Analysis of push-out test data based on interfacial fracture energy

Push-out testing is frequently used to assess the interfacial shear strength developed at a bone-biomaterial interface during in vivo experiments. The aim of the present research was to assess the in vivo performance of a novel substrate/coating combination and to introduce a more rigorous fracture mechanics analysis of the push-out test data. An adhesively bonded hydroxyapatite (HA), and a Ti-6Al-4V alloy plasma sprayed with HA, were implanted in female New Zealand white rabbits for up to 6 months in duration. After death, push-out tests were carried out and the shear strength was calculated in the conventional way, together with microscopical examination of crack paths. A finite element model was drawn up representing four potential failure mechanisms. The measured "failure shear strengths" in conventional analysis were approximately equal for the two coatings. However, JC at failure calculated from the model was 210 J m(-2) at the novel adhesively bonded HA/bone interface and 5 J m(-2) at a conventional titanium/plasma-sprayed HA interface. The conventional shear strength approach is strongly test dependent, and we believe that the fracture energy approach represents a more rigorous analysis of the real failure criterion in the implant/host tissue structure.