Spooky Box. That is the name of our project and I am very glad that we are finally finished. So, I think it is bets to write this blog post in order of what we did and how we went about. First, we designed the box. Second, we made a prototype of the box (this kind of failed but more on this later), and then we made our box. As a final little step, we finished our box, but this will be included in the same paragraph.
First we designed our spooky box. Truth be told, our process began with picking mechanisms and then constraining our system to those mechanisms. There were many hurdles along the way, but we began with the constraints in sight. The first thing we did was pick our group of mechanisms. We chose mechanism 24, 502, 505, and 96. I list 502 and 24 below because we ended up only using these two in our final model. Really, we chose so many such that we would be able to cut out mechanisms if they ended up not working. Remember, here, we were constraining our design, this wasn’t final. Because we were comfortable making boxes with the makercase software, we also chose to make all of the gears within the box, giving our box a jack in the box feel to it.
After choosing the constraints for the box, we designed the box roughly in procreate. We designed our box to be sectioned as mechanisms, where the mechanisms were separated into layers. This allowed us to visually explore what decor and theme we chose, as well as line up how our mechanisms would function. Here, we had no idea what to do for the theme, so we chose the most relevant theme possible: Halloween. We settled on a box that would wave to you (this would change but it was our original idea). We wanted an item that children could walk up to (when trick or treating) and play with. Maybe grab some candy and play with a cool engineered box, so this is what we went for. This said, things did change during our design process, and we knew that going in.
Above are our first few designs made in procreate. We separated each layer into individual drawings to best understand them.
What did we do next? Well, we knew that we wanted to cut the box in the following week after we designed our box, so we quickly CAD modeled our designs in fusion 360 and then ported the dxf to adobe. In fusion 360, our designs fit great with each other and the whole box worked, however, this is in a computer. In reality, when we ported the files, they distorted in scale, which we didn’t check, so our first model was very silly, with many big and small gears. This said, we were able to check that some of the models functioned together.
Here are the CAD designs that we created for each model.
Here is our comically large spiders and the pieces that we first cut out. We were able to test that they fit well together within their specific layer, however, there was zero intra layer interplay
Moving forward from this silly mistake, we scaled our whole design together to match the 1 cm hole width and then we cut out the pieces again. From here, we checked that all layers would work, at least prior to gluing them and making permanent changes. We did this by laying out the layers and checking them. What we found was that our 3 gear setup and our tombstone cam were going to require a lot of fixing, both were quite rough. We had to choose which, so we went with the three gears, because our metal pieces were attached to the spider gears.
Here is a photo of our gears once cut out. We chose to double up a lot of our gears for more bulk and surface area between them. Our thought process was that, yes, we are increasing the friction, but also that we will have a potentially stronger gear if a tooth chips in years to come. We want our product to be used outside on Halloween night, so it can take a beating.
Above is a photo of our functioning layer test. We used small dowels to assemble our piece and test it out
Here is a photo of our metal spiders. These are glued to a set of wooden spiders, which glue to the rotational gear.
Once we chose what we were going to work on, we cut the box out to hold the dowel together. Here, none of our project was glued, but we could check that it all worked together and held tight. Silly enough, we only cut 4 sides of the box at first. We really wanted to make sure that our box would work, so we assembled only what we needed. We were going slow and steady and (as it turns out) that’s okay.
All of the gears sitting on the primary axel. This was our first test to mark and line up where the gears would sit. At this point, we had chosen to cut the tombstone, however, the original cam for the tombstone (middle left) would help us line up where to put the tombstone. At this stage, we also measured our box and used makercase to get a rough file for our box.
The box files are shown here. Why are there two different files? This is mainly due to a communication error within our group. We have to admit to these things and they happen. To clarify, we never cut the full box on the left, however, in measuring the box, we realized our final 2 sides were too large, prior to cutting. With this in mind, we redesigned the sides and cut them out.
Our tombstone roughly assembled with 4 walls (and lots of painters to keep it up right). We should also address this because it is in the photo. There was miscommunication on the orientation of the box leading us to cut the box inlays twice. That is why (in the photo) you see the bottom piece of wood cut, as well as a hole for another dowel in the side. Due to the large size of our box, we didn’t feel comfortable wasting so much wood, so we opted to roll with the mistakes and put the bad cuts in areas the user would never see them: the bottom.
So our box was put together, what were the issues? This is the funny part of the story where the box rapidly changes based on our time constraints and fix what we had to. So, we figured out that we didn’t have enough torque in our model, so we had to fix our crank; we figured out our 3 gears were stuck due to an angle, we had to fix this; and our skeleton hand (from the original design), was not feasible unless we dangled the design off a ledge. This was an option, but we decided having Frankenstein rotate was just as fun for kids during trick or treating, it wasn’t like they would stay at the house for too long. A big note here, we switched the side of the handle from facing towards the user to the other way (towards the candy giver). Why? Well, our box was large and, although we accounted for torque, was very unsmooth due to our unfixed angle. While our box got smoother, we decided that taking the handle and putting it in the back towards the candy giver would allow children to have a first introduction to the scary box festivity, then play with it if they want. This way, the person giving out candy has another fun decoration. Really, we doubted that children could rotate the box when our three spider gears were stuck. In the interest of time, we made this change.
The box with a a fifth door on it. We used a rubber mallet to pressure fit it into the box. Yes, we also used wood glue but it was so satisfying to give the wooden side a slam and hear it close shut.
Fixing the designs. The first thing that we fixed was the handle for the box. The way that we fixed this was add wood circles to the handle, giving it more surface area to adhere to the main axel. This worked really well, and frankly we got lucky. We were really worried about this fix. Additionally, we slightly sanded the gears to make their action smoother. Also, a big change. We took out our planetary gear. We really liked the planetary gear and it was cool, but it made our design function too rough. Our design was for children and if it took both Louis and I a lot of muscle power to muster a crank, we decided it wasn’t worth it. Sacrificing some cool mechanics for functionality was our choice. The next fix, our three spider gears didn’t turn. It turned with two gears attached but when all three gears were present, the system would lock. We honestly thought that we had run out of luck. We took a lunch break and came back later. What we thought was that the dowel on our third gear was stuck, this might be partially true, so we sanded the dowel down a little. While this helped, our gear would still get stuck. After fidgeting with our setup, we found that the angle of the gears was slanted, causing the top gear to spin off the setup and lock up the system. To fix this, we created a little circle and glued it in front of the gear. In effect, our gear could no longer slip. After testing it, we found our system to be much, much smoother. The last change. Our last big change was the skeleton hand and the Frankenstein rotational gear. We really wanted Frankenstein to wave to the kids as they passed, but this was likely not feasible because the skeleton arm would get stuck. Because we were close on time and already had 3 moving pieces, we decided to leave out the skeleton arm and have a fun rotation Frankenstein. This would be both fun for the kids and more consistent.
Here is the finished tombstone name tag (we eventually replaced this name tag for a few reasons) and the spider gears. This photo displayed the fully finished spider (with the metal glued to the wood spacer) sitting on the gear.
Here is the box with our 6 walls on it. Here, you can also see the bat stoppers that we had to limit the movement of the skeleton arm. We decided to keepx these (even though they serve no purpose) because they match the Halloween aesthetic.
We wanted to show this off at some point in the blog post but we really tried to sand the project. We used 60, 120 and 180 (in that order) sand paper to get clean finished on our pieces. Here is our handle while it is being sanded.
How did we make an aesthetic box. So, after sanding and gluing the box, the original idea was to really emphasize that this is a Halloween box. We wanted to push the envelope. What we first thought of was that our bottom layer had to be black. So we did this. We used 2 small foam paint brushes and covered the box in one layer of black. To add to the rustic feel, we had a wet paper towel and would mop up sections of black, giving an eerie worn look. Next, we wanted to add decorative paint. At one point, we were going to print a stencil that said Happy Halloween but it was a little too much, so we decided to do blood splatter. In short, we sprayed red spray paint on our gloves and then did a snap motion to fling the paint onto the box. This created a bloody box. We liked the look of this effect so much, that we stuck with it and covered the box. This look was originally only intended for the handle, but it looked to cool not to do on the tombstone, spider gears and sides of the box. After we let the box dry, we cut out a vinyl sticker that says “SPOOKY”. This sticker is printed in orange and kind of adds to the Halloween theme. To protect our box, we used two layers of clear coat. We did these really light and they shouldn’t be noticeable.
Above is the black paint and red spray paint on our box.
Here is the final box with the vinyl cut “SPOOKY” sticker on it.
Let’s talk about final mistakes and large take aways. I will discuss three here. First, planning. This was incredibly challenging for us. What quickly became apparent was that we had to cut out aspects of our final design. Yes, we had planned for this from the beginning but I am wondering if going with a more simple design would have led to more saved time in the end. No, cutting the project didn’t take much time, but it did require thought and I think putting in the effort on the front end and not the back end could’ve lead to less of a headache. For example, if we were to do this project again, we would likely put the gears horizontal, not vertical, because there was a lot of slippage of gears on the same rotational plane. Horizontally aligned gears have less of an issue. Planning is important so maybe spending 2 more hours here could’ve helped. Next, the assembly of our design. The assembly went well but this goes alongside with planning. Some of our pieces are harder to move than we intended. This kind of adds to the whole feel of the Halloween box, sure, but in the end, we would have enjoyed a smoother box from the assembly. Less post processing. This would have saved a lot of time trouble shooting but I think some of this comes with experience. Here’s an example. We had a planetary gear in our original handle design but the teeth of the gears kept getting stuck. Why? Well, what we found was that the planetary gears, although similar in spacing, don’t have as much depth between their teeth, for this exact reason. We didn’t know this going in and designed away. Next time, we believe designing for direct assembly would save a lot of time. The last thing, our design. We love our design and we are proud of it. I guess what I would change is maybe its functional design. I really wish we could put a box of candy on it, or maybe in it. I think this is a really missed opportunity. Additionally, it is frustrating that the handle got moved to the back. In the future, I would want to fix the smoothness and mechanical advantage so a child could spin the crank.
For the most interesting part of this project. The cost.
This is our respective cost breakdown. Under these conditions, the box would end up costing $491.18. Does this make sense? Would I pay $491.18 would this box? Of course not. In reality, I think I would maybe pay $50 for this box, and that is it. The box is designed as a Halloween, seasonal item. It is not going to be getting higher than that $50 point. Maybe it could be pushed to $80 but then your consumer is the crazy Halloween enthusiast and not some engineer that thinks this would be cool to show off to kids. In all honestly, it would be great to lower the cost, to around $40. At this price point, I think the item is really approachable, because of how interesting it could be when giving candy. So what do we have to do to get it here. Well, all of the operator time has to be cut. Much of the process has to be streamlined such that an assembly worker cut pop it all together with glue and send it off. How could we do this and how would it cut costs? I think the most efficient way would be laser cut, process and paint, assemble. Yes, there is no final process this way, but this process is much faster. Additionally, time sink pieces, like the sticker, could be cut from the design. It takes too much time. Maybe, a branding sticker could be added to keep brand name present, but the sticker would have to be a little easier to take off the transfer paper and be far less intricate. What about our one time costs? Well, art is a one time cost, sure. Also, the design is also a one time cost as well. These costs do count, but they become much more trivial as more boxes sell on the market. By this notion, we can assume these costs would become less apparent overtime. Lastly, the one that is always silly in these blog posts. The materials we used. Materials are always much greater of a cost because they cover all of the can of paint or all of the spray paint, when only a very small portion was used. By this notion, we would be able to produce many more boxes, not just this one. Thus, we would make a profit back overtime.
