This week’s group assignment covers testing the design rules for the Prusa Mk4 in our lab. I was already familiar with 3D printing and testing 3D printers capabilities before this group assignment. But, I still learned a few new things. We’d gotten the Prusa MK4 recently, so I hadn’t actually printed with it before. This assignment was a good introduction to it’s capabilities and settings. In addition, I got to test the “Input Shaper” in Prusa Slicer, which reduces printing times by quite a lot.
I started this week by printing one of the test pieces provided by Fab Academy. It was the clearance piece. I first sliced it in Cura, and I decided to try the “Normal 1.5 mm” resolution setting. I printed with supports, and I was curious as to how the clearance would turn out.
The print ran without issues.
Printing the clearance piece
Although there were problems with the output. The clearances were close to their real-world values, since the 1mm was flowing freely and the 0.1mm was basically fixed in place. But due to the layer height being 1.5mm, it actually lost it’s dimensional accuracy. On top of this, the surface quality was really poor, especially on the sides with supports.
With this knowledge in mind, and knowing that I got familiar with the specific 3D printer in the lab, I moved onto my custom piece.
A workspace I’ve always wanted to explore further in Fusion was Generative Design. I messed around with it for a bit in the past, but I did not utilize it to create actual prototypes. In the course of Digital Fabrication, I want to explore it further and actually create things with it.
The process starts with us creating the base pieces in the design environment.
Then, we move on to the Generative Design environment. Here, the workflow follows the icons on the top from left to right. We assign certain pieces as “Preseve Geometry” (green) and some as “Obstacle Geometry” (red).
Fusion will try to create a design based on the parameters we will give it, connecting the green pieces, while avoiding the red pieces.
Now comes the part where we add the “load case”. We basically give Fusion where and how much load is going to be applied to the piece, for it to use in the simulation. For my example, I chose 1kg of load to be carried by this piece.
Then, we can give it more constraints like “fixing” a piece in place, to simulate it being mounted on a wall for example.
In my example, I want the big horizontal circle have a pipe running through the middle. That’s why I added a red obstacle geometry there, and also fixed it.
Also, we have to define the “objective” in the top panel. This is what Fusion will aim to do with the design. For example, I chose “minimal mass”. Now the program will try to design a piece that can hold the specific load case I’ve given it, while using the less amount of material. Basically designing a strong but light part.
We also choose a manufacturing method in this stage. I’ve chosen “unrestriced”, since I will be 3D printing it. But there are options to choose Milling, or Additive Manufacturing for example.
After this, we choose the material and the study will use. I chose “Nylon 11” for this, since it the closest option in Fusion to a 3D printed plastic. But I was aware that this will not give me very exact real world conditions.
After we start generating, the results start rolling in. It generates a few different results for us and asks us to choose. Here is the option I chose.
I sliced it in Cura and started printing. Since my previous trial with 1.5 mm resolution didn’t go as expected, I chose 1mm (Fine) and 0.8mm (Extra Fine). So I printed two different options to try and see the differences between them.
After printing and cleaning up, here are the results. The one on the left is 1.0 mm(Fine). The one on the right is 0.8mm(Extra Fine)
Cleaning up both was very hard. Since it had to use supports, the surface quality in these parts were very bad. I could not find a way yet to make the supports not destroy the surfaces.
Although the 0.8 mm print was better than the other one, in terms of dimensional accuracy.
Since this design was created with the purpose of carrying 1 kg of load. I wanted to try and see if Fusion and I were successfull in creating a part that lives up to this expectation.
I used a water bottle as a weight. I first started with a smaller weight of 860 gr.
The part succesfully carried the load without failing.
860gr Load Test
I increased the load with an improvised contraption, to see if it fails around 1kg.
The part passed this test too. It can actually carry more than 1kg.
1100gr Load Test
Since it was apparent that the part was succesfull in carrying the expected load, the next thing I wanted to see was where it would fail when applied enough force. So I applied as much force as I could, until the part snapped.
My aim was to determine the weakest point of the structure.
Structural Failure Test
Interestingly enough, the weak point was not the generatively designed connection that Fusion designed, but the initial circular geometry that I had given it.
I suppose this means that Fusion succeeded in designing a structurally sound part.
For 3D scanning, I tried the new Creality Ferret we got in the lab. Compared to other 3D scanners like the Artec Leo, Creality Ferret is a much cheaper and lighter alternative. And it works by downloading an app on your phone and hooking it up to the device. I was excited to try and see how well this works.
The device is stored as disassembled. The assembly process takes a few minutes.
Once assembled, I placed my phone in the holder and plugged it in. To use the scanner you have to download Creality’s scanning app. The app gives you a quick intro on how to choose proper scanning settings, and what you can scan. According to this, organic shaped, small to mid size objects with non-reflective surfaces(human face, statue, carving etc.) are suitable. Large objects, objects with too much flat faces and shiny objects are not suitable.
It was a sunny and warm day at Helsinki so I went outside to look for objects to scan. I set the scanner to geometry mode. The first object I tried was this wooden box with things inside (bee-related stuff?). However, the scanner had difficulty due to all the through holes and complex geometries in the front face. I believe shadows might also have played a part in it. It was not picking up any geometry in the front face.
Then I moved on to a bird house. However, this one had too much flat surfaces. The app prompted me to try objects with rich surfaces.
After a few more failed tries, I settled on a tree. It had rich surfaces and I wanted to try the texture scan mode. I chose that in the app and started scanning. After going around the whole tree, the scan went though nicely.
After the scan is completed, the following processes are done in the app:
Below is the final result. There are some holes in the mesh and I could have scanned it better by spending more time in specific faces and getting more angles in. But I think as a first experiment, it works well.
I also wanted to try the geometry scanning mode. So I chose an object with more geometric features and less surface texture: the card reader at the door. After scanning, here is how the mesh turned out:
The scan turned out well in general, but it had one big problem. There were unwanted geometries protruding from the right of the card reader. I could have fixed these later after exporting, but there is also a way to do it in the app. After the optimization phase, you can use the eraser to get rid of any part of the geometry. It is a bit hard to use, since it depends on what angle you are looking at the object, but it’s a good tool to delete large parts from a scan.
After cleaning up, I continued with meshing and color mapping. Here is the final result:
This was a good exercise in experimenting with the Creality Ferret. I was impressed by it’s ability to scan complex geometries, considering it’s size and price range. The process is a bit slow, as the meshing and optimization depends on the performance of your phone. The scannable object range isn’t as big as an Artec Leo for example, but I think for the price it gets the job done.