My personal design goal was to start experimenting with flexures and living hinges. I chose a (too) ambitious goal for this: The Umbilic Torus.
This shape has a single continuous surface that twists around mobius-style to form a 3-dimensional donut with a cycloid profile. This served as the inspiration, but not literal guidance for my press fit kit. I (originally) wanted to make a recognizable 3d shape from a single continuous piece.
On shallow joints, chamfering is very important. Unchamfered slots were either too tight to reliably insert into, or too loose to hold the piece after insertion.
Pieces narrower than the width of the corrugation wave are essentially two floating strips of paper waiting to be destroyed. These designs did not survive reinsertion well.
Several geometries involved pushing a pin with a bulge through a hole to create a locking fit. It was very common for the bulge to get stuck between the two cardboard layers instead of going through to the other side.
To achieve a 60 degree angle, I made a hook'n'hole joint with extra hook length. Ideally it would be short enough such that the hook engages at 60 degrees, but this made it too difficult to assemble. A few iterations settled on a hook length that allowed assembly without too much extra wiggle.
I added a meander path to give the strip torsional and bend flexure. Originally I did this by cutting elliptical notches out. Unfortunately, this led to a miscommunication somewhere in the translation: The elliptical portions were interpretted such that the laser spent many extra vector cuts on each one, repeating over those path segments upwards of 8 times. This took extra cut time and lit the material on fire. Changing these to slots made it look less cool, but also reduced how on fire it was.
Singulating the strips was much harder than I expected. The meander path creates delicate binding points that require a lot of effort to individually remove. I tried a more parallel / less manual process of punching the middle pads and then lightly shaking the entire thing. This took care of ~80% of the probem, and left the rest for the manual process. It also attracted the nose of Lola:
Creating the torus from a single continuous piece is infeasible due to the size constraints of the stock and the laser cutter. Chopping this long piece into shorter segments proved to be the main failure point of the project - the butt joint fails catastrophically.