Direct-Drive Recumbent Bicycle

Direct-drive recumbent bicycle


The direct-drive recumbent bicycle is front-wheel-drive and has a crank axle aligned with the front wheel axis. There is no chain, and gear ratios are obtained by a transmission within the front hub itself.

This results in a very simple bicycle design, with full-size wheels ideally positioned for an even weight distribution, giving excellent low-speed balancing, a smooth ride, and low rolling resistance. The front wheel is sufficiently well forward for full braking without going over the handlebars. Since the seating position is not too low, the rider’s eyes face horizontally forward without neck strain. The rider sits at a similar height to that of a car driver, giving good interaction in traffic. With the low crank position, the rider can readily place feet flat on the ground, making the bike user-friendly and well suited for in-town and recreational riding. Also, the cargo carrying capacity of the direct-drive recumbent is impressive due to the amount of space made available under the seat by the long wheelbase and absence of chain. Advantageously, the weight of the cargo is centrally located and low down, not disturbing the handling of the bike.


Optimizing the design


The recumbent design challenge

The position of the front wheel of a recumbent bike is a challenge for the designer. In the recumbent riding posture, the rider’s feet are in front, positioned where the front wheel would be best located. From a design perspective, this causes a conflict between the ideal front wheel position and the location of the pedals on the bicycle’s frame. This conflict has strongly influenced recumbent bike design, leading to a categorization based on how this wheel-pedal interference is avoided (Stegmann (2002)). The long-wheelbase (LWB) recumbent avoids interference between the front wheel and the pedals by placing the front wheel ahead of the pedals; the short-wheelbase (SWB) by positioning the front wheel behind the pedals; and the compact long-wheelbase (CLWB) by positioning the pedals above the front wheel. A tadpole recumbent tricycle addresses the conflict by having two front wheels and placing one on each side of the pedals.


The direct-drive solution

The direct-drive recumbent design transforms this wheel-pedal interference problem into a solution. Since the pedals are best positioned at the ideal front wheel location, why not simply drive the front wheel with the pedals directly? By attaching cranks, pedals, and a hub transmission to the front axle, the design problem is solved. In addition, the long cumbersome chain is eliminated.


Pedal force feedback

However in directly driving the front wheels with the pedals, the pedaling forces are transmitted to the front fork instead of to the bicycle frame. Since the front fork does the steering, there is pedal force feedback to the steering that the rider must resist through the handlebars. Therefore to design an effective direct-drive recumbent bike, there needs to be a clear understanding of how the pedal force effects the steering and the best design practices to address this effect. The direct-drive recumbent is not the only type of bike that has this pedal force feedback effect. In fact all bikes where the pedals are mounted on the front fork assembly will have some pedal force feedback. This type of bike is known generally as a moving bottom bracket (MBB) bike, because the bottom bracket (crank set) moves (turns) with the steering. However the direct-drive recumbent has more pedal force feedback than other MBB bikes (see Garnet (2008) Fig. 20), and so it needs more careful design to prevent the force from being excessive.

It is worthwhile to note that pedal force feedback to the handlebars is experienced, on occasion, in a regular bicycle. When a bike rider is pedaling when standing off the seat (which is often when the pedal loads are high), the rider is resisting the pedal force by the hands through the handlebars. So some pedal force feedback is not unknown to the everyday rider, and should not be a reason to dismiss the direct-drive recumbent design solution.


Best design practices

To address these design issues, Velotegra’s proprietor investigated the pedal force feedback and handling of direct-drive recumbent bicycles in a study published in 2008 (Garnet (2008)). The study recommends an optimal frame and fork design to reduce pedal force feedback, improve handling, and augment the overall user-friendliness of the bike. With this recommended geometry, the pedal force that is fed back to each of the rider’s hands can be as low as 12 percent of the total applied pedal load. The study highlights the following design aspects:

Handlebars: The handlebars are best arranged in the above-seat steering position with the rider’s arms outstretched ahead, allowing the arms the greatest mechanical advantage. An inverted V or W shape to the bars gives the greatest leverage to the rider’s arms for a given overall width, while providing the necessary knee clearance.

Optimum head angle: Inclining the head angle reduces the pedal force feedback. However, this is not due to a closer alignment of the steering axis with the applied pedal forces, because there is a corresponding decrease in the mechanical advantage of the rider’s arms in resisting these pedal forces. Rather it is due to an increase in the fork offset which gives the outward lateral pedal forces greater leverage in cancelling the main pedal forces. However there are handling difficulties with shallow head angles, so the head angle should not be too low. The study indicates that although there is an advantage to reducing the head angle, there is no benefit in reducing it to less than 56 degrees to the horizontal.

Seat height: Increasing the seat height decreases the pedal force feedback by aligning the pedal forces more closely with the steering axis. Providing the rider’s arms remain at the same angle (i.e. providing the handlebars are raised the same amount as the seat) there will be no corresponding loss of arm leverage. However, the seat should not be so high as to lose the recumbent comfort. The feet should still be able to plant firmly on the ground without too much seat cushion depression, and the centre of gravity should remain low enough to prevent over-the-handlebars mishaps under full braking. All this means that for a rider of about 1.8 m (6 ft) height, the seat height should not be more than 560 mm (22 in).

Trail: The trail is the horizontal distance from the steering axis to front wheel ground contact point. Trail causes the front wheel to steer in the direction of the tilt of the frame, restoring balance automatically – even if the bike is ridden hands-free. The amount of trail determines how quickly the steering reacts to a lean of the bicycle frame. In a regular bike, the trail is about 50 mm (2 in). A direct-drive recumbent should be designed with less trail due to a heavier weighted front wheel and a more inclined head angle. However, negative trail should not be used due to instability from ridges and other road surface imperfections. A formula for the trail is presented in the study. The formula indicates a trail of 25 mm (1 in) for a head angle of 56 degrees to the horizontal.

Centering spring: A centering spring is a spring that returns the steering to the center, i.e. to the straight ahead position. A centering spring is required for a direct-drive recumbent bike because the mass of the front assembly (everything that turns with the steering) is much larger than in a regular bike, due to the weight of the rider’s feet on the pedals. Without a centering spring, the steering would respond too strongly to a tilt of the frame or a slight turn of the steering, resulting in an over-correction of the imbalance, and ensuing instability. A centering spring controls these effects, restoring stability and user-friendly handling. The study presents an equation for determining the required spring constant for the centering spring. This spring can be a simple torsion bar placed in the head tube.

Wheelbase: For compactness, it is tempting to place the rear wheel immediately behind the seat back. However this will result in a front wheel which is too lightly loaded, leading to traction problems particularly when going uphill. It is best to aim for a 50/50 weight distribution between the front and rear wheels with the rider on board. This result in a wheelbase ranging from 1300 to 1550 mm (51 to 61 in), depending on the size of rider.

Rigidity: In a regular bicycle, the torsional rigidity of the frame is important to ensure that a portion of the rider’s energy is not lost in frame flexure. For a direct-drive recumbent bike, it is rather the torsional rigidity of the front assembly that counts, particularly from the handlebars to the front wheel axle. This ensures that energy is not lost in structural flexure as the rider’s hands resist the pedal forces. The torsional rigidity can be improved by a fork with a direct connection from the front wheel to the handlebars, with the headset offset behind. This avoids the weakening that results from a relatively narrow steerer tube portion passing through the head tube.

There are other design advantages not included in the study. For example, the centering spring not only improves handling but also makes the bike much easier to walk. One can simply hold the top of the seat back and walk the bike with one hand, leaning the frame to turn with no risk of fork flop. If a tighter turn is required, one simply holds the handlebars with the other hand and forces a tighter turn against the force of the centering spring. In more confined spaces, one can easily place the bike vertically of the front wheel due to the light rear assembly. Simply hold the seat base with the right hand, while holding the handlebars with the left hand with the front brake handle depressed to prevent the front wheel from turning, and lift the bike vertically on the front wheel. Once upright, release the front wheel brake and the bike can be maneuvered with ease. The centering spring keeps the rear assembly upright with no difficulty.

Because front assembly torsional rigidity is required rather than frame rigidity, there are more design options for allowing some frame flexibility to improve the ride, without needing suspension. The 50/50 weight distribution already improves the ride by effectively halving the height of each bump, and careful attention to the frame’s vertical flexibility can improve the ride still further.



History of the direct-drive recumbent


The bicycles of the 1860s, particularly those designed in France, were essentially semi-recumbents (Hadland and Lessing (2014)). Since they were front-wheel-drive and direct-drive, these bike were essentially the first direct-drive recumbent bicycles, albeit semi-recumbent and lacking a transmission hub or freewheel. The illustration below shows an example (from US patent 59,915 to Lallement).




To achieve higher gearing, the size of the front wheel was gradually increased – leading to the well-known high wheeler bicycle, also known as the ordinary or penny-farthing. The large front wheel required the rider to have an upright posture, rather than semi-recumbent, so as to prevent interference between the rider’s legs and the circumference of the large wheel when steering. This upright riding position, combined with the height and forward position of the rider, resulted in a bicycle that was very unsafe, prone to throwing the rider from great height over the handlebars when a large bump was encountered, or under heavy braking.

To address these safety concerns, the safety bicycle was introduced in the mid 1880s. This design, which has not changed significantly in layout to the present day, moved the rider to a lower, more rearward, position by employing chain-driven rear-wheel-drive. This reduced the pitch-over risk, but retained the same upright rider position of the “ordinary”.

The alternative idea of addressing the safety issue by reducing the size of the front wheel and gearing-up the drive system was proposed in several designs, most notably the Crypto Bantam bicycle of the 1890s, shown below. As with the safety bicycle, this design placed the rider lower and slightly further back than the “ordinary” design, with ensuing safety benefits. However, the safety bicycle was already well established and the Crypto Bantam design did not catch on in significant numbers.


Bantam_Bicycle 4

1893 Crypto Bantam. Front-wheel-drive with step-up planetary transmission.


Shortly after this, in the early 1900s, bicycles with a true recumbent position – rather than merely semi-recumbent – were introduced (see below). However, these were chain-driven and rear-wheel-drive, a testament to the popularity of the “safety” bicycle.


Jarvis 4

1902 Jarvis recumbent: Early bike with full recumbent position


Surprisingly, the idea of combining a direct-drive front hub with a recumbent riding position was not proposed until about eighty-five years later. In 1987, Dirck Hartmann of California patented a direct-drive hub and front-wheel-drive recumbent, shown below from US patent 4694708. Hartmann’s also patented many other direct-drive hub designs.


Hartmann recumbent_Page_2 (2)

1987: Direct-drive recumbent proposed by Hartmann.


In the late nineteen nineties, Thomas Kretschmer of Berlin developed a direct-drive recumbent bicycle having a very shallow head angle (45 degrees). This bike (shown below, from German patent DE19736266 (A1)) included a centering spring for the steering to improve the handling. Kretschmer also proposed a multi-speed hub design (German Patent DE19824745 (A1)). Details of Thomas Kretshmer’s direct-drive hub and related bikes were also published in the journal Human Power (Kreschmer (2000)).


Kretschmer 4


1999: Thomas Kreschmer’s direct-drive recumbent. Low head angle and centering spring.


In the early 2000s, John Stegmann of South Africa proposed a direct-drive recumbent bicycle design that takes full advantage of the cargo-carrying capacity afforded by the long wheelbase and absence of chain (Stegmann (2002)). Stegmann also emphasized the need for a viable hub for direct-drive. Stegmann’s design proposal motivated Velotegra’s proprietor to build a direct-drive recumbent bike to test the viability of the direct-drive recumbent (Garnet (2003)). The bike, illustrated below, used an adaptation of a Schlumf Speed-Drive™ bottom bracket gear to provide a makeshift single-speed direct drive hub.



2003: Jeremy Garnet’s direct-drive recumbent – first frame


A year later the original frame was replaced with a better quality frame and components, see below.


second frame

2004: Jeremy Garnet’s direct-drive recumbent – second frame


In 2008, a study (Garnet (2008)) was published to determine the required frame and fork geometry to minimize pedal force feedback and provide user-friendly handling. As part of this research, a variable-geometry frame was built, see below. A summary of the results of is given above in the “Optimizing the design” section.


Variable geometry bike

2008: Variable-geometry frame to test different head angles and front wheel trail.


In 2018, to test the Velotegra hub first prototype, Jeremy Garnet constructed a fourth direct-drive recumbent – shown below. The frame is made of epoxy-bonded and riveted 6061-T6 aluminum. The bike includes an “in-line” headset configuration, where the headset is positioned in-line along the down tube. A centering spring is concealed within the tube. The head angle is 56 degrees to the horizontal, and the wheelbase is 1500 mm.

direct-drive recumbent 2018



Hadland and Lessing. (2014). Bicycle Design: an Illustrated History. Cambridge, Massachusetts. The MIT Press. p. 473.

Stegmann, John. (2002). “Chain of Thought”. Velo Vision, no. 8, pp. 22-25.

Kretschmer, Thomas (2000). “Direct-drive (chainless) recumbent bicycles”, Human Power, no 49:11-14.

Garnet, Jeremy. (2003). “Delving into Direct-Drive”. Velo Vision, no. 12, pp. 18-20.

Garnet, Jeremy (2008). “Ergonomics of Direct-Drive Recumbent Bicycles”, Human Power eJournal, Article 17, Issue 5.