A mathematical model and computational implementation of a digital twin for a bioresorbable polylactide (PLA) bone scaffold with a functionally graded triply periodic minimal surface (TPMS) architecture are developed. The model comprises four state variables (polymer mass, osteoblast population, growth factor concentration, and newly formed bone mass) and a system of reaction–diffusion equations adapted to the 3D geometry. For the first time, it is shown that under certain physiological conditions, a simplified local (spatially homogeneous) version of the system exhibits chaotic dynamics resembling the Rössler, Aizawa and Rucklidge attractors. This finding allows interpreting unpredictable fluctuations in resorption and osteogenesis rates as a manifestation of deterministic chaos, which should be taken into account in clinical prognosis. A sigmoidal spatial porosity gradient is proposed to mimic the cortical–trabecular transition, and stable dynamics are demonstrated for the corresponding full 3D spatial model.
Science Journal 