Matthew Gutierrez and Anas Yousaf
Overview
Although we originally planned to create a watch with one of the motions seen below, we ultimately moved away from that idea due to the fact that we thought it wouldn’t be feasible to cerate the 3 arm movements that we wanted. We then decided to pivot to a new idea using the mechanical motions 34 and 158, where we would create a Pokémon trainer throwing a Pokeball to catch a Pokémon. We would accomplish this by essentially using motion 158, however instead of directly moving the wheel ourselves, it would be controlled by a gear system like that in motion 34.
Figure 1 & 2: Mechanical Movements 34 and 158
Low Fidelity Model
To begin our low fidelity model, we started by first designing our mechanism in illustrator to get an idea of what pieces we would need. We started out with cardboard to test the gear teeth, and it worked surprisingly well on our first try. Once we found that the cardboard fit well together, we moved on to creating a simple rectangular base that would hold our components in place so they could rotate.
Figure 3 & 4: Illustrator Model and First Cardboard Cut
We also tried to start creating arms that would play into our second motion, so we could get a general understanding of how they would work and how long they would need to be. Initially we used the radius of the circle to constrain the little arm, however during testing we found out that it would not work, and the small arm on the wheel needed to be increased in size to allow for the arms to reach a full range of motion. We also began to measure the length of the wheel arm at the position in which the long secondary arm would be at full extension, which gave us a better understanding of our design and motion.
Figure 5: Full Cardboard Model with Small Arm
Figure 6: Full Cardboard Model with Proper Arms
Medium Fidelity Model
We began to move on to higher fidelity materials, in this case, wood. The differences we started to make were to increase the layers of the design, as we made another layer of the outer gear to allow for more contact area between the gears, as we doubled up the small inner gear as well into two layers. We also added a spacer layer so that the wood would not be rubbing against the other wood layer.
Figure 7: Rotating Plate Layers for Wooden Model
We also fiddled with the arm lengths once again, as the arm length we had chosen was simply too long, as they moved below the plain, which would not work in the final design. We moved to shorten the arms once again, this time keeping the primary arm between the radius and diameter of the circle. This fixed the issue and gave us the range of motion that we wanted for the arms. This is also around the time when we decided to make extensive use of bearings in our design to allow for a clean range of rotation in our parts for ideal pivot points, which helped us reduce friction in a lot of our parts such as the arms.
Final Product
In our final design, we changed our back and front plate into acrylic so that the gear motion could be observed. We also made use of the vinyl cutter to put a sticker on our gear, for our Pokémon, and the Pokeball. The gear was turned into the Pokémon Klink, the Pokémon we would “catch” Zubat, and the Pokeball on the end of our arms. We spray painted the inner wooden gears black to create a cave aesthetic for our Zubat, and all other pieces were given a clear coat to help protect the wood. In our final design an oval base was created along with triangular support to allow our mechanical motion to stand up right and give our device stability. As seen on the left, we laser etched a Pokémon trainer as well to “throw” our Pokeball. We also have a handle in the back of the design which allows for easier turning and more ease of use for the user to turn our gear.
Figure 8: Final Model
Difficulties Faced
Throughout the course of the midterm, there were several difficulties that we encountered. The first issue that we encountered was making the gears work properly. Although the initial gears that we made from the gear generator (link) fit well together, the cardboard laser cut had some issues working. However, after cutting the gears out of wood, doubling up the gears, and adding a gap between the back plate and the gears, we were able to make our gear mechanism run smoothly. Another minor issue that we encountered was that our medium fidelity base was not properly aligned so the back plate that had everything on it was not aligned with the extra base pieces at the bottom. After learning how to use the scissors tool in Adobe Illustrator (thanks Douglas), we were able to adjust the extra pieces and align the bottom and dowel holes properly.
A big issue that we encountered was getting the correct proportions for our arms. At first, we created a small arm around the length of the radius of the wheel along with the long arm attached at the base. We quickly realized that the length of the arm attached to the wheel was completely incorrect and that we had to make that arm at least as long as the diameter of the wheel. When we prototyped the arms a second time, we realized that we had overcorrected the arm lengths and ended up making them too long. This meant that at the lowest point, the arms would go lower than parallel to the ground. After spending a couple more hours on Illustrator, we were able to figure out arm lengths that worked properly.
Figure 9-11: Iterations of Arm Proportions
Finally, the biggest issue that we have with our final product is the dowel that is attached to the internal gear that is cranked. Since the bearings that we decided to utilize had slightly smaller holes than the dowels we used, we had to sand the dowels down to insert them into the bearings. However, since the dowel attached to the gear on one end and the crank on the other hand had to go through the bearing, it was physically impossible to sand the entire dowel down and still fit it into the gears/crank. Therefore, we decided to create two dowels that were sanded down on one side and glue the dowels together inside the bearing. Although this mechanism worked initially, it created several problems for us. The first issue was that the bearing was misaligned meaning that the gear was slightly slanted when attached. After messing around with the gear and adjusting it slightly, we were still able to get a pretty smooth gear motion. However, after messing with the bearing too much, the adhesive holding the two dowels inside the bearing wore down, causing the connection to become really weak. As a result, the crank that we attached does not actually turn the gear when cranked.
Lessons Learned
I think one of the biggest lessons that we learned was to use the proper materials and tools for the type of prototyping. Since we utilized super glue for the two dowels connected inside the bearing, the connection eventually weakened and broke, meaning that our crank does not work. If we had utilized a stronger and more appropriate adhesive such as epoxy, we would not have had this issue and had a fully functioning prototype.
Another lesson learned from this project is to properly take care of our pieces so that the final product can look clean and well put together. Since we removed the blue covering off of the acrylic as soon as we laser cut it, the clear acrylic got really dirty over the course of our assembly. Although we tried to clean it as best as possible, there were some glue marks that we could not remove. As a result, our acrylic pieces look a little dirty in some areas. Furthermore, some of the wood that we clear-coated got blemished by other paints and marks during our assembly. Overall, we should have been more careful with our pieces so that our final product would have looked more presentable.
Figure 12: Blemishes on Acrylic and Wood
The major lesson that can be taken from this is that we should have started to assemble parts of our final prototype, especially the most difficult areas, earlier so that we could troubleshoot any errors that we might have had. Although the weekly check-ins helped keep us on track, we made a lot of changes from the medium to the final prototype. If we had started assembling our final prototype a little earlier, we might have been able to address some of the issues that we encountered such as the weak adhesive and the small blemishes on areas of the project.
Cost Analysis
For this analysis, we will look at the cost of: cardboard, wooden dowels, ¼ inch wood sheets, ¼ inch acrylic sheets, super glue, bearings, wood glue, and labor costs.
We used around two sheets of cardboard which, priced at around $2 per sheet, comes out to $4. In total, we probably used one dowel which comes out to $1. For the wooden sheets, we used approximately 6 sheets of wood (due to some failed cuts) which comes out to $30 since it is $5 per sheet at the OEDK. For the acrylic, we only used one sheet of ¼ inch acrylic which the OEDK sells for $5. We used around one big bottle of super glue which can be approximated to $5. We utilized around 10 bearings (some testing and some implemented) which costs around $7 per pack of 10. One bottle of wood glue is around $4 but since we did not utilize an entire bottle, we can approximate the price to around $1. Finally, we both put in around 35 hours each for this project. For Anas’s pay of $10 per hour, his labor was worth $350 and for Matthew’s pay of $12 per hour, his labor cost was $420. Overall, the total cost of this project was $823. It should be noted that this cost does not include the cost of materials and tools readily available at the OEDK that were not used in excess. This includes: laser cutter, vinyl cutter, vinyl paper, clear coat, belt sander, sand paper, tape, paint, and Adobe Illustrator.
Links
