HCDE 451: 3D Printing
Introduction
This week, to gain experience with Rhino3D for 3D modeling, we were given an open-ended project to build anything we’d like, as long as it included at least an extrusion, revolution, or boolean (adding or subtracting one object from another) and remained within a “reasonable size.” Our object was then to be converted to an STL file and printed using PLA in MakerBot.
Description of Idea
I wanted to make something cute that would also be useful to me in some way. After some deliberation and image-searching, I found inspiration in a charming ninja incense holder on Google.
Having chosen an incense holder as a concept, I wanted to make the idea my own by changing the idea. Instead of a ninja with an incense sword, how about a body builder with an incense barbell, or a wizard with an incense wand? I also toyed with the idea of cats or humanoid faces as alternatives. Some of my sketches are below. I eventually decided on the wizard.
Design Process
Before settling on the wizard, I actually started to build an incense-holding cat in Rhino. Since I missed our in-class crash-course on 3D modeling, this was my first time attempting to 3D model in Rhino. While watching a tutorial, I spent a few hours trying to create a cat on top of a photo before admitting it might be too complicated for my first project.
You can see in the image below that I got some basic shapes in for the body, and that I initially planned to include a base as part of my incense holder. It was when I realized that I had no idea how to take what I’d started and rotate it to the other side of the frame that I decided to scrap the idea. Maybe I’ll come back to it for fun in the future.
When I started over, since I could not find a simplistic wizard shape on Google, I decided to dive right in and make a general outline that I thought would be suitable when revolved. The most telling feature, I figured, would be the pointy wizard hat.
Below you can see the first version of my wizard in progress. I started out with one rotated line for the body and two more rotated lines for upper and lower arm pieces. Frustratingly, I could not manage to connect to the arms to the base in Rhino. Instead, I just placed the arm pieces near the base and hoped for the best.
Despite how rough it looks, I was happy to find that the MakerBot successfully printed the wizard, connected arms and all, despite the arms technically being separate pieces.
At this point, even though I knew the wizard I had printed would be suitable as a deliverable, I was determined to make it better. After all, all I’d proved was that this figure could print — it was still far too small and lacking in a proper hole in which to place incense.
The Final Product
For my third and final 3D modeling iteration, I started out in much the same way as I started version 2. One rotation for the body, this time including shoulders.
While adding shoulders seemed like a great idea from the front-facing perspective, I soon realized that the shelf I had created made no sense rotated around the entire body. After at least an hour of trying — and failing — to cut the shoulder bits off the back and front using trim, cut, boolean, and every other conceivable method, I resorted to to re-drawing my original outline sans shoulder-shelf and making a new body. The shoulders would eventually be added as standalone pieces between the arm and body, as shown below.
Note that my wizard actually looks solid this time ! A kind woman at the CoMotion MakerSpace pointed this feature out to me (in the Display tab, you can choose to view your object shaded or rendered) and it definitely helped me visualize how well my pieces were coming together compared to the wireframe viewport.
When I was satisfied enough to send my wizard to the printer, it looked like the image below, but I would end up having to delete the nose I added, as I it would not combine with the base and rather than proceeding as my V1 Wizard had done, Rhino was giving me “bad object” errors to prevent me from converting to STL.
Behold, the final product below! I could not smooth out all the edges, particularly around the arms and shoulders, and sadly, the MakerBot was not accurate enough to leave a sizeable hole for incense to fit into in the wizard’s hand. But I’m still satisfied with how much I learned in the process.
Further pictures of version 1 and version 2 of my wizard side-by-side below:
Analysis
Before this design challenge, my only 3D modeling experience was with a proprietary software called DELMIA which is used at Boeing to model airplanes. It took me weeks to learn the program and I remember being frustrated by the lack of documentation available. Still, I eventually learned it and I found myself missing some of the features of DELMIA when transitioning to Rhino.
My biggest challenge in moving over to Rhino was that placing objects near, on top of, or inside one another is comparatively unintuitive and I repeatedly found spaces between objects (either because I was only aligned on one plane or because I wasn’t zoomed in enough to see a half mm of space) when I intended there to be none. This challenge is what made removing the shoulder-shelf in version 2 of my wizard impossible after an hour of trying.
The second challenge was that the measurement system in Rhino is still confusing to me. In DELMIA, possibly because manufacturers need to be sure of exact specifications at all times, it was very easy to set and edit measurements. In Rhino, this was not intuitive and I could not find a way to edit measurements once I set them. I’m sure it’s possible, but when I had to make modifications, I found it easier to just delete and start over. On a higher-fidelity project, I would not want to take this route, and it is my opinion that Rhino needs more documentation, more tutorials, or that there needs to be a resident Rhino expert with a sign saying so in the CoMotion MakerSpace, as finding help when I needed it was difficult.
Lastly, this is not the fault of Rhino or Makerbot, but I ran into a challenge related to other users on both the laser cutter and while converting my Rhino files in the CoMotion space. Someone, in both instances, had changed default settings and while I caught this before starting the laser cutter, with the 3D printer, I was never given instruction on how many 0s were appropriate for our conversion. Someone had added a couple of 0s to the default before I got there, and as a result, I spent about 40 minutes trying to convert my Rhino file to STL before an employee came over — because they were closing for the night — to troubleshoot with me. (I actually asked a different employee for help earlier, but she didn’t notice the 0s either.) It took about 8 minutes a pop for each failed conversion due to the ultra-precise settings, and I believe the conversion was failing when the flash drive ran out of space! I feel like this shouldn’t happen. Perhaps there is room for improvement in the MakerSpace to keep people from running into time-consuming problems like this as a result of another person’s input.
