3D Scanning & Printing

Group project portion. We kept it simple and hands-on. For printing, we mainly chased two questions: how far can we push overhangs and how much does infill vs. wall count change the feel. Everyone printed small test sticks, swapped slicer settings, and compared by touch. For scanning, we tried a turntable and quick phone captures—fun, fast, sometimes messy. Below are my notes and photos in the order I made them.

3D Printing

My checklist this week was very practical:

Overhang ladder: when the angle got too steep it sagged unless I slowed down and pushed more fan. Part orientation made the biggest difference.
Infill vs. walls: for small parts, adding perimeters changed stiffness way more obviously than cranking infill from 10% → 30%.
Fit chips: tiny step gauges helped pick a clearance that “just snaps” without sanding.

Print notes and jigs from the bench — 1. Early print notes and test jigs spread across the bench.

Tolerance bar set, first pass — 2. First tolerance bar set to check basic clearance values.

Tolerance bar set, refined — 3. Refined version with more useful step sizes.

4. Numbered clearance chips for quickly testing snap fits.

Clearance from 0.10 to 1.00 — 5. Clearance chips from 0.10 to 1.00 mm, used as a physical reference.

6. Detail coupon exploring thin walls and tiny features.

Hole spacing probe — 7. Hole-spacing probe to see what still prints cleanly.

Overhang comb—where it starts to droop — 8. Overhang comb showing exactly where unsupported angles begin to droop.

Vent pattern test — 9. Vent and perforation pattern tests.

Snap fits and tiny grids — 10. Tiny grids and snap-fit experiments printed in one go.

Assembly mockup on the desk — 11. A small assembly mockup combining several of the test pieces.

After that warm-up I revived my home printer. I used a tiny pumpkin as the “hello world” to check motion and clearances—low risk, high joy.

12. First successful movement after bringing the home printer back to life.

Fresh pumpkin figure — 13. Freshly printed pumpkin model straight off the bed.

14. Articulated pumpkin legs—clearance landed right in the sweet spot.

Then a thin-walled lampshade. Goal: let light pass through, not brute strength. I went with 0.2 mm layers, steady fan, and very light infill.

Lampshade render — 15. Lampshade concept render with a gentle spiral pattern.

On the mat before printing — 16. The lampshade flattened onto the build plate in the slicer.

Printing ultra thin walls — 17. Printing ultra-thin walls—beautiful, but the cooling needs constant attention.

18. Finished lampshade glowing softly in a dim room.

19. Quick night test of the lampshade with real light.

Next I modeled a one-piece articulated finger. Clearances are built into the CAD so it prints in one go and bends right off the bed.

Finger mechanism CAD — 20. CAD model of the articulated finger mechanism.

Exploded joints — 21. Exploded view of the joints and hinge clearances.

Printing in place — 22. Finger printed in place, hinges already separated.

23. Fresh off the printer and ready for range-of-motion tests.

24. Flex test—full range of motion with no sanding needed.

I also played with a self-locking bracelet made from repeating links. Printing in place mostly worked; the first layer drifted a bit but it still clicks together.

Bracelet link idea — 25. Early sketch of the bracelet link geometry.

26. Interlocking form that lets the bracelet self-lock.

Spiral study — 27. Spiral and curvature study for how the bracelet wraps the wrist.

Printed ring, orange PLA — 28. First ring of links printed in orange PLA.

Large ring on the floor — 29. Larger ring layout on the floor to check overall length.

Trying it on — 30. Trying the bracelet on—fit is close but workable.

Desk shot with bracelet — 31. Finished bracelet resting on the desk.

Support fail, noodle city — 32. A classic fail: old dragon model printed without supports → instant spaghetti.

3D Scanning

I bounced between a turntable scanner and Polycam on my phone. Turntable looks neat but struggles with plush and the underside; Polycam is forgiving and great for quick context. Not perfect, but fast enough to be fun.

Turntable scan of a Miffy doll — 33. Miffy on the turntable—pose is cute, underside still occluded.

Turntable scan preview — 34. Turntable preview showing patchy coverage on the bottom.

MIT beaver plush scan — 35. MIT beaver plush: recognizable shape, but fur detail gets smoothed away.

Head scan assembly preview — 36. Quick head-scan preview assembled from several passes.

Live head scan; tracking drift — 37. Live self-scan with noticeable tracking drift—funny but not printable.

38. Fragmented head mesh: still recognizable in a glitchy way.

39. Polycam reconstruction of a small scene, processed on the phone.

Polycam room capture with me in scene — 40. Room-scale Polycam capture with me in the scene—fast and robust enough for notes.

Quick printer settings that worked for me

PLA · 0.4 mm nozzle · 210 °C / 60 °C · 0.20 mm layer · 3 walls · 10–20% gyroid. For overhangs: slow down + more fan; for stiffness: add walls before adding infill.