The latest documented construction combines bamboo veneer tubes with 3D-printed PA12-CF lugs. That hybrid approach matters: not every part needs to use the same manufacturing process.
A hybrid frame
The long structural members are built from bamboo veneer tubes. Complex interfaces are concentrated in printed lugs around areas such as the head tube, bottom bracket, seat cluster and dropouts.
That gives both materials a clear role. The tubes provide the long connections through the frame, while the printed parts can combine angles, fits and several interfaces in a single component.
The question is therefore no longer how much of a bicycle can be 3D-printed. The more useful question is where the geometric freedom of FDM actually adds value.
Bamboo veneer tubes
The prototypes use MOSO N-Vision bamboo veneer tubes. Diameters of 25, 30 and 40 mm have been explored in the project.
Using tubes keeps the concept recognisable as a frame structure, while the printed joints allow much more freedom than a conventional welded frame.
At the same time, those joints remain one of the main engineering concerns. Fit, bonding area, tolerances and alignment all influence whether the separate parts behave as one structure.
PA12-CF lugs
The functional lugs are FDM-printed in PA12-CF. The documented workflow uses a 0.6 mm hardened nozzle, a heated enclosed print environment and settings aimed primarily at consistent layer bonding and functional strength.
For parts like these, material choice alone is not enough. Print orientation, wall construction, seams and cooling affect how a component can be loaded. A lug that looks strong in CAD may still be poorly oriented relative to the print layers.
The goal is therefore not only to optimise the geometry, but also to account for how loads travel through the printed part.
Bonded joints and tolerances
The bamboo tubes are bonded into the printed lugs with epoxy. The prototypes showed how much small design decisions matter here.
More insertion depth creates more bonding area. Better fits make the joint more predictable. Loose tolerances ask too much of the adhesive layer, while a fit that is too tight makes assembly and alignment harder.
Later iterations therefore explored deeper joints, tighter tolerances and West System 105/206 epoxy.
Parametric in Inventor
The frame is built parametrically in Autodesk Inventor. The aim is not to lock down one perfect bicycle size, but to model the relationships behind the frame.
When the frame geometry changes, tube lengths, angles and interfaces should be able to move with it. The lugs then become part of a product family rather than isolated CAD files.
That also makes a future configurator interesting: not simply selecting an STL, but defining a frame geometry and generating the matching parts from it.
What is not yet proven
Openframe remains an engineering experiment. The combination of FDM parts, bamboo and bonded joints still needs much more systematic testing for fatigue, creep, impact, moisture, temperature and repeated loading.
Failure mode matters as well. A part surviving a short static test does not make it a safe bicycle component.
I therefore treat the material and print choices not as a proven recipe, but as a working research direction that needs to be understood better with every iteration.