Week 10: The Last Leap

Katie b -

As my project is starting to wrap up, there are some final steps I have made.

After receiving my 3D-printed toe box, it turned out super cool! It looks like an actual pointe shoe. However, of course it was barely too small to where I could not fit my foot inside to properly test it out. 🙄 But I am not afraid of a little trial and error. In order to create this box, I had to play around with new tools on SOLIDWORKS that I did not need to use when creating my shank. Since I was figuring it out by myself, I was destined to make a few mistakes. The first one being my design was way too thick. In order to create a three-walled shape, I used the “shell” tool and set the thickness to 3mm. I didn’t realize this meant 3mm inwards and 3mm outwards. So, when I got my printed model back it was 6mm thick. Oops! I opened my SOLIDWORKS design back up to make some changes. I decreased the thickness of the walls to 1.5mm which will help to widen the box. I also adjusted the roundness of some edges to make it less sharp and pokey so it doesn’t dig into my foot. I have sent these updates back to the 3D-printing guys, and I will hopefully have a new model by my presentation next week. For now, here are some pictures of the original model.

3D-printed box angle 13D-printed box angle 23D-printed box angle 3

The second thing I conquered this week was some data analysis of the SOLIDWORKS simulations from a few weeks ago. After another meeting with one of the guys at Gore, I was able to get a more detailed understanding of the simulation software and some screenshots to support our findings for each material. We simulated a total of 4 materials: one control and three experimental.

The first material we tested was steel. This was our control group so that we could ensure the software was working and performing as expected. A shank made out of steel should barely bend/move when a body weight force is applied, making it a good baseline measurement. This is exactly what we found when measuring the displacement. In the image below, the red color represents areas with the largest total displacement, and this was only 0.1865mm.

Steel shank displacement

The second measurements we looked at were the von Mises stress values. Von Mises stress is a value used to determine if a given material will yield (permanently deform) or fracture under a certain load. Once again, a shank made of steel should certainly not yield under a body weight load, making it a good comparison. Thankfully, this is what the computer found as well; the yield strength of steel exceeded the maximum von Mises stress in the simulation. 

Steel shank stress

Now that we had confirmed our simulation was performing realistically, we could move forward with testing more materials. The three materials we simulated were PLA, PETG, and Nylon 6/10. The images below show the displacement and von Mises stress for the PLA simulation. I am only going to insert the screenshots from this simulation for the sake of not making this post extremely large, but I will share the quantitative results we gathered for the other materials as well. As you can see, the point of maximum displacement for PLA was 35.75mm. This was not surprising as I knew it would bend more than steel, but it is still a relatively stiff plastic.

PLA shank displacement

Now looking at the stress, the PLA performed similar to the steel; the yield strength exceeded the maximum von Mises stress meaning that the shank would not have permanently deformed or fractured.

PLA shank stress

Finally, here are our findings for the last two materials. The Nylon 6/10 performed similarly with a maximum displacement of 59.35mm, and it did not yield. However, the PETG had a much larger maximum displacement of 106.5mm, and it did yield under a body weight load. This was very surprising to me because I have tested a PETG shank in person, and I did not find any evidence of it being permanently deformed. Overall though, I am super happy with how the simulations turned out, and I am excited to have some quantitative data to add to my presentation.

Moving into this final week, I will be working on cutting time off of my presentation since it was slightly too long when I practiced with the Senior Project Committee. I am also hoping to get my new toe box model soon so that I can quickly try it out over the weekend. But… I am super excited to finally get to share my project with more people next week, and I can’t wait to hear everyone else’s presentations!

More Posts

Comments:

All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.

    liam_k
    Hi Katie! That's really cool you are able to get this quantitative data and compare it to your real world experiences with the shanks. Is there any reason for the discrepancy in the PETG shank you can think of?
      katie_b
      Hi Liam! In all honesty, I am not sure why there is such a discrepancy between the simulations and my real world experiences for PETG. My first thought was that the way in which we applied the virtual load was not a good replica of a real dancer's load, but when I compare how the PLA shank performed in both scenarios, it is nearly identical. My best guess is that there is a very minor error in the how the simulation was set up, and this error shows through the greatest with PETG for some reason.
    riley_b
    I'm excited to hear your presentation! I hope you can get the toe box in time.
    catherine_f
    Hi Katie! It looks like your project is almost complete. I look forward to seeing your presentation!
    cindy_h
    Hi Katie! I'm excited to see your presentation, everything you've presented here is really cool! Do you think the von Mises results about the material yielding under body weight might change with use? For instance, for Nylon 6/10, maybe after lots of wear and tear, the von Mises test would show stress?

Leave a Reply to katie_b Cancel reply

Your email address will not be published. Required fields are marked *