interval into a syncopated sound more like…more like a Ducati V-twin.

In its early years, Ducati built small-displacement sporting singles whose power came from high rpm. In the 1960s the company tried to scale them up to follow the market. Bad idea! Singles without mechanical balancers shake, and the bigger their pistons, the harder the shaking. Its 350 desmo was a fast motorcycle, but the dominant characteristic of the 450 desmo was vibration. On a paved surface, left idling on its stand, it would slowly creep.

Next try was a British-style parallel twin, but such engines shake just like a big single, caring not whether its builders drink tea or cappuccino.

Chief engineer Fabio Taglioni knew the solution. By adding a second cylinder at right angles to the first he made a 90-degree V-twin. Placing counterweights on the crank able to balance 100 percent of one cylinder’s reciprocating weight (the piston, rings, pin, and the small-end of the con-rod) produced a twin with zero primary shaking force. This meant that unlike parallel twins, vibration was not a barrier to high-rpm operation. In 1972 when the late Paul Smart arrived for the Imola 200, he exclaimed in disbelief, “They’re telling me this thing makes peak power at 9,750 rpm! Is that even possible?” Ducati would go on to develop giant Superbike racing twins that were capable of more than 12,000 rpm despite giant bore sizes far over 100mm.

Moto Guzzi’s present engines are 90-degree V-twins, but with the crankshaft axis fore-and-aft rather than transverse. Earlier, Guzzi had built a 120-degree V-twin that Stanley Woods used to win the 1935 Senior TT. Its two staggered crankpins, with a main bearing between them, were set at 120 degrees to give the engine an even firing order. The forward cylinder lay horizontal as in that company’s singles, while the other cylinder angled upward, above the gearbox behind it.

Honda would have a 1980s flirtation with staggered-crankpin V-twins, like its RS750 dirt-tracker, 500cc Ascot, and 650 Hawk production bikes, as well as many of its Shadow cruisers. Staggering the crankpins allows zero primary shaking force provided that the crankpin offset angle, plus twice the cylinder V angle, equals 180. Unlike the Guzzi, these engines had no center bearing; the staggered crankpins were instead joined at the crank’s center by a relatively thin disc. All was well at the planned-product rpm, but when RS750 and Hawk engines were modified for roadrace use, crankshaft life could be interrupted by cracking and failure.

BMW’s long-serving line of horizontally opposed boxer twins originated at the pen of aircraft engine designer Max Friz. Knowing the ill effects of engine vibration only too well, he gave it two crankpins set 180 degrees apart, making all piston shaking forces, primary and secondary, self-canceling because its pistons always moved in opposite directions. A number of similar engines had been built after WWI to power light aircraft, their cylinders projecting to the sides from a narrow fuselage to receive excellent equal cooling. Those qualities would serve it well as a motorcycle engine.

Because of the boxer engines’ 180-degree crankpin spacing, their cylinder centerlines are offset from each other, causing an oscillation around the vertical yaw axis. This causes the familiar “BMW buzz” that grew in strength with each increase of piston weight or rpm. An added balancer finally calmed it.

During WWI, a terrible simplified flat-twin engine was designed for a flightless trainer called “Penguin.” Having only one crankpin, its pistons moved together, giving it a heinous level of vibration.

A form of twin less often seen is the tandem, in which two single-cylinder crankshafts are geared together. Kawasaki engineer Nagato Sato drew the 250 and 350 roadrace twins with a pair of contra-rotating cranks, and years later, the late Dan Gurney, needing a very narrow but smoothly rpm-capable engine for his feet-forward Alligator motorcycle development, did the same. Sato’s first try used 180-degree firing and vibrated dreadfully, but that was soon corrected to simultaneous firing, and the bikes went on to win world championships. Velocette’s Roarer was likewise of this type.

Because the motorcycle has taken so many forms, it is impossible to identify a “best” engine architecture. Engineers and product planners choose the form best suited to their application, and it’s up to us to get on and ride.