Maker Chair Workshops

A one-week workshop asking a simple question: can you 3D-print a chair on an 18×18×18 cm build volume?

role: Group work · concept author (Organic Joints)
location: Warsaw Faculty of Architecture
org: KNAI · Information Architecture Research Group
tutors: Krzysztof Nazar
tools: Rhino · Grasshopper · Prusa Mini (3D printing)

Overview

The project started from the idea of building a chair that could not be made with traditional woodworking or sheet-metal bending. The only constraint: every part had to be 3D-printable on our Prusa Mini (18×18×18 cm build volume). A chair is roughly five times that in every direction, which is precisely what made the brief interesting. With one week on the clock we split into three groups and spent the first two days on research, each group taking a different theme.

Group 1 studied topological optimization in architecture (outside of engineering and construction). Group 2 explored modular, discrete architecture and new prefabrication methods, drawing on Philippe Block (ETH Zurich), Gilles Retsin and Kengo Kuma. Group 3 looked at Industry 4.0 in architecture (Thomas Bock), mass customization and open-source design.

My contribution was the first concept, Organic Joints. The logic: keep the wooden sticks uniform in length and let the 3D-printed joint absorb the curvature of the seat. Curves were divided by a sphere to distinguish joints (inside) from pipes (outside); cutouts on the intersecting parts formed the connection between them.

Group 1 · Organic Joints (my concept)

Curves were first sorted by length: 30 curves in <3.5, 5>, 23 curves in (5, 7.5>, and 55 curves in (7.5, 10>. Dividing the pipes into three lengths, a compromise on the original single-length idea, made the script more reliable and reduced structural instability. The final seat adds a denser top layer of wooden branches for comfort, supported from below by identical sticks meeting at 3D-printed organic nodes.

Group 2 · Polygon + screw connection

A surface-based design with a hexagonal pattern projected onto the seat geometry. Six-sided 3D-printed polygons are joined with bolts and nuts; sockets on both sides of each polygon hide the fasteners and prevent movement. Boundary cells outside the surface were trimmed. The approach worked cleanly until the surface bent drastically, at which point the projected pattern could no longer be applied consistently.

Group 3 · Puzzle intersecting structure

A puzzle-like assembly reinforced by intentional deviations on the intersections, raising friction between parts. The first attempt failed: pieces slipped apart and only held once glued, which rather defeated the point of a friction joint. The second attempt used a list-based scripting approach (sorting, replacing and swapping elements) and became the final design for this group.

Group 2 · exploded diagram of few layer panels