Numerous mechanobiological algorithms have been developed in recent years to simulate biological processes such as bone remodelling and tissue differentiation. However, the implementation of such algorithms for the design and testing of orthopaedic devices has yet to be fully realised. In this thesis a mechanoregulation algorithm based on a combined strain-damage remodelling rule together with the integration of a tissue differentiation algorithm is proposed. Such a mechanoregulation approach was tested using a finite element model of a uncemented femoral hip prosthesis under idealized bonding characteristics. Three stem materials (iso-elastic, titanium, and CoCrlvIo) were investigated and compared to previous numerical and clinical observations to corroborate the predictive power of the algorithms.
Results of the bone remodelling algorithm predicted similar remodelling trends to those observed clinically. Namely, that while the use of an iso-elastic stem reduces proximal bone loss because stress shielding is prevented; proximal interface resorption is increased due to damage stimulated resorption. On the other hand, a stiff cobalt chrome stem increases both proximal strain-stimulated resorption and damage stimulated interfacial resorption at the distal tip. Simulations for the titanium stem were predicted to minimise both strain and damage related remodelling. The inclusion of the tissue differentiation simulation predicted bone regeneration for all resorbed interfacial bone through either intramembranous or endochondral ossification, which can be corroborated by clinical observations of implant stabilisation through osseointegration. However, it was predicted that some soft-tissue formation persisted at the proximal medial interface of the iso-elastic stem, following the trend of what is observed clinically; early failure of iso-elastic stems due to excessive proximal soft tissue formation.
In conclusion, by integrating different mechanoregulation algorithms simultaneous predictions of remodelling and differentiation can be achieved. This opens new avenues for computational pre-clinical testing of orthopaedic implants.