Multi parameter computational and experimental investigations into the robustness of cementless total hip replacements

Traditional computational and experimental assessments of implant performance are determin- istic; each computational (usually finite element (FE) based) simulation or experiment describes a single situation. While useful information can be gained from these analyses, when the number of variables involved increases, experimental simulations becoming increasingly time consuming and complex. In these cases, computational simulations are increasingly relied upon to predict implant performance. However, even when employing computational simulations to look at the effect of a large number of variables via sweep simulations for example, the problem can become computationally expensive and unfeasible in terms of time required. In the present work, stochastic cornpu tational methods are employed to assess the effect of multi- ple variables on the performance of the cernentless hip replacement. To verify the computational simulations, at each stage of the project, selected scenarios were tested experimentally. To assess implant performance, the following metrics were used: (i) implant micromotion and migration: excessive micromotion and migration are believed to be related to the most common cause of implant failure, implant loosening, and (ii) bone strain; excessive bone strain can result in bone fracture. An initial study on a neutrally positioned stem showed good correlation between the experimental results and the computational predictions. Mesh morphing techniques were employed to allow implant position to change throughout the simulations and assess how this altered the output metrics; it was observed that micromotion and strains generated in the cortex were most sensitive to varus/valgus angle. To further reduce computational expense, a surrogate modelling technique was used to assess the effect of both loading and implant positioning, on micromotion. The surrogate model was verified by selected FE models, placing confidence in the model, and again highlighted that in addition to vertical load, the varus/valgus angle affected the micromotion of cementless implant. Experimental investigations were carried out to corroborate the results obtained computation ally. The novelty of the experimental tests was in the use of an optical system, called digital image correlation (DIC), to measure implant motions and bone strains. This technique enabled non- contact three dimensional measurements to be made. While some qualitative relationships were obtained with FE outputs, good quantitative corroboration between the strain gauge and DIC suggests that DIC is a promising technique for the evaluation of implant performance in vitro.EThOS - Electronic Theses Online ServiceGBUnited Kingdo