Making something big(ger)

Originally, I wanted to work on the gear design for my ornithopter in a larger scale format, but after finding the thesis of Jackowski and seeing how the choice of components and their sizing was critically chosen, I decided that the gears would not be very useful in large scale. Instead, on my way back home, I realized that I needed to bring my plants back inside (it's getting cold) but I don't have anything to hold them nicely. Thus I decided to create a piece of furniture that would allow me to hold multiple plants. As it happens that I drink a lot of iced tea, I accumulate cups, which I can use to grow plants.

I searched for indoor garden ideas, and seemingly the web is full of it! From dedicated shops, hanging structures, furnitures, recycled shoe bags, ladder structures, so many on google! There are also interesting modular designs and eventually I found plant stands which I decided to use as a base for my idea.

Incremental Design

My initial sketches on paper were targetting a cuboid with shelfs that would have holes to hold cups / plant holders. However, given the little time available for cutting this week (at most 1h30min at once, because of the slots and the fact that we would use the ShopBot). The EDS does not have any such machine (at least not yet) and thus we are doing it at IDC near the MIT museum. However given the unique slot, I decided to focus on a simpler design using standing columns, which could be joined using horizontal panels.

I created a taller version using linear repetition and added a plate connecting the two columns with holes to hold cups.

Prototype on Laser Cutter

I first laser-cut a small version of it with the top plate to check the design. Unfortunately, I did not try to laser-cut a tall version at that point, which would have been very useful since this tall design is not fabricable! How do you connect the two column sides? They are prevented from doing so because of the multiple layers!

CNC Milling

To enable CNC milling, I originally planned to remove extruded circles (diameter D of the endmill) at the main corners. After discussion with the TAs, it was suggested to instead use a shifted circle that would allow a continuous path for the given diameter since my circles would require a retraction and hole drilling, which our endmills are not made for. I designed two small variants:

• A circle shifted towards the outside, in contact with the corner. This can be done by using a similar corner circle with diameter $\sqrt{2}$D for construction, which generates two intersection points with the adjacent edges, and then create a circle passing by these two points as well as the corner itself. This was used for most corners except the slots that were too small.
Note: in 3D mode, we can use a translation (here would be by D/2 diagonally), but in sketch mode, there is no history, so I did not want to use the translation and rely on geometric features instead which would still be visible as a history.
• A circle shifted by a radius D/2. The simplest construction is to use a first construction circle of diameter D, and then the real one centered at the intersection with the edge. This does not touch one of the edges, which makes it better for narrow slots where only the tall edges can be carved.

The cutting did not result in any major issue. The only problem was that I used the original material thickness of 7/16 inch for my slot, but the sheet I used was slightly thinner, which led to the joints being too loose. I had planned ahead by creating variables that would have allowed an easy change of the material thickness, but sadly did not update it! (arghk)

The previous group had used the 1/8 inch endmill for small features but given that my slots were made for 1/4, we switched the endmill to half the cutting time. Note that pass depth should usually be half the endmill, but with our material, it was said to be at most the endmill diameter, enable only two passes to cut through with the 1/4 inch endmill.

I added a pair of 3d tabs per toolpath component to avoid having components shaking or flying away during cutting. The original tab depth was quite small, and we increased it slightly, but we could have increased a bit more as one of the horizontal plate circles did go loose during cutting. It didn't create any problem fortunately.

Then comes the sanding to make the edges more comfortable! (and allow holding / grabbing the actual structure, which was painful originally)

And finally, we assemble all of it into and stack them (together with the original cardboard prototype). Initially, I wanted to introduce slots to allow a real stackable piece of furniture, but this ended up not being possible because of the short time slot.