This section focuses on the refinement of our manufacturing process. It begins with a critique of our preliminary method and then transitions to our final process. Lastly we conclude with our lessons learned.
Preliminary Method
Our initial method involved laser cutting acetal delrin (Fig 4.) Acetal Delrin sheets (12"x12"x1/4") for our long, planar links. We reserved 3D printing was reserved for parts with intricate geometries such as our custom gears.
<CAPTION: PRINTED W/ IN FILL OF ?? AND RES ?? USING ?? PRINTER>
. Those of which, could not be readily fabricated by the means we had at our disposal. These parts included the modified spur and bevel gears, and the mount for our gripper. See Figure 2 below.
Upon inspection of our laser cut links, we discovered visible heat damage and material expansion. See the image belowFigure 1. (Below)
1. These Photos show the lack of precision and visual aesthetic that was achieved when using laser cutting.
Furthermore, the laser cut holes in our delrin Delrin were slightly angled and lacked concentricity. This complicated the press fits on our rods and bearings while adding additional alignment issues. We did expect this complication; however our method for expanding those holes with a drill press did not yield desired results. The accumulated error from this process prevented assembly of our first prototype.
2. Initial Gear Designs and Prints
After this prototyping mishap, and considering time constraints, we decided that instead of making a "final prototype" , that we would make a "working prototype" and this proved to be the most beneficial approach. Though we changed our approach, we still needed a means for appropriately cutting our stock material.
Final Method
Upon reevaluation and much consideration, we concluded that CNC milling is would be the best method to machine our parts. Not only is the CNC mill more accurate and precise than the laser cutter, but it also did not leave burn scars marks or expand our material. Moreover, the precision offered by the CNC mill enabled us to design for interference fits which allowed us to press fit our rods and bearings securely into each specified link. See Figure 5. And although we changed our cutting scheme for the links, we decided that the gears that we had made in the preliminary stages would suffice in terms of strength and ductility. The only modification that we made was adding more teeth and changing the geometry of our spur gears to provide for a smoother motion.
3. 3D Print Setup - All our gears as well as our parts for actuation were printed in PLA at the highest possible quality
After we knew that we had all of our links and gears appropriately designed and manufactured, it was now time to start looking into assembly and testing.
4. Special Care had to be taken when manufacturing the Pins as we did not want them to interfere with other links in the assembly. With this in mind we used a drop saw to make the initial cut and then used the manual lathe to remove material and ultimately achieve our precise required lengths.
For our motor mount, we knew that there would be a considerable amount of torque imparted on the motor. Therefore it needed to be sturdy and able to withstand such torque. Extruded aluminum from 80/20 corp. allowed us to achieve this goal. Furthermore, we gained full adjust-ability in terms of placement.
5. Motor Mount using Extruded Aluminum from 80/20 inc.
We would like to thank and acknowledge our classmate, Michael Bettatti for lending us his expertise and advice in CNC milling, holding extra machine shop hours, and just answering general machining questions we had. By doing so, he facilitated our manufacturing , design, and assemblyprocess.
Lessons Learned
There were a variety of challenges we encountered during the manufacturing process. The most infuriating was receiving bent stock delrin sheets. We quickly learned that our full assembly would only be as good as our raw parts; and unfortunately, the effect of the bend carried through the entire design. Had we set out to CNC our parts initially, we would have opted for aluminum sheets instead. Additionally, the 3D printed gears proved to be weak parts and susceptible to chipping and bends. Furthermore, the printed part tolerances were variable between printers and between print jobs. We could have eliminated the need for 3D printed parts had we selected standard gears off of McMaster and designed our project around them.
5. Preliminary Method: Laser Cutting
6. Final Method: CNC Milling