Hello dearest reader,
I’ve spent the past 15 days working on my midterm project: a working mechanical model! Looking at the previous semesters’ projects, I really liked the ones were the motion had an obvious function, like a ladybug’s wings opening or a guy surfing. I decided I wanted to make mine a very familiar action as well, so I decided to do mini-golf: who doesn’t love some good putt putt, right?
I initially selected Motion#131 from 507 Mechanical Movements: a rack and pinion.
I spent the first few days of the project sketching my design up in SolidWorks, since the laser-cut box project taught me how bad I was at Adobe Illustrator. I also imported my CAD files into one large assembly, which I then applied mates to to simulate motion. This step of creating the assembly file, though it took a LOT of time, was priceless in helping me get a grasp of how slight changes to dimensions impacted the operation of my final design.
Back of SolidWorks model: rack and pinion
Front of SolidWorks model: mini golf!
A video of the moving rack and pinion in the SolidWorks assembly can be found here.
You may notice that the teeth of the rack and pinion don’t align very well in the assembly. That is because I designed them more on whim than on science. During the first checkpoint assessment in class, I was encouraged to use an automatic gear generator online to make sure my gears would mesh.
I ended up using this website to generate my rack and pinion. I laser cut the automatically generated SVG file from cardboard to make sure the website’s results were correct.
Perfect mesh from the generated rack and pinion!
I then converted all of my SolidWorks files to DXF files so I could open them as 2D vector files in Illustrator. I laser cut my entire assembly from cardboard, incorporating the tooth dimensions from the generator website.
At this stage, I also made some changes in the dimensions of holes on my prototype, as I realized it would be much more efficient to design around the dowels the OEDK had in stock. I grabbed one of each dowel in stock 0.5″ and smaller and rounded the dimensions of every hole to the nearest dowel size I had.
Since I never intended to laser cut the dowels, I hadn’t given serious consideration to how long they should have been. Instead, I assumed that if they were long enough to hold all of the pieces they needed to hold, that would be sufficient. With this train of logic, I intentionally oversized all of my dowels before hot gluing them to my cardboard prototype together.
Low-fidelity cardboard prototype: front side
Low-fidelity cardboard prototype: back side
Almost immediately I noticed a few glaring issues with my prototype.
Having unnecessarily long dowels was not the move. As you can see in the above images, all of the extra length of the dowels was on the back end of the prototype, not distribute evenly about the backboard. Because the wooden dowels were so heavy compared to the cardboard they were supporting, the long dowels would tilt, interfering with the system’s motion. In fact, the golf ball that was meant to roll down the incline wouldn’t even move at all because of this tilt.
The second issue that presented itself was the width of the long protruding bit of my pinion. After a few spins of the system, the cardboard started to bend. This was a significant problem since in this type of system, if the driving piece doesn’t stay rigid, it won’t push the next piece along.
Cardboard bending
Perhaps the largest issue of all was the fact that I hadn’t thought about the “layers” of my prototype. While I was operating in the SolidWorks sphere, it was easy to mate objects to be collinear to certain reference points to keep them in line. However, upon having to physically prototype this system, I realized I had put no measures in place to keep these parts the correct distance from the board. This issue was especially apparent at the connection between the rack and the pinion. The pinion moved freely back and forth on its oversized dowels. The rack was held in place by two cardboard U-shaped rack supports; however, these too were oversized and so the rack moved not only back and forth within these clasps, but would also tilt and rotate because its motion wasn’t sufficiently restricted. The rack and pinion both having these unpredictable motions meant it was almost impossible for the two to align and actually carry out the desired mechanical motion.
Having identified so many problems in the low-fidelity model, I decided I wanted to make another prototype before our second class checkpoint. My main goal now was to have the motion actually work, so I decided to upgrade from cardboard to wood.
I also decided I didn’t want to have to glue pieces to each other because I noticed it had greatly impeded my ability to disassemble if I needed to swap out parts in my low-fidelity prototype. Instead, I went about making a press-fit wooden prototype.
The laser-cut box project taught me well, so I knew my first step was to measure the kerf of the laser cutter. (Also can I just take this moment to say that the new laser cutter is absolutely AMAZING)
I cut a few small squares to identify the machine settings that would successfully cut. I found this to be 20% speed, 100% power, 10 frequency.
I then designed some sample joints with different kerfs accounted for and chose the fit I liked the most.
Sample joints to determine kerf
I also separately determined the kerf for concentric connections, as I had many pieces connecting to dowels instead of flat bits. Ultimately, I decided I needed 0.03″ overlap on my flat connections and 0.02″ overlap on my circle connections.
After making the appropriate changes to my Illustrator files, I laser cut and assembled my medium-fidelity wooden prototype (I actually sized out how long my dowels needed to be this time!) Because I had the correct kerf settings, the prototype was super easy to assemble!
Medium-fidelity wood prototype: front side
Medium-fidelity wood prototype: back side
This time, the rack and pinion motion actually worked! I increased the width of the long protruding part of the pinion that had bent in the cardboard version, and that helped a lot. That being said, there were still few issues encountered.
While I had cut down the dowel on the golf ball, I had accidentally made the inclined hole it travelled down a press-fit as well. Needless to say, the ball didn’t roll.
You may also notice a white spot on the rack in the image. This is where I sanded off one of the teeth of my rack because it was hitting up against the support. (If you’re wondering why I sanded this down instead of reprinting a new modified rack, it’s because prototype checkpoints were in 10 minutes and both laser cutters were occupied). In my cardboard prototype, I had simply hot-glued my rack supports where I wanted them on the board; however, this time, I thought ahead and cut out little rectangular holes in my backboard where I wanted to press-fit the supports. I guess the location of the support changed slightly in this process, which is what cause the rack tooth to start hitting up against it. Sanding off this tooth solved that problem.
You may also notice some duct tape on the ends of my dowels. I realized I needed some means of making the moving pieces stay on their dowels. For this medium-fidelity prototype, quickly applying a few layers of duct tape took care of that problem, but looking towards the future, I decided I was going to laser cut press-fit dowel caps.
The last issue I noticed was more of an aesthetic one. My mini-golf game wasn’t that engaging, and there was a lot of empty space on my board. I decided I wanted to add a windmill that would also be turned by the original crank motion.
Because I wanted to fill the empty space on the left side of the board, I placed a large gear between my original crank and my windmill gear. I also wanted the windmill to spin multiple times per “game” of minigolf, so I incorporated a 3:1 gear ratio to obtain this difference in speed.
While I originally planned to add the driving gear of the windmill system onto the crank dowel, I soon realized that I could swap out the driving circle of the rack and pinion for the gear, as long as I placed a hole in this gear to insert a dowel to drive the pinion. By doing this, I was able to drive both mechanical motions using only one piece opposite the crank.
Unfortunately, in turning the circular piece into a gear, I had to move the smaller hole closer into the center so it wouldn’t interfere with the meshing of the gears. I couldn’t simply increase the size of the gear or it would hit the fixed dowel of the pinion. This slight change in placement of the dowel driving the pinion ultimately required the operator to use an unexpectedly large amount of force to drive the system.
The next step was to sketch up my windmill system pieces in Illustrator. This was also when I finally sketched my dowel end caps to stop pieces from moving away from the board. To stop pieces from moving in towards the board, I decided to cut some pieces (like my rack) multiple times and glue them together, so the outer piece would mesh with the pinion). I laser cut these new pieces along with the pieces from my medium fidelity prototype.
These are all of the wooden components of my final project!
There are a few slight changes from the medium fidelity pieces to the final ones: the slot for the rack-driven golf club is smaller since I realized the slot only needs to span the golf club’s path, not the rack’s; a few more holes have been added to hold the windmill system gears; and the dowels are equally and intentionally sized.
Since this was the final project, I had aesthetics in mind from the start. Before assembling the pieces, I decided to stain all of them mahogany. (This was my first time using stain and it was surprisingly easy!) After staining a test piece to make sure I liked the color, I stained all of my pieces (including my dowels cut to size).
Stained pieces!
I left the pieces to dry overnight before assembling. At this point, I had enough wooden pieces to complete the entire assembly, and I hadn’t yet decided which wooden piece I was going to swap out for a waterjet cut aluminum piece.
My first thought was to make the driving gear of both mechanisms (the one that was originally a circle) the waterjet cut piece; however, when I attempted to assemble I realized I had not measured the kerf of the waterjet cutter. I had instead incorrectly used the same measurements as I had for the laser cutter. As it turns out, the kerf of the waterjet cutter is significantly less, so the gear did not fit tightly onto the dowel but instead there was a visible gap. I tried to resolve this by adding a dowel cap which I epoxied to the surface of the aluminum gear in an attempt to make the crank turn the metal gear; however, this dowel cap improvisation meant my pinion was resting upon the dowel cap on the gear and not on the gear itself. The pinion’s long side would slide off this protruding cap and not be able to complete its cycle since it could not get back on.
Pinion long piece rests on dowel cap instead of on metal gear
I realized I was better off switching my metal piece over to the middle gear of the windmill system instead (two of the three windmill system gears were the same size, so this did not require cutting any additional pieces). However, because I had already epoxied the metal to the dowel and adhered the dowel to the backboard, it was impossible to remove. I resorted to drilling out the dowel until the metal piece popped off. This process left a hole in the dowel that I then filled with wood filler and re-stained.
Hole in dowel that was later filled with wood filler
Speaking of drilling, I also encountered a different issue that also required drilling to solve. I had accidentally placed my hole for the windmill dowel in the backboard too far from the other gears, so the windmill system gears didn’t actually mesh as desired. Because I had already stained my backboard and assembled so many pieces onto it, I was reluctant to start over from scratch, reprinting the backboard and all of the attached pieces; just the thought of the wasted time and wood made me shudder. Instead, I decided to drill the hole for the dowel where I wanted it to be in the backboard.
Hole vs marked dot where the hole should be
Clamps set up to drill new hole
I didn’t bother filling in this hole with wood filler, since it was too large and it was mostly covered by the windmill arms anyways.
In my attempt of disassembling my mechanisms to swap the metal and wooden gears, I ruined my press-fit dowel caps, so I returned to the laser cutter to reprint those necessary pieces. I then had to stain them so they would match the rest of the assembly. If only I had had enough forethought to print and stain spare parts at the start…
Laser cutting the replacement dowel caps
Let me take a moment to talk about adhesives. I learned a LOT about adhesives during this project. Between the different fidelities of prototypes, I used hot glue, super glue, wood glue, and epoxy. Most of the time I referred to thistothat.com to determine which adhesive to use; however, there was one instance when I was trying to glue two small wooden pieces to each other and, even though this website told me wood glue would work, the two pieces would not stay stuck. After reading up about other people who had had a similar issue on Google, I learned that super glue was actually better for small connections, so I made the switch. It also saved me a lot of time and effort since super glue dries much faster!
Since my medium-fidelity prototype, I shortened the rack a bit because I didn’t want it to protrude past the edge of the backboard and be visible to the user during motion. As a result, it was now just short enough to fall out of its support every cycle. I experimented using some taped cardboard underneath the rack to support it and hold it in place. Upon seeing this method worked, I ended up replacing this cardboard with wood. I also ended up adding a similar guide rail on the top.
Cardboard piece taped under rack to keep it from falling down during motion
The device was finally fully assembled! A video of the motion can be found here.
You will notice the motion was very rough. There was a lot of friction between my gears and sliding wooden pieces, so I applied lubricant. At the time, the only lubricant the lab assistants were able to find was from the machine shop, so it was actually meant for heavy duty machines, but it worked great for me.
Lubricating the gears
Now that my device was functional, I devoted the rest of my time to making it more aesthetic. I designed stickers to fit over my windmill arms so they would actually resemble a windmill while spinning, as well as a white circle to make my ball look more like a golf ball. I then cut them out with vinyl cutter and used contact paper to remove the unwanted pieces.
Shape Building the Illustrator file for the windmill arm stickers
Using the vinyl cutter
Contact paper for transferring the sticker
I decided to make a large nameplate to fill up the empty space in the bottom right corner of my board. I liked how the engraved text matched the mahogany stain of my board and the contrast between the unstained and stained laser-cut wood, so I chose not to stain the nameplate.
Laser cutting the nameplate
I loved the dark stain I’d chosen, but it didn’t really match the fun and engaging mini-golf vibe I was going for, so I decided to paint a landscape onto the front with bright colors. Because my stain was so dark, the only paint I could find that would show up was spray paint, but the idea of pointing a can of spray paint at my beautifully stained wood terrified me, so instead I sprayed the paint into large pools on scrap cardboard and then transferred it over to my project. While I started out doing this with a brush, I didn’t like the brush lines, so I ended up finger-painting with my gloved hands.
Applying the wood stain, the lubricant, and now the spray paint with my bare gloved hands amounted to a lot of finger-painting!
Taping off sections for spray paint in an attempt to get clean lines
After all of the finger-painting, I decided I wanted to add a little note for people walking by to let them know they had my permission to interact with my device, so I vinyl cut a little “Spin Me!” note that I stuck onto the crank. Unfortunately, I chose to do this in red, thinking it would match nicely with the windmill body and flag. Instead, it came out barely legible.
Final project: front side
Final project: back side
A video of the final project in motion can be found here.
Cost Estimations:
Labor: 40 hours * $10/hr = $400
Wage based on OEDK Lab Assistant pay
0.09″ x 12″ x 12″ Aluminum Sheet: $9.72
3/16″ x 12″ x 48″ Wood: $17.70
1 Large Cardboard Box: $3
Vinyl: 0.25 square foot * $7/ square foot = $1.75
1/2 Pint Mahogany Gel Stain: $11.13
Waterjet Time: 1hr * $24/hr = $24
2 Cans of Spray Paint: 2 * $4.98 = $9.96
Lubricant: $7.15
Wood Glue: $3.98
Super Glue: $2.86
Epoxy: $14
Hot Glue Gun with Glue Sticks: $12.49
Latex Gloves: $19.99
Estimated from reported cost of abrasive * 4.33 (since abrasive was 75% of total hourly cost)
Laser-Cutter Time: 1 hr * $15/hr = $15
Vinyl Cutter Time: 0.5 hr * $15/hr = $7.50
Note: I couldn’t find an hourly rental rate for the vinyl cutter, so I just assumed the same as laser cutter
Adobe Illustrator: 1 month * $20.99/month = $20.99
SolidWorks: 1 license * $810/3 month license = $810
Note: 3 months is the shortest SolidWorks license available
Total (excluding software subscriptions): $560.25
