A Master's project to create a simulation that would optimise the volume fraction distribution of a lattice structure knee implant design, such that bone remodelling capability was maximised subject to the strength requirements of in-knee loading conditions being met. Three methods were explored: 'non-penalising stiffness-maximising topology optimisation'; 'looped stress limiting process'; 'non-penalising stress-constrained topology optimisation'. Designs were achieved from the first and second methods. The looped process gave different designs subject to initial conditions and iteration numbers. The designs were compared to a 'control' design with perfect bone remodelling capability but no strength consideration. The failure load of the 'control' design was ~600 N. The failure loads of the designs achieved from the simulations were much better, at 800 - 1000 N. The best design was 'stress limited control' - a single iteration of the looped process upon the 'control' design. The percentage of implant elements predicted to be at risk of in-knee failure fell by ~44.9% upon the 'control' design with only a 12% average deviation from ideal-remodelling modulus targets.