The bicycle, a simple, reliable form of human powered transportation, has evolved over many years. During the glory days of cycling, almost a century ago, many variations, both aesthetic and mechanical, were produced. Inventors also attempted to power the bicycles with steam, compressed air, electricity, and internal combustion engines. Most of these early attempts met with little success. The internal combustion engine proved the exception, eventually emerging in the form of the larger motorcycle. Presently, there are many styles of cycles pedaled through the streets of the world, but little attention is given to self-propelled bicycles.


Depending on how one defines "bicycle", it may have originated hundreds, or even thousands, of years ago. The wheel itself is commonly thought to have been invented at least five thousand years ago. A relief adorning the Place de la Concorde in Paris, dating back to 2700 BC, shows a figure with a two-wheeled, bicycle-like machine. Another primitive human-powered vehicle, but with only one wheel, is depicted in an Etruscan illustration from approximately 1300 BC. [Van der Plas, 1983] Unfortunately, such early records are scarce and provide very few details.

Beginning in the fifteenth century A.D., there is much more evidence of human- powered vehicles, with various number of wheels, being simultaneously developed in several countries, including England, France, and Germany. By 1791, the Comte de Sivrac was reportedly seen riding a "c‚l‚rifŠre" through the gardens of the Palais Royal in Paris. This heavy "wooden horse", subsequently called a "v‚locifŠre", consisted of two wheels connected to a rigid frame. It was propelled by pushing the feet directly against the ground and steered by lifting or dragging the front wheel to either side. For a brief time, this was seen as a great sport, with several clubs forming and running races [Oliver and Berkebile, 1974].

In 1816, Nic‚phore Niepce revived interest by introducing an improved v‚locifŠre. His model was considerably lighter and had larger wheels to provide a smoother ride [Oliver and Berkebile, 1974]. Shortly thereafter, Baron von Drais, of Sauerbrun, Germany, made the revolutionary advance of steering the front wheel. Von Drais patented his wood and iron machine, known as a "draisenne", and it quickly gained popularity. In 1818, he brought a version to Paris and patented it as a "v‚locipŠde". The user, as seen in figure 2.1, could expect to go "8 to 10 miles per hour" [Sharp, 1896, p. 147]. The invention spread throughout France and over to England, where similar vehicles were produced as "hobby horses", "dandy horses", "swiftwalkers", and "pedestrian curricles".

On June 26, 1819, W. K. Clarkson Jr., of New York, was granted a US patent, possibly the first for a velocipede. Unfortunately, the patent record was destroyed by fire, so the actual design is not known. According to Oliver and Berkebile [1974], there is little evidence of early American interest in velocipedes.

However, there were more developments elsewhere, such as the hand-operated velocipede, designed by Lewis Gompertz of Surrey, England in 1821. As shown in figure 2.2, the steering lever was connected through a ratchet system to power the front wheel. Unfortunately, such improvements received little attention during the next two decades.

The first "bicycle", defined in modern dictionaries as having two wheels, a seat, steering, and pedals, was produced by Kirkpatrick MacMillan in 1839. A skilled Scottish blacksmith, he used levers to connect the pedals to the rear wheel in a reciprocating treadle fashion. The vehicle, depicted in figure 2.3, was made from wood and iron, had a 30 inch front wheel with handlebars for steering, and a 40 inch rear wheel, sized for mechanical advantage. Although innovative, the treadle style never caught on, possibly being overshadowed by the railroads developing at the time.

The lull in activity continued until 1862 or 1863, when pedals were directly connected, via crankshafts, to the front wheel of a velocipede. Most accounts credit Pierre Michaux of Paris with this invention of what some consider to be the first true "bicycle". According to these accounts, Pierre Lallement was briefly employed by Michaux before coming to the United States in 1865. In 1866, he received US patent 59,915 for the velocipede shown in figure 2.4, with financial backing from James Carroll of Connecticut. The patent was sold several times, with the Pope Manufacturing Company of Boston being the final owner before it expired.

However, some feel that Lallement was the true inventor. In particular, the Lallement Memorial Committee, based in Boston, has gathered significant evidence which paints a much different picture. According to an article in The Wheelman Illustrated [Pratt, 1883] and other documents of the time, Lallement conceived of the pedal arrangement on his own while in Paris. After demonstrating his idea in France, he came to New York, and later Connecticut, where he improved his invention and received the patent with the support of Carroll. Lallement returned to Paris to find Michaux's company, along with others, producing "v‚locipŠdes a p‚dales" in increasing numbers, as the bicycle rage began.


Once interest was sparked in the United States, it spread like wildfire, prompting many innovations. In 1868, the Hanlon brothers, from New York, were granted U.S. patent 79,654, which included the use of rubber "rings" on the wheels for a smoother ride. By 1869, over 1 in 75 U.S. patents granted related to velocipedes.

During this time period, the first self-propelled bicycles were made. Using a custom two-cylinder, charcoal fired steam engine, Sylvester H. Roper, of Roxbury, Massachusetts, produced the vehicle shown in figure 2.5. At about the same time in France, Michaux added a Perreaux steam engine to one of his velocipedes [Wilson, 1973, p. 83]. Shortly thereafter, Louis G. Perreaux secured a U.S. patent covering a flywheel-aided velocipede. His preferred version, seen in figure 2.6, was powered by a small steam engine, fueled by oil stored in the hollow tubing of the frame. For those wishing to use human power, Perreaux designed a spring-loaded version, to be used with a treadle system. Although he noted that the flywheel could assist in climbing hills, his greatest concern was apparently to provide an extremely uniform riding speed.

Meanwhile, in England, James Starley developed light-weight, radial spoke wheels, and in turn produced the "Penny-Farthing". He and other designers had quickly realized that enlarging the front wheel was a simple way to achieve a higher gear ratio. In 1874 James Starley again made an important advance, this time with the introduction of tangential spoke wheels. Complete with eyed and threaded nipples for adjusting tension, this is the same system still in use today. "Ordinaries" or "high-wheelers", such as in figure 2.7, made use of these improved, light-weight wheels and quickly gained popularity in many countries. The front wheel could be over 80 inches in diameter, depending on the leg length of the rider. To date, bicycle gear ratios are still often expressed in terms of the equivalent front wheel size.

Due to the position of the rider directly over the front wheel, "headers" were all too common, causing many would-be users to turn instead to tricycles. In 1879, Harry J. Lawson introduced a rear-wheel drive "bicyclette". The seat and pedals were positioned midway between equal-sized wheels, and a single, endless chain connected the crankshaft to the rear wheel. This first "safety" did not receive much attention, and many other designs were tried. These versions often still powered the front wheel and required a cumbersome arrangement of multiple chains or gears and split crankshafts.

The "Rover" safety, created in 1885 by William Sutton and John Kemp Starley (nephew of James Starley), was the first rear-drive to be popular. The original version, as seen in figure 2.8, used a chain drive mechanism similar to Lawson's. The third revision, available one year later, added direct steering, and the beginnings of a diamond-shaped frame. In 1890, Humber & Co. produced a safety with a longer wheel base and a true diamond frame. As can be seen from figure 2.9, this design was extremely similar in both form and function to most bicycles in use today.

Also in 1890, pneumatic tires, created by the English inventor John Dunlap in 1888, began rapidly replacing solid tires. Coupled with the design of the safety, a more comfortable and faster ride resulted, creating the first generation of "modern" bicycles.

Utilizing the now-accepted diamond frame, more methods for powering bicycles were explored. Figure 2.10 shows a compressed air motor as applied to a bicycle. Designed by Theodore Cummins in 1893, a tank of air, charged to 200 psi, powered two slow moving pistons. By using a long rack and multiple gears, Cummins claimed to achieve a gear ratio so high that each stroke could propel the bicycle a half mile or more on flat ground. He also apparently found that enough power was available to climb hills, as no pedals were provided for human assistance.

Two years later, Nelson Hopkins received what may have been the first patent for a "moped", although the term was not then used. He added a small, two-cylinder internal combustion engine to a standard safety, leaving the pedals on for start-up and hill climbing. Figure 2.11 shows the pistons on either side of the rear wheel, and the spark coil, batteries, and gas tank mounted to various parts of the frame. Mopeds are still in use today, but have been overshadowed by the more popular motorcycle.

By the late 1800's, human powered versions, with diamond- shaped frames, had achieved great commercial success. There were also experiments with several primitive recumbent designs prior to 1895 [Ballantine, 1987, p. 177, Whitt and Wilson, 1982, p. 23]. The Brown Recumbent, as illustrated in figure 2.12, was invented in America and brought to Britain at the turn of the century. It appears to be surprisingly similar to many of today's semi- recumbent, "ergonomically-correct" bicycles.

By the 1890's, many improvements in the drive train had been made. New hub designs, replacing a directly attached gear, allowed for "free-wheeling". Combined with ball bearings, already in use for several decades, the rider could now easily coast long distances. Further additions to the hub resulted in weather-proof coaster brakes, which allow the rider to slow down by applying reverse pressure on the pedals.

As an alternative to chain drives, which require frequent cleaning, shaft drives were often experimented with, and even marketed by many American companies in the late 1890's. These were not successful, due to additional cost, greater weight, and a loss in efficiency (up to 8% [Whitt and Wilson, 1982, p. 298]). However, the inefficiency may have resulted from the poor quality of bevel gears available at the time.

By 1902, H. Sturmey and J. Archer had developed their classic 3-speed internal transmission hub, using epicyclic, or sun and planet, gears. Derailleur systems, where the chain is simply moved from one size gear to another, were in commercial use by the 1920's. Most bicycles sold today rely on a derailleur, which can provide up to 21 speeds, but Sturmey-Archer hubs are also still used.


In 1895, Ogden Bolton Jr. received one of the first patents for an electric bicycle. His novel idea, shown in figure 2.13, involved the use of a custom rear wheel, with the motor as an integral part of its hub. The patent primarily covers construction details, but does note that the motor would be wound to use high currents at a low voltage. Bolton's example of 100 amps at 10 volts would have resulted in over 1 HP of output power, assuming a relatively efficient motor. This, as will be seen later, is adequate for most bicycle travel. However, since no pedals were provided for assisting with start-up and hill climbing, and since there was no reduction gearing, the motor must have been able to produce a very high torque. To control the motor, Bolton simply connected a single battery via a rheostat. By considering the types of batteries then available, and also considering the amount of power dissipated by the rheostat, it is safe to conclude that the driving range before recharging was rather short. It is interesting to note that 100 years and 5 million patents later, electric car conferences still include displays such as the "Electric WheelTM" [Edwards, 1992, Riezenman, 1992].

In 1897, Hosea Libbey of Boston received three consecutive patents for electric bicycles. The first, shown in figure 2.14, used a motor in place of the pedals to drive a double rear wheel via cranks and connecting rods. The other versions used two motors, either in place of the pedals, or mounted over the rear wheel and connected by standard chain. All three had a double wheel either on the front or rear, probably to help stabilize the extra weight. By using multiple motors, he hoped to provide a back-up in case one motor failed. In all versions, Libbey used a split battery, so as to offer a two-speed control system.

A curious drive mechanism was patented in 1898 by Matthew Steffens. As can be seen in figure 2.15, a motor under the seat drove a pulley which in turn was connected to the rear wheel via a cupped belt that actually went around the entire outside of the tire. In addition to reducing the possibility of slipping, this setup provided a very high gear ratio, allowing a small, high speed motor to be used.

In 1899 John Schnepf received a patent for a motor unit, shown in figure 2.16, designed to be added to any standard bicycle. Like many others, he positioned the motor over the rear wheel and used a simple friction drive, adding a mechanism to allow the rider to raise the unit out of contact if desired. To control the motor, a simple on-off switch was provided. More importantly, Schnepf noted that if the unit was left turned on and in physical contact with the wheel while on a decline, the motor would act as a generator and recharge the battery.

In another patent issued just fourteen months later, Albert H„nsel of Germany took this concept one step further. He realized that the battery could be recharged anytime the rider desired to slow down by "employing the electric motor as a brake" [H„nsel, 1900]. This was possibly the first American patent to specifically recognize the concept of regenerative braking for an electric bicycle. Regardless of the precise origins, it is clear that not only were electric bicycles being built by the turn of the century, but also that the importance of efficiency and regenerative braking had been recognized.


Unlike the original electric bicycles, most patents issued in this century centered on only assisting the rider. In turn, some inventors have found ways, other than using electric motors, to help the rider sustain a relatively constant speed. That is, methods were developed to average out the power required to propel the bicycle.

One intriguing variation, patented by Vincent Murray in 1975, used two large solenoids to provide auxiliary power to a bicycle. One solenoid, or "linear motor", was placed on either side of the bicycle and connected to the pedal crankshaft via long levers, as shown in figure 2.17. The rider would use simple on-off switches to activate the solenoids when assistance was desired, although correctly timing the sequence undoubtedly required great practice. Since the solenoids could not easily be used to produce electric power, Murray choose to add a separate generator. Unfortunately, this use of additional parts for regenerative braking increases weight and decreases efficiency.

In an earlier patent [Pecci, 1972], a generalized motor using multiple solenoids had already been presented. Although a distributor system was provided to eliminate timing control problems, the inability to be easily used as a generator was inherent.

Over one hundred years after Perreaux invented his flywheel-aided, steam driven velocipede (see figure 2.6), Eugene Large patented a human- powered, flywheel-aided bicycle, as shown in figure 2.18. The flywheel was geared to revolve over one thousand times faster than the rear wheel to which it was coupled. A slip clutch was included for the selective engagement of the flywheel.

Another mechanical regener- ative braking system was patented by Mark Brent and Jim Papadopoulos in 1988. In their design, a long elastic cord is used to store energy. To brake, a cable is coiled up, stretching the cord. The rider then controls a clutch to release the energy as desired.

Engineering reviews [as cited by Whitt and Wilson, 1982], have shown that given the weight and size limitations of a bicycle, such mechanical storage systems are not practical. Furthermore, even a modest motor/generator and battery set would double or triple the weight of a typical bicycle. Whitt and Wilson concluded that a rider interested in energy-storage bicycles might be best off with a "battery-powered motorcycle".


Although the focus of the remaining electric bicycle developers had largely shifted to electrically assisted units, general interest has been renewed over the past twenty- five years. In 1969, with the U. S. Environmental Protection Agency being formed and the first "Earth Day" on the horizon, Garfield Wood, Jr. received a patent for an electric bicycle. In this he notes that: "with the current emphasis on the control of air pollution and on the use of motors which do not release toxic vapors, there is a need for a self-propelled bicycle which does not release toxic exhaust fumes." Of technical interest, Wood expanded on Libbey's 1897 application of dual motors by using multiple subfractional horsepower motors. However, he still used a simple rheostat for speed control. As seen in figure 2.19, the front wheel was connected via a friction drive to four low-cost permanent magnet motors "of the type used in many toys".

Throughout the 1970's and 1980's, various patents were granted for bicycles that used electric power to assist the rider. Although a few designers did experiment with frequency and duty cycle control of motors on these bicycles, none were marketed.

A 1981 market review of "Motorized Bikes and Mopeds" by Consumers Union found that a limited number of electric bicycles were commercially available. Several add-on units, with list prices ranging from 170 to 285 dollars, were tested, along with one complete electric bicycle, with a list price of 410 dollars. All had top speeds of under 15 mph and cruising ranges of 7 to 10 miles. Also, no innovative technologies were demonstrated, as all used small motors, ranging from 1/3 to 3/4 HP, connected to a lead- acid battery via an on-off switch. However, it was noted that the rider could leave the power turned on almost continuously since the severe lack of power made going too fast basically impossible. The lack of power was also extremely evident on even slight inclines, where "energetic pedaling" was required to keep the bicycles moving.

One of the latest patents, from 1989, takes the additional step of including solar cells to recharge the batteries. A standard bicycle is retro- fitted with a small motor, mounted transversely above the pedals, and the control unit shown in figure 2.20 attached to the cross bar of the frame. The solar cells are mounted on small panels which fold down, so as to not interfere with the riders legs while the bicycle is in use. The inventors claim that the solar cells can recharge the battery in six hours, and the battery will then last for two hours of use " combination with a manual pedal driving system". The small size of the unit, which includes the batteries and weighs less than seven pounds, and the use of a simple rheostat for speed control, leads to questions of how much power is actually provided by the unit.

In both the 1990 and 1991 American Tour de Sol races, motorized mountain bicycles were entered by Paul Butler of Team Rosebud from Vermont. The 1990 model was based on a DC motor, and successfully completed the race, but did not employ regenerative braking. As can be seen in figure 2.21, the motor was mounted in the middle of the frame, and coupled with standard bicycle chain to an oversize gear mounted on the bottom bracket along with the original pedals. Thus, the motor was connected to the rear wheel via a transmission consisting of the original derailleur mechanism. "Rosebud 2", Butler's 1991 version, used a Delco AC brushless 600 Watt, 28 volt truck alternator, mounted over the rear wheel and connected via a friction drive to provide both propulsion and regenerative braking. However, due to controller problems the vehicle completed the race only under human power. Both years, standard deep-cycle lead-acid batteries were used, with limited additional power produced by a solar panel worn on the rider's back.

Unfortunately unable to make the race, UFX Enterprises of Pennsylvania had planned to enter a solar-electric moped in the 1991 Tour de Sol. Based on a modified moped frame, this vehicle weighs 175 pounds, and includes a 120 watt solar array, as seen in figure 2.22. With a total load capacity of 250 pounds, a top speed of 21 mph, and a range of 25 miles was claimed. A 1250 watt-hour battery bank powers a Hoover Electric D-1000 DC compound motor, capable of delivering 950 watts continuously, with peaks of 1400 watts. A commercial MOSFET controller was chosen, which provides regenerative braking in addition to mechanical drum brakes. Although this vehicle has substantially more power than Butler's motorized mountain bicycles, it still has pedals available for additional rider supplied power.

According to John Dunlap, the South Coast Air Quality Management District in Del Monte, California received two electric mopeds in October of 1991. These production units are from "a company in Taiwan", and will be used for on-site deliveries once properly registered and insured. Additionally, an international student at Tufts University reports that electric bicycles, with top speeds under 15 mph, are already commonplace in China. These models apparently cost three times as much as an ordinary bicycle, for a total equivalent to an average person's monthly salary. Unfortunately, further data, such as the ratio of electric power to human power required, is not known for these units.

Along with an early version of the electric bicycle presented in this thesis, the 1992 American Tour de Sol entries included a small-size motorcycle converted by the Central Connecticut Solar Electric Racing Team. Shown in figure 2.23, the 400 pound "EnviroCycle I" uses lead-acid batteries to power a 4 HP Advanced DC series motor with a Curtis pulse width modulation controller. The three batteries, weighing roughly 150 pounds, provided 900 watt-hours for a total range of 50 city miles, at a 35 mph efficiency of 126 Wh/mile [S/EV 92 Proceedings].

A full size motorcycle was demonstrated after the finish of the 1992 race by the Vehicle Development Group of New Hampshire. Able to reach 85 mph, this is an unusually large 2-wheeler, and as is obvious from figure 2.24, it has superb farings to help reduce aerodynamic drag. Although it is able to maintain highway speeds, it is also cumbersome, with a total weight of 750 pounds. This is due in large part to the 8500 watt-hour battery bank needed to power the 11 HP Solectria motor and matching controller [S/EV 92 Proceedings].

In 1993, the Envirocycle I was again entered in the American Tour de Sol along with the "Envirocycle II", a slightly bigger, 3-wheeled version with similar performance. In addition, a recumbent 3-wheeler was entered by Electrifying Times of Oregon. As shown in figure 2.25, it uses standard bicycle wheels mounted in a custom steel tube frame. With essentially the same motor and controller of the Envirocycle I, a range of up to 100 miles is available, due, in part, to the use of developmental batteries from BAT international. However, at 250 pounds, the reduction in weight could account for much of the additional range, diminishing the credibility of the claimed 6000 Wh battery capacity [Northeast Sustainable Energy Association 1993 American Tour de Sol guide book].

Until recently, the only commercially available, completely self-propelled, small- scale electric vehicles were those marketed to the elderly and handicapped, such as electric wheelchairs and other personal mobility products. For example, Palmer Industries of New York produces the two passenger electric tricycle shown in figure 2.26. This has a top speed of 14 mph, and a range of 18 miles, or up to 50 miles with additional batteries. Not including the batteries, the vehicle weighs 155 pounds, and costs 2950 dollars. A permanent magnet motor with gear reduction is used, which provides dynamic braking in addition to mechanical disc brakes.

In 1993, Unique Mobility Inc. of Colorado signed a memorandum of understanding with the Kwang Yang Motor Company of Taiwan. This agreement, strongly encouraged by the Taiwanese Government, calls for Kwang Yang to produce electric motor scooters using Unique's high efficiency permanent magnet motor technology. The 180 pound vehicles are expected to run 35 or 40 miles between recharges in urban traffic at speeds up to 35 mph. Presently, Taiwan has 8 million gas-powered scooters producing a vast amount of pollution.

Also in 1993, Real Goods, a California mail-order company, began offering the "EcoScoot" electric scooter seen in figure 2.27 for a little over 2000 dollars. It is advertized as having a 24 volt motor "with an electronic speed controller". Powered by two 12 volt, 30 amp-hour lead-acid batteries, this scooter is supposed to have a range of over 20 miles and travel up to 30 mph. Although sold as a "motorized bicycle", Massachusetts law would require that it be registered as a motorcycle as it can surpass 25 mph and has no pedals. This appears to be the first commercial 2-wheeled electric vehicle that provides electronic control of the motor.

As the original bicycle craze began over a century ago, inventors were quick to try to completely replace human power with electric power. However, technology then available forced makers to settle for electrically assisted bicycles. Although the technology is now available to realize the original goal of complete self-propulsion, few American companies have commercialized. It is important to realize that electric bicycles, and electric vehicles in general, are not the technology of the future, but are already feasible, using available, time-tested technology.


Go onto Chapter 3: Motor Selection

Return to Table of Contents


Design of an Efficient and Economical Electric Bicycle /