Reflection on Group Work and Personal Learning

This week allowed me to reinforce the importance of regular machine maintenance and proper calibration, as well as the correct configuration of printing profiles. These aspects are essential to push the limits of this type of technology and achieve reliable results. Thanks to my previous experience with these machines, the activities were relatively straightforward for me, but these exercises helped me to revisit and not forget these critical practices.

One area I realized can always be improved is the preliminary testing and cleaning processes. During one of the group tests, I encountered issues with the supports because I neglected to properly adjust them and failed to clean the machine beforehand. This oversight caused the supports not to adhere properly, resulting in a failed print and wasted time.

This experience highlighted that, even with experience, skipping essential steps can lead to errors. It also reinforced the value of documenting these situations to avoid repeating them and to optimize the workflow. These lessons were shared and discussed with the team, strengthening our good practices and awareness inside the lab.

Design and Print of a Custom Part

---

This week, I decided to keep the individual work simple and used Tinkercad, a very versatile tool that lets you design basic models in a short time.

In the past, I tried to create an object inside another using traditional methods, but it was nearly impossible. With 3D printing, this is achievable. I decided to design a cube with a sphere inside. I first created a conceptual sketch using AI to visualize the idea before starting the modeling.

To begin in Tinkercad, I dragged a sphere into the workspace and left it at the standard size of 20x20x20 mm, as I hadn’t yet decided on the cube dimensions.

Next, I added a cube and made it 4 mm larger than the sphere in every dimension, making the edges 2 mm thick. I began adding rectangular shapes to create holes through each axis of the cube, aligned at 20 mm spacing.

Using Tinkercad’s Boolean functions and the alignment tool, I set the rectangles to “hole” mode so they’d be subtracted from the cube. After repeating the process for all sides, I achieved the desired hollow cube.

Later I realized the original 20 mm sphere was too small, and it might fall through the openings. So, I scaled it up to 23 mm in all directions. This way, the sphere would remain inside but still be smaller than the cube’s interior.

Once the design was ready, I exported the model in STL format and used the Bambu Studio software for slicing. I used the standard 0.2 mm profile and selected the BambuLab A1 printer, which is fast and well-calibrated.

I chose a bright orange PLA filament. The slicer estimated a print time of 26 minutes, so I let the machine run.

I was concerned about the bridges since I printed without supports. Based on previous group tests, bridges up to 20 mm were manageable, but I kept a close eye on the print at that stage.

After completion, I removed the print carefully. The walls seemed thin, and I didn’t want to break the piece. This is something I could improve in a second iteration.

After reviewing the piece, I noticed a slight artifact in the vertical edges. It appears to be caused by a change in speed between the bottom layer and the vertical walls.

Despite these minor issues, the final result matched my expectations.

3D Scanning Workflows

---

3D scanning has always seemed more complex to me, but this week I followed a structured approach to make it easier:

1. Decide what to scan
2. Use a reference or concept tool
3. (Optionally) print a sample
4. Scan using Revopoint POP 2
5. Print with resin for detail

Following this route, I wanted to try something complex. Though I know iPhones/iPads can scan objects using built-in sensors, their results vary.

Instead, I wanted to try a larger scanner, so I attempted to scan a bust. Since scanning myself is tricky, I discovered a tool on MakerWorld that creates busts from a single image.

MakerWorld Bust Generator

I uploaded a personal photo, and after a short processing time, the tool generated a 3D model. While it can't perceive depth like a real scanner, the result was impressive.

I liked the result so much that I printed it using the Elegoo Neptune 4 Pro at a high resolution of 0.08 mm layer height. The result was amazing.

Using the Revopoint POP 2

For a real scan, I used the Revopoint POP 2. I chose to scan a colleague using a PC setup (for higher quality compared to mobile).

I verified mesh detection before starting. Despite documentation mentioning issues with dark colors, we tried anyway. The results came out great.

The filament print worked fine, but I wanted more detail. So, I printed it in resin, where every feature is captured beautifully.

Personal conclusions from this week

This week was one of my favorites, as I’m very passionate about 3D printing and own several machines. The modeling part was straightforward thanks to my previous experience. However, I’d like to improve the aesthetics of future models, especially wall thickness and overhang strength.

Regarding scanning, I’m amazed at how AI tools are evolving to generate 3D geometry from simple images. Still, dedicated 3D scanners like Revopoint offer unmatched detail and performance. Mastering this tool took me weeks back when I first started, but now I was able to get results on my first attempt.

3D Models

• 🔗 Tinkercad cube and sphere STL
• 🔗 MakerWorld generated bust STL
• 🔗 Scanned 3D model optimized STL