For several months, I’ve been making and trialling a new recumbent
bike. Its unique feature is a rear fork suspension with rubber blocks. The
blocks rest inside the rectangular frame and compress over bumps. This is a new
concept, but the rest of the bike – seat, tailbox, front end - is based on
front wheel drive leaning trikes I have been riding for years. As with all my
cycles, aims with this bike were to be aerodynamic, light and have good load
capacity.
The rear fork has no pivot point and finds its own position.
The fork stem is from brazed steel tubes stacked as a rectangle. With 3d printed plastic spacers each side,
a pin to hold position, and the rider’s weight, the fork can only rotate in one
plane and acts as suspension. The suspension blocks are cut down portions of
Mckay industrial dampers. Although other rubber material can be used (pedal
rubbers, cycle tyres and tubes), the industrial dampers are designed for
suspension of this type, so are stable over a wide range of temperatures.
It’s taken a long time to come to this design as I started
building front wheel drive bikes with the rectangular hollow frame that could
use it in about 2012. Here
is a blog post from back then.
The 2012 bikes and some of the trikes I’ve built since have used Capral
rectangular hollow 82.6 x 28.6 x 2.3 wall aluminium beams with a combined tailbox
and seat. The tailbox is aerodynamic and has about 50 litres of storage. It
clamps to the beam and slides up and down the beam to adjust for leg length.
It’s removable and doesn’t need any frame holes or bosses for support or grip.
The trick to using this aluminium section is that with the front wheel drive I use, there’s
no frame chain stress, so stress is mainly from body weight. This means a
relatively deep and thin section of tube can take the stresses and not flex too
much.
|
Bike rear triangle circa 2012
|
|
2012 bike on Murray tour
|
The first frames of this type were fabricated by Michael
Rogan from MR Components in Hastings, and
they included reinforcements and pivot bosses at the back. But since then, I’ve
worked out how to do leaning trike frames at home with some custom cast parts,
some 3d printing, handsaws and a drill press. A plan for the leaning trike is on thingiverse
at https://www.thingiverse.com/thing:4201871
. This latest rear fork borrows from the design of the front frame join: body
weight under gravity helps keep parts together.
So far the new bike works well and I’m happy with it. For me
it takes more shed time to produce than the equivalent leaning trike, but the
bike back end weighs less and should have less rolling resistance. A nice
modularity is also possible, with the same tailbox and front section accepting a bike
or leaning trike rear end. The rear wheel and fork detaches quickly from the
bike by removing one screw, then some jiggling. That should be good for fitting
the bike in trains or cars.
The bike’s tailbox is still a work in progress, but I am
confident I will end up with a light design. At the start my shoulders were too
far back, and I’ve made a temporary
wedge that pushes my shoulders forwards but is also a small glovebox for keys,
wallet, phone, camera etc.
Here are the parts for the back of the new bike and the
equivalent parts for a leaning trike.
Bike back frame & wheel
Parts: 2 Custom rubber buffers, 1 custom fabricated fork, 4
3d printed guide plates, screws, custom wheel build includes dynamo (4 sets of
parts)
Frame drilled one place only
Rear frame & parts weight: 4.6kg
Trike back frame and wheels
Parts: Unicycle cranks, YST bottom bracket, 2 custom bearing
housings, 2 custom wheel axles, 2 standard pedal prix trike wheels from MR
components. (5 sets of parts)
Frame drilled 4 places.
Rear frame & parts weight: 5.2kg
Forces and movement of back wheel
With 2 separate elastomers working at the same, the rear wheel dynamics aren't immediately obvious. But each elastomer can be considered a pivot point as well as compliant suspension, and when this is done some motion / suspension equations can be worked out. In the diagrams above, the top line is the resultant of the two lower motions. The simple version has equal spring rates and distances, and the general version has unequal distances and spring rates.
Regards
Steve Nurse