The requirements for the full sculpture are very similar to those of the mini sculpture, with the exception of a few additional elements. However, the main requirements remain the same:
One sensor
One actuator
Human input/interface
Fully autonomous
The additional elements that needed to be incorporated into the full sculpture included:
Elevator
More tracks
Error free
Additional elements such as the ones above were to be integrated as well as possible into each team’s current mini sculpture in order to develop a bigger, hopefully better, and most likely more complex, full sculpture.
Since the full sculpture is an extension of the mini sculpture, here is just a quick reminder as to what the mini sculpture was so that one can more easily understand the development of the full sculpture from the mini sculpture.
The mini sculpture consisted of two trapdoors, a speed sensor, a trampoline and a basket. Both the trampoline and the basket would move to catch the marble based on the velocity of the marble detected by the speed sensor.
The sketch below is the first sketch that this team had drafted of their full sculpture. While it is not the neatest, nor is it the most finessed, it gives the general ideas to what this team was thinking of in terms of design for their full sculpture. Everything drawn in blue is "old," as in it was part of the mini sculpture already. Everything drawn in red is "new," as in it is what was added onto the mini sculpture as an extension to the mini sculpture.
While the sketch above was just a rough sketch of the final, full sculpture, the sketch below is the official final sketch for the full sculpture. One can notice that there are several elements that changed from the rough final sketch to the official final sketch. There were definitely some elements that were removed such as the little loop for the marble to run through. There were a lot of challenges that the team was faced with every time they attempted to add the loop into their structure. Some elements that were added to the sketch were the new trapdoor system at the very top, as well as a basket beneath it, and another, additional track. The official sketch is much cleaner and definitely more refined than the rough final sketch.
One of the main obstacles for the full sculpture was automation. The most important criterion for this final project was to make the sculpture reliably automated while maintaining the ability for the user to alter the course of the marbles. While the mini sculpture also required human input, it allowed the user to set the marble before giving an input. In the full sculpture, Team HR utilized an elevator for the marble to be carried up, which then goes through the sculpture and back into the elevator. The challenge here was ensuring that the marble doesn’t pull up at the top of the elevator, waiting to go into the sculpture. To solve this, they developed a system to redirect the marbles that overflow after the elevator. Throughout the process of building the structure, questions of reliability occurred most often. For example, trampoline and basket calculations may not be precise enough, the positioning of the uppermost ball dropper may be skewed such that marbles would only go down our alternative track, marbles may fall out of the elevator or bounce off tracks, and much more. Time was another challenge. Since they attempted to not just automate the mini sculpture, but also add another system to the whole structure, the three days that they were provided with in the last week was another constraint.
Other challenges that the team faced throughout the process of designing, building and programming the full structure included:
The elevator constantly breaking down
Marbles having too much speed --> constantly bouncing off the track or not falling into the correct places
Marbles hitting speed sensor instead of going through --> slowed marbles down
Space between NXT motor trapdoor and track trapdoor (xylophone ball trap) was too large --> sometimes would hold 2 marbles instead of only 1
Ball dropper at top of structure wasn't always positioned correctly --> sometimes 2 marbles would get stuck, sometimes no marbles would stay
Communication errors. Sometimes the ideas weren't communicated correctly, causing some misinterperation of the designs.
Shown below is an image of the final version of this team's full sculpture. To simply describe what is happening in the image, the marble will ride the elevator from the bottom of the structure to the top of the structure, fall through a dropper to right in front of a trapdoor. If there are too many marbles pooled up in front of the trapdoor, they will instead fall through a door beneath the dropper and onto a different track, going through a series of twists, turns and loops until it reaches the main track again to return to the elevator. If the marble goes down the main track, one of the two trapdoors will open, releasing it to go through the speed sensor, and once the trampoline and the basket read the velocity and do the calculations, they will move accordingly to catch the marble. Once the marble is caught in the basket, it will fall through and fall through one of the three green funnels below the basket, thus falling into the main track below the funnels and returning to the elevator.
In regards to how this team reached their final decision on what to do for their full sculpture, everything started from the mini sculpture. The process of designing, building, simulating and programming the mini sculpture was a lengthy process that consisted of lots of sketches, a pugh chart, a red-filled risk analysis table and plenty of communication (and sometimes miscommunication). To learn more about the team's mini sculpture, click the button below to view the mini sculpture page.
FLOW CHART
Shown below is the flow chart for the team's final design. The flow chart is meant to help visualize what the program/code is supposed to get the final full sculpture to do. It is also a means for students to look back at whenever something goes wrong with the code, so that they can visualize what might have gone wrong, what was supposed to happen, and what needs to be done to achieve that solution and fix the problem. Aside from our flow charts from the mini sculpture, we needed to code the trapdoors, or gates, to be automated.
This was a project done by Hikaru. During the last week, Jessica, Giselle, and Kylie worked on integrating and automating the mini sculpture into the full sculpture. Meanwhile, Hikaru designed, 3D printed and programmed the delta kinematic. This was planned to be part of the full sculpture but it unfortunately did not make it. However, it does function by itself.
The delta kinematic, more specifically the linear delta kinematic, is often used in 3D printing. Utilizing 3 vertical axis, it looks vastly different from other kinematics used in 3D printing.
This portion was a challenge for Hikaru himself. It started out with the idea of creating a system that moves a marble in three dimensions. For that purpose, Hikaru looked at mechanisms used in 3D printing.
KINEMATICS
There are 3(4ish) main kinematics that are widely used in 3D printing. Cartesian, CoreXY, and Delta. Cartesian can come in two different types.
Cartesian can come in two different types. One types consists of two components that move in either the x or the y axis, as well as one of the components also moving in the z axis. Another type is one component moving in the x and y axis, while the other move in the z axis.
CoreXY is similar to the the second type of Cartesian, but two motors move the component diagonally at the same time, removing the need for two motors to work individually in one axis.
Delta has three vertical gantries that move up and down. Two parallel arms extend from the gantries to the head. By combining the three different heights, the position of the head can be determined.
DECIDING FACTOR
Delta was chosen as the system for the project because it allowed the head to move in 3 dimensions. Cartesian and CoreXY both only allow the head to move in the x and/or y axis. On the other hand, delta can move in all three axis. This fit our purpose the most since the goal was to move the marble across a 3D space.
Since the provided parts for the final sculpture did not have the key components for the kinematic, parts were designed in Fusion 360 and printed with the 3D printer available.
PROGRAMMING
A program was written to calculate the height that each gantry of the delta has to be at, and move the motors accordingly.
MATH
By using the Pythagorean theorem, the distance from the head to the gantry was calculated, and the length of the arm was given. Using the Pythagorean theorem again, the height of the gantry was calculated.
This video shows the simulation created in Geogebra by Hikaru. This follows how the code functions, created in order to ensure that the program works as intended before building the whole structure.
CHALLENGES
FRICTIONS
In the first prototype, the gantry had too much friction and also was too short. Being short caused it to lock in position if given uneven pressure. In the second prototype, the clearance was made wider, the gantry was made longer, and a smaller contact surface was created.
TENSIONING THE BELT
A fishing line was used to connect the gantry to the motor which spins to move the gantry up or down. There were adjustments needed to the design of the motor so that it had enough tension for the motor to have a grip on the fishing line.
CATCHING AND RELEASING THE MARBLE
One of the most difficult design was the design for catching and releasing the marble. Since the head cannot support or have space for a whole NXT motor, the mechanism had to run without additional motors.
FINAL RESULT
Although the delta system did not make it to integration into the full sculpture due to time constraints, it does function on itself.