Whenever a ball moves between the stages of the display it goes through the elevator to get there. Below the stages are lifted and each of the four arrows represents a time the ball must go through the elevator.
I was between a few designs for how to lift the balls up from the bottom to the top. I toyed with the idea of using friction and a belt to roll the balls up a wall. I also thought about using air and blowing the balls up a tube. I ended up designing an augur with multiple channels. My first augur elevator prototype had a large diameter. (Second Image Above) I only realized after this how much space the elevator would actually need. The first Draft was very clunky and would not have fit inside the display. I also realized that a lot of the geometry that determined the path of the ball would be very finicky. The balls jammed or managed to skip into different channels because the features did not come out well on the printer. Below is the second elevator prototype.
I had thought I was done with making the elevator, but I was not. Once I got to integrating the elevator within the system, I realized how much of an issue corking would be and did not want to have to put a disruptor onto every channel of the elevator. I realized it would be best to make it so the bottom of the elevator had an integrated disruptor that could unjam balls at the base. The finicky nature of this problem forced me to do a lot of iterating of the augur and auger barrel. Below is an example of corking at the base of the elevator and some of the prototypes.
Corking Example
Prototype 4 with moving lip
Adding Fin to Tap the balls away that were corking.
Better Picture of Fin
Removed Fin. It was too subtle to work consistently. So, I switched to a barrier that would fully obscure the opening.
cross section with new barrier
The inner feature preventing the channels from mixing at the bottom was too fragile and snapped.
Increased barrier size to remove that dividing feature
New problem. If more than 1 ball goes up at a time, the top feature breaks.
Final Fix To Augur Bottom
The top geometry was less finicky and only required 3 iterations. The issues I had was making a geometry at the top that did not rely on the elevator being perfectly vertical. If the elevator leaned more than a few degrees pellets would get stuck at the top. A couple adjustments to the auger and auger barrel fixed this problem.
After getting the prototype done, I changed one dimension and a friend printed me the full height version. Below is the final part next to the shorter final prototype.
