The Velassi


Edit:18 sept. 2023, Cre:18 sept. 2023

Velassi V 1

The story

The Velassi ‘V1’ is the first version of the Velassi that I built in April 2019 after a long phase of study and preparation.
In 2022, this bike covered just over 5400 km on the road without any problems and just over 400 km on a ‘home trainer’ during the first confinement.
It proved to be a good all-round bike, which is unusual for a recumbent, but its configuration with a long wheelbase and a fairly upright seat had a lot to do with it. The V2 performs even better thanks to a more stable steering system, a more suitable tire set-up and air suspension (rubber block suspension on the V1).

I’d never built a bike before, nor even ridden one horizontally, so it was a big gamble, which I managed without any major problems, but only after a long period of study.

The choice of wood as a construction material posed particular engineering challenges, due to its great flexibility and heterogeneity. This too was a gamble. At first, I had to make a few reinforcements to improve its torsional rigidity, but the bike has never had any behavioral problems or failures. I did, however, have to redesign and rebuild the handlebar joint in early 2021, as it was beginning to show serious weaknesses.

However, as with all epoxy resin work, this bike should have been given at least one more coat of epoxy and varnished to reinforce its protection against UV and humidity. Having decided to build a new one, I never took the time to dismantle it to apply another coat of epoxy and varnish, and a few traces of moisture penetration through the resin appeared in a few places, and the ‘head tube’ had problems swelling after a very wet ride, necessitating a loosening of the steering.

Unsurprisingly for a first design, it does have a few flaws and I quickly decided that there would be a new version and, as I’m getting on in years, that this future version would be electric.
But all things considered, the bike wasn’t doing so badly, so the new version, the ‘V2’, took almost 2 years to emerge. When I built the V2 in December 2021 and January 2022, it had been over a year since the plans were completed, the components sourced and the aluminum parts made.
I ordered the engine in July 2020, received it at the end of September 2020, but didn’t test it until June 2021. During these tests, following an ill-timed trial run, I burnt out the engine control board, which cooled me down a bit. In practice, changing this card (in December 2021) turned out to be very simple and relatively inexpensive, even if it seems that the new card was buggy (I’ll come back to this in a specific page about this engine).

I hesitated for a long time to build the V2 in riveted aluminum, but I figured that changing the material would delay the construction even further, and I finally took the plunge at the end of 2021 for a new wooden construction, but in pine and fir instead of the oak of the V1 version.

Images



To that:



From that:

while going through:



Construction details





Specifications

Bicycle specification (small size, for 73 to 80 cm inseam):
Wheelbase: 1650 mm
Front wheel diameter: 20” (rim 406)
Rear wheel diameter: 26” (rim 559)
Seat 520–540 mm off the ground (can be adjusted by adjusting suspension support with a spacer)
Crankset 120–140mm lower than seat (it’s 400mm higher)
145mm cranks (at the time of manufacture 152mm): This is a recumbent bike for smaller riders, and replacing the 152mm cranks with 145mm has been very beneficial.
59mm caster → to know how to choose the caster see a online calculator in English
Wheel flop (lateral tilt): 26.4 mm (too much), due to the large inclination of the head tube.

Single-plate transmission on 8-speed Nexus hub. Very short development (5m in 8th gear), so no pedaling on descents.

I’ve modeled several frame types, including a single-beam, but in practice, two scissor-type trellis frames, one in wood and the other in steel, seem the most practical, while remaining relatively light. The stresses on a long-wheelbase bicycle frame are quite considerable, and I have little confidence in an amateur-made single-beam frame (aluminum bikes made by pros are heat-treated after welding, and steel ones are a bit heavy)

Main problems

The bike is heavy

This is not specifically related to the wooden construction. A recumbent bike is always heavier than an ordinary bike, because of the seat, the shape of the frame and the very long chain. What’s more, this bike is equipped with indirect steering, which means it has an additional control linkage and an extra bearing for the handlebars. It also has a very long frame. My canvas seat is also significantly heavier than the ‘shell’ seats found on performance recumbents. This resulted in a weight of around 22kg once all accessories were fitted. Pneumatic equipment with puncture-proof tires doesn’t help keep the weight down.

Performance is not good

Using a gearbox with very modest power reduces efficiency, especially as I installed a double-roller tensioner to absorb the length variation due to the suspension. This was a design error, and a single-roller tensioner with an extra roller at the bottom should improve efficiency. The handy hub dynamo also absorbs quite a lot of power (maybe 9W), as it’s not a dynamo specifically for small wheels (my front wheel is 20”) and it turns faster than the manufacturer intended. While it provides power at very low speeds, it absorbs more of the rider’s energy. These two elements, combined with a choice of tires adapted to the ride, lead to a fairly low output.

The development range of the Nexus 8 gearbox (306%) is insufficient.

While this range is sufficient for riding on flat terrain, as I’ve adopted a fairly low gear ratio, in off-road use, on steep climbs, I’ve clearly missed shorter gear ratios. One solution is to have a double chainring (made possible by the double-roller tensioner, whose original function is exactly this).

Stability at low speed is not very good

Since I didn’t know whether I’d be satisfied with a low-handlebar steering system, and since I had the opportunity to buy a second-hand Rans Stratus fork, I adopted the same geometry as the Rans, with a very raked steering system that would also allow me, if need be, to have n high handlebars instead of the indirect low-handlebar steering system. Such a geometry poses stability problems at low speeds, but I was able to improve the situation by adding a steering spring.

And with a motor?

The first three shortcomings disappear if the bike is fitted with a motor. The weight only becomes a problem if you have to negotiate stairs or ramps, which is quite rare. Efficiency becomes irrelevant, and a motor greatly reduces the range of speeds required. As for the lack of stability, this has been corrected by the new geometry of the Velassi V2 and, in any case, the engine’s power means that the minimum stable speed can be reached more quickly.

A gearbox for a touring bike

A gearbox is a practical and comfortable piece of equipment, even more so on a recumbent bike where it’s very difficult to restart if you’ve forgotten to downshift when you stopped, because you can’t get into a dancer’s stride to help you start.
But performance is not very good. The curves show a 2–3% drop in efficiency compared with a derailleur drivetrain. Unfortunately, these curves are given for power ratings of 150W or more. For touring use, the power you develop over the course of a day is much lower, especially if, like me, you’re not particularly robust. I estimate it at 70W, and at this power, the gearbox efficiency drops significantly, as the loss in the gears (greased on a Nexus 8) does not decrease proportionally to the power transferred. Efficiency then falls below 90%, which is very low.
The speed range, adequate for urban use, turned out to be too short for my use, which included quite a lot of unpaved sections, sometimes with very steep gradients.

(c) Pierre ROUZEAU
Privacy - Vie privée - Imprimable - Rechercher
Page mise à jour le 18/09/2023 16:21