Source: Cycle World
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/)Who remembers Honda’s NR500, the oval-piston V-4 four-stroke whose intensive and costly five-year development was intended to stop the two-stroke tide in FIM Grand Prix roadracing? Many more remember American Freddie Spencer (1983, 1985) and Australian Mick Doohan (1993–1997) later dominating Grand Prix bike racing on Honda two-strokes. What unseen force could have triggered such a switch by a company so publicly devoted to the four-stroke principle?Look even further back, to the period 1959–1967, when Honda established its name worldwide by winning 16 world roadracing championships. Its tools were daringly designed multicylinder four-strokes from 50cc to 350cc, some revving as high as 21,500 rpm. At the end of 1967, Honda stopped racing. Its signature megaphone exhausts went silent as the company devoted all its R&D resources to entering the auto business (resoundingly, with the Civic of 1972).Honda’s 1979 NR500 featured groundbreaking technology—most notably a V-4 engine with oval pistons and eight valves per cylinder. (Honda/)Through the following decade, Honda assumed the role of “the General Motors of motorcycling,” growing so big and so successful that it had nothing more to prove.That ended when Honda’s new president, Kiyoshi Kawashima, chose the occasion of the CBX 1000 launch in October 1977 to announce Honda’s imminent return to Grand Prix motorcycle racing.How To Defeat the Two-Stroke?How would Honda overcome the dominant two-strokes, even as the last four-stroke holdout, the Italian MV Agusta team, was giving up the fight? MV had been the architect of 38 world titles, 18 of them in 500, the class Honda had never won.Honda’s project was given the name NR 500—NR for “New Racer.” A special organization, NR Block, was created within Honda’s Asaka R&D Center. Famous names led the project; Shoichiro Irimajiri, whose designs had won Honda’s last 1960s championships, was overall leader. Engine leader was Suguru Kanazawa, chassis leader Satoru Horiike—both later rose to directorships of HRC. And a cast of hundreds.The three major determinants of engine power are displacement, rpm, and net, stroke-averaged combustion pressure. Because the last is limited by atmospheric pressure, and engine displacement is fixed by rule (500cc, or 30.5ci), that left rpm as the path to competitive power—as it had been in Honda’s 1960s racing. That path appeared to be blocked by the 500cc class limit of four cylinders (otherwise the company would have immediately begun development of a V-8).Inspiration StrikesThe official story is that one evening, Irimajiri was commuting home, wondering why pistons had to be round. If they could be made oval and still seal, many more valves could be fitted, able to supply fuel-air mixture at fabulous rpm. This would be equivalent to a 500cc V-8 whose pistons had fused together in pairs, assuming oval form. It was proposed that such an engine could make 130 ps at 23,000 rpm (ps was metric horsepower, now replaced by kilowatts in European power measurement).The first step was to test the sealing ability of oval pistons and rings. A simple four-valve test engine, designated Koo, was improvised from an XL250 bottom end. Methods of manufacturing the unusual parts had to be devised. Upon assembly and test in April 1978, this single validated the concept, but making only 10 hp.This is the 0X oval-piston V-4 that first made 88.7 hp but eventually made 115 hp. (Honda/)Next came a dedicated test engine. Koo employed two connecting rods to better distribute acceleration loads into its single wide piston. It would be topped with the planned eight-valve head.In development they encountered mysterious “missing piston events” when the test engine stopped running and its head was removed, just the two bare con-rods were visible. Where was the piston? In NR testing, cyclic crank twist above 10,000 rpm had caused piston tilting, breaking the piston.By October—six intensive months later—this engine had been developed to reach 15,000 rpm, making 20 hp. Problems showed up as revs rose: loss of ring seal at 10,000, valve spring failures at 12,000. To each problem were attached engineers and technicians, devising, testing, and building possible solutions—lots of them.The First EngineIn November 1978, results so far justified attempting a full V-4 build. For this work the design team isolated itself in a closed-for-the-season hotel, 115 miles away up in the Nasu highlands. One of the many NR engineers, Toshimitsu Yoshimura, recalled that “…we wrapped ourselves in blankets as we drew layouts, because the heat[ing] wasn’t working. I recall our excitement at finally having completed the drawings.”By April of 1979 the result, called “0X,” was making 90 ps (88.7 hp) on the dyno. The next, and extremely ambitious, goal was to test complete motorcycles against the two-strokes in one or more actual GPs that year.The pressed-together and heavily counterweighted crankshaft and its eight one-piece con-rods turned on needle rollers, just as in so many of Honda’s 1960s GP winners. The front cylinder bank was offset to the left, and the cylinder Vee angle was 100 degrees to provide room for four flat-slide carburetors, each with two 30mm throats. The cam drive was at one end of the crank. Bore and stroke were (93.4 x 41) x 36mm (36mm = 1.417 inches). A round piston of the same area would have a 66.5mm diameter.Two connecting rods were used on the oval piston to better distribute forces over the wide area of the oval piston. (Honda/)Would magnesium pistons work? Weighing only 60 percent as much as aluminum, they could reduce bearing loads. Nope, they broke up at 17,000 rpm. How about two-piece piston rings, made in the form of J-shaped elements? Kanazawa eventually overcame poor sealing (smoke!) by changing piston shape from flat faces joined by semi-circular ends, to a more-nearly-elliptical section with continuous curvature. It is the flat-faced type that can be seen in Honda’s museum at Motegi.You can see the curved piston face on the production NR750’s pistons instead of the flat front originally found on the early NR500 test units. (Honda/)Meanwhile, the bike to carry this engine was equally heavy with innovation, having a thin aluminum monocoque chassis, an “upside-down” Showa fork, carbon brakes, and side-mounted radiators. Many production bikes today share technologies originally developed for NR500.First TestFirst track test was at Yatabe Proving Ground in May 1979. Problems appeared, not seen on the dyno, such as breakage of the cam drive gear train. This is the “measles” of young race engines. Crankshafts cannot rotate smoothly because they are driven by combustion thumps, while the camshafts are either slowing as they lift valves against their springs or galloping ahead, driven by spring expansion during closing. These “force lumps” can add up to peak torques many times greater than estimated by calculation. Development of a compliant element to protect the cam drive took time and testing before the breakages ceased. Another problem was inadequate cooling with the NR’s dual side-mounted radiators.Back to Yatabe a month later, they had 105 ps (103.7 hp) and a weight of 276 pounds. This time the prizes were oil leakage, valve failure, and piston breakup. Frantic redesign and unrelenting pressure on suppliers gradually trampled problems to death.At the next test at Suzuka Circuit, rider Takazumi Katayama learned that feel was no substitute for the tachometer’s advice. He, having won the 1977 350 championship on a Yamaha, was used to two-strokes, whose steep power drop-off after peak power signals when to upshift. The NR just revved, giving no tactile warning as its tach swept into the red zone, scattering an engine.First RaceJune and July 1979 were solid with testing and trampling. The goal of all this sleepless effort was the deadline for shipping completed NRs to England for the British GP at Silverstone. With 0X engines, the bikes were run in Silverstone free practice by riders Mick Grant and Katayama.Despite problems crushed, the bikes weren’t ready to race. As the GP annual Motocourse put it that year, “…they sounded fascinating and had the look of extremely intelligent engineering, but they simply got in the way. They were slow down the straight and slow around the corners.”Yet the team had arrived with all the trappings of success: giant transporters, hospitality tents, and teeming personnel. It was maximum humiliation. Grant’s bike was slow to push-start (running engine starts were not adopted until mid-1984), it tipped over at the first corner, and burned. Katayama was shortly out as well.<i>Cycle World</i> Editor Allan Girdler reported on Mick Grant’s and Takazumi Katayama’s less than impressive debut on the NR500. (Cycle World/)Grant later revealed, “They didn’t want me and Katayama coming last and second-to-last so they told us they only put in enough fuel for a few laps, and to put on a show, then stop.”Problems ContinueThe French GP was worse: Both NRs failed to qualify. Despite this, the team rolled them onto the start grid, filling the positions of two qualifiers who had somehow disappeared. Race officials ordered them removed.After this they tested at Donington with rider Ron Haslam, identifying needed improvements. Engine response was jerky, slowing corner exits. The cause? The effort required at roll-on to unstick the eight throttles from the closed position where the engine vacuum was firmly holding them. Mechanics called this “the bang.” Differently shaped throttle pulleys were tested, increasing mechanical advantage at the beginning of throttle slide lift.The NR also shared a problem that had hindered the MV fours toward the end: engine-braking on closed throttle was dragging or hopping the rear wheel, interfering with braking and corner entry. For this they created the slipper clutch, found on so many bikes today. When the rider closes the throttle from high revs, the rear wheel drives the engine. Ramps built into the drive applied this reverse torque to reduce pressure on the clutch plate stack, allowing it to slip, stopping the hop. The slipper’s second application would be the special 1,000cc V-4 bike called “FWS,” designed specifically to win Daytona (it led repeatedly but ate its tires).Continual RefinementIn November, the next NR version, 1X, entered the layout stage. Its weight grew from measures necessary to improve reliability over 0X.In April of 1980 it was clear the monocoque chassis made engine access too slow for the pace of a racing weekend. This was despite Honda’s special NR service trolleys, one for the engine and rear wheel, the other for the chassis and front end. After removing the many 6mm screws joining the two, they could be rolled apart for such routine activities as spark-plug access and carburetor jetting. Too slow. A conventional steel-tube chassis was designed by Ron Williams to provide normal service access and raise stiffness. This English chassis would have a conventional front-mounted radiator and single unit rear suspension.Development of the 0X continued, reaching 115 ps at 17,500.By January 1980 the 1X version had reached 120 ps at 18,000 rpm, now with a central cam drive gear train and a 90-degree Vee angle, made possible by engineering more compact carburetors. Weight had risen to 320 pounds.In October 1980, layout of a 2X version of the NR engine began. Think of the foundry rush jobs, pouring aluminum and magnesium, the high-priority machining, the quality control measures.Near the end of 1980 Mr. Irimajiri received a proposal from engineer Shin’ichi Miyakoshi, then developing two-stroke motocross engines in MX Block. His idea arose from his recent trip to the GP at Assen, where he noticed that race times of 500s and 250/350s were not far apart. What if, he proposed, Honda were to build a simple two-stroke concept he called “a 100 hp 250″? Being light, it would accelerate well and be kinder to tires than the high-horsepower four-cylinder two-strokes.During the 1981 season I would see and hear the NR at Monza, Italy, under the grand old shade trees there. Upon start-up, it sounded exactly like a big muscular Superbike of twice its displacement, deep and powerful. What a contrast on the track and at race revs. It was hard to tell it from Suzuki’s RG500 square-four two-strokes! Why? Because the four-stroke’s peak revs were close to twice those of two-strokes, while the two-strokes fired twice as often per cycle.Despite Honda’s immense commitment of R&D resources, NR won no races and earned no GP points.The NR500 did not accomplish the goal of a GP championship. (John Ulrich/)NR work continued—the 2X version reached 125 ps in 1982, and 1983′s 3X variant reached the original goal of 130 ps at more or less 19,000 rpm. Two basic roadblocks remained: Two-stroke power was now rising steeply, and second, the NR was still 60 pounds heavier than its competition. Two-stroke power growth was coming from subtle changes to cylinder ports and exhaust pipe shape, while four-strokes had to raise rpm to boost power—an expensive process.Honda’s urgent weight reduction program was partly successful, yet the 40 extra pounds of the NR’s machinery-packed cylinder heads remained. Jorg Moller, designer of Franco Morbidelli’s title-winning 125s, said at the time, “Ideas remain important in the design of two-strokes. But for four-strokes, it is only money.”Extrapolating from the 19,000 rpm Honda had needed to reach 130 ps, the 190 hp of the last two-stroke 500s would have required an NR reliable at 28,000. Other inconvenient facts remained: that friction loss tends to rise in proportion to rpm, squared, and that at the originally planned 23,000 rpm maximum, peak piston acceleration would have been 30 percent above anything then current in F1.Was the NR a Failure?The decision came at the end of 1981 to continue oval-piston research while racing Miyakoshi’s two-stroke triple. Freddie Spencer as rider would be essential. The NS500 was making just 108 hp when it started in its first GP in 1982, and Spencer was able to win the very fast Belgian GP at Spa. With the force of Honda R&D behind it, the NS was soon making over 120 hp at 11,000. Freddie rode it to the 1983 500 World Championship. Honda and rider Mick Doohan would ride Honda four-cylinder two-strokes to five consecutive 500 titles.Honda’s NR750 prototype endurance racer. (Honda/)Many years later, engineer Yoshimura was to say, “When I look back…I’m not sure if we were experimenting with cutting-edge technologies or obsessed with foolish ideas.” In his desk at work he kept a small collection of exotic failed parts. He said they reminded him of the many chilly hours spent in the dead of winter at that empty hotel in Nasu.Because at the time FIM rules permitted supercharged 250s to compete in the 500 class, development continued on an oval-piston turbo twin. By October 1984 the turbo project had reached 153 ps on the dyno. It was never raced.Was NR500 a failure? Judged solely on its stated aim to win the 500 title, yes.But when large numbers of creative people are given freedom and resources to explore in all directions, valuable innovation results. Not only were many concepts originated that are in wide use today, but NR gave Honda a new direction in engine design: the V-4. It was immediately ordered into production as Magna (1982) and Interceptor (1983), remaining to this day a living engine architecture. Think of Bell Labs in its free-wheeling days, leaving researchers alone to follow their interests. Who can place a value on their most significant result: the transistor?Just 300 street-legal production 1994 NR750 units were built. (Brian Blades/)Oval pistons were banned from Formula 1 use, but Honda did produce some 300 examples of an NR750 production bike. Like so many limited-production exotics, they were gobbled up by collectors and now repose in temperature- and humidity-controlled rooms.It’s interesting to note that two cylinders from a present-day 1,000cc V-4 MotoGP engine would give just under 150 hp at 20,000 rpm with round cylinders, four valves, and plain bearings. The impossible becomes the possible, and what once appeared magical becomes commonplace.From the NR500 (right front) came the 1981 Suzuka 200-Kilometer winning NR500 (2X) ridden by Kengo Kiyama (right rear), the NR750 Endurance Racer (left rear), and the 1994 NR750 production model (left front). (Honda/)
Full Text:
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/)
Who remembers Honda’s NR500, the oval-piston V-4 four-stroke whose intensive and costly five-year development was intended to stop the two-stroke tide in FIM Grand Prix roadracing? Many more remember American Freddie Spencer (1983, 1985) and Australian Mick Doohan (1993–1997) later dominating Grand Prix bike racing on Honda two-strokes. What unseen force could have triggered such a switch by a company so publicly devoted to the four-stroke principle?
Look even further back, to the period 1959–1967, when Honda established its name worldwide by winning 16 world roadracing championships. Its tools were daringly designed multicylinder four-strokes from 50cc to 350cc, some revving as high as 21,500 rpm. At the end of 1967, Honda stopped racing. Its signature megaphone exhausts went silent as the company devoted all its R&D resources to entering the auto business (resoundingly, with the Civic of 1972).
Honda’s 1979 NR500 featured groundbreaking technology—most notably a V-4 engine with oval pistons and eight valves per cylinder. (Honda/)
Through the following decade, Honda assumed the role of “the General Motors of motorcycling,” growing so big and so successful that it had nothing more to prove.
That ended when Honda’s new president, Kiyoshi Kawashima, chose the occasion of the CBX 1000 launch in October 1977 to announce Honda’s imminent return to Grand Prix motorcycle racing.
How To Defeat the Two-Stroke?
How would Honda overcome the dominant two-strokes, even as the last four-stroke holdout, the Italian MV Agusta team, was giving up the fight? MV had been the architect of 38 world titles, 18 of them in 500, the class Honda had never won.
Honda’s project was given the name NR 500—NR for “New Racer.” A special organization, NR Block, was created within Honda’s Asaka R&D Center. Famous names led the project; Shoichiro Irimajiri, whose designs had won Honda’s last 1960s championships, was overall leader. Engine leader was Suguru Kanazawa, chassis leader Satoru Horiike—both later rose to directorships of HRC. And a cast of hundreds.
The three major determinants of engine power are displacement, rpm, and net, stroke-averaged combustion pressure. Because the last is limited by atmospheric pressure, and engine displacement is fixed by rule (500cc, or 30.5ci), that left rpm as the path to competitive power—as it had been in Honda’s 1960s racing. That path appeared to be blocked by the 500cc class limit of four cylinders (otherwise the company would have immediately begun development of a V-8).
Inspiration Strikes
The official story is that one evening, Irimajiri was commuting home, wondering why pistons had to be round. If they could be made oval and still seal, many more valves could be fitted, able to supply fuel-air mixture at fabulous rpm. This would be equivalent to a 500cc V-8 whose pistons had fused together in pairs, assuming oval form. It was proposed that such an engine could make 130 ps at 23,000 rpm (ps was metric horsepower, now replaced by kilowatts in European power measurement).
The first step was to test the sealing ability of oval pistons and rings. A simple four-valve test engine, designated Koo, was improvised from an XL250 bottom end. Methods of manufacturing the unusual parts had to be devised. Upon assembly and test in April 1978, this single validated the concept, but making only 10 hp.
This is the 0X oval-piston V-4 that first made 88.7 hp but eventually made 115 hp. (Honda/)
Next came a dedicated test engine. Koo employed two connecting rods to better distribute acceleration loads into its single wide piston. It would be topped with the planned eight-valve head.
In development they encountered mysterious “missing piston events” when the test engine stopped running and its head was removed, just the two bare con-rods were visible. Where was the piston? In NR testing, cyclic crank twist above 10,000 rpm had caused piston tilting, breaking the piston.
By October—six intensive months later—this engine had been developed to reach 15,000 rpm, making 20 hp. Problems showed up as revs rose: loss of ring seal at 10,000, valve spring failures at 12,000. To each problem were attached engineers and technicians, devising, testing, and building possible solutions—lots of them.
The First Engine
In November 1978, results so far justified attempting a full V-4 build. For this work the design team isolated itself in a closed-for-the-season hotel, 115 miles away up in the Nasu highlands. One of the many NR engineers, Toshimitsu Yoshimura, recalled that “…we wrapped ourselves in blankets as we drew layouts, because the heat[ing] wasn’t working. I recall our excitement at finally having completed the drawings.”
By April of 1979 the result, called “0X,” was making 90 ps (88.7 hp) on the dyno. The next, and extremely ambitious, goal was to test complete motorcycles against the two-strokes in one or more actual GPs that year.
The pressed-together and heavily counterweighted crankshaft and its eight one-piece con-rods turned on needle rollers, just as in so many of Honda’s 1960s GP winners. The front cylinder bank was offset to the left, and the cylinder Vee angle was 100 degrees to provide room for four flat-slide carburetors, each with two 30mm throats. The cam drive was at one end of the crank. Bore and stroke were (93.4 x 41) x 36mm (36mm = 1.417 inches). A round piston of the same area would have a 66.5mm diameter.
Two connecting rods were used on the oval piston to better distribute forces over the wide area of the oval piston. (Honda/)
Would magnesium pistons work? Weighing only 60 percent as much as aluminum, they could reduce bearing loads. Nope, they broke up at 17,000 rpm. How about two-piece piston rings, made in the form of J-shaped elements? Kanazawa eventually overcame poor sealing (smoke!) by changing piston shape from flat faces joined by semi-circular ends, to a more-nearly-elliptical section with continuous curvature. It is the flat-faced type that can be seen in Honda’s museum at Motegi.
You can see the curved piston face on the production NR750’s pistons instead of the flat front originally found on the early NR500 test units. (Honda/)
Meanwhile, the bike to carry this engine was equally heavy with innovation, having a thin aluminum monocoque chassis, an “upside-down” Showa fork, carbon brakes, and side-mounted radiators. Many production bikes today share technologies originally developed for NR500.
First Test
First track test was at Yatabe Proving Ground in May 1979. Problems appeared, not seen on the dyno, such as breakage of the cam drive gear train. This is the “measles” of young race engines. Crankshafts cannot rotate smoothly because they are driven by combustion thumps, while the camshafts are either slowing as they lift valves against their springs or galloping ahead, driven by spring expansion during closing. These “force lumps” can add up to peak torques many times greater than estimated by calculation. Development of a compliant element to protect the cam drive took time and testing before the breakages ceased. Another problem was inadequate cooling with the NR’s dual side-mounted radiators.
Back to Yatabe a month later, they had 105 ps (103.7 hp) and a weight of 276 pounds. This time the prizes were oil leakage, valve failure, and piston breakup. Frantic redesign and unrelenting pressure on suppliers gradually trampled problems to death.
At the next test at Suzuka Circuit, rider Takazumi Katayama learned that feel was no substitute for the tachometer’s advice. He, having won the 1977 350 championship on a Yamaha, was used to two-strokes, whose steep power drop-off after peak power signals when to upshift. The NR just revved, giving no tactile warning as its tach swept into the red zone, scattering an engine.
First Race
June and July 1979 were solid with testing and trampling. The goal of all this sleepless effort was the deadline for shipping completed NRs to England for the British GP at Silverstone. With 0X engines, the bikes were run in Silverstone free practice by riders Mick Grant and Katayama.
Despite problems crushed, the bikes weren’t ready to race. As the GP annual Motocourse put it that year, “…they sounded fascinating and had the look of extremely intelligent engineering, but they simply got in the way. They were slow down the straight and slow around the corners.”
Yet the team had arrived with all the trappings of success: giant transporters, hospitality tents, and teeming personnel. It was maximum humiliation. Grant’s bike was slow to push-start (running engine starts were not adopted until mid-1984), it tipped over at the first corner, and burned. Katayama was shortly out as well.
<i>Cycle World</i> Editor Allan Girdler reported on Mick Grant’s and Takazumi Katayama’s less than impressive debut on the NR500. (Cycle World/)
Grant later revealed, “They didn’t want me and Katayama coming last and second-to-last so they told us they only put in enough fuel for a few laps, and to put on a show, then stop.”
Problems Continue
The French GP was worse: Both NRs failed to qualify. Despite this, the team rolled them onto the start grid, filling the positions of two qualifiers who had somehow disappeared. Race officials ordered them removed.
After this they tested at Donington with rider Ron Haslam, identifying needed improvements. Engine response was jerky, slowing corner exits. The cause? The effort required at roll-on to unstick the eight throttles from the closed position where the engine vacuum was firmly holding them. Mechanics called this “the bang.” Differently shaped throttle pulleys were tested, increasing mechanical advantage at the beginning of throttle slide lift.
The NR also shared a problem that had hindered the MV fours toward the end: engine-braking on closed throttle was dragging or hopping the rear wheel, interfering with braking and corner entry. For this they created the slipper clutch, found on so many bikes today. When the rider closes the throttle from high revs, the rear wheel drives the engine. Ramps built into the drive applied this reverse torque to reduce pressure on the clutch plate stack, allowing it to slip, stopping the hop. The slipper’s second application would be the special 1,000cc V-4 bike called “FWS,” designed specifically to win Daytona (it led repeatedly but ate its tires).
Continual Refinement
In November, the next NR version, 1X, entered the layout stage. Its weight grew from measures necessary to improve reliability over 0X.
In April of 1980 it was clear the monocoque chassis made engine access too slow for the pace of a racing weekend. This was despite Honda’s special NR service trolleys, one for the engine and rear wheel, the other for the chassis and front end. After removing the many 6mm screws joining the two, they could be rolled apart for such routine activities as spark-plug access and carburetor jetting. Too slow. A conventional steel-tube chassis was designed by Ron Williams to provide normal service access and raise stiffness. This English chassis would have a conventional front-mounted radiator and single unit rear suspension.
Development of the 0X continued, reaching 115 ps at 17,500.
By January 1980 the 1X version had reached 120 ps at 18,000 rpm, now with a central cam drive gear train and a 90-degree Vee angle, made possible by engineering more compact carburetors. Weight had risen to 320 pounds.
In October 1980, layout of a 2X version of the NR engine began. Think of the foundry rush jobs, pouring aluminum and magnesium, the high-priority machining, the quality control measures.
Near the end of 1980 Mr. Irimajiri received a proposal from engineer Shin’ichi Miyakoshi, then developing two-stroke motocross engines in MX Block. His idea arose from his recent trip to the GP at Assen, where he noticed that race times of 500s and 250/350s were not far apart. What if, he proposed, Honda were to build a simple two-stroke concept he called “a 100 hp 250″? Being light, it would accelerate well and be kinder to tires than the high-horsepower four-cylinder two-strokes.
During the 1981 season I would see and hear the NR at Monza, Italy, under the grand old shade trees there. Upon start-up, it sounded exactly like a big muscular Superbike of twice its displacement, deep and powerful. What a contrast on the track and at race revs. It was hard to tell it from Suzuki’s RG500 square-four two-strokes! Why? Because the four-stroke’s peak revs were close to twice those of two-strokes, while the two-strokes fired twice as often per cycle.
Despite Honda’s immense commitment of R&D resources, NR won no races and earned no GP points.
The NR500 did not accomplish the goal of a GP championship. (John Ulrich/)
NR work continued—the 2X version reached 125 ps in 1982, and 1983′s 3X variant reached the original goal of 130 ps at more or less 19,000 rpm. Two basic roadblocks remained: Two-stroke power was now rising steeply, and second, the NR was still 60 pounds heavier than its competition. Two-stroke power growth was coming from subtle changes to cylinder ports and exhaust pipe shape, while four-strokes had to raise rpm to boost power—an expensive process.
Honda’s urgent weight reduction program was partly successful, yet the 40 extra pounds of the NR’s machinery-packed cylinder heads remained. Jorg Moller, designer of Franco Morbidelli’s title-winning 125s, said at the time, “Ideas remain important in the design of two-strokes. But for four-strokes, it is only money.”
Extrapolating from the 19,000 rpm Honda had needed to reach 130 ps, the 190 hp of the last two-stroke 500s would have required an NR reliable at 28,000. Other inconvenient facts remained: that friction loss tends to rise in proportion to rpm, squared, and that at the originally planned 23,000 rpm maximum, peak piston acceleration would have been 30 percent above anything then current in F1.
Was the NR a Failure?
The decision came at the end of 1981 to continue oval-piston research while racing Miyakoshi’s two-stroke triple. Freddie Spencer as rider would be essential. The NS500 was making just 108 hp when it started in its first GP in 1982, and Spencer was able to win the very fast Belgian GP at Spa. With the force of Honda R&D behind it, the NS was soon making over 120 hp at 11,000. Freddie rode it to the 1983 500 World Championship. Honda and rider Mick Doohan would ride Honda four-cylinder two-strokes to five consecutive 500 titles.
Honda’s NR750 prototype endurance racer. (Honda/)
Many years later, engineer Yoshimura was to say, “When I look back…I’m not sure if we were experimenting with cutting-edge technologies or obsessed with foolish ideas.” In his desk at work he kept a small collection of exotic failed parts. He said they reminded him of the many chilly hours spent in the dead of winter at that empty hotel in Nasu.
Because at the time FIM rules permitted supercharged 250s to compete in the 500 class, development continued on an oval-piston turbo twin. By October 1984 the turbo project had reached 153 ps on the dyno. It was never raced.
Was NR500 a failure? Judged solely on its stated aim to win the 500 title, yes.
But when large numbers of creative people are given freedom and resources to explore in all directions, valuable innovation results. Not only were many concepts originated that are in wide use today, but NR gave Honda a new direction in engine design: the V-4. It was immediately ordered into production as Magna (1982) and Interceptor (1983), remaining to this day a living engine architecture. Think of Bell Labs in its free-wheeling days, leaving researchers alone to follow their interests. Who can place a value on their most significant result: the transistor?
Just 300 street-legal production 1994 NR750 units were built. (Brian Blades/)
Oval pistons were banned from Formula 1 use, but Honda did produce some 300 examples of an NR750 production bike. Like so many limited-production exotics, they were gobbled up by collectors and now repose in temperature- and humidity-controlled rooms.
It’s interesting to note that two cylinders from a present-day 1,000cc V-4 MotoGP engine would give just under 150 hp at 20,000 rpm with round cylinders, four valves, and plain bearings. The impossible becomes the possible, and what once appeared magical becomes commonplace.
From the NR500 (right front) came the 1981 Suzuka 200-Kilometer winning NR500 (2X) ridden by Kengo Kiyama (right rear), the NR750 Endurance Racer (left rear), and the 1994 NR750 production model (left front). (Honda/)
