products books and linksHow To and DIY guides and articles

Bicycle Technology: Unconventional Bicycle Designs

Guide to Bicycle Technology (article index)

It is not unreasonable to say that the bicycle has essentially remained unchanged since the safety bicycle was introduced in the 1880’s. Although the overwhelming majority of all bicycles on the road are based on this same familiar concept, using the same standard parts arid built to the same set of standard specifications, different designs are introduced from time to time.

Two reactions are common when people discover how old the bicycle’s design is: They either consider it time for some drastic changes, or take it for granted that it must be the optimum solution that simply could not be improved upon. Both reactions are partly incorrect: on the one hand, materials and construction methods, as well as insights in the physics and mechanics of cycling, have not stood still during the last 100 years. On the other hand, the bicycle is so close to the limit of what can be achieved with human power, that at best only marginal improvements can be expected from even the most thorough redesign.

In discussing this subject, it is often claimed that the restrictive rules of the UCI, the international sanctioning organization for bicycle racing, which applied until the fall of 1990, should be blamed for the lack of innovation and development in the industry. I find that hard to believe, just as the rules of Formula 1 motor racing haven’t stopped the manufacturers of road cars from improving them. In fact, most don’t even feel a need to participate in the expensive and irrelevant business of racing. Similarly, it is not likely that any major manufacturer would hesitate introducing a different machine just because it could not be used in sanctioned races. After all, the overwhelming majority of bicycles are neither used nor intended for racing.

17.2. The most unconventional bicycles that keep the rider sitting the way Nature meant him/her to be are those designed by Alex Moulton. This is the AM 2, or City version, with two-speed hub gearing.

17.1. This 1922 Jarray recumbent bicycle not only features an unconventional rider’s position, but also a cable- operated treadle drive.

Although you wouldn’t think so looking at the bicycles on the road today, many different concepts have been introduced at times — both for the bike as a whole and for individual components. That most of these innovations never became universally accepted is not in itself proof of the conventional bike’s superiority. Quite like lv. some were basically sound ideas whose time had not come — or perhaps they just did not get enough publicity to become commercial successes. With recent developments in materials and production techniques, some of the earlier ideas may now be easier to realize. Just don’t expect any miracles.

In the remaining sections of this section, the various real or presumed problems that most designers of alternative bikes and systems seem to address are presented, referring to the solutions that have been offered so far. It should be noted that many of the problems addressed only exist in the eyes of non-cyclists. Others find non-solutions to real problems — such things as bent cranks have been seriously suggested in the past, although they don’t solve the problem of the rider’s limited force which they supposedly set out to correct, and equally absurd designs are sometimes proposed today.

17.3. Different bicycle de signs allow different postures. But most commercial recumbents do not really allow a lower profile than conventional bicycles.

Seating and Riding Position

Leaning back lazily is the ideal in seating comfort in most modern cars, so to those not used to cycling, this seems desirable on bicycles too. Ever since the early 1920’s, recliner and recumbent designs have been proposed and sometimes even marketed in the hopes of making life more comfortable for the cyclist.

17.4. Comparison of frontal area arid drag coefficients on regular bicycle.

Fig. 17.3 compares the posture for a normal bicycle with that for a recumbent. The most frequently quoted reasons for the leaning-back posture are better aerodynamics and the ability to apply more force. One argument is as fallacious as the other. Let’s look at the supposedly increased force first.

True, sitting on the floor with your back against the wall, it is easier to apply the available force of the legs, so that is the way to move a heavy cabinet. The force that can be applied with the legs is not determined by the position but by the presence or absence of a restraint in the back. But on the bike, your output is not limited by how much force can be applied, but by how much power can be delivered. Pushing a heavy cabinet slowly along the floor does not really prove you have more power than you have in another position. In fact, without a restraint, the bicycle’s gearing takes care of this.

Actually, the normal position is the more effective. Here, the legs are hanging down naturally, while isometric work has to be performed in order to keep the legs in the horizontal position dictated by the recumbent pose. Although this problem can be largely overcome with the use of clipless pedals, the fact that the legs have to be forced into a position that is not natural reduces long-term comfort as well as output.

Let’s look at the aerodynamic advantage next. On a racing bike, and in fact even on most other convention al machines, the body can be lowered considerably when battling a headwind, effectively reducing the wind resistance. On recumbents, the rider is stuck with the position forced on him by the bike. On most commercial recumbents, that is not low enough to provide less wind resistance than in the full tuck, or racing, position on a regular bike. It is unfortunate that the obvious potential of these machines has never been exploited commercially, because, even though I personally favor conventional bikes, a little diversity on the road would be welcome.

The other related argument for the recumbent is the possibility to enclose machine and rider so as to reduce air drag even more dramatically. An additional advantage of this enclosure would be weather protection. Of course, there have been enclosures for normal bicycles as well, and in races between enclosed varieties of recumbents and normal bikes, the enclosed conventional bikes just as often run away from the recumbents as the other way round.

17.6. Unconventional alright, if not very aerodynamic. This is a modern version of the Dursley Pedersen bicycle with its unique triangulated frame design. It is ridden here by its builder Jesper Søling.

17.5. Wind resistance as a function of speed and posture.

Officially sponsored international races between aero dynamically enclosed bicycles have been held since 1914 (that’s right: aerodynamics have been around longer than most people realize), when the Dutch rider Piet Dickentman defeated the German Arthur Steibrink in Berlin. They used regular bicycles with torpedo- shaped enclosures. In recent times, both recumbents and regular bikes have been used in such races, and although the recumbent position seems to allow better aerodynamics, the regular shape has the advantages of more convenient dimensions and better long-term com fort.

A serious drawback of the recumbent posture is that the rider cannot unload the seat by shifting his weight when riding over “unevennesses” in the road surface, as can be done on any conventional bike. Consequently, the only way to make a recumbent comfortable is by integrating a rather sophisticated suspension, adding both weight and complications. Sure, it can be done, but then most of the bicycle’s inherent advantages are lost. Just the same, I hope developers keep working on the concept of the recumbent, because having the choice between really different bicycles makes cycling more interesting.


Even though the introduction of the pneumatic tire brought hitherto unknown comfort to cycling, most cyclists agree that a little more suspension would be nice. The real need for suspension is not so much in road riding and for comfort, but rather in off-road cy cling for traction: an effective suspension keeps the wheels on the ground. Actually, the pneumatic tire, perhaps unbeknown to its inventor John Boyd Dunlop, put the suspension in exactly the right place. The theory of suspension, which was largely developed since that time, recognizes that the unsprung mass should be minimized. That’s another way of saying the closer to the road, the more effective the suspension. You can’t get any closer to the road than the tires.

Most other suspension systems ever proposed or realized have been mere compromises compared to the pneumatic tire, which can even be adjusted by regulating the inflation pressure to match the road conditions. Not only do all of those other systems leave some portion of the mass unsprung (namely whatever is between the spring and the road), most of them also affect dimensions that should remain constant for effective cycling. Although the tire’s problem is the very limited depth by which it can deform without negatively affecting rolling resistance on a smooth road, most of the other suspension devices have other disadvantages that are at least as serious.

17.7. The effect of dampening

17.8. Allsop mountain bike with dampened, sprung saddle mount. This is purely a downhill bike, since the suspension is in the wrong place to keep the distance between pedals and seat constant for efficient pedaling.

17.9. Off-road Flexstem. This is probably the lightest method of suspension, but it merely keeps vibrations off the hands — rather than stabilizing the bike.

For the efficient transfer of power to the pedals, the distance between the rider’s seat and the crank axle should remain constant. Any suspension system in the seat, the seat post or the seat tube completely ignores this requirement. Oddly enough, such designs recur on a regular basis. Three other dimensions that should remain constant are the wheelbase, the amount of trail, and the head tube angle. That rules out all but the cleverest suspension systems in the front end. Finally, the suspension must be dampened to be satisfactory. That means that after contracting, it should come to rest in the original position with a minimum number of oscillations, as shown graphically in Fig. 17.7. Off-road, the need for a well dampened suspension becomes most apparent, even if it is bought at the price of more weight. It vastly increases wheel contact in rough terrain, improving both traction and steering accuracy, while still allowing the high tire pressures that keeps rolling resistance down on smooth surfaces.

Fortunately, sometimes people with more insight and fewer illusions tackle the problem, and then interesting designs result. But that clearly involves doing more than adding a spring to a conventional design. The most prominent and successful approach has been that of Alex Moulton, the ingenious British engineering consultant whose small-wheeled bicycle designs have toppled conventional thinking twice in the last thirty years, because he looked at the machine as a whole, drawing the conclusions that were necessary to make his suspension work.

Moulton’s bicycles, the current design of which is shown in Fig. 17.2, have small wheels to provide room to take the displacement of the suspension and to minimize the unsprung mass. The 17” wheels have only about 65% of the mass of regular size wheels. The front suspension is designed in such a way that it does not noticeably affect the bicycle’s trail, head angle and wheel base. The rear suspension is arranged around the bottom bracket in such a way that it does not change the distance between seat and crank spindle, nor does it affect the wheel base. Besides, both suspensions are adequately dampened to avoid the see-saw effect of many other sprung components.

Drivetrain Alternatives

The bicycle’s inventor Von Drais propelled his machine by intermittently pushing off with the legs on the road. That turned the intermittent motion of walking into the continuous one of rolling, which represented a mile stone in transport efficiency. It took until about 1860 before pedal-drive bicycles gained acceptance and this was another step forward due to the continuity maintained in propulsion. Another twenty years went by before the chain-driven bicycle was introduced, which has remained the standard for well over a hundred years now. Not surprisingly, some inventors feel it is time for a change.

17.10. Toothed belt drive. Although it is a lot cleaner than the familiar chain, it precludes the use of derailleur gearing.

Although the chain transmission has a high efficiency (about 95% in clean, well maintained and lubricated condition), it is by no means without drawbacks. In the first place, its efficiency suffers dramatically from neglect, soon reaching a mere 80% if allowed to run dry, dirty and rusty. The other disadvantage is the fact that it is an inherently messy system that cannot easily be protected, nor can the rider’s clothes be protected against this moving, greasy, dirty component, certainly not on a bicycle with derailleur gearing. On the other hand, the alternative transmission systems don’t allow the use of derailleur gearing.

The various alternative concepts for bicycle transmission that have been proposed and sometimes actually used in the past fall into three categories: belt drive, shaft drive and treadle drive. All of them are old hat. Shaft-driven, so called acatane, or chainless, bicycles have been introduced at intervals ever since the 1890’s. In fact, the famous Black American bicycle racer Major Taylor set some of his early time trial records on such machines.

17.11. Crankset with an internal mechanism that increases the gearing during the downstroke. This one dates back to 1924. The same idea is reintroduced under different names from time to time, most recently as the Bio-Cam drive.

17.12. Allenax drive system. Instead of a circular leg motion, it relies on a reciprocating one — without any notice able benefit to the rider.

Obviously, this system can work — until you realize that the chain has been improved more than the shaft and bevel gearing systems used for the chainless bicycles since those days, not to mention price and weight: the chain drive method is incomparably cheaper and lighter. Most recently, the German manufacturer Fendt introduced a shaft-driven bicycle in the early 1980’s, but this model had relatively crude bevel gears that are noisier and less efficient than the ones built 100 years ago.

Belt-drives have certain advantages, especially if one considers the recent perfection of the toothed belt de signs, using high-tech materials such as aramid (Kevlar) and synthetic rubber that make them both light, flexible and unstretching. The Japanese Bridgestone company, though now fully independent from the company by the same name that makes such belts, successfully uses belt drives on some of their folding bicycles. Their replacement requires the frame to be separated (except on frames with raised chainstays) and they cannot be used with derailleur gearing. Just the same, for the utility bike without gearing, belt drives may well provide a workable alternative.

The third method of transmission is by means of treadles that rock back and forth. In fact, this design was the major characteristic of the bicycle by Gavin Dalzell, probably built around 1860 (based on purely speculative inference, this design is often claimed to be the work of Kirkpatrick Macmillan and to date back even further). More recent attempts to make this kind of mechanism work include Jarray’s revolutionary recumbent introduced in 1921, and the Allenax introduced in the 1980’s. Jarray used cables instead of rods, arid the Allenax drive uses springs and chain sections, with derailleur gearing. It appears at one bicycle trade exposition after the other without ever having caught on. Perhaps it just doesn’t work as well as it is claimed to.

17.13. Variable crank length drive systems in crease the leverage on the down stroke.

Long Cranks and Other Tricks

One concept that never seems to die, despite all proof of its inadequacy, is the use of cranks that are either longer or vary in length depending on their orientation in the pedaling cycle. The underlying assumption is that long cranks offer more leverage, allowing more power to be done with the same force, or the same power with less force. How silly: the bicycle’s transmission ratio can take care of that and the rider’s output can’t be increased over what his physical abilities determine.

Sure, long cranks are probably more effective for riders with long upper legs, but so are short cranks for those whose legs are shorter. The 170—175mm crank length used on most bikes is an excellent choice for the vast majority of riders, and the option to replace them is there for those who feel they need it. Ideally, the bicycle should be designed for the crank length actually used: the bottom bracket should be slightly higher if very long cranks are installed to keep the pedals from hit ting the ground.

17.14. A recent attempt to revolutionize the drivetrain with variable length cranks. This is the Canadian-German STS Power-Pedal.

More subtly wrong is the concept of variable length cranks shown in Fig. 17.13, which is the underlying principle of a number of special drive units introduced from time to time. Here the crank effectively gets longer on the down stroke, increasing the leverage, to shorten again on the rest of the pedaling cycle. It’s all in vain:

the supposed advantage of the longer leverage on the downstroke is not real, because the total distance between top arid bottom position of the pedals stays the same — and on designs where it does not, anything said above for longer cranks applies.

Oval and other non-round chainrings are all based on the same erroneous kind of reasoning as the concept of longer or variable length cranks. In the 80’s, Shimano spent a fortune on (successfully) advertising its egg- shaped BioPace chainrings to convince the public there would be an advantage to this shape due to the distribution of forces during the pedal cycle. They put up a smoke screen of highly scientific looking graphs and formulas — and other companies followed suit with their own non-round designs the public started demanding.

It took Shimano five years to come out with a new, more efficient form of chainring that was nearly round, finally followed by the ultimate High Power BioPace, which turned out to be perfectly round. But I suppose millions will be fooled again the next time a big company tries to sell them such a bill of goods.

Other developments based on the same fallacy as the variable length crank and the oval chainring include cam-assisted devices that reduce the length of the power stroke and increase that of the recovery phase in each crank revolution. This is based on the physiological fact that short powerful muscle contractions, followed by long recovery phases, allow a higher output to be maintained. Other manufacturers, who have heard something about the concept but don’t even under stand the theory behind it, have done similar things to achieve the opposite effect. Sure, if you believe in the benefits, the placebo effect will be enough to give you the feeling of greater efficiency, but self-hypnosis would provide the same benefit.

One concept that is perhaps more promising than any of the tricks covered so far is pedaling backward. After all why should we have to pedal in such a way that the leg is extended forward and withdrawn in an almost vertical position. Doing it the other way round is not particularly difficult, and some tests indicate that it is perhaps marginally more efficient. It remains a mystery that this simple trick, which any handyman can teach his own bike by running the chain differently, has not been pursued more extensively.

17.15. Most infinitely variable transmissions for bicycles rely on this principle.

17.16. Deal Drive automatic transmission.

Automatic and Infinitely Variable Gears

This is another dream of most non-cyclists: gears that adjust themselves automatically to the conditions, or at least allow easy selection without having to shift in distinct steps. Actually, since the introduction of indexed gearing, which takes most of the guesswork out of gear shifting, especially in conjunction with Shimano’s Hyperglide tooth design, which allows shifting under load, this demand should become less prevalent. Even so, at tempts to provide an automatic transmission still continue to be pursued.

The wonderful thing about modem-day index gear is that it works with otherwise normal components: you need neither a different frame, nor a different crankset, nor anything else beyond a special shifter and matching cable. And it seems even the novice can handle most systems without problems, which is not always the case with the more complex automatic or infinitely variable gearing systems that have been introduced from time to time.

All such systems require major modifications to the bike, meaning that they can’t succeed unless several bicycle manufacturers agree to design their production models around them. The problem with most automatic systems is that the human factor is overlooked: the cyclist feels differently one day than another, even from one minute to the other. While an engine is fully predictable and can work well with an automatic trans mission, the cyclist may want to go uphill in one gear today, in another tomorrow — yet the automatic system would force him to ride in the same gear each time.

The most recent, and probably most perfected, automatic transmission introduced so far was the model known as Deal Drive, invented in the 1980’s by a French engineer working in Britain (see Fig. 17.16). Like all other such devices, it was rather heavy, but it worked like a charm and its efficiency was about as high as that of normal gearing systems.

Most infinitely-variable gearing systems are based on a system by which the individual chainrings in the front are replaced by concentrically arranged separate smaller toothed wheels that can be moved equally far in or out, as shown in Fig. 17.15. In the retracted position, they form what is effectively a small chainring; when extended out, they provide the high gearing achieved with a large chainring. Excel’s patented Cambio Gear of the mid eighties and the Tokheim design of the early seventies were both based on this principle, and neither of them ever made it commercially.

17.17. Raleigh’s 1974 edition of the original Moulton design of 1962— with small wheels and suspension both front and rear.

Alternate Frame Designs and Materials

Bicycle frames don’t have to be made with steel tubing joined together in the familiar shape. Although it is hard to understand what is wrong with the convention al frame, I can sympathize with those who want to travel different paths. The technology of materials and joining methods in industry has changed over the years, offering more options, and recently there has been a spate of revolutionary designs using high-tech materials and construction methods.

The idea of a plastic bike has been around for a long time, early versions dating back to the fifties. A commercially produced plastic bike was introduced in the 80’s under the name Itera. Like most of the earlier ones, it was not a success, probably because it did not address the whole concept of the bicycle, but merely re placed the frame with something that looked more like a motor scooter than a bicycle. Although this provided slightly better protection to the rider’s clothing, the use of plastic did not really solve any problem that was inherent in the conventional design.

The argument that plastic does not rust is rather insignificant, because the frame is about the last part of a bike that will rust — long after rims, spokes, hubs and cranks, chains, chainrings and handlebars have be come corroded. Add to that the fact that the plastic frame reverberates, amplifying any sounds generated as the bike bounces along the road, as well as the in creased weight combined with reduced rigidity, and you have a concept that is sure to fail, as the Swedish state, which had financed the whole idea, learned the hard way.

Recent years have brought both aluminum and titanium tubing, followed by carbon fiber reinforced composites and magnesium. The latter materials are no longer restricted to the conventional tubular shapes joined together by conventional methods. One-piece frame designs using these materials have not only been shown, they are now commercially available. Some of them may have certain virtues for certain applications, but none have so far hinted at being competitive with the conventional frame construction method on all counts.

An entirely different way is still open: the use of thinner tubing in a fully triangulated design. This method can achieve a lighter and very rigid structure if done right. Take a look at many modern structures such as tall building site cranes, and you get an idea of the concept. Moulton’s present generation of small-wheeled bicycles is built this way: expensive, but beautiful and sound from an engineering standpoint.

17 18. Comparison between dimensions of regular bike and Moulton’s small- wheeled machine.

17.19. Probably the best folding bike available: This Montague bicycle is based on conventional bicycle geometry and hinges around the seat tube.

The first bike built with thin tubes was the Dursley Pederson, designed around the turn of the century by a Danish engineer working in Britain. This design is characterized by more than the thin tubes: it also uses cables instead of rigid members to take up the tension forces that result in certain members due to the fully triangulated design. The problem of this design was its inadequate lateral stability, largely due to the fact that the cables were only tensioned by the rider’s weight.

A more recent and more satisfactory design on the same basis is one introduced by the Dutch industrial designer Frans de la Haye in 1979. This design, recently adapted in essence by California mountain bike frame builder Joe Breeze, uses a cross-shaped frame held under tension by tie cables. Unlike Breezes design with its continuous cables, the original was fully collapsible and was perfectly rigid under all kinds of static and dynamic loads.

Folding and Parting Bicycles

Bicycles are often too big to transport easily, as anyone arriving at an airport with his machine will soon find out. Most US airlines now punish the cyclist with such exorbitant excess luggage charges that it may seem cheaper to buy another bike at the destination. This reasoning led to the development of folding bicycles.

Folding bicycle designs have been around a long time: I have found patents dating back to the last century. The first practical folders were developed during the first twenty years of the twentieth century with military use in mind. Not until the 1950’s did the idea surface again seriously. This time, the manufacturers used smaller wheels, the prototype of most of these designs being the Japanese Star bicycle.

The first generation of Moulton bicycles picked up on the idea of the small wheels but distinguished itself from the others by including all the details that were needed to turn this into a fully satisfactory bicycle. It was unfortunate that this concept was eventually killed by its own popularity: Moulton sold out to Raleigh who stripped it bare of its virtues and retained only the name and the small wheels. During most of the sixties, the small-wheel bicycle, despite its discomfort, dominated a large part of the European market — setting bicycle technology back by at least as many years.

More recently, interesting compact designs have been introduced. Not only did Moulton return with a second generation, as was mentioned earlier in this section, but a number of other manufacturers also came up with more or less satisfactory designs. On the one hand, most of these are easier to disassemble or fold than the Moulton, which does require rather extensive nuts-and-bolts work to separate or join the two parts. On the other hand, they are compromises, few of which are really rigid, light arid adjustable enough for serious use, especially by tall cyclists. The very light Bickerton and the extremely compact DaHon do deserve special mention under the compact machines, while the Montague should not be overlooked because it is essentially a full size bicycle when it comes to riding, even though it does not fold down to quite such a small package.

17.20. Bickerton folder with lightweight aluminum frame arid tiny wheels. Collapsed, it is small and light enough to be carried as luggage on just about any bus, train or plane.

17.21. Simple coasting test set-up. Use this method to evaluate the relative efficiency of bikes and components.

Home Testing

The relative virtues of different bicycle or component designs can often be tested objectively by the interested cyclist. This has the advantage that it is possible to test for those properties that seem personally relevant. In this section a few simple testing techniques will be suggested.

Many properties can be established by a simple rolling test, using a set-up as illustrated in Fig. 17.21. Make sure the bike starts with the rider’s center of gravity at the same point each time and let it roll down a known ramp over a standard surface. The bike that rolls over the longest distance has the best rolling properties. Even more accurate would be a similar test using an accelerometer, a $400 device that measures how much the bike is retarded.

Another interesting method allows testing the drive- train components, but also such factors as riding position, seat height and crank length for any particular rider. To do this, use a heart rate monitor and a cyclo-simulator, which is a turbo trainer equipped with a wattmeter to measure the output. Realizing that the cyclist’s pulse is a pretty reliable indicator of relative effort, the combination of this device with a pulse rate monitor gives you the closest thing possible to a work physiology laboratory on a reasonable budget. You can compare the pulse rate that results from reaching the same output under different conditions — different bikes, different components, different gears, or the same bike adjusted differently. Whatever combination requires the lowest heart rate is the most efficient.

17.22. Weird and wonderful. Designer’s answers, such as this Strida bicycle from England, are rarely efficient — or even comfortable.

Guide to Bicycle Technology (article index)
Top of Page Prev: Tandems Next: The Bicycle of the Future HOME