The Process

Part 1: Planning

Inspired by a recent delve into a YouTube black hole, Amanda and I decided to look into how we could bring the legendary Loch Ness monster, Nessie, to life in our art. For our midterm assignment, we decided to create a mechanical model of Nessie swimming in her beautiful lake. Using the Cam movement #138 from 507 Mechanical Movements, we planned to utilize 5 cam points along an axis to power the bobbing motion of 5 segments of Nessie’s body. The central axis would be turned by the user using a lever, which is cleverly designed to look like her tail. The bobbing segments would be attached to a stem that interfaces with a reinforced circular piece attached (and fixed) off-center and perpendicular to the single central axis. Her head would be laser cut, and remain a static attachment to the piece; this will allow her to lock in on her prey more easily.

A few of our preliminary sketches can be found below.

Amanda was able to mock this system up in SolidWorks as a full assembly, which helped us quickly optimize part dimensions and measurements.

We exported these parts separately, which were then imported on a central Adobe Illustrator (AI) file for full display and easy triplication, where appropriate.

Part 2: Low-fidelity Prototyping

Now that we had our AI files ready to go for the laser cutter, we organized all the pieces on the same artboard, and selected a large piece of cardboard to begin cutting the pieces of our low-fidelity prototype. We planned on making three of each of the following items to later glue together in post-processing stage for more part volume: all 5 cams, 5 stems, 5 “above water” body segments, and Nessie’s head. In this stage, we also printed 3 tail parts out of cardboard, 1-2 of which will be replaced by metal in the high-fidelity/final model.

After all the parts were cut out, we removed them from the laser cutter and transported them to our workstation. Here, we glued the necessary pieces together using wood glue and left them to dry. We retrieved a wooden dowel that was 0.5″ in diameter to serve as our central axis, and slipped the 5 cams along the axis, spaced 1.625-2.215″ apart (depending on CAD measurements) and secured with more wood glue. We assembled and attached the outer box pieces, excluding the top, using super glue. After placing the wooden dowel between the two circular holes at the ends of the box, we stacked the stems atop the cams. We then lowered the top of the box over the stems (making sure they went through the square holes), and secured the top with super glue. We super glued the semicircular body segments to the top of the stems, the tail piece to one of the dowel ends, and Nessie’s head to the opposite box end outer wall.

Part 3: High-fidelity Prototyping

We made some slight adjustments to the AI file after noticing some imperfections in our first prototype. Some of the first things we did was to make the square holes on the top of the box smaller in size (slits > square holes)

and added a small “tab” to the end of the tail piece. To incorporate our metal piece, we decided the tail could be a good point to include this aspect. We deleted 2 of the tail copies in our laser cutting file, with the idea to laser cut only one tail piece with a “tab” piece added to the inserted side, and plasma cut the other to later be joined with the wooden piece.

To plasma cut the other tail piece, we isolated our tail piece to its own .dxf file. After firing up the machine, we imported the file, ensured the correct dimensions, and plasma cut the steel tail piece.

For post-processing, we used the angle grinder the remove the dross and smoothen the edges of the piece. To make the rest of her body, we used hunter green spray paint to paint the steel tail piece.

Recognizing it would be difficult to glue plywood to itself at the edges of the box, we looked to other solutions for assembling the outer structure of the high-fidelity model. Using MakerCase, we created a finger joint box that was 6.5″ x 6.5″ x 12″ in dimension. After importing this file into AI, we added the necessary features to the appropriate side of the box, including the 5 stem holes at the top of the box, two circular cutouts for the central axis on the sides, and a long cutout window (where the acrylic overlay will be fixed) to a long side of the box.

With the adjusted laser cutting files, we laser cut our pieces using 0.128″ thick plywood. Similar to the low-fidelity prototype, after removing the pieces, we glued three of the same part together for the cams, stems, body segments, and head using wood glue. We clamped these for several minutes while they dried before going outside and spray painting the body parts with hunter green spray paint. For the box sides, wanting to make this look like water, we used a blue wood stain to color the wood. We retrieved another 0.5″ diameter wooden dowel and, using a handsaw, cut it to measure 12″ in length. This, too, was stained blue. After wiping off the excess stain, we left these pieces to dry for about an hour.

To create the box window, we used the laser cutter to cut a rectangle of acrylic material. Instead of leaving this plain, we decided to add waves to the acrylic piece. To execute this, thin wave figures were hand drawn in AI and cut using the vinyl cutter. Using the transfer paper, we applied the vinyl stickers to the side of the acryl window piece.

Finally, it was time for the assembly of our high-fidelity prototype! We had press-fit the sides of the box together, excluding the top and of the the short side pieces (tail end). Using a handsaw, we created a small notch in one end of the dowel, which would be used to help attach the wooden tail piece (reprinted to have a small tab that would fit into the dowel notch). We slid the cams onto and along the axis, and secured their positions with wood glue. After drying, we positioned the axis with the cams attached into the box’s side circular holes, with the notch side facing the tail end. Using wood glue, we inserted the wooden tab of the tail into the notch and allowed this to dry. We press fit the remaining side piece into the finger joints of the rest of the box, and only the top remained open. We placed the stems atop the cams and lower the box top onto the model, threading the stems through the now-larger square holes. This box top was press fit into the rest of the frame, and the body segments were glued to the top of the stems using super glue. Nessie’s head, the plasma cut tail piece, and the acrylic window were attached at the appropriate locations using super glue, as well.

Now it was time to test our prototype. Immediately, we noticed the stems traveled to much in the z-axis (plane going into the model from Nessie’s side profile). To mitigate this, we added 0.5″ x 3″ wooden pieces to the bottom of the stems, which were cut, sanded, spray painted, and super glued to the stems. Unfortunately, this did not allow enough space for the largest cam to rotate freely within the box, so the box of the model had to be disassembled so the largest cam could be sanded down to a smaller size. The sanded parts were repainted before the model was fully assembled again. Vaseline was applied to the cam surfaces and the inferior aspect of these new wooden pieces attached to the bottom of the stems, which greatly help limit friction by lubricating the dynamic pieces. Lastly, the stems appeared too long, as would often bump and catch by balancing on the edge of the square holes. To address this, we cut off the body segments, cut off ~0.75″ off the upper parts of the stems, and temporarily reattached the body segments with hot glue to ensure proper height. After confirming the new stem height, we removed the body segments once more and definitively attached them back to the stems using super glue.

After making these changes, we were able to achieve our wanted motion! And she is beautiful.Video Demo

Clean Up

Before leaving the OEDK after each work day, we ensured that our workspace was clean and in better shape than we left it.

Reflection

This project was extremely rewarding, and we learned a lot! From the start we had a clear idea of the theme and motions we wanted to see in our mechanical model. Our first drawing seemed clear but after examining the actual mechanical movements we quickly learned we were overcomplicating the options and could achieve the movement we wanted with cams rather than a complex series of gears with multiple possibilities of failure. 

During our first assembly of the low-fidelity prototype, we found that it was difficult to emulate the movements we wanted with cardboard because it was too light and the rough edges created too much friction such that the model did not completely rotate as we hoped. This issue showed that some ideas may be possible and feasible, both as a design and a high fidelity prototype, but material constraints may cause issues. 

The final assembly of our high-fidelity final prototype was another small struggle. We assumed the assembly process would be quick and easy since we had been so successful with our low fidelity model. We soon learned that in order to produce a good quality model, we had to work more slowly and monotonously. We initially thought we would have much more time to add small aesthetic details to our model, but eventually had to pivot towards ensuring the major motions of our model were functional. Similarly to the first assembly, some pieces were stiff at first and did not move the way we had hoped. We had to make some adjustments, like using Vaseline to lubricate the cams and cutting the stems down so everything would rotate as we intended. As stated before, this showed again that while an idea may work well in a CAD file or with ideal materials, it is difficult to know if a device will function exactly as you intended without using the exact materials or those very similar to them.

As a whole, this project was fun and successful, despite some minor setbacks. In the future we hope to have more time to customize the model and play with different mechanical movements.

Cost Estimate

Cost Type	Cost	Price	Source	Quantity	Total
Materials	¼” Plywood, 2ftx4ft	$17.98/ piece	Lowes.com	1 pc	$17.98
	Aluminium Sheet, 12×12”	$11.47	HomeDepot.com	1 sheet	$11.47
	Cardboard Sheet 24 x 48”	$12.19/ 5 Sheets	Staples.com	2 sheets	$4.88
	Blue Wood Stain (half-pint)	$5.48/can	Walmart.com	1 can	$5.48
	Wood Glue (4 oz)	$5.50/tube	Walmart.com	1 tube	$5.50
	Super Glue (15g)	$5.98/tube	Walmart.com	1 tube	$5.98
	Hot Glue Sticks (Mini)	$2.99/ 25 Sticks	Michaels.com	1 pack	$2.99
	Spray Paint	$5.98/can	HomeDepot.com	1 can	$5.98
	Vaseline(1.75 oz)	$3.29/tub	Walgreens.com	1 tub	$3.29
Labor	Laser Operator	$19.26/hr	ZipRecruiter.com	N/A	$0
	Machine Shop Operator	$22.61/ hr	ZipRecruiter.com	½ hr	$11.31
	Prototyping Engineer	$22/hr	Personal Wage	20 hrs	$440
Overhead	Machine TIme (Plasma Cutter)	$0.15/ kw/hr	hypertherm.com MarketWatch.com	¼ hr	$0.03
	Machine Time (Laser Cutter)	$0.36/hr	accurl.com	1 hr	$0.36
	Quality Control	$19.88/hr	ZipRecruiter.com	N/A	$0
Total Cost					$515.25

The final total cost for our mechanical model was $515.25. Since this was our first time utilizing 2D methods to make a 3D mechanical model, our prices were much higher than they could be if our process was optimized. The highest contributor to cost was our personal wage which would certainly be reduced after creating a prototype and iterating. This will greatly decrease cost although there might be a slight increase due to quality/management wages. Additionally, having dedicated employees for specific parts of the process will cause less confusion or time spent explaining choices, since actions will be preplanned. This means time will be saved and therefore pricing will be lower. Working in bulk will also help to lower costs. By buying material in bulk rather than using single packed items, the cost per piece will be much lower. In terms of material, thicker wood pieces can be used meaning that multiple cuts and gluing/sanding of layers is not necessary therefore more time and money will be saved in both materials and man hours. Finally, changing the materials and using something like plastic could be cheaper and easier to assemble; for example, 3d printing pieces or injection molding for large scale production.