General
Figure 1. Fin Support Issue
One factor that should be considered in the final assembly is ensuring the PVA support material sets on the first layer of the PVA. We had trouble sometimes with the small fins getting enough support material and sticking underneath them in the correct place. This may be due to outside conditions like humidity and not the ultimaker. Still, the ultimaker performed as necessary and gave the best print for all of the sides with the only minor hiccup being the inconsistent support of our small fins.
Joystick
Figures 2 and 3. Final Joystick printed with TPU
For the joystick, there were many iterations that were made in order to get to the final product, but the result ended with a joystick that was not very flexible. At first, only a 20% infill was used so it had the flexibility to move, but had problems with delamination. The second iteration used 100% infill but then the joystick was so rigid it broke off fairly easy along the base. The last iteration was in the middle, using a 50% infill, chamfered top and bottom of the stick and a 2 mm shell size to increase the strength. While the side does work, it is not as flexible as intended and could use some additional work to have less infill, more height in the CAD model, and possibly adjusting the shell size for better fidgeting functionality. Currently this is the best working setup with the TPU which had its own set of problems.
The TPU had an issue while printing where the PVA was not supporting the TPU properly so TPU was just used as the support material. While it gave the best print quality, there was a lot more post processing that was needed to remove the supports. For our purposes this worked the best and would be recommended for future prints with TPU. Unless there was another printer than can use flexible material on campus, the FDM ultimaker is the quickest and most accessible option for this print with just some minor adjustments that can be made to have the best result.
Dumbbell Slider
Figure 4. The Functional Sliding Dumbbell Side
For the final iteration of this side of the cube, the space between the dumbbell and the housing was increased, and the infill density was increased. In the previous iteration, the flat, bottom side of the dumbbell fused to the housing, and when it was moved, the dumbbell broke in shear. To compensate for this, extra layers of soluble support (PVA) were added around the figure. To further avoid this issue, the infill density was increased from 20% to 75%. The print time rose from 5:20 to 5:45 (only 25 minutes) and the increase in printed material was almost negligible, so this was considered to be a justified fortification.
Button Side
The idea for this side was to give someone who likes to click pens something to fidget with while not creating a noise that some people find annoying and potentially aggravating. The original design was to utilize the dual extrusion capabilities of the Ultimaker 3 and print a portion of the design with PLA and the actually buttons with a flexible material, TPU. Through a couple of tests doing this, it was found that printing TPU is more difficult. Because of multiple failed attempts with printing PLA and TPU within the same print, the design was then changed to utilize just PLA. This design was successful in being manufactured, but the buttons were extremely stiff and took away from the original idea of the design. With the reduced functionality of the buttons, a new design was looked into. This new design would be able to be printed with the buttons separately from the face plate and then assembled in a way that they could not come apart. This design used a “pin” and “barb”. It can be seen below how the two pieces fit together in the end and how the barb overlaps onto the buttons securing them in place.
Figures 5 and 6. New design of the buttons.
Roller Ball Assembly Side
For the final iteration of this side, the clearance between the ball and the back plate was increased to the maximum possible while still containing the ball within the back plate. Up until this print, every print prior had resulted in the ball becoming fused with the backing plate. At the end of this print, the ball did not appear to be fused, however, it did not roll. As a post processing process, a drop of acetone was placed in the visible gap. Acetone was used because it does not dissolve the PLA print material that was used. This drop allowed the ball to rotate freely in the backing plate. It is believed that the acetone acted as a lubricant and allowed enough reduction in friction for us to break free any imperfections and create a working part.
Figure 7. Final print for the roller ball.
Conclusions
After many design iterations and trials and failures, the Ultimaker 3 3D printer was found to be a great option for manufacturing this Fidget Cube. Multiple sides of the cube as well as the frame required removable supports, and the printer’s ability to extrude the water-soluble PVA made this possible. Furthermore, this printer is able to extrude a variety of materials, such as the flexible TPU, which made possible the creation of the flexible “Joystick” side. The tolerances and quality kept by the printer were of a high caliber, and cumulatively it made the engineering process of this product a success. If all additive manufacturing technologies were available for use, the best process for most of the parts would be some form of vat photopolymerization like CLIP. The reason for this would be not weakness between layers and also the ability to print with different materials to achieve the best properties possible. Utilizing the CLIP printing technology, it would also be possible to greatly reduce the print times for the product.
Figures 8 and 9. The final assembly of the fidget cube.