Week 3: Computer-controlled cutting

Individual assignment: parametric hex snowflake kit

Task and why a snowflake kit

The individual brief asks for a parametric construction kit that accounts for laser kerf and can be assembled in more than one way without glue. I wanted something small enough to iterate on one sheet of plywood but still teach me the same lesson as a big press-fit project: if board thickness or kerf changes, I should change one variable, not redraw every slot by hand.

I chose a six-fold snowflake: an inscribed hexagon hub with six rectangular spokes that slot into matching partners. Chamfers on the tabs make the first push-fit less brutal. After the cut I could stack the pieces flat, fan them into a star, or build the full 3D snowflake shown in the hero photo—it is the same kit, different assemblies.

How this page meets the assignment

Mapped to the Fab Academy computer-controlled cutting checklist:

Parametric design in Onshape

I modeled entirely in Onshape as a single sketch driven by variables. The screenshots follow the order I actually built it: hex hub first, then spokes sized with kerf, then chamfers, then export.

Variables I used (names match my feature tree):

• diameter — controls the inscribed hexagon (hub); I started at 40 mm for the inner geometry.
• kerf — width of each rectangular spoke tab, tied to laser kerf compensation from the group tests.
• length — overall length of each rectangular spoke arm.
• chamfer_distance1, chamfer_distance2 — drive the chamfers on joint edges so press-fit assembly is easier than square corners.

Step 1 — inscribed hexagon hub

I sketched a regular hexagon from an inscribed circle and bound its size to diameter so the whole hub scales together.

Step 2 — six spokes with kerf and length

I added six rectangles for the snowflake arms, set width = kerf and length = length, then constrained each rectangle to the hex hub with coincident and center constraints so the pattern stays symmetric when I nudge parameters.

Step 3 — chamfer variables and geometry

Square slot corners fought me on the first dry fit, so I introduced chamfer_distance1 and chamfer_distance2 and applied chamfers to the joint edges instead of redrawing the slots.

Step 4 — DXF export for the laser

From the finished sketch I exported a flat DXF for the lab laser workflow (same file linked under Design files). I checked scale and units in the export dialog before walking the file to the cutter PC.

Laser cutting

I transferred the DXF to the laser workstation, picked cut settings for our approved plywood (power/speed from lab notes and the group characterization table), and ran the job. The clip below is the cut on the machine. Useful when I need to remember engrave-vs-cut order and how long six duplicate spokes take on one sheet.

Assembly and finished snowflake

Off the bed I deburred tabs with a knife, then pressed pairs together starting from the hub. Chamfers helped more than I expected. The square-edged trial in CAD felt fine until real kerf tightened the slots. The assembled star and the final 3D snowflake are different views of the same kit.

Finished 3D snowflake

Design files (download)

Flat layout exported from Onshape. If the browser opens the DXF instead of downloading, use “Save as”.

• Hex snowflake construction kit (.dxf)

Reflection

Building the kit in Onshape forced me to treat kerf as a real dimension, not a note on the group page. Linking slot width to kerf meant when the group comb test nudged my value, I could regenerate the DXF instead of hand-editing six slots. Chamfers were the other lesson: parametric sketches only stay pleasant to assemble if you plan for entry angles, not just nominal width.

What I still owe this week is the vinyl cutter part of the assignment and tighter numeric rows in the group table (power/speed/kerf numbers next to this exact plywood lot). For the final project I may reuse the same hub-and-spoke pattern for decorative panels once I know the machine settings by heart.