Harley-Davidson VR1000 - God's Own Voice

It is 1986, and Harley-Davidson is in the midst of a rebirth. After years of struggling under AMF ownership and suffering through poor quality, lagging sales, and a tarnished reputation, the 1980s have offered a new era of prosperity for America’s perennial motorcycle manufacturer. Following the purchase of the works from AMF by a group of investors led by Willie G. Davidson in 1981, a major restructuring has restored solvency to the marque. And now the company is looking to recapture some of the racing successes that had driven their brand for decades. The XR program led by Dick O'Brien in the early 1970s had given The Motor Company a strong base for success in American racing, but it was limited to dirt track and a few notable but fleeting wins in European road racing with Renzo Pasolini and Cal Rayborn aboard the XRTT. With the coffers finally filling after the dark days, Harley's reputation improving, and production steadily climbing, the mid-1980s seemed like the ideal time to begin a new program that would lead to the development of the most potent, most modern motorcycle HD would ever create.

This is the story of the VR1000, the Superbike contender that was hoped to put Harley-Davidson back on the road racing podium. This is a story you might think you are familiar with, but the truth of the matter is that you haven't heard the real story of the VR, how it came to be, and how it came to end.

Following the investor's purchase of the company from AMF, Harley-Davidson was a shell of a company on the brink of bankruptcy. Operating with strictly just-in-time manufacturing and staffed by a bloated workforce despite years of sliding sales, at the time there appeared to be little hope for the once-storied brand. Painful and drastic reorganization was needed – the first order of business for the new ownership was to reduce the workforce by a staggering 40% and stem the tide of losses while new models were developed. In 1983 Ronald Reagan's notorious 700cc-plus 45% import tariffs were interfering with Harley's overseas competitors and allowing the company to gradually reclaim the US market. By 1984 things were on an upswing, with the introduction of the new generation of Evolution big twins and the creation of the popular Softail range, two important developments that would drive the company's success in the coming decade.

Private road racing successes began to make headlines. Harley dealership owner and legendary tuner Don Tilley, backed by Dick O'Brien's racing department, developed the XR-powered series of machines that became known as Lucifer’s Hammer. Operating outside the factory with sponsorship from the Harley Owner’s Group, these thundering warriors began earning wins in AMA's Battle of the Twins category from 1983 onward. Gene Church and Jay Springsteen were doing well aboard Tilley's machines, which used pushrod, undersquare 997 cc XR powerplants in modified XL Sportster chassis (and with Lucifer's Hammer II, an Erik Buell frame design), a combination that seemed unlikely to win against the highly refined European offerings – but win they did, and with this success came the idea to begin an in-house Harley road racing effort in earnest.

Willie G. Davidson had been a strong proponent of racing as an integral part of Harley's history, and pushed hard for a return to racing once the books were balanced. HD CEO Vaughn Beals had attended Laguna Seca in 1986 and voiced his concern about how few Harley riders were present in the crowds. The conception for an all-new, entirely modern road racer emerged in this fruitful period, with pencils hitting the drawing board in 1987.

Mark Tuttle, Vice President of Harley-Davidson engineering, organized a group within the company to develop the new machine. A meeting was organized by Tuttle in 1987 to discuss development of a new platform. The initial plan was to develop a modified air-cooled twin based on XR750 architecture with oversquare dimensions and a five speed transmission.

Erik Buell, then operating independently after working as an engineer for Harley from 1979 to 1983, suggested developing a modern liquid-cooled V-twin to compete with the four cylinder opposition, citing the better tractability and torque of such a platform compared to the 750 fours. He felt that an air cooled engine wouldn't be competitive in road racing and that the best course of action would be to adapt existing 500cc cylinder designs from an automotive racing engine. Overhead cams and four valves (or five, an idea pursued in early prototyping) per cylinder would be employed. Ducati's contemporary resurgence in Superbike racing with their 851 desmoquattro was a clear indicator of the viability of such a configuration. An initial configuration of a 92mm bore (the same as the 851) with a 75mm stroke to reach 998 ccs was proposed, closer to the 1000 cc ceiling for twins in AMA Superbike than Ducati was then using.

The resulting machine would be unlike any Harley-Davidson and would be a far cry from the crudity of Lucifer’s Hammer, which relied on antiquated technology made competitive against more advanced machines through force of brute engineering. The mandate came to be for a liquid-cooled, overhead cam V-twin housed in an entirely new chassis. It would not have any base in production Harleys and would be developed from the ground up as a racing machine, aimed at competing in AMA Superbike where 1000 cc twins went toe-to-toe against 750 cc fours.

Buell was hired to design the chassis while HD designer Mark Miller was tasked with developing the engine around Buell's blueprints. Engineering input was sought from Cosworth and Jerry Branch from Branch Flowmetrics, who had personally built the heads for the production XR1000 and would work on some of the early cylinder head designs of the VR.

The first meeting's spec for a 92x75mm design were abandoned in favour of a 95x70mm ( 992 cc) configuration sketched by Buell and based on the Cosworth BDG 2000 cc four cylinder engine, laid into a narrow-angle 60-degree V-twin – in fact the prototype engine was expected to use standard BDG pistons as a starting point. Miller designed a unit construction bottom end with dry sump lubrication based on Buell's specs. The first gearbox design was based on a modified XL five-speed transmission.

A narrow Vee angle would give compact external dimensions and a short overall length that would make chassis layout easier. Ducati and other marques have often struggled with the length of a 90-degree twin, which places the front cylinder too close to the front wheel and the rear cylinder (and its exhaust plumbing) in the same space you'd traditionally want to stick a monoshock suspension. Keeping the wheelbase tight and the weight distribution correct requires a lot of fettling that has led to some odd solutions, like the infamous rotary damper fitted to Suzuki's TL series, or Bimota placing the rear shock above the engine next to the throttle bodies on the TL-powered SB8, or Buell's underslung shock that mounts below the crankcases. With a narrow angle twin you solve most of the packaging problems at the expense of the primary balance of the engine, an issue that ended up plaguing the early VR engine.

Buell RR1000 Battle Twin.

While sketching the prototype frame, which was initially to be a steel tube design similar to his XR1000-powered RR-1000, Buell consulted with Cosworth and had a revelation. The gurus at Cosworth noted that the best way to make power in a four-valve design is to feed as much air as possible using a large airbox design. A back-of-the-envelope calculation showed that a 24 litre airbox volume would be ideal – the problem was where to stick it, when the intake for a narrow angle twin needed to occupy the same space as the fuel tank above the engine. So Buell came up with an ingenious solution he would revisit 16 years later with his XB series: get rid of the tank entirely and make the frame do double-duty as a fuel cell, carrying the fuel inside the hollow beams of large alloy spars. While not an entirely new idea (the Pierce Four carried its fuel and oil in the frame tubes as early as 1909) Buell's design was the first application of this innovation in a modern motorcycle chassis.

Image Source

Given the inherent vibrations of a narrow-angle twin, Buell specified a version of his uniplanar rubber-mounting configuration for the engine, isolating the vibration from the twin-spar chassis. The crankcases and heads of the prototype engine were designed to suit – a pair of large mounting brackets secured the rear of the crankcases to the rubber mounts at the base of the frame, with a small forward mount on the front cylinder head supporting the other end. Split radiators were mounted flat against either side of the frame. Wheelbase was a tidy 54 inches, quite a bit shorter than the Ducati 851 which was closer to 57.

Image Source

By this time Ducati had been saved from the brink of bankruptcy by their production Desmoquattro design. Despite falling short of the 1000 cc limit of the Superbike series, the 851 proved to be a force to be reckoned with on the track. First tested in 1985 but officially introduced into World Superbike in 1987, the 851 would see continual development through 2001 with the 996 RS before being superseded by the Testastretta architecture in 2001. With its Cosworth-inspired cylinder head and liquid cooling mated with a good chassis, backed by millions of dollars of factory support and bleeding-edge tuning, the Ducatis quickly became the benchmark for twin-cylinder racers. Desmoquattros won the World Superbike championship in 1990, 1991, 1992, 1994, 1995, 1996, 1998, 1999, and 2001, took the 1993 and 1994 AMA Superbike championships, as well as the 1995, 1999, 2000 and 2001 British Superbike championships. Harley appeared to be on the right track.

Image courtesy Denis McCarthy.

Once work was underway on the engine design, it became clear to the Harley engineers that they were out of their element trying to develop a competitive liquid-cooled, quad-cam engine. It was at this point, in 1989, that H-D made the fateful decision to contract Roush Industries in Livonia, Michigan to help develop the top end for the VR. Steve Scheibe was entrusted with the task of designing the cylinder head and the electronic fuel injection system. An employee of Roush since 1985, Scheibe was a Michigan-based engineer and experienced road racer who already had some Harley projects under his belt, in addition to automotive projects for Chrysler and Ford and work for Mercury Marine – he had helped developed a four-valve pushrod head for the Evolution engine, dubbed Abacus, as well as a prototype fuel injection system for production Harleys.

First-generation uniplanar-mount VR engine. Note the bosses on the rear of the crankcase and the shape of the stator cover. Image courtesy Steve Scheibe.

Using Mark Miller's bottom end Scheibe helped complete the first generation VR engine. To reach a proposed rev ceiling of 11,000 RPM Scheibe rejected the BDG dimensions and proposed a then-extremely oversquare design with a 98mm bore and 66mm stroke giving 996 ccs. Wiseco pistons were used, along with Carillo connecting rods. The initial design used a chain-driven idler for the dual-overhead cams added outboard of the alternator on the left side, for a total of three cam chains. The crankcases featured the rubber mounting system specified for the first Buell chassis. Racing redline was set at 10,800 RPM, with a "street" redline of 10,200. Initial dyno testing with a Weber fuel injection system yielded around 140 HP - and cracked crankcases at 8000 RPM due to bolt bosses that proved to be inadequate at that output.

It was a start, but far more development was needed to make the VR engine reliable and competitive. Chief concerns were the lack of a six speed gearbox (and no provisions to fit one into the tight confines of the crankcases designed around a five speed), the unacceptable vibration levels without a counterbalancer, the considerable weight of the powerplant, and what proved to be a destructive cam drive layout; with three inches of unsupported crankshaft sticking out past the alternator to drive the primary cam chain, resonance was setting in at high RPM and promptly destroying the primary cam chain tensioners and snapping the chains themselves.

Early crankshaft.

Three of these first generation engines were built for dyno runs and initial chassis testing, while Scheibe set to work addressing the shortcomings. A primary-frequency counterbalancer gear, driven at engine speed off the crankshaft, was engineered - no mean feat as it had to be compact enough to fit within the existing crankcases. A new 6 speed gearbox with a modern shift linkage was developed to replace the XL-based unit. The primary cam drive was redesigned, using a bigger primary chain inboard of the alternator to eliminate the resonance issues, enabling reliable performance up to the 11,000 RPM target. Later development would replace the primary chain with a three-gear system engineered to fit within the existing space, which along with other valvetrain updates gave reliable valvetrain performance up to 13,000 RPM. Rigid mounts were incorporated into the castings. With the basic architecture of this second-generation engine then determined as the basis for the production VR, the remainder of development was a slow process of massaging more power, increasingly reliability, and paring down weight.

Buell's second fuel-in-frame chassis for the VR. Image Courtesy Steve Scheibe.

Buell continued to refine his chassis design and adapted it to suit the proposed rigid-mounted second generation engine, but these prototype chassis was never fitted with a running engine (years later, after Harley had purchased Buell, they installed a production VR engine into Buell's chassis for testing). Meanwhile Harley turned to a York, Pennsylvania factory employee named Mike Eatough to provide an alternative design. Eatough had considerable experience in the field of chassis design, having worked for UK-based Armstrong Industries. In the 1970s Armstrong had built a formidable team of chassis designers through the purchase of Clews Competition Motorcycles (CCM), Cotton Motorcycles, and Barton Engineering. The company's designs had done well in various categories into the 1980s, with wins at the Isle of Man, the British Championship, and in Grand Prix racing. In addition to designing several Grand Prix frames, Eatough had the distinction of helping design the first carbon fibre chassis for the Carbon Fibre Armstrong in 1983.

Prototype with Harris-built chassis and first-generation engine. Image courtesy Steve Scheibe.

Eatough had come to Harley following H-D's purchase of Armstrong's military and off-road division in the late 80s and was an ideal candidate for an in-house chassis designer for the VR. Harley brass, in particular Willie G., were keen to bring all the expertise into the factory to make the VR a truly home-grown racing effort. It was for this reason that Scheibe was hired from Roush by HD in 1991 as the new manager of racing, with an office at the Juneau Avenue facility in Milwaukee, Wisconsin. He was informed at the time that his program was intended as a test bed for future development that would trickle down to production Harleys, a way of introducing new ideas and refining them in a flagship racing effort – ideas like liquid cooling, overhead cams, modern cylinder heads, and fuel injection.

Image courtesy Steve Scheibe.

Buell's fuel-in-frame design was abandoned in favour of a more conventional twin-spar design penned by Eatough, which sacrificed some of the airbox volume offered by the fuel-in-frame layout. Harris was contracted to build the prototype frame. New mounts were added to the first generation crankcases to suit; with the second generation motor still under development, the first VR would use the unbalanced prototype engine. By 1991 a running prototype using the Harris chassis was assembled and delivered to Blackhawk Farms Raceway in South Beloit, Illinois for initial testing. Kawasaki ZX7 fairings, repainted in the corporate black and orange colour scheme, were slapped onto the test mule which was ridden by Buell employee Scott Zampach as well as Scheibe himself. Extreme vibrations were noted from the unbalanced engine at high RPMs, but the tests were successful.

Testing the Harris prototype at Blackhawk Farms. Image courtesy Steve Scheibe.

A Harley Owner's Group representative who was present at the Blackhawk test told Scheibe the VR "Sounds like God's own voice."

Further testing was done by the late Fritz Kling, who rode the Harris prototype at Grattan Raceway in Michigan and would go on to become one of the factory VR1000 riders in the 1994 season.

Image courtesy Steve Scheibe.

To be legal in AMA Superbike racing Harley needed to produce a series of ostensibly street-legal machines for homologation. AMA Rules at the time were as follows: four cylinder machines from a small manufacturer required a minimum of 500 units to homologate, while four cylinder machines from a major manufacturer required 1000 units. V-Twin machines only needed 50 examples to be homologated, a rule that favoured Ducati's limited production numbers and penchant for producing cost-no-object homologation specials that often stretched the definition of "street-legal" to the absolute limit. Harley discovered that two could play at that game. Nowhere did the AMA specify where the machines had to be legal. A certain executive at the company put it bluntly to the team:  "Go somewhere in the world and get us a homologation certificate." The Harley production engineering department was never significantly involved in the VR program, as it was viewed as a Race Department project, and there were no real provisions for making the machine a street legal motorcycle beyond slapping on lights, turn signals, and a starter. In 1993 a single machine was quietly pushed through the German Technischer Überwachungs-Verein (TÜV) by one of the company's contacts who helped certify production Harleys in Germany. That lone VR1000 was given a certificate of homologation for sound and emissions. That proved to be good enough for the AMA.

You may have heard that the VR was only legal in a certain Eastern European nation that borders Germany. This isn't true. At some point a rumour began in the press that the VR was homologated in Poland, an idea that is presumed to have started after a journalist saw an example wearing a Polish plate and made a rather large jump to that conclusion. Harley executives saw this as a perfect cover story to protect their contact at the notoriously strict TÜV and did nothing to quell the rumour.

Cycle World cover bike. Image courtesy Steve Scheibe.

At this point Cycle World expressed interest in doing a cover story on the VR1000 for their March 1994 issue. Kevin Cameron was sent to interview the team and gather technical data. While the chassis and engine were mostly ready for primetime, the bodywork was not. Willie G. was in charge of styling, and wanted to clothe the American machine in bodywork that was distinctly unlike the Asian and Italian competition, something that would appear distinctive on the track. It was his idea to paint the two sides of the bike in contrasting orange and black, bisected by a white racing stripe down the middle. Scheibe served as the ergonomic model, and a body was built in time for the Cycle World photo shoot – look closely and you’ll see the cardboard panels spray painted black to fill in the unfinished areas. To keep within the American-made theme, the Cycle World bike used Penske forks and rear shock, while the brakes and master cylinders were provided by Wilwood. Later testing showed the forks to be problematic and prone to stiction; after the 1994 Daytona introduction more proven Ohlins forks were installed. Racing machines used 6 piston AP Racing calipers and AP master cylinders (replaced by Brembo racing hardware in 1998) while production bikes retained the six-piston Wilwood brakes, with Ohlins forks and Marchesini wheels. All production bikes, and at times the racing machines, used a Penske shock working through a rising-rate linkage on a braced box-section alloy swingarm.

Cycle World cover bike. Note the black cardboard insert. Image courtesy Steve Scheibe.

Production began in 1994. The first 10 examples were built at Roush, with the remainder assembled under Eatough’s supervision at York. All subsequent production engines would be assembled by Roush. Alloy fabrication and extrusion specialist Anodizing Inc. in Portland, Oregon built all the production frames. Once the first run of machines was built AMA inspectors visited the factory and gave the final OK for homologation. The AMA was less concerned with street legality than it was that any racer who desired to purchase an example would have access to one. In this respect the VR1000 never had any trouble – at $49,000 USD a pop (later reduced to $34,000) for a then unproven racing machine, it is unlikely demand would have ever outstripped supply. By the end of the year a total of 55 complete machines were built, with an additional 8 (or so) frames made as spares. Exact production figures are difficult to pin down given that these were racing bikes - factory machines were regularly torn down and reassembled with parts from other bikes or bits from the spares bin, and damaged frames were replaced with new items bearing the same serial number.

Production lineup at York, Pennsylvania facility. Image courtesy Steve Scheibe.

Despite the presence of lights and an electric starter, the production VR was far from a street bike. A Motorrad magazine road test revealed the box-stock bike weighed 451 lbs and produced 116 hp at the wheel (versus a claimed 135 hp at the crankshaft), which was pretty impressive for a V-twin sportbike in the mid-1990s. But this wasn’t a highly-polished mass-produced machine with a warranty; it was a privateer race bike with lights slapped on. Look at a VR up close and you’ll note rough edges that reveal its true nature – the weave of the carbon fibre bodywork is apparent through the paint, the engine components are sandcast, the plumbing and wiring are haphazard with no attempts to hide the functional bits for cosmetic reasons, and the finishing isn’t as polished as you might imagine. The testers noted that the alternator was so weak that it couldn’t keep the battery charged on long rides, first gear was extremely tall, the exhaust was ridiculously loud for a “stock” bike, and by the end of the test the engine and fork seals were weeping oil. However they found no faults with the chassis or powerband, noting impressive shove above 5000 RPM and solid handling.

Miguel Duhamel and Steve Scheibe. Image courtesy Steve Scheibe.

The VR hit the track during the 1994 AMA Superbike season with Miguel Duhamel as the factory’s star rider, lured away from his career in Grand Prix racing with a substantial salary offer from Harley, with the bike debuting at the Daytona 200. The machine had been developed within the expectations of releasing the VR around 1990-91, and it would have likely been competitive at that time. But four years was an eternity in racing development and by the time it was released the VR was short on power, still producing around 140-150 hp which put them at least 10 horses behind the competition. The factory effort consisted of four machines with five sets of spares.

Image courtesy Steve Scheibe.

Daytona proved to be an ignominious start. Duhamel’s machine lagged on the banking due to its lack of peak horsepower, but Duhamel rode well and made up ground on the infield.  Then a weld failed on the counterbalancer gear and the engine blew. The VR’s first trial by fire ended with a DNF.

VR production at the York facility. Image courtesy Steve Scheibe.

While the VR had trouble keeping up on the straights, nobody complained about the handling (aside from some media pundits and at least one journalist who managed to snag a test ride aboard the “street” version and noted some wayward characteristics). Duhamel called it the best handling bike he had ever ridden. Following his move to Honda in 1995, it was heard through the grapevine that the HRC engineers were sick of listening to Duhamel tell them how good the Harley handled. Not every rider had such high praise for the VR; word was that Doug Chandler had some complaints, and race commentators were quick to note from their armchairs that some riders might not have been suited for the VR’s style of power and handing – the traits inherent in a bigger, meatier V-twin powered race bike, which are quite distinct from “traditional” four-cylinder superbikes.

Following the unsuccessful intro at Daytona, early results seemed promising though plagued by bad luck. Duhamel led the pack at Brainerd International Raceway in Minnesota, but ran wide in a turn on the final lap and was passed with only three corners left to go. He led once again at Mid-Ohio, but lost when a shifter bolt came undone due to heat from the exhaust seizing one of the Heim joints. This pattern of promising results dashed by bad luck became the modus operandi of the VR1000 race program – always tantalizingly close to a victory despite a power deficit, but plagued by niggling issues and minor mistakes that intervened before the win.

Privateers who purchased production VRs and raced alongside the factory team became a significant influence on the program – good and bad. Tuners who thought they could do better than Harley began tinkering with the engines and reliability suffered, reflecting poorly on HD despite the fact they had nothing to do with many of the issues. Privateers struggling against the multi-million-dollar Ducati, Honda, Kawasaki, Suzuki and Yamaha teams served as a unique market and a source of valuable feedback for further development of the VR. That and a laundry list of shortcomings following the introductory ’94 season lead to the establishment of an independent firm to run the VR1000 program. Gemini Racing began in 1995 as a way for privateers and the factory team to pool their knowledge and solve problems together while taking the burden away from HD proper, making for quicker turnaround and more independent problem solving and parts fabrication. From 1996 onward engines were leased from the factory to the private teams – the little guys had access to the same powerplant pool as the Harley big shots, and would benefit from the same development process.

Gemini Racing. Image courtesy Steve Scheibe.

The decision to setup Gemini Racing as an independent team was inspired, in part, by the VR1000’s dry clutch. The early models suffered issues with their clutch basket design and the weak paper material friction plates they used. Scheibe modified the basket and used Ducati plates as a temporary fix, then sought help from technicians in the Harley tool room. He was dismayed to learn the hierarchy of priorities – production line issues and production bikes came first, then a few other departments, with the VR sitting somewhere around number seven on the list. With Mark Tuttle’s support, Gemini was created as a way of outsourcing problem solving and taking some of the effort away from Harley-Davidson engineering, which was well occupied with the upcoming Twin Cam 88 and fulfilling the enormous production demands during Harley’s mid-1990s heyday. Under Scheibe’s direction, Gemini was established at a new location in Mukwanago, Wisconsin, 30 miles from Juneau Avenue.

Don Tilley once again became a significant figure in Harley road racing, fielding his own team of VR1000s and working closely with Gemini. He became a strong supporter of the program after one notable incident. During practice at an event Tilley noted that his rider was down a few miles per hour on the straight compared to the factory team. He hinted that the factory bike might be a ringer and that he was getting the short end of the stick. The next day Scheibe invited Tilley into the Gemini trailer. He told him to pick any one of the engines under the workbench, and he would install it into Tilley’s VR. He would then take Tilley’s old engine and put it into the factory bike. Tilley agreed and the next day the results were exactly the same – Tilley’s rider was slightly slower, because the factory rider was getting on the gas harder out of the corner and carrying his speed through the straight. From that moment on Tilley, a member of the Harley Dealer Advisory Panel, became a powerful ally of Gemini and a key supporter of the VR program.

Chris Carr. Image courtesy Steve Scheibe.

During the 1995 season Doug Chandler was hired to replace Duhamel, but sat out most of the season due to injury. He was supplemented by Chris Carr, a former dirt tracker who earned the VR its only pole position at Pomona, California.

Windtunnel testing with Chris Carr. Image courtesy Steve Scheibe.

Given the slim possibility of customer bikes being used as sort-of street machine ridden by collectors, durability testing was performed in 1995 by Harley test rider Dusty Smitherman at the Talladega Test Facility in Alabama. Riding at the facility and on the surrounding backroads, Smitherman earned the distinction of likely being one of the most prolific riders of a VR1000, with about 5000 miles under his belt.

Dusty Smitherman. Image courtesy Steve Scheibe.

In 1996 Tom Wilson was the lead rider with Carr as his teammate, and earned the VR its only victory – sort of. Wilson apparently won the Mid-Ohio race, but a red flag was called and the race was put back one lap, making Pascal Picotte aboard a Yoshimura Suzuki the official winner. The VR’s weak clutch became a liability when it led to a loss at Sears Point – while Carr and Wilson were running second and third, a crash prompted a red flag and a restart. Knowing the clutches were weak, Scheibe crossed his fingers and advised the riders to start gently off the grid. They did, and moved up through the pack again… Until there was another crash and another red flag. The clutches didn’t survive the second restart and the duo finished 4th and 5th.

Privateer Ron McGill. McGill rode the VR in the BEARS world championship in 1995, dicing with Andrew Stroud and Stephen Briggs aboard their respective Britten V1000s. Image courtesy Steve Scheibe.

Season after season, race after race, good results were stymied by bad luck. Tom Wilson and Chris Carr rode the 1997 season, with Wilson scoring a second place finish at Mid-Ohio that year. A string of big names rode for Harley over the following years, but victories remained elusive. Mike Smith, number 911, served as the factory’s “emergency” rider from 1998-2000 and rode for Harley full time in 2001, when he lived up to his number when his VR burst into flames after a crash at Road Atlanta. Pascal Picotte rode the VR from 1998 to 2001, with Scott Russell as his teammate in 1999 (taking over from Wilson after he was seriously injured in a crash at Loudon in 1998). Russell managed to set lap times that would have been competitive in World Superbike, but again wins were not in the cards. Russell, five time winner of the Daytona 200, was favoured to win at Daytona again in 1999 but a bar brawl on Saturday night left him with a broken jaw that had to be wired shut. Smith substituted for him in the race on Sunday, where Picotte rode exceedingly well, moving his way through the pack despite his shortage of power on the banking and taking the lead for several laps. Then, with victory tantalizingly close, a bungled pit stop cost him several positions. Bad luck intervened, as per usual.

Pascal Picotte. Image courtesy Steve Scheibe.

Once the bike was homologated, development work was limited by the letter of the rules – the basic chassis and engine crankcases could not be modified. The VR1000 program became one of steady evolution of the original design; while the competition was able to work in 2-3 year model cycles that allowed significant changes to be made to remain competitive, the VR had to stick with the homologated chassis and crankcases whose specs were finalized in 1993. The VR was at a disadvantage from the start, but Gemini slogged through regardless and managed to find power through steady development.

Late VR engine. Image courtesy Steve Scheibe.

Bearings were made smaller and the liberal application of titanium, magnesium and carbon fibre components shaved off significant amounts of weight – the final engines were 45 pounds lighter than the first prototypes. Scheibe exploited helpful contacts in Formula 1 suppliers who became keen participants in the development of the VR’s powerplant. New crankshafts and pistons came from F1 suppliers - a new, lighter Mahle slipper piston netted 6 hp alone. After electric pumps proved unreliable, a 10 BAR F1 mechanical fuel pump was fitted, feeding a dual injector system with shower injectors up top for high-RPM duty and port injectors to preserve low-RPM drivability. Scheibe managed to squeeze a six-speed transmission into a space designed for a five-speed (while leaving room for the counterbalancer), and had a lighter and simpler shift mechanism fitted. Valves grew in size and lost weight through the use of titanium. Titanium Pankl rods replaced the steel Carillo items used in the production VR. The fragile clutch of the early bikes was replaced with an AP Racing carbon slipper item that proved to be virtually indestructible. Because no starter was fitted and the slipper clutch prevented bump or roller starting, a keyway to the end of the crank was added on the left side of the engine  - VRs were started using a production starter motor fitted with a pair of handlebars, a solution that became commonplace in racing later on.      

Meanwhile the politics at Harley-Davidson were beginning to change. The mid to late 1990s saw an explosion in Harley’s production, with demand skyrocketing to the point of creating lengthy waiting lists and premiums over retail. Production exceeded 100,000 units for the first time in 1995, clearing 180,000 by 1999. New bosses began to apply more automotive styles of management and the pressure began to increase on Gemini. The will to race was diminishing as production models became the focus and the racing budget, however modest it might have been compared to the factory efforts from Ducati and the Asian contingent, began to seem like a liability with limited returns. The lack of victories and the long-in-the-tooth VR platform didn’t help the case for continuing the program.  

Late and early crankshafts.

Scheibe resigned from his post as Technical Director in January, 2001. By this point Gemini was an eight man team and had expanded their focus to include in-house carbon-fibre fabrication. Scheibe was succeeded by John Baker, a curious choice given Baker had no race management experience. However Baker had been involved with the VR program from an early stage – he had coordinated parts procurement during the early development of the VR and had experience in engineering and business planning since 1993. By this point the series of constant refinements to the VR engine had netted 175 HP with the rev ceiling raised to 12,300 RPM, while still maintaining a respectable 400 mile overhaul interval – though even with the substantial increase in power, the VR still had trouble keeping up with the competition on long straights, sometimes lagging 10-15 MPH on the top end. Tilley had once infamously noted that Lucifer’s Hammer had posted higher trap speeds at Daytona than the VRs had, a tidbit that no doubt irked the factory. With star riders like Picotte and Russell on board, things looked as promising as they ever had, but the budgets were still modest despite the aims – competing head-to-head with the unstoppable factory teams required far more money and effort than Harley ever appeared willing to spend.

The amount of development that did occur was made all the more remarkable considering the modest budgets compared to the competition, nevermind the impressive longevity of a platform that was already outclassed when it had been introduced seven years prior. The VR earned a reputation as an underdog; word was that rival team members were happy to lend a hand to HD riders in the paddocks, hoping that Harley’s continued presence in AMA Superbike would lend some prestige to the series. With Baker at the helm public statements became more cautious and realistic: they were there to race and to finish as best they could while developing the platform, with little hope voiced for winning outright.

Under Baker’s direction Gemini was moved to a new location near the Buell factory in East Troy, Wisconsin, a spot seen as a better site for development work. Progress was no doubt aided by the proximity of Buell’s talent pool, who along with Cosworth Racing and Ford Racing served as engineering support for the program (though details of their exact contributions were kept secret). Buell had been working on improving the handling of the VR by making some significant revisions – chassis stiffening was performed, the steering head was made steeper to shift more weight over the front wheel, split radiators flanking the frame were installed, and the intake was modified to flow more air. The resulting machine, piloted by Shawn Higbee, managed to post faster significantly quicker lap times than the factory VR.

Mike Eatough was brought back to help and seek feedback from racers and teams on how to make the VR more competitive. A move was made to bring more of the development work back in-house to Harley after years of Gemini operating more or less independently. With Baker under significant pressure to get results, it was proposed that the VR would require complete redevelopment – a new engine and chassis would be required to make it competitive.

Then it was all over. In August 2001 at the end of the AMA Superbike season Harley-Davidson withdrew funding and announced the end of their Superbike program, and the suspension of privateer support for the VR platform. Gemini was purchased by Michael Jordan (yes, that Michael Jordan) and served as the racing arm of Michael Jordan Motorsports, who fielded Suzuki GSXRs in AMA Superbike and Supersport from 2004 until 2013, while continuing their work in the field of carbon-fibre fabrication that had begun within the VR program.

Steve Scheibe riding Picotte's VR at the 2005 Goodwood Festival of Speed. Image courtesy Steve Scheibe.

The lasting legacy of the VR1000 was, much like Harley executives had mentioned to Steve Scheibe upon his hiring in 1991, the application of modern technology to production Harleys. The VR engine was handed off to Porsche for reworking into a reliable, street legal engine. They came back with a design that retained the basic dimensions and 60-degree architecture of the VR, but shared no parts in common: the 1131cc 100x72mm Revolution engine announced in the summer of 2001 was the direct descendant of the VR, and a strange rebirth for the platform just before the death of the racing program. Producing a claimed 115 HP and 65 LB/FT of torque through a five-speed gearbox in its initial guise introduced in 2002, the VRSC V-Rod combined a powerful, modern motor with handsome muscle-bike styling and promptly became the finest Harley-Davidson that nobody bought. It was a curious end to the long and arduous VR1000 program that had fought valiantly against the odds, but it was a result that suited The Motor Company’s aims – as Scheibe discovered in the HD tool room in 1995, production, not racing, was the number one priority.  

Interesting Links
Motorcycle.com review of the VR1000
Motorrad review of the VR1000
Erik Buell's patent for a fuel-in-frame motorcycle chassis
Mike Smith joins the factory team
Beginning of the 2001 Superbike season
Interview with John Baker
Superbike Planet Obituary for the VR
Harley-Davidson's summary of their racing history