h a l f b a k e r yi v n i n seeks n e t o
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Conventional electric-assist bikes have motors that power the bicycle chain via a sprocket. The rider pedals as normal, but the motor supplements the power by turning.
The problem is that while the motor runs with a constant power output, the rider's legs do not. The rider pushes down, developing
more power on each down stroke than when the feet are at the apex and nadir.
As a result, the drive power is not stable. It feels like you are cycling downhill or with a tail-wind, but (wo)man and machine are not working as one.
I suggest that instead, the motor should drive the chain via a pair of eliptical gears, each at 120 degrees leading and lagging the footpedals. There would be three phases of drive force per pedal cycle: two electrical, one human.
In the same way that three phase electricity generates a (nearly) smooth power output in industrial motors, this drive mechanism will deliver easy and complementary power.
[link]
|
| |
I remember some bicycles had an elliptical chainwheel for the rider. Didn't catch on. The electrical way of doing this is to regulate the speed of the motor :) |
|
| |
Since your bike is electric anyway put a torque sensor on the hub or a strain sensor on the chain. Then use that signal to control the torque of the motor. Saves weight and complexity. (There is still complexity in the electronics but you buy that as a chip). |
|
| |
/(wo)man and machine become one./ |
|
| |
Wire the batteries directly into the cyclists leg muscles. |
|
| |
Why not put a set of toe clips on the pedals or a set of clipless pedals, and learn how to spin? As one leg pushes down, you pull up with the other. |
|
| |
I think the point is that even if you learn the pull-while-pushing technique, you still generate much more torque while your feet are moving vertically than you do at the top/bottom of the stroke. It would be at that point that the electrical assist would be most appreciated, especially when going up a steep hill. (+) |
|
| |
A pedaling person will generate two power pulses per revolution, seperated by 180 degrees. This holds regardless of if the person is using one leg at a time or two (pulling while pushing). Motor supplimental power could be applied in two pulses per revolution at 90 degrees from the human powered input, resulting in a 4 phase system, 2 human, 2 electrical. |
|
| |
In this simple form the motor cannot put in more power than the human. If it did then the two phases powered by the motor would make the drive power 'unstable' again. This makes it difficult to provide large amounts of supplimentary power. |
|
| |
A more sophisticated system would target a constant power output, monitor the human generated input power and subtract that amount of power from the motor output, resulting in consistent, smooth output power supplimented by the human. |
|
| |
On modern bicycles with elliptical chainrings, the drive is designed to draw maximum power from the rider on the up/down stroke, and to give their legs a brief rest at the top/bottom. My applying a motor-assist to this part of the cycle, you shorten this rest and could conceivably add to the fatigue of the cyclist. |
|
| |
Perhaps as an alternative, you could drive the system *in phase* with the cyclist. They'd get the same rests, and you'd be able to apply the benefits of the research into non-circular chainrings, but the cyclist would feel superhuman. Kinda like that exoskeleton freight loader in Aliens... |
|
| |
At the bike dealer or at a central location, place a very adjustable stationary bike. Potential customer rides bike after getting seat pedals and handlebars just the way he likes them. |
|
| |
During each customers ride, the special bike tries various levels of timing and assistance. Customer votes on each offering and eventually leaves with a custom programed chip for his bike. Maybe chip/rider have the ability to make some small adjustments while in use as grade and conditions change. |
|
| |
So humans create oscillations in torque. You want to
cancel this out by having artificially oscillating electrical
torque. |
|
| |
Hmm, why not have a damper? Actually, bicycles already
do this. When the amplitude of torque oscillations is high
and frequency low, for example climbing a hill in too high
a gear, you can see the speed of the bicycle change with
the angle of the cranks. This effect must happen
constantly, but as the frequency goes up and the
amplitude goes down it becomes progressively less
noticeable. I think this is because of the damping or
buffering capacity of the bicycle. A moving bicycle can
actually absorb a little of the torque differential, so that
the change in bicycle speed is not always directly
proportional to the input torque. |
|
| |
The mechanisms behind this are multiple and compound.
Apply torque to the pedals and a few things happen. The
pedal axle bends a little, the crank bends a little, the
whole frame flexes.. the chain will stretch a little. Then
the tourque reaches the wheel where it is applied at the
hub. Here, the hub may twist as the torque is applied
asymmetrically. Then, the whole spoke-rim combination
may deform. Lastly the wheel will revolve inside the tyre
a little (or a lot, tearing off the valve stem if you're a
little overenthusiastic and under-inflated). When you
apply much lower torque at a different crank angle, all
this flex will unwind and apply stored energy to the
forward motion of the bicycle. Neat. |
|
| |
The other aspect to the damping is the flywheel. IC
engines utilize flywheels to damp their own torque
oscillations, by exploiting the non linear relationship
between input torque and rotation rate. Bicycles
constitute an interesting 2 layer momentum storage
system where input torque pulses are converted to
angular momentum in the wheels rotation and linear
momentum in moving both the bicycle, wheels and rider
forward. |
|
| |
In short, I'm not sure such fine motor control is necessary
considering the buffering capacity of the system. |
|
| |