How the marble elevating wheel works
Coming up with the idea for the marble elevating wheel mechanism was my inspiration for Marble Machine 2.The problem with the marble pump mechanism is not so much the pumping, but having an intake that essentially serializes the marbles. If too many marbles are at the intake, they may cluster around the spot where the opening constricts to one marble in such a way that they jam. Picture a funnel with a hole just big enough for one marble. If you fill it with marbles, it is almost sure to jam before all the marbles fall out. The only reason a hourglass is unlikely to jam is that its opening is actually wide enough for several grains of sand across, making it less likely to jam.
The wheel mechanism is inspired by what I saw when I examined a corn seeding machine (way back, when I was a 9 year old kid). The machine had a disk with holes in it, that rotated through the bin of seed corn. Each hole was big enough for one seed corn, and would almost always, after having been dragged through the bin, have one seed in it.
The marble elevating wheel works on a similar principle in this regard. It consists of a disc with 16 holes drilled in the circumferance, although this disk is oriented vertically, and the holes do not go all the way through the disc.
|
With the marble elevating wheel, I can have many marbles (20 or more) at the bottom of the wheel,
and be sure that it will not jam the input, because the marbles do not all have to make it into a fixed
hole. As the wheel rotates, the holes move past, and at some point, a marble will be directly in front
of a hole, and roll in. This gives many places where a marble can go into the wheel, eliminating the
possibility of input jams. However, if a lot of marbles are at the input, the wheel will cause them
to stir around quite a bit, making a lot of noise. This noise is less interesting than the other
noises produced by the machine, and thus not necessarily a good thing.
Because the holes are angled to face slightly towards the axis of rotation, marbles have a tendency to roll into the holes at the bottom, and out of the holes at the top. As soon as a hole is above the axis of rotation, the holes faces slightly downward. This necessitates a barrier to keep the marbles in their holes until they reach the top.
|
Throughput optimizations
The wheel mechanism has much higher throughput than the marble pump. However, as the wheels are turned faster, it gets more difficult for the marbles to roll into the holes as they whiz by. To improve this, I shaped the openings of the holes to allow marbles to roll into them in a smoother path. However, as the wheels turn even faster, centrifugal force can prevent the marbles from rolling out at the top.
On my machine, optimal throughput is achieved at about 25-30 RPM. This gives me a capacity of roughly 6 marbles per second, with some of the holes not always catching marbles. Unfortunately, I only drilled 16 holes into the wheels. There would have been room for 24 holes, had I spaced them differently. This would have given me a capacity of about 9 marbles per second.
In comparison, the best I can do with a pump is about 3 marbles per second, but the motor on marble machine 1 is gared for just two marbles per second.