Hello again! This blog post will be the longest one I’ve written yet as I will be covering all the steps I took in completing the Midterm Project. The basis of this project was to take one of the 507  mechanical movements and make a working model of it. If you can remember, one of the earlier assignments was about drawing one of these movements in Adobe Illustrator, and that same movement, as seen in Figure 1,  is the one I modeled for this project.

Since I already knew what movement I wanted to use, the next step was to figure out the general setup of the model, i.e. what the base would look like, how the user would interact with it etc. While I was sketching this out, I faced a dilemma. The original movement had two clearly documented components, the circular disk and the L slot. However, the the slot is shown to be connected to some moving bar, but it is not shown what this bar looks like, nor what it is connected to. So, after some brainstorming, I came up with the idea to use a spring, such that it can stretch to accommodate the strange movement of the L  piece, but can also fairly easily be anchored to a base. Once I had this idea, it was easy to sketch out my intended design, which can be seen in Figure 2 below.

After I had a general idea of the design I needed to make, I began creating a 2D drawing of the model in Adobe Illustrator. The first iteration of this was simply modifying the file I already had to create a completed “front” view of the device, showing the original components, the pieces that help them up, and the connection points. This original 2D drawing can be seen in Figure 3.

At this point, I had general designs for all of my pieces, so in order to make sure that the work I had completed through this point was successful, I needed to make a low fidelity prototype using  my initial designs. I chose to use cardboard because of it laser-cuttability and because its very easy to find and modify in preliminary prototypes. Before cutting however, I first added some slots to the support pieces so they could all easily connect together and I cut off the long sides to make assembly faster. Once this design was finalized, I cut it out of cardboard and hot glued it together. This first prototype can be seen in Figure 4.

Coming out of completing this prototype, I became aware of many adjustments I needed to account for. The biggest issue was that having only one spring, and having this spring placed slightly offset from the L piece mean that the L was being pulled away from the back support in a way that is not intended. This movement can be seen in Figure 5. In order to lessen these effects, my new plan was to incorporate 2 springs, one on each side of the L such that their forces in the outward direction were balanced. Besides this issue, I had made no mechanism for securing each of the pieces in specific locations on the dowels. I was able to make do with this prototype through creatively using hot glue, but I knew I needed to design pieces to serve this purpose.

Now that I had clear improvements to make, I began on my second iteration. In order to account for the placement of the support panels, the spacing and attachment pieces and the springs, I drew a side view of the model with the distances between each component clearly recognized.  In the drawing, each bar is the size of the thickness of the material used, and can be adjusted depending on the material. I first adjusted my design to be able to work with cardboard, so that I could check to see if any other components needed to be adjusted before beginning to work on the final iteration.

Once the design was finalized, I again cut it out of cardboard and assembled it using hot glue. This second iteration can be seen below in Figure 7.

This prototype was a big success (mostly). Adding the spacing/attachment pieces really made the model look more finished and helped when assembling, and the two spring mechanism worked just as intended. However, adding the second spring increased the forces felt by the L piece by a sizeable amount, so soon after it was fully put together, the L collapsed outward, out of shape. This is why there’s the seemingly random dowels glued onto the side. Besides this material limitation, the model worked very well, and at this point I was ready to move onto high-fidelity — wood.

I first began by plasma cutting the L piece. The L piece was a great fit to be laser cut because of the high forces it would be experiencing (metal is generally stronger than wood) and because I could easily paint it so that the L stood out in the final product. I used 1/8″ aluminum to provide added strength, and once it was cut, I used the wire brush and files to take off the slag, and I also sandblasted the entire piece. This L bracket can be seen in Figure 8. After sandblasting, I applied primer and two coats of teal paint to each side, letting it dry in between.

Next, I needed to focus on the wooden pieces, while the previous prototype worked well, there were a few changes I wanted to make to the design. Up until this point, I had simply glued down each of the supporting pieces onto another piece of cardboard, but this method would not work for the final design. So, I took some inspiration from the laser cut box from last assignment, and designed a two layer base where the top layer had slots where teeth on the supporting pieces could slot into. I used a tolerance of .01″ as that was what worked for my laser cut box. After adding the teeth, all of the support pieces were ready, but I also adjusted the design for the circular crank pieces such that they were also two layered, with the outside layer having a slot where the dowel will more easily be attached. At this point I was ready to laser cut.

However, this is when a semi-major issue came up. On and off toward the due date for this project, the laser cutter would stop being able to cut all the way through the wood, even after multiple passes. In order to work around this, I talked to one of the other lab assistant, and she helped me learn how to use the muse, which is a mini laser cutter we have out on the floor. While I wasn’t able to cut all of my pieces in one run using the muse, I was still able to cut my pieces, which was the important part. The only issue I ran into with the muse was that the tolerances were a bit too tight for this laser. So, I adjusted the design such that the base piece would use a tolerance of .005″ and the cut would run twice through. With these settings, the pieces fit together perfectly. These cut pieces are shown in Figure 9.

At this point,  I sanded/dremeled each of the pieces so that there were no burn marks or overlaps, and I used wood glue to attach each of the two layer components. I let that dry for a day, and then began applying a finish on the  wood before attaching everything else. For this step I used Feed-n-Wax which is a wood polish made with beeswax and orange oil. I used this poolish because the warm tone it gives compliments the teal L, and the set time for it is quite fast at 10 minutes. The polished pieces can be seen below in Figure 10.

Once the polish was applied and the excess was wiped off, I etched my name, the class, and the year into the base.  I then began assembling all of the pieces, using wood glue where needed (in the base slots, attaching the dowels where needed, etc.). After completing this, I had finished! The final device can be seen in Figure 11.

Overall, I had a lot of fun working of this project. I really enjoyed the interative process and solving the issues I faced as they came up.  I am very happy with my final model, and I’m especially glad I got to use the movement that matched the first letter of my name.

Finally, I will do a quick cost estimate. I worked for around 20 hours and at $20/hr, labor costs total $400. Material costs include laser cut table plywood, which you can purchase for $13.73 for a 4’x4’ square from Lowes. For the plasma cut piece, I found a metal supplier where you can purchase a square foot of 1/8″ aluminum for $34.36. I used two tension springs, and you can get a pack of 10 for $21.46. I also used around $5 worth of dowels. Thus, the total for this project comes to $474.55 — my most expensive yet!