Fabrication#
Data becomes matter again — the round trip’s return leg
Recap#
- 3D printing and laser cutting: what each is good at, and where each fails
- designing for the machine’s constraints rather than designing on screen and hoping
- a fabricated part as the physical end of a pipeline that started with tracking, a rule system, or a sensor
- studio week next: two parallel clinic tracks before the final
Printer or laser cutter, decided first#
A laser cuts a 2D path through flat sheet material. A printer builds up an actual 3D volume, layer by layer. A design suited to one is often completely unusable on the other, so this is the first decision, not an afterthought once you’re at the machine.
Get the file checked before it runs#
The lab is supervised only. Getting your file checked by a tutor before any job runs is the real gate here — not a formality you can skip because you’re confident it’s fine.
Designing around the machine’s own constraint#
A laser needs a kerf allowance built into any slotted joint — cut the slot slightly narrower than the tab that meshes into it, or the pieces won’t hold together. A print needs enough wall thickness, plus the right orientation and supports for anything that overhangs.
From checked file to finished part#
Run the checked file and collect the part — then check it against the digital model before calling the job finished.
What this can build toward: a laser-cut kinetic mechanism#
Gears, a linkage, or a simple moving toy cut from cardboard or thin ply — the kerf allowance built into the joint is what decides whether the pieces actually mesh once they’re off the bed.
What this can build toward: a 3D-printed enclosure#
A housing built around a sensor or small circuit from your physical computing work — wall thickness and orientation decide whether it survives being printed at all, let alone snaps together afterwards.
Next week#
Studio week — two parallel clinic tracks, code clinic with Idris and make clinic with Marisol, to shore up your weaker half before the final