5. 3D Scanning and Printing¶
Weekly Assignment:¶
Group Assignment:
- Test the design rules for your 3D printer(s)
Individual Assignment:
- Design and 3D print an object (small, few cm3, limited by printer time) that could not be made subtractively
- 3D scan an object (and optionally print it)
3D Printing¶
Coincidentally, Mr. Dubick taught a lesson on advanced 3D printing settings in his Biomedical Engineering class the day we started this week of Fab Academy. I found the content he covered to be very relevant, so
3D Printing Process:
- 3D model file: Begins with a digital 3D model of the design
- Slicing: The model file (.stl) is then split up into very fine layers in a software program called a sliver. The output from the “slicer” is gcode, which tells the 3D printer how to move and where to lay down material
- 3D Printing: Once loaded, it will begin the production process one layer at a time
- Post-Processing:
Types
- FFF (Fused Filament Fabrication): The most popular type done by melting plastic filament and depositing it onto a print bed, where it solidifies. Additional layers are printed on top of each other until the model is complete. Some plastic types include: PLA, ABS, and composites that combine plastic with wood, copper, bronze, and several others
- SLA/DLP (Stereolithography/Digital Light Processing): Uses light sources to cure a liquid photopolymer. A print bed is submerged into resin and the light is used to solidify certain areas. Once a layer solidifies the print bed will pull it off the bed and move, allowing the light to sure the next layer
- SLS (Selective Laser Sintering): Works in a similar manner to SLA/DLP tech but the light source is used to fuse powdered material together. Most common material is nylon
Design Considerations:
- Overhangs: Needs at least a partial surface or layer to print on top of, so there are problems when printing steep overhanging features. The larger the overhang angle, the poorer the surface quality
- Bridging: Essentially an overhand but the printer bridges across two points. The larger the bridge, the poorer the quality
- Tolerances: Determines how accurate or precise you need to be and is defined by you or your application
- Accuracy: How close a measurement is to true-value
- Precision: Measures the repeatability or consistency
- Wall Thickness: Wall thickness may vary (only 20-30% of the inside of a print is PLA)
Troubleshooting:
- Not extruding at the start: If the first level is already warping and curling or is not flat, stop it early
- First layer not sticking to bed
- Under extrusion: Does not extrude enough plastic, gaps between perimeters and infill
- Over-extrusion: Printer extrudes too much plastic, prints look messy
- Gaps in top layers
- Overheating leads to deformation
- Stringing
- Layer Shifting: Looks like it shifted halfway through the print, it literally did
- Curling or rough corners
- Blobs and zits
- Scars on top surface
- Warping–did not get good adhesion on the first layer, then it gets pulled off the bottom when it gets hotter on top
FFF - Fused Filament Fabrication¶
I first looked to Angelina's documentation and she did a circle encased by a dodecahedron. I was intrigued by the general, thing-inside-a-thing idea and came accross Ermal's Jade Ball documentation. I also came across the Chinese puzzle ball Sydney Richardson did. I ended up going with something more like the Jade ball.
Generally, nested structures cannot be made subtractively. Subtractive manufacturing involves removing material from a solid block using cutting or carving. Thus, they require an entry point. But since the inner spheres are entirely enclosed by outer layers, there's no way to access and carve the inner layers using conventional subtractive tools without breaking the outer sphere first. That's why traditional Chinese artisans use a rotary carving technique.
On the other hand, additive manufacturing (what 3D printers do) is a process of creating objects by adding material layer by layer based on a 3D model. The object is built from the ground up, adding only the material needed for the final shape.
Onto the design process, I first made the inner ball, giving it a diameter of 40mm.
To make the outer shells, I set the gap from the prevous outer diameter to the next inner one as 20mm.
Then, I sketched a circle with diameter 30mm on the XY origin plane and extruded it symmetrically so that it could cut through the entire width of the sphere.
The hole felt too small, so I increased the diameter of the sphere to 50mm and the diameter of the cut hole to 40mm. Becauase of parametric design, it automatically updated. I then repeated my process with the other two axes.
I did the same, but with an even larger sphere, so that it totalled 3 pieces.
I faced an issue when trying to have them be within each other, floating. When I adjusted them in Bambu, they would automatically snap back to the plate.
In retrospect, I should have assembled in Fusion then exported the .stl instead of doing that manipulation in the slicing software.
I thought I found the way when I went into Assembly View, but it did not affect the objects placement on the plate. I ended up ditching the outermost layer and just doing a ball within the second shell.
And the print went well.
SLA - Stereolithography¶
At our lab we have a Formlabs resin 3D printer that uses Stereolithography. It works by a process called photopolymerization where liquid resin hardens (or cures) when exposed to UV light. I was told they were able to produce smooth-surfaced objects, so I had high expectations for my print.
Mr. Budzichowski had printed an overhang test for as an example. It was definitely very smooth, and was not subject to some of the inconsistencies I have faced using a 3D printer.
I first had to download PreForm. Then, I found a file of what I wanted to print out. I decided to go for this cute charm. I thought it would not need supports because it could be completely flat. Pre-form auto generated supports for me and I was initially confused. But Mr. Budzichowski explained that it was because the pieces needed to be slanted so that there was space for the resin to drip out. If not, it would pool and that would ruin the print.
With everything set up, I sent it to the printer. When I went to press start on the printer, I saw that it had already started to prepare for the print, so all I had to do was wait.
Note: The preparation time is not included in the estimated print time. The estimation itself is also severly inaccurate. It takes around 20-30 minutes for it to start printing. Then the print itself is usually an hour or more than what you're told it will be. The screen on the printer might tell you '18 minutes left' but for me it was stuck on '<1 minute left' for 5 minutes...
After 2 hours, I checked in on the print and this (left) is how far it got. When it was finally done after four-ish hours, the plate moved to the top (right).
We moved it to the Wash Station (IPA Wash or Ultrasonic Cleaner) – This machine uses isopropyl alcohol (IPA) or another cleaning solution to remove excess uncured resin from the print. When we took it out of that machine, the drips were gone and there was only our intended designed on the plate.
After 20 minutes washing, we moved it to the Curing Station (UV Curing Chamber) to fully harden the resin. This ensures the print reaches its final mechanical properties and improves its strength and durability. After curing for an hour, we could take it out.
Just like the printer, it was inaccurate in estimating when we could take it out. Angel and I put it in a 7:09 and it told us we could take it out at 8:09, but we only actually took it out at around 8:20. Not as egregious as before, but still annoying.
Then, I removed the supports. Because it's resin, it was hard to take out and I needed to use a clamp. There were also imperfections in the points where there were supports. I assembled the two pieces and they were perfect.
I really like the design because it can rotate and turn.
I originally wanted to hang it on my backpack as a pendant, but since I scaled it down so much, it was the perfect size for an earring. I used to have an obsession with making wire jewelry. I found my old kit where I had earring hoops, and turned the pendant into an earring, which was super cute.
3D Scanning¶
Qlone¶
I started 3D scanning on Saturday during a debate tournament, so I had to be resourceful throughout the process. I started by downloading Qlone through the App Store on my phone . The app had helpful instructions for how to use it. I didn't want to pay for the premium version, so I used the mat.
I also did not have an ink printer on hand with me but I still wanted to be productive, so I scanned it with the QR code open on my Ipad. It worked surprisingly well, in my opinion (especially after seeing the terrible outputs I got even using the paper QR code later that night).
But it was still imperfect and didn't really get the shape correct, which I blamed on the fact that I didn't print out the code.
Once I got home, I tried scanning one of my brother's chess pieces, but it turned out oddly shaped and discolored.
I thought it was because I did not do the 4 layers of scanning Qlone wanted in the right order. I would go 2,1,3,4. That definitely affected the projection of what the object looked like. I changed my ordering to 1,2,3,4, top to bottom.
I also thought it might be because the figure was too complicated, so I switched to a different, simpler object. It was literally a cylinder, but the scanning app made it a cone somehow.
After all this, I decided to ask chat about what I might be doing wrong. Here is the downloadable conversation. Originally, I was worried it was because my phone was not stable enough. I had been holding it using my hands because I didn't have a tripod or anything to stabilize it. Chat agreed but thought it might also be because my surroundings were too cluttered and that was affecting what the software considered to be part of the object.
I moved to a cleaner table and had my brother help by holding my phone (this did not help because his hands were shakier than mine). I tried this hand sanitizer, but I also got the weird cone top.
Then, I tried going even simpler, with an object that did not have that much height. I also found out that it was better to from bottom to top: 4,3,2,1 by what I defined as layers. It turned out much better. Though, for comparisons sake maybe its better to judge using the same object rather than using different ones after each previous one fails.
I also tried a highlighter cap to see if a little height would work when something much taller than it was wide would not. It turned out okay.
Finally, I did my hairbands and this result was probably the best out of everything I've done so far.
Polycam¶
I also downloaded Polycam onto my phone. Its instructions were also pretty much self-explanatory, though it did give a brief tutorial. I scanned my image using photos. I think how it works is that while you are moving your camera with the photos option started, it takes images at a certain interval and compiles them to generate a 3D model. I was surprised it did not require much else, but not surprised that it turned out a bit off. I think the more photos you use the better.
Group Assignment¶
This week's group assignment is to test the design rules for your 3D printer (Bambu A1 mini and Formlab resin 3D printer). My group focused on a torture test, infill tests, and overhang tests for our 3D printers, using FFF 3D printers. Our documentation can be found here.
Individual Contribution¶
I worked on identifying and defining key terms related to 3D printing. I also did the all-in-one torture test for the 3D printer and discussed my findings.
Terms¶
Design Considerations:
| Term | Definition |
|---|---|
| Dimensions | Measurements of a print in the X, Y, and Z axes |
| Clearance | The space between parts to ensure proper fit |
| Wall Thickness | The thickness of the outer walls of a print |
| Anisotropy | A material property where strength differs in X, Y, and Z directions |
Printer Components:
| Term | Definition |
|---|---|
| Extruder | The assembly that feeds and melts filament before deposition |
| Hotend | The heated part of the extruder that melts the filament |
| Nozzle | The tip of the extruder that deposits melted filament |
| Build Plate | The surface where printing occurs |
| Cooling Fan | Blows air onto the filament to solidify layers quickly |
| Filament | The material used in FDM printing (e.g., PLA, ABS, PETG). |
| Resin | A liquid photopolymer |
Print Layers & Structure:
| Term | Definition |
|---|---|
| Layer Height | The thickness of each layer in a print; smaller heights yield finer resolution |
| Perimeters (Shells) | The outermost layers of a print |
| Infill | Internal structure of a 3D print; affects strength, weight, and print time |
| Bridging | Printing filament in mid-air between two points without supports |
| Overhang | Any shape that extends outward, above the previous layer, and doesn’t have direct support |
Support & Adhesion:
| Term | Definition |
|---|---|
| Supports | Temporary structures that stabilize overhangs and bridges, removed post-printing |
| Tree Supports | A branching support structure that reduces material usage and is easier to remove |
| Bed Adhesion | Methods to keep the first layer attached to the build plate |
| Raft | A thick, sacrificial base layer that improves adhesion and levels the print |
| Brim | A thin perimeter around the base of the print to enhance adhesion |
| Skirt | An outline printed around the object to prime the extruder before printing |
Print Speed & Motion Control:
| Term | Definition |
|---|---|
| Print Speed | The speed at which the nozzle moves while printing |
| Travel Speed | The speed of non-printing movements |
| Acceleration | The rate at which printhead speed increases |
| Retraction | A setting that pulls filament back during non-printing movements to prevent stringing |
Common Print Issues:
| Term | Definition |
|---|---|
| Over-extrusion | Too much filament is extruded, causing blobs and rough surfaces |
| Under-extrusion | Too little filament is extruded, leading to weak and incomplete layers |
| Stringing | Thin strands of filament between parts due to improper retraction settings |
| Ghosting/Ringing | Vibrations causing repeated echoes of a feature |
| Warping | The curling of a print’s edges due to uneven cooling |
All-in-One Torture Test¶
For my work, I referred to Angelina, Collin, and Kabir's group work as well as Richard and David.
I downloaded the .stl file for the all-in-one torture test here. I was originally thinking of using a Prusa printer, so I opened the .stl file in PrusaSlicer.
I sized it down 80% before slicing it, but I got an error about overhang, which was to be expected.
Using the Prusa, I would constantly get issues with my print, such as the terrible adhesion that lead to clumping because the nozzle would drag the filament that was already laid down.
So, I switched to a Bambu printer. Because they are faster, I was less limited by time and chose to retain the original size of the design. My print got through 60% until it ran out of filament. I thought I had enough, but the machine gives an alert when there is around a meter left, and does not print the rest.
I switched out the filament spool with another one that was the same color, but full and would definitely be able to finish my design. However, the new layer with the new filament did not add smoothly to my design, also clumping after a few passes.
I restarted the print on the Bambu again, and it went perfectly!
There were three main parts on the torture test that I noted. First was the overhang, how as the angle increases, the the surface becomes less smooth because the lines are spread farther apart. It made sense, because the printer layers the filament in horizontal layers. At around 50 degrees, I saw a notable difference in the surface quality.
Next, also on the overhang, I noticed that some of the filament underneat the sections that are very angled would sag and create a bump on the bottom surface. It was all fine until 80 degrees. This could also be attributed to the widening distance between each line and less surface of a layer connected to the one below it. Also gravity.
On the bridges, I also saw some sagging especially on the longer ones that have more distance between each of the poles that offer supprt. It was at around 0.86 mm that I started to see the filament dropping.
Overview¶
This week was pretty light again. Like vinyl cutting and CAD, I went back to a practice I was already familiar with. I was alreay familiar with subtractive and additive manufacturing, but it helped to have an assignment where I had to clearly differentiate the two and create something that transcends the abilites of subtractive manufacturing using additive manufacturing.
My favorite part was definitely making my SLA print something I could use in daily life. Most of the other things I have created for fab have just been to fill in a weekly requirement or build to my final project that I don't really apply in my personal life. I also originally thought 3D printing was limited to plastic filament, but I was blown away learning about more types. The only thing that was really annoying was having to wait so, so long. It really does not make sense waiting hours for something I can fit on my fingernail.
And then scanning was also something new I did. I honestly expected it to be much harder to scan and I doubted how accurate it could be at first. Qlone (7/10 I'd say) worked pretty well compared to Polycam (4/10), though neither were perfect. It was interesting how Qlone is exclusively a phone app, which I feel like I haven't seen for Fab so far.
For the group project, I already knew about, and have been victim of, various 3D printing issues like unsmooth surfaces, warping, etc. It was helpful to make a reference object to simulate what certain overhangs and bridge lengths would work and when the machine would be pushed too far.