Before arriving at our final product, our mechanism required several stages of rapid prototyping to test tolerances, fine tune interfaces, and evaluate material choices in an effort to minimize the amount of friction and slop present in our mechanism. We utilized the following manufacturing technologies to significantly reduce the turnaround time between design revisions, allowing us to rapidly prototype all subsystems in a matter of two or three days.
Manufacturing technologies utilized:
- FullSpectrum 48x36" 140W CO2 Laser
- Bambu Lab X1 Carbon FDM 3D Printer
- Raise 3D Pro3 FDM 3D Printer
- Horizontal Bandsaw
At first, we decided to laser cut all of our linkages out of ⅛” thick plywood for our first prototype. Unfortunately, because we had several cantilevered linkages, we found that ⅛” plywood was far too flexible and would result in sagging, out-of-plane linkages. For our final product, we decided to switch to ¼” thick acrylic. We found that acrylic was noticeably more rigid than its plywood counterpart, and the extra thickness (¼” instead of ⅛”) significantly reduced the sagging as well. A downside of using acrylic instead of plywood is its tendency to crack and shatter when subjected to internal forces. We had planned to press-fit flanged bearings into these linkages, but the brittleness of the acrylic meant we would need to be incredibly careful when doing so. To account for laser perf, we found that under-dimensioning each ¼” acrylic part by 0.2mm from nominal dimensions was perfect in allowing us to press-fit in a bearing without cracking the acrylic.
While most of the parts we used on the final assembly were manufactured using the laser cutter, there still were quite a few parts that we 3D printed using a personal Bambu Labs X1 Carbon printer and the Raise 3D Pro3 printers supplied by Texas Inventionworks. Many of the parts we 3D printed were critical interface parts, and they were crucial in creating the joints between two linkages by holding one linkage stationary relative to that shaft. These parts, shown below, went through several iterations to ensure that they had enough clamping force to prevent shafts from slipping while remaining as compact as possible. We found that most joints required very little clamping force, so smaller interface clamps (shown on the left) were used. However, we found that the locations transmitting high torque, such as the driven gear and input linkage, were prone to slippage and therefore required significantly more clamping force. Thus, after several iterations, we arrived at a dual-part clamp design with several stages that gave us a much more robust interfacing solution for high-torque applications (shown on the right).
In addition to interface clamps, we also utilized 3D printing for parts such as our paper/ink pad holders, as well as support brackets and feet for our “shelves”. We opted to use PLA for all of our parts and we ensured that all critical ID dimensions were given a + 0.3-0.4mm tolerance from nominal to account for the layer expansion of each FDM 3D printer.
Finally, we decided to use a horizontal bandsaw to cut our steel shaft to length. It was critical to ensure that our shafts were cut to proper length, as the assembly allowed for minimal error in the shaft lengths due to how compact the mechanism was. Furthermore, it was critical to ensure that the shafts came out perfectly straight, as any bend would render them useless for our application. Finally, we used a belt sander to debur the edges of each cut to ensure that each shaft would be able to snugly fit inside the bearings.
For assembly, we used M3 hardware for every part and 6mm diameter shaft for the joints. We felt that M3 screws were more than robust enough for this relatively low-stress application while also offering a smaller package size than their M4 counterparts. We designed each interfacing part with assembly in mind, ensuring that the assembly process would be as pain-free as possible. First, we used captive nut inserts and counterbored cap head screw clearances for every part. The captive nut inserts, when dimensioned correctly (~5.8mm diameter for M3 nuts), ensured that nuts would not spin freely when captured, and therefore guaranteed properly torqued screws without needing to reach into awkward locations with a pair of pliers. Furthermore, to eliminate virtually any out-of-plane slop in the joints, we created a small collar that would clamp behind the very bottom linkage at each joint. These collars ensured that each linkage was fully constrained vertically. Lastly, we also created small, PETG “caps” for the end of the steel shafts in cantilevered joints. Despite using ¼” acrylic, we found that we still had a considerable amount of sag in cantilevered linkages. To remedy this issue, we extended the steel shaft in these cantilevered joints to be in contact with the base plate. Unfortunately, this constant rubbing between the wood baseplate and the steel shafts introduced a significant amount of friction into the system. The PETG “caps” were placed on the end of these steel shafts to glide along the baseplate, which we also changed to acrylic to minimize friction.