For our ENGI 210 midterm, we were tasked with creating a physical mechanical model, based on one of the examples of the 507 Mechanical Movements. I chose the “ordinary crank movement”, which was the movement I drew for my 2D drawing assignment. In the initial stage of the project, I struggled with brainstorming an idea that involved this movement, but still had an element of creativity. I was interested in incorporating sound or music in my model, and found inspiration in a piano’s mechanism. I decided to use the idea of a piano hammer, moved by the crank movement, to hit a string, key, bell, or some object that would produce sound.

Stage 1- Low and medium fidelity prototypes

I already had the drawing for the crank drawn on Adobe Illustrator from my 2D drawing assignment, so my first step was to laser cut the drawing onto cardboard. I wanted to identify if there were any mistakes with the scales and dimensions of the parts that I didn’t realize on the computer. 

After laser cutting the parts, I assembled them using hot glue and hard straws.

From this first prototype, I realized some flaws in my drawing. Primarily, there were too many unnecessary cuts (especially in the wheel), and I was missing a few pieces, such as a support for the wheel or the hammer. 

I edited the drawings to include a support for both the wheel and the translational movement part of the model. With regard to the piano hammer and the musical key, I prototyped without a drawing. My idea was to have a hammer on a hinge (separate from the crank arm) that would be pushed into a suspended musical key.

From this prototype, I made some adjustments to the drawing. One of the main changes was extending the length of the crank arm. As I tried to replicate the crank movement, I noticed the parts kept getting stuck. I realized this was because the arm of the crank was too short and didn’t allow for the square piece in the non-wheel component to fully extend. By lengthening the arm, I was able to completely turn the wheel and fully extend the sliding square piece.

I was also skeptical about the effectiveness of having a hinged hammer from the base of the model, so I began to consider alternative ways that I could make the hammer hit a key. I decided to have the hammer suspended from a support.

My fourth iteration of the prototype addressed the flaws in the previous prototypes through corrected drawings for the parts: 1. I enlarged the moving squares attached to the arm to prevent the parts from moving around too much and getting stuck. 2. I extended the length of the support for the translational movement all the way to the base of the model. 3. I made slits in the base to correspond to the dimensions of the supports so that they could be directly inserted into the base.

The movement in motion.

Although this prototype demonstrated mechanism worked well, I still needed to further develop the ‘musical’ component of the model. I decided to keep the idea of a suspended hammer, but decided to change the idea of a key for a bell. I thought a bell would work better than a key because less force would be necessary to hit it in order to make sound. With regard to the hammer, I struggled in thinking about how to build and position it for the following reasons: 1. I had to find the right distance between the sliding piece, hammer, and the bell (so that the hammer would be close enough to be hit and the bell was in the range of motion of the hammer). 2. I had to think about ways that the hammer would remain upright before the sliding piece reached it. 3. I had to think about how to make the hammer return to the initial position after it was hit (I figured using a spring would be the best way to go about doing this).

Additional challenges I faced with this model were that the slits in the base were too small, I was missing ‘stoppers’ on the dowels to prevent the parts from sliding off, and I found that I had to lower the height of the hammer’s support. However, for the most part, the model seemed to work correctly.

Stage 2- Creating the parts for the final model

Since I had a working model, I was ready to move onto the creation of my parts for the final model. My first step was to use cut out the wheel in aluminum 0.045 inches thick, which I chose to do using the water jet cutter. The cutting went fairly well, but I had to run the cut twice since the water jet didn’t cut through the material all the way the first time. 

I post processed the wheel by first angle grinding it to remove some excess material and later sand blasting it. I wanted the wheel to be shiny so I decided to polish it as well. The polishing process turned out to be much more tedious than I had anticipated. I started using the polisher in the machine shop, but the material kept releasing a black color (Adulfo realized it might be because of aluminum’s low melting point). I tried applying a polishing paste on the wheel to see if it was more effective, but the difference was minimal. I reverted back to the polishing wheel and found with patience, I could polish part of the wheel and then wipe off the black with a cloth. Even though I spent approximately 40 minutes polishing the wheel, I wasn’t able to get the shiny look I was hoping for. In hindsight, I would have preferred to cut the wheel in steel rather than aluminum.

My next step was to cut a part for the crank in aluminum using the plasma cutter. The cutting process was quite straightforward. For the first two cutting attempts, the machine’s current was set at 45 amps and I chose to cut using a voltage of 120 then. I realized that the current was too strong because it was destroying the holes for the part. In the third cut, I lowered the current to 30 amps and used a voltage of 120.

Like with my water jet cut piece, I also angle grinder and sand blasted this piece. I considered polishing it as well, but decided against it because I hadn’t as successful with the wheel and I figured the matte texture from the sand blasting would add more texture to the model.

The final parts for the model were all laser cut in wood using the cutting settings 3 speed, 10 frequency, 100 power. I didn’t struggle with this part of the process, because I had previously done test cuts to ensure I was using the correct cutting settings.

Stage 3- Assembly and finishing

Since I had all of my parts ready, I decided to assemble them to make sure they fit together correctly.

Everything fit well, so I moved on to paint the parts. I’m a big fan of the neoplastic art movement so I took inspiration from there. I used Adobe Illustrator and my 2D drawings to test potential color combinations. I liked that the model seemed almost like a child’s toy.

For the actual coloring, I used a combination of acrylic paint and spray paint depending on what was available at the OEDK. When spray painting, I had to redo several pieces because the paint kept bubbling. I realized this was happening because I was holding the can too close to the material.

Once the parts were all painted, I began assembling them. At this point, I ran into a problem because the paint had thickened the parts, causing them to no longer fit together. This was particularly noticeable in the translational sliding part. From the friction, the sliding part had also stained its support. I recut the parts and repainted them. This time, however, I didn’t paint the inside of the sliding piece. 

The rest of the assembling was simple, but simply took time. At times, I struggled with making the wooden dowels fit through the parts, but I resolved this issue by sanding down the dowels with sandpaper. I used wood glue for attaching wooden parts together and epoxy for attaching wood and metal parts together. Lastly, although the sliding piece was able to move pretty well, I added vaseline to the support to make the movement smoother and easier.

Cost Estimate

The final step was to complete a cost estimate for the project:

Labor

• I worked for 26.5 hours, assuming a rate of $15/hour = $397.5

Materials

• Plywood: ¼ in. x 4 ft. x 8 ft. sheet sold by Home Depot for $28.82. I used approximately ¼ of that amount, so $28.82/4= $7.205
• Aluminum: 36 in. x 36 in. sheet sold by Home Depot for $31.98. I used approximately ⅓ of that amount, so $31.98/3= $10.66
• Dowels: 3/8 in. x 3/8 in. x 48 in. sold by Home Depot for 98 cents a piece. I used approximately 2, so 0.98 x 2= $1.96
• Paint: One can of spray paint sold by Home Depot for $4.27. I used approximately half a can, so $4.27/2= $2.14

Equipment

• Laser cutter used for 2.5 hours, water jet cutter used for 50 minutes, plasma cutter used for 40 minutes= approximately 1.3 hours of machining. Assuming $15/hour to use the machines = ~$20

Total cost: $397.5 + $7.21 + $10.66 + $20 + $1.96 + $2.14 = $439.47

Done!

After spending the last four days almost exclusively working on this project, I’m glad to finally be finished. There are aspects of the model I would still like to improve, but I’m happy with the end result.

The final functioning model.