I’m sure that after my last blog post, you’re waiting with bated breath to see how I could possibly top it in this one.

And I guess you’re going to be disappointed because I’m not even going to try.

Because that last post – although I am still ridiculously proud of it – took about 4 hours, 30 pictures, 4000 words, and an 83-page Google Doc containing every Marianas Trench lyrics ever written. And considering it’s currently 2:45 in the morning, I just don’t have that kind of motivation in me at the moment.

So I’m going to give you this confession instead.

Confession: I was terrified of this project. I had been ever since I saw all of the midterms sitting neatly on the ENGI 210 table first semester freshman year. I was supposed to be able to build one of those? A working mechanical model? On my own? From scratch? And it was supposed to work?

So I had been dreading this for a while. And when Dr. Wettergreen assigned it, I knew what was coming, especially from all my other friends who had taken this class previously. I knew I was in for long nights and long days spent fighting for the laser cutter and crying in frustration when it was 12 hours before the deadline and my prototype didn’t work. It’s fine I’m fine.

I didn’t want to do this project. And considering Dr. Wettergreen assigned it when I still had two other outstanding deadlines for him, it took about a week for me to actually touch the project. That was probably a mistake, but considering the amount of time I spent over midterm recess trying to finish both my box and my plasma-cut throwing diamond, I think it was justified.

When it came time to actually start thinking about the project, I originally planned on doing the gears I had designed for my 2D drawing assignment a couple of weeks prior. I had spent extra time cutting cardboard prototypes during that project making sure it actually worked, so I figured that I had a leg up on the project by starting with a model that I knew semi-worked.

That being said, when Dr. Wettergreen got wind that I was attempting to do this model, he tried desperately to get me to change my mind, mainly because those gears are so exotic that we actually don’t know anything about the dimensions that the teeth need to be to get it to work. I wouldn’t listen on a point of pride, but after speaking to a couple of other past ENGI 210 students who told me that, no, Mechanical Movement #38 is actually physically impossible because the animation literally phases through itself, I decided I wasn’t in the mood to break physics and began looking for something else to attempt.

I really love the aesthetics of gears and steampunk mechanism, and although my friends had told me to avoid them, I still wanted to work with them, especially because most of the available animated mechanical movements involved gears. So I dug through the 507 mechanical motions website again and Dr. Wettergreen came to talk to me about choosing a project. This led to the following exchange:

Dr. Wettergreen: “At this point, you need to just make a decision and commit to it. I’ll give you until 2:00 t0 decide.”
Me: *Glancing at clock* “Um, it’s currently 1:47.”
Dr. Wettergreen: “Yes.” And he walks away.
— Seven minutes later —
Dr. Wettergreen: “So what do you want to do?”
Me: “Sleep.”
And with that, he got the idea that I was not in the mood for this right then, and he let me be.

However, I did find Mechanical Movement #429 online, which called to my love for steampunk gears because it’s a design for a steam-powered double elliptical engine. This was similar to my original plan because it still utilized exotic gears, although these were less crazy because they were mostly normal with some interesting indentations. Upon seeing them, I had a cool idea for making a space-themed model with gears painted with the sun and moon with metal axles shaped like stars (cue memories of Caz’s Solar Lunar project), so I decided that this was the movement I wanted to pursue.

So now that I had a movement and a design in mind, I had to actually start designing it. I followed basically the same processes that I did for the 2D drawing activity, but this time I got smart and used a template from Gear Generator. I then ported it into Illustrator to add the large spokes and the divots in the gears. This process involved using a lot of the Shapemaker tool and also physically cutting out a screen capture of the animation to try to figure out how many teeth were actually on these gears. It also involved a lot of looking nervously at the computer while attempting to trace imprecise dimensions and talking to Caz, who was at the computer next to me trying to casually revolutionize the 3D printing industry.

Tracing the gears was overwhelming (partially because I was so sleep-deprived when trying to do it), but I knocked out the majority of the base design in one night. This is also definitely not the most important part of this project (and this happened three weeks ago at 2 in the morning, so I don’t remember a lot of these details), so I’m going to skip over most of the details of this because I talked about this tracing process a lot more in a previous post.

At long last, I had a base design for the gear. It wasn’t perfect, but it was enough that I could make a drawing to show to Dr. Wettergreen the next day. It was also enough to attempt to laser cut it out of wood, make a sort-of working prototype, and knock Dr. Wettergreen’s socks off the next day at the progress evaluation.

So the next day in the computer lab, Dr. Wettergreen and I were chatting about my magnum opus – a certain blog post in which I got petty revenge for not being in a music video by spending a large percentage of time quoting relevant Marianas Trench lyrics – and he made the throwaway comment, “If it makes you feel better, I was in a music video once. It wasn’t all it was cracked up to be.” I just kind of stared at him in shock, so he elaborated. “Yeah, I kind of had to look like I was having the time of my life for 30 seconds at a time… about 40 times. Not the most enjoyable way to spend my time.”

I just kind of stared at him in shock, but Christina at least had the state of mind to ask the important questions: Title? Artist?? Details??? And amazingly enough, he actually told us the title and artist of the song and walked away.

So Christina immediately starts googling, and she very quickly finds the music video. She watched through it, at one point pausing it and tapping my shoulder. “Hey, do you think that’s Wettergreen?” She pointed at the image of a man whose head was perfectly cut out of the video frame. “He’s wearing a blazer, so it must be him, right?”

I glanced up from paint shopping on Amazon to see what she was pointing at and nodded affirmatively. “Yup. That is definitely Wettergreen’s neckline.” And then I went back to shopping for paint for my project because at that point, I fully intended to paint my wood and the OEDK paint selection leaves much to be desired.

A couple of minutes later, Dr. Wettergreen came back around to our spot in the computer lab. “Ladies, how are you progressing?” And, although he undoubtedly meant ‘how are you progressing in fleshing out your designs for this project?’ Christina immediately pulled up the video, found the shot of decapitated-blazer-man, and pointed at the screen.

“Is this you?”

Dr. Wettergreen looked closely at the screen, chuckled, and admitted that it was him. Then he got a wistful look in his eye and commented, “Yeah, that wasn’t their best song. I should know. I managed them for a while.”

And then he walked away.

I paused. “Wait.” When I got no response, I turned around to face him. “Wait-”

And I swear to god, Dr. Wettergreen made direct eye contact with me, smiled mysteriously, and walked out of the computer lab, never to be seen again.

Okay, that’s a lie. I saw him in class two days later when I was frantically trying to get my wooden prototypes to work. But every other part of that story is 100% true.

I tried a couple on scrap wood pieces (wood is a much better material for testing gears than cardboard because the corrugation gets caught on itself), and when each one failed, I tried to fix the file and try again. I may or may not have been making progress, but it was slow going until I looked back at the file more closely and realized that my hand-drawn spokes and divots weren’t really symmetrical on the gears. So I took a day to fix that by copying the gears on top of themselves and using Shapemaker to make them rotationally symmetric. When I cut that model out, I noticed improvements, but they still weren’t turning perfectly.

Part of the problem may have been that I was trying to use masking tape to tape wooden axles to a wooden base, which is generally a bad idea because it slips. I tried to mitigate that by cutting holes in the the wooden board, but if I were off by mere millimeters, then my gears wouldn’t work. So I started testing it by just taping the axles directly to the table. And that worked a lot better, especially once Kelly suggested that I sand down some of my large spokes because they were getting caught in the divots. Once I did that, I had a pair of gears taped to the table that worked perfectly. I could move on with my life and focus on the aesthetics of my project because the mechanical motions worked great!

Of course, that’s when I got cocky. If I could get two gears to work, then could I get three gears to work? Pride was part of the motivation for this, but I admit that the allure of more moving parts – and thus a higher score multiplier – allured me, especially because I had a mere 2 gears in my project while other people had like 10 extendable rods that contributed to their score multiplier.

So I tried it with three gears, again taping them to the table and sanding them down because if it worked for two gears, then it would work for three, right? Obviously, there was no way this could go wrong.

But go wrong it did. For some reason I couldn’t get the gears to move as smoothly when there were three of them. This is probably because I now had to worry about double the points of contact, so if I did have issues in my design, even if they were minor with only two gears, those issues would magnify when another gear was added. It was also probably because I had nothing holding the gears to their axles, so they would tilt when being tested. It was also probably because I didn’t gauge the distance between them correctly when I taped them down. So, to fix this, I followed Dr. Wettergreen’s advice and cut out a cardboard slot for the gear axles to sit in and slide in so I could stop the gears from tilting into each other and test different distances between the gear axles and make it work.

Very quickly, I was able to determine the best distance for the gears, and I put a piece of tape along the back of the cardboard to prevent the gears from sliding out of phase. Now I could focus on getting them to move more seamlessly among each other. 

I went downstairs to cut a wooden handle for my project, and Dr. Wettergreen motioned me over to talk to me, so I went to show him my prototype. He was impressed with the work I had done, and he gave me some tallow to use to lubricate my gears because, in his words, “We use animal fat to make our machinery work better. We are very sophisticated in ENGI 210.” I just kind of looked at him funny when he said that. “You know, it’s kind of like this old cooking material that people used back in the 80’s called Crisco-“.

“Uh, Dr. Wettergreen, Crisco definitely still exists. I have definitely used it in my mother’s kitchen.”

This led to a relatively extensive debate over whether Crisco still existed, during which we conducted a quick Google search and Dr. Wettergreen admitted his defeat.

We then spoke about my progress on the project, with him claiming that I’m making progress and me expressing my doubts. I admitted that I was scared of this project and having to design and build a working model from scratch. He was confused but ultimately happy to know that previous students’ work was good enough to intimidate me. Which, yeah, I guess that’s a good thing, but also. He then repeated the same thing he’s been telling me since the beginning but I’ve been brushing off because, oh yeah, hubris: “This should be one of your easier classes. I want this class to be fun and enjoyable. You should calm down and relax and enjoy this process.”

I just kind of stared at him. “That’s really funny, Dr. Wettergreen.”

And then I heard laughter from behind me. I glanced over to see Caz, the ENGI 200 TA, laughing at me.

Needless to say, when a prototyping TA starts laughing at you, that is not a good sign.

So I left and went back upstairs. I applied the tallow to my gears and got them to work perfectly. Now that I actually had a working mechanical model I was satisfied with, I was free to design and prepare the rest of the project.

Conveniently enough, Dr. Marquez had cancelled class that Friday, so I came in early in the morning and spent about two hours laser cutting my parts. Overall, laser cutting went without error, except for the fact that the base of my project, which has a gear design and my project information etched into it, somehow took an hour to raster. I don’t know how it took that long considering I used fairly normal settings for it, but I did know that no matter what, I was not recutting that piece of wood if I could help it. I was not about to waste another hour in the laser cutting room.

I also specifically grabbed a thinner piece of wood to use for the back wall for my gears with the intention that the thinner wood would provide enough clearance on my axles (cut out of slightly thicker wood) for the wooden gears to spin more freely. I still stand behind this as being a good idea, but I quickly realized the implications of what I had done when I was sanding the parts down. Remember which other type of wood was thinner than the normal stock of wood we use and gave me hell on the laser cutting project? That’s right. The crappy red underlayer wood. It was back to haunt me again.

At this point, I was ready to apply color to my wood. Although I had originally wanted to paint the wood to look like celestial objects, I quickly realized that paint applies a nontrivial surface coating to the wood, meaning it adds a thick enough coating to the wood that could completely mess up my gear motion because I was working with tolerances of millimeters and I barely had it working as it was. So I decided to avoid that problem by instead staining the wood, which doesn’t really provide a surface coating but more like provides color that soaks into the wood without adding any extra thickness.

Dr. Wettergreen had just bought some more stains (in addition to the wide variety of three stain colors that we had in the OEDK), so I spent a bit staining rejected prototype parts to try to pick which colors I liked best. From this experimentation, I decided to use the mahogany and black stains because they provided the wood with really nice, bright colors and contrasted nicely with each other.

 

I also coated the back panel with linseed oil in an effort to make it slicker and easier for the gears to slide on it. I hoped that the oil would darken the wood enough for it to blend in with the base, but because I was working with crappy red underlayer wood, it only succeeded in making the wood panel turn orange. At that point, it definitely didn’t match the base, but after consulting with Christina and Kalen and Mikaela, I applied oil to the base and axles as well in an effort to get them to match. Because they are made of a different type of wood, they still don’t match, but at the very least, they’re sealed now, and that’s never a bad thing.

Because I ended up oiling and staining different parts of the project at the same time, a lot of my experience consisted of me accidentally getting black stain on the oiled pieces, freaking out, and desperately trying to sand off the stain. So that was fun.

I did eventually get everything stained correctly though. Including my hands.

At that point, it was time to plasma cut the required pieces for my project. Because my entire design consisted of gears, I had gotten previous permission from Dr. Wettergreen to not plasma cut a working component of my project. That conversation went like this:

Me: “Dr. Wettergreen, you say that I need to plasma cut a working part of my model.”
Dr. Wettergreen: “That is correct.”
Me: “My design is made entirely of gears.”
Dr. Wettergreen: “That is correct.”
Me: “It’s a really bad idea to attempt to plasma cut gears because of the inherent imprecise nature of the plasma cutter.”
Dr. Wettergreen: “That is correct.”
Me: “…”
Dr. Wettergreen: “That would make your job harder than necessary.”
Me: “…So what do I do?”
Dr. Wettergreen: “How about you get some sleep and come up with some ideas later?”

Which was insanely helpful.

So I intended to plasma cut the tops of the axles for the gears in a star shape, in homage to my original design. Although they aren’t technically a moving part of my project, they are still important because they hold the gears onto the back and allow them to turn. (The axles also do actually turn on the finished product, but that’s nonintentional and also beside the point.)

So I plasma cut my stars out of steel, which was super fun, especially because I kept having to get Joe to troubleshoot the plasma cutter for me. (The grounding was wrong? The shield was at the wrong height? We’re not really sure, but we did get it to work.)

I then post-processed the stars by angle grinding the dross off of them, and then I filed down all of the edges to make them less sharp (I realized later that I should have just asked Michael how to use the rotating wire brush, but oh well.)

I then went to go use the sandblaster to give my stars a nice even finish, but I was stopped by Kelly. This led to the following exchange.

Me: “I’m going to sandblast my stars.”
Kelly: “…The sandblaster’s not working.”
Me: “The new one?”
Kelly: “It’s not hooked up yet.”
Me: “…The one I used two days ago?”
Kelly: “… oh, I guess it is hooked up.”
Me: “I’m going to go sandblast my stars now.”

Once I had finished sandblasting the stars, I sprayed them (and my gears) with an acrylic coating so the delicate finish wouldn’t be damaged by scratching and dust.

Once all my components were coated and protected, I carefully glued them together, clamped them to the table, and left them to dry overnight, partially to ensure they stayed glued, and partially in a valiant attempt to try to overcome the inherent warpedness of the wood. It didn’t really work, but the final glue job was functional enough to move on.

Then it was time to attach the steel stars to the wood. Nick Escobar had been raving about JBWeld for the past couple of days, so I tried to get him to show me where it was and how it worked. He found some at the table, handed it to me, and turned to walk away.

“Wait, Nick, how do I use this?”

“Oh, it’s just epoxy.” He smiled and shrugged in that self-knowing way of his, and turned away.

“But Nick, I have never used epoxy.”

At this, the entire class turned around and gasped, looking at me funny.

I just kind of held up the two bottles taped together. “So, anyone know how to use this?”

For my handle, I was originally going to use a long handle on the edge of the middle gear and have people grab that to turn the mechanism. However, Dr. Wettergreen stopped me before I did that, warning me that in the entire course of this project, I had never once tested my gears in that way. Using a long handle attached to one end of a gear would not work as a good handle mechanism due to physics and mechanical advantage, so I needed a different plan. Perhaps I could somehow make a mechanism to attach to the middle of the middle gear so I could spin the entire gear and not just one long handle?

So I thought about it and came up with a star shaped handle (in keeping with the theme of star-shaped axles) that went over the top of my gear that people could grab and turn. The problem was that I still needed an axle cap for the gear so it wouldn’t fall off the vertical panel it was mounted on. So I designed a star axle cap to go on the axle, and I made a larger star that would pass over the axle cap and act as a handle to rotate the star. If it worked correctly, the axle cap would stay stationary while the outside star handle would rotate around it. If it worked, it would be a pretty badass centerpiece for my model.

I then glued the new axle and handle to the middle gear and prayed that it would work because I was running out of time.

At this point, I ran into Thomas (the self-professed Canadian rock star) and his girlfriend, so I asked her for some aesthetic advice, especially because I’m friends with her and she was sitting on her laptop at the open-use table next to me. So I held up the base of my piece to her and asked if she could read the engravings.

She just kind of squinted at it. “There are words on there?” She took it from my hand and held it close to her face. “Oh, I see them now. I guess I can read them?”

Well, that more than answered my question.

It was very difficult to read the engravings on the base of my project because I had previously covered it in oil, which reduced the contrast between engraved parts. So I solved that problem by filling in the text with a fragment of 0.5 mm pencil lead I borrowed from Scout. The lead was small enough to fit into the indents with the words, and it’s light enough that it makes the words visible but doesn’t scream for attention.

Then I slid the back panel in the base because I ultimately decided not to glue that in. It fits well enough without glue, and now, I can easily remove the base from the stand to more easily transport the model should I need to.

I gave the model a couple of test spins to find that it spun really easily Is this a benefit of being vertical vs. horizontal? I don’t know, and I really don’t care because I was finished! And it actually works!