What follows is a description of a proposed modification to Flevo tilting trike rear suspension. It is a radically simple idea, but would it work in practice? I have not tried it yet, but I did build a model.

Flevo Tilting Trike

This is a very popular trike in the Netherlands, where it is often used for transporting heavy loads. A large boot for carrying luggage mounts between the rear wheels. It is front wheel driven, the rear wheels are mounted independently on a fixed axle, and the seat and front wheel tilt relative to the rear assembly as shown in the diagram below. The track width is only 60cm, and the center of gravity (CG) is relatively high, but the tilting design gives this trike reasonably good rollover stability.

Although extremely simple and effective, the design involves certain compromises: Firstly, the Flevo trike has a rear suspension, but only the seat is suspended. The luggage boot is mounted to the rear axle and is thus not suspended. Secondly, the fixed rear axle and luggage boot, combined with a heavy load may cause the rear part to tip over during hard cornering even when the rider is leaning far into the turn. Incidents of this kind have been reported, sometimes resulting in a catastrophic rollover. Moreover, the fixed axle design puts a lot of lateral stress on the rear wheels if cornering under heavy load.

The Fleovo tilt mechanism is shown below.

One obvious improvement would be to keep the basic Flevo steering arrangement, but to make the rear wheels and luggage compartment tilt in conjuction with the front wheel. There are many ways to do this, but in keeping with the elegant simplicity of the original Flevo design, and with the added requirement of an improved suspension system, how is this to be achieved?

Can it be improved?

I built a 1/5 scale model that tilts on all three wheels. It is a highly simplified model: The wheels cannot rotate - the two triangular plates represent the rear wheels. It was designed only to illustrate an alternative rear suspension. Click on the pictures to see larger images (the real Flevo trike is shown at the left).
The simplest solution I could think of started out from the common design in which a leaf spring acts as one of the suspension arms in a parallelogram arrangement. The stub axles are mounted to the ends of the leaf spring. Taking this idea to its extreme: what if the spring forms a circle, with the stub axles attached to its perimeter? The sketch below illustrates this schematically, where A-A' are the stub axles.

Such a "circular leaf spring" might be constructed out of carbon fibre and made strong enough in the right places to hold the stub axles and at the same time be flexible enough to act as a spring and to allow tilting. As the diagram shows, tilting is possible through deformation of the circular spring. The load is attached at two points shown as black circles in the sketch. Tilt is induced by shifting the load, e.g. leaning the body to one side.

The basic principle can be illustrated if you roll a sheet of paper into a cylinder and flatten it slightly by pressing it between your hands. Compressing the roll is the "spring" action. If you then roll the paper between your hands while holding them parallel, there is no resistance. This is the tilting action. The "leaf" would have to be quite broad so as to keep the wheels parallel in a fore-and-aft direction and to bear up under loads due to braking torque. The wheels are kept parallel in the vertical plane by virtue of being attached at opposite points on a circle: Tangents at opposite points on the circumference of a circle are parallel. Deformation of the circle, such as flattening it, will not upset this basic relationship. However, assymmetrical deformation does lead to loss of alignment. On a human-powered vehicle this might be tolerable. Thus, we have a tilting suspension design in which there are no ball joints or other pivots, and no tie rods or suspension arms.

Model of a concept

To find out how it would work in practice, I built a scale model (seen above). The circular spring was made from a tin can 10cm in diameter and .20mm wall thickness. It is slightly flattened to an oval, so it models a full scale track width of about 60cm. The spring would be about 1mm thick at full scale. I soldered on a "seat" with backrest, and a crude front end to represent the steering assembly.

Initially, I thought this "tin can spring" could in fact serve a dual function as suspension and luggage container. Later I abandoned that idea in favour of a second container suspended inside the "spring". In the model this was realized as a second tin can that is about 8cm in diameter, closed at one end. The seat and front end of the trike are attached to this inner can, which thus has a structural role: It both carries the weight of the luggage and bears the load of the rider in the seat above. How the model tilts is shown in the composite photograph below. The picture also illustrates how the inner container is attached to the spring.

The picture clearly shows the deformation of the spring. To check the parallelism of the "wheels" I joined the top and bottom of each with a string. When a string goes slack this indicates the distance between the top or bottom of the wheels has changed. In fact, this does happen, even though the strings themselves actually help to keep the "wheels" parallel. It seems this fact is the main disadvantage of the "tin can spring" - which could lead to a high degree of tyre scrub in practice.

Note that the attachment points of the inner can to the spring are modelled as pivots, but that these could just as well be fixed in the same as the stub axles are fixed. There is very little actual movement in these pivots, and deformation of the spring would be sufficient to allow tilting.


The model has demonstrated that a practical implementation of this idea is possible, but not highly effective in maintaining wheel alignment. Given the cost of carbon fibre materials, I am not going to try building a full-scale version of this, even though it might be possible to perfect the design.

November 1999

The views and opinions expressed on this page and those linked to it are strictly those of the page author/owner and do not reflect those of the University of Cape Town, nor have the contents of this page been approved by the University.