Assignment items

Group assignment
Test the design rules for your 3D printer(s)

Individual assignment
Design, document, and 3D print an object
3D scan an object

Fab Academy Rubric — Have you?

The criteria evaluators look for this week.

Linked to the group assignment page and explained what you learned testing the 3D printers
Documented how you designed and printed an object, and why it couldn't be made subtractively
Documented how you scanned an object
Included your original design files

Tools

OS: Windows
Software: FreeCAD, Revo Scan, Bambu Lab
Machines: Revo Point camera scanner kit
Material: PLA filaments

The process

I have been using 3D printers for over 6 years. I learned many things from YouTube and trial and error. I was comfortable going through this week's assignment.

Group assignment

Musaed AlKout is the one who had the full process documented in detail; I will cover some highlights on my page.

01: 3D printer's design rules

01| From the MakerWorld website I searched for the test prints of the Bambu A1 3D printer

02| The outcome was satisfying

03| The A1 is able to print angles up to 60 degrees. The design looks messy above 75 degrees

04| And hanging bridges up to 24 mm can be built nicely without support

This is good

Feedback: The outputs of the Bambu A1 and PLA filaments are not bad.
Challenge: The rules change for different filaments and printers

Individual assignment:

For this assignment, I will be designing and printing an object that I need for WRO — particularly a prototype of a trophy that I will be giving to our partners and sponsors as a gift and a token of appreciation.

01: Previous 3D designs

01 | 3D design of a sponsor black

02| 3D print of a sponsor black

01 | 3D design of a building unit

02| 3D print of a building unit

01 | 3D design of a gears for conveing prototype

02| 3D print of a gears for conveing prototype

02: Design 3D object and print

As I'm progressing in the projects, the design is getting simpler. After my experience creating a shoulder pad from cardboard in Week 03 — Computer-Controlled Cutting, I thought of making it even simpler — a clip-on device that can be connected to anything we wear, either with a clip or a magnet. This is the form factor that became the housing for my final project.

01| I designed a box to house the PCB and the main sensors → a sliding-in LED cover → clips with magnetic pockets — which, looking at it now, I can see was illogical 🙂 since the magnets weren't secured.

02| Bambu Studio recommended printing at a 45° orientation to reduce support material. I think that was partly because I was using a basic CAD tool to design the shape — TinkerCAD doesn't give you the precision or feature control that Fusion 360 or FreeCAD offer (and I switched to FreeCAD for the parametric clip-case work in W15).

03| By mistake, while taking this picture, I hit the printer's pause button on the touchscreen. After I resumed the print, I noticed the box's shape was distorted and there were weak points along the layer interface — most likely caused by the unplanned pause.

04| This is how it looked at the end of the print. Another reason the distortion may have happened: printing the whole design at a 45° orientation — that's why the highlighted part of the case came out more rounded than the model.

05| The reason these layer lines aren't attached is most likely the temperature change in the filament during the pause — the new layer didn't bond to the previous one because the surface had cooled below the inter-layer bonding temperature.

06| For the magnet pockets, the next revision needs a top cover, and the slicer needs a scheduled pause layer so I can drop the magnets into the clip before the print closes over them — this pause-and-place technique is one of the features that's only possible with additive manufacturing; it can't be achieved subtractively. It became central to the final-project clip case in W15 — System Integration.

07| There's a clear mismatch between the size of the clip and its placeholder. A more precise design tool like Fusion 360 or FreeCAD would have prevented this — especially because both offer an intersection / interference check feature that TinkerCAD doesn't have.

08| This is the lip that the LED cover should slide into. Because the original box was distorted, the cover couldn't slide in fully. I'll also need a locking mechanism so the LED cover doesn't slide back open easily during wear.

09| The size is generally good, but I'll likely change the depth of the box and align the pieces inside in a better way. It all depends on how the PCB ends up looking and on the smallest case size I can reach — finalised later in W15 once the PCB layout was locked.

Not bad for the first prototype

Feedback: It feels good to see my idea come to life, and even though I need to work more on the design, I feel I'm on the right track.
Challenge: The magnet should be strong enough to carry the weight of the box with all the parts in it

03: 3D scan an object

01| I used Revo Point camera for this assignment

02| I first connected the cables to the camera, turn table and the PC

03| Then I open Revo Studio App to start the process

04| I selected Benchy Boat as an item to be scanned

05| I fixed the distance between the turntable and camera as well as the angle of the camera

06| My target is to reach an indicator of good to excellent distance and object clarity

07| I turned on color scanning to make sure the color of the object is taken into consideration, and the camera started to flash

08| Then I started the scanning process and the camera started to build up rendering frames as the turntable kept rotating the object.

09| While I was doing this assignment, my daughter came into the room and accidentally hit the table, moving the object and camera, and the rendered 3D shape was compromised after nearly 6000 rendered frames

10| As a second trial, I chose one of her toys as the object, since it also had fewer cavities and would be easier to render.

11| First I fixed the distance between the object on the turntable and the camera angle

12| turned on the color scanning

13| And after 1900 rendered frames, when I saw the object complete, I stopped the scan

14| As I turned the shape, I saw some holes

15| before filling them, I meshed the shape to create clearer surfaces

16| Here I started to fill in the holes by first selecting them as I kept turning the object in all directions

17| I then exported the file as an STL file to import it to Bambu Studio and prepare it for printing

17| I then exported the file as an OBJ file to import it to Bambu Studio and prepare it for printing while keeping the colors. An STL file would not keep the colors of the object.

18| We fixed the layout of the object

19| After slicing and preparing the object to be printed, the duration of 12 days was unreasonable and only justified by the need to change filament and multi-color print

interesting but limited!

Feedback: It was an interesting process and can be very useful for reproducing spare parts
Challenge: The angle of the object, color, and surroundings will affect the results. When meshing, we do not have the same control over the object as when designing it

Reflection

What worked

Six years of 3D printing experience meant the design-rules testing and printing went smoothly.
Simplifying the final-project housing down to a clip-on box was the right call — this form factor carried through to Week 08 — Electronics Production and the final assembly.
On the second scan, exporting as OBJ kept the object's colours — STL would have lost them.
The pause-and-place magnet idea — drop the magnets into the printed pockets during a scheduled slicer pause, then resume the print — was prototyped here and became the central assembly technique for the final-project clip housing in W15 — System Integration.
Realising the additive-only features (sealed magnet chambers, internal wire channels) that simply can't be made with subtractive methods — this is what justified using FDM for the clip housing over laser-cut MDF.

What didn't

Hitting the pause button accidentally mid-print caused layer-bonding failure — the filament cooled below inter-layer bonding temperature, so the new layer didn't stick to the previous one. The box's shape never recovered.
The first scan was compromised when the turntable got bumped after ~6000 frames.
The multi-colour print estimate came back at 12 days — not realistic for this object.
TinkerCAD's precision limits showed up clearly: the clip and its placeholder ended up mismatched. TinkerCAD doesn't have an intersection / interference check, so size conflicts only became visible after printing.
The LED sliding cover couldn't fit because the original box was distorted by the unplanned pause — and even without distortion, the cover needs a locking mechanism so it doesn't slide back open during wear.
Printing the whole design at a 45° orientation introduced rounding artefacts on features that needed sharp edges — a wrong-default choice for parts with sliding-fit interfaces.

What I'd do differently

Keep the print area undisturbed — don't touch the screen or the table during a print or a scan.
Pick a scan object with fewer cavities and a stable setup from the start.
Check the slicer's time estimate before committing to a multi-colour print.
Use a precision tool (Fusion 360 or FreeCAD) from the start for any case that needs interference-free fits — TinkerCAD's fine for blocking out an idea but not for parts that snap, slide, or hold a magnet pocket.
Override the slicer's default orientation when sliding-fit interfaces are involved — a 0° / 90° print orientation will preserve sharp edges much better than the auto-recommended 45°.
Add the LED locking mechanism in the parametric model itself, not retro-fitted after print.
Plan the pause-and-place magnet layer at design time so the chambers, the pause height, and the magnet polarity are all decided before slicing.

Key learnings

3D printing design rules change per printer and per filament — they have to be tested, not assumed.
3D scanning is sensitive to angle, colour, and surroundings, and meshing gives less control than designing.
File format matters: OBJ preserves colour, STL does not.
Pause-and-place is a manufacturing primitive, not just a trick. Once I understood it, sealed magnet chambers, captive nuts, internal cable channels, and dual-material features became thinkable — all things subtractive methods can't do in one pass.
The choice of CAD tool isn't aesthetic — it's functional. For parts that have to interlock or slide, you need a tool with interference checks; for a quick block-out, TinkerCAD is fine. Picking the wrong tool for the job costs a print.
Inter-layer bonding depends on temperature continuity. An accidental pause or a cold draft can break a print at the layer boundary — and once broken, the part has to be reprinted, not rescued.
This week is where the pause-and-place magnet method that defines the final-project clip housing was first prototyped — even though the W05 print failed cosmetically, the technique it tested became one of the most important assembly decisions in the project.