Synthetic bone analogues under load

The brief

Which plastic (HDPE, acetal or nylon) best matches the femur of an average woman under load, as a material for 3D-printed bone replacements? The project was motivated by the underrepresentation of women in orthopaedic research.

What I did

I planned and programmed the tests on an Instron ElectroPuls E3000: compression to yield, 1,000-cycle cyclic loading and 40,000-cycle fatigue tests at a 1.5 kN load range. Each sample produced more than 50,000 data points, which I analysed in MATLAB. Early samples were ejected from the machine under load, so we reduced their length from 3 cm to 1 cm.

I proposed adding finite-element analysis and modelled scanned female femurs in each material to check whether the small-sample results said anything about a whole bone. We were the first group to include FEA in this project.

Measured yield points were lower than published values for all three materials. I identified three systematic causes: localised buckling, surface irregularities that remained after sanding, and friction between the sample and the plates, which creates a triaxial stress state. We also carried the instrument uncertainty through into our stress and strain values.

Result

None of the three materials approached the strength of cortical bone. The project won the UCL Instron Prize for Best Group Research Project 2025 (89%), and we presented the poster to staff and to engineers from Instron. The following year the department asked me to help mark the next cohort on the same project.