# Week 4: 3D Printing and Scanning - A Tale of Two Rings <a href="../index.html">Back to Home</a> Files for download: * <a href="files/Swing_straightbars.stl">Swing, modified to have a straight bar</a> ## Ideation For this week, we had to 3D CAD and print a structure that could not be made by, say, milling from a couple of different sides. Essentially this meant printing "things in things", or something with moving parts. I thought about doing a chain, but according to my TA Zach that's been done a lot before; really wanted to do a fractal tree and maybe try out using Grasshopper, but couldn't figure out how to make it truly unable to be made subtractively. So I decided to print a swing! Something decently challenging, but not beyond my ability to CAD, and interactive to boot. ## CADing Even though I had heretofore used Fusion, I decided to switch to Rhino for this project. The main influence behind that was the fact that Zach was well-versed in Rhino and could give me a one-hour intensive on how to 3D CAD in it. This was 100% the right decision; Rhino proved to be much more intuitive than I remember it being the last time I used it, and more intuitive than Fusion. Even though I knew having an arc for the swing's bar wouldn't allow it to swing as much as having a straight line, but I liked the aesthetic of the arc, and thought that it might make it easier to print on the Sindoh, in terms of reducing the amount of supports needed. (Spoiler alert: it didn't really.) In hindsight, the better thing to do would've been adding increasingly smaller, 50 degree segments to support the corners. With a few hours of using simple tools like truncated cone, boolean difference, pipe, and trim, I came up with this: <img src="images/swing_rhino.jpg"/> (I made the swing seat bucket-shaped because I had a notion to 3D scan my stuffed cow that I've had since I was 2, but, as we shall see, that didn't pan out - at least not yet.) ## Converting to GCode I knew the rings of my swingset were floating, but I hoped that the 3DWox software for Sindoh would be able to put in the supports automatically. My support settings were Everywhere -> Zigzag, my infill settings were Crystal (1), which was a triangular infill, for stability (the honeycomb patterm is supposed to have the best strength:material ratio, but 3DWox didn't have that option). There was a lot more support than I had hoped for, but Zach assured me they would snap off fine. (They do.) ## Printing I was away for the weekend, but Zach was kind enough to load my gcode files to the Sindoh for printing. Unfortunately, it was not as straightforward as one would have hoped for. ### Attempt the First <img src="images/failed_supports.jpg"/> (Material: PLA, took about 8 hours to print) So that didn't work. Zach said it looked like the components got shifted during the print, causing misalignment, and that the rings could use some more manually-added support. I added a raft to connect the two end pillars to provide the stability. I was worried that my manually-added supports wouldn't be easy to remove, so I just shifted the swing so that the rings nearly rested on the arc. <img src="images/swing_rhino_raft.jpg"/> ### Attempt the Second... ...never occurred, because the printer apparently pooped out. At this time, Zach realized that 3DWox was not adding the necessary supports for the rings, so asked me to send him the STL file for Ben to try it on the Stratasys (how many TAs does it take to help a clueless and hapless How to Make student?). The Stratsys deals with prints that require supports by printing the supports in a different material that dissolves readily in a heated citric-based acid solution. Which leads us too... ### Attempt the Last In the end, it worked! Unfortunately it was a lot smaller than I wanted because I'd forgotten to scale my model in the STL file, but that may have been for the best anyway - bigger print would have meant more time to dissolve the excess material. Ben said it took around 2-3 hours to print, and the supports took less than 2 hours to dissolve. <img src="images/swing_success.jpg"/> The arc did, alas, snap in one area when I was loosening the rings, but it still swung: <embed src="images/swing_video.mp4" height="640" width="360"/> ### Bonus I verified that the Sindoh supports were indeed easy to snap off, and decided to keep parts of the failed print because...why not? <img src="images/failed_separate.jpg"/> ## 3D Scanning Since the software we were using, Sense, was extraordinarily sensitive to movement, Zach made us a turntable for the object to be scanned so we could stably rotate it 360 degree for the camera to process. I took the first thing I saw on the table, a flashlight, and scanned it. Placing it precisely in the middle of the turntable helped, since it minimized translation of the object position in the camera's field of view. Still had to turn the turntable agonizingly slowly for the camera to not lose track of the object. There was unfortunately one section that the camera just would not pick up, but you can sort of make out a flashlight. <img src="images/flashlight_scan_screenshot.jpg"/>