Computer simulation methods were developed to study time dependent mechanically induced biological processes in bone around orthopaedic implants. The thesis shows that the methods can be used to pre-clinically test prosthetic designs, to guide the design process of new implants and to guide, or perhaps sometimes replace, animal experiments, to study the effects of fixation of orthopaedic implants. Comparison of the results of the numerical simulations with experimental configurations, clinical trials or animal experiments, were promising or even impressive in some cases (Chapter II).

The thesis is focussed predominantly on total hip arthroplasty (THA), and it is shown that mechanical aspects (by means of bone-remodeling processes or proximal bone atrophy) do have an important contribution in the failure mechanism of THA components (chapters II, III, IV and V).

Chapter II presents a comparison between an animal experiment, studying the effects of adaptive bone remodeling around non-cemented THA and a computer simulation of this experiment. The numerical simulation features a three dimensional finite element model, integrated with a quantitative adaptive bone-remodeling theory and including adaptations at the bone surface (geometry) and in internal bone morphology (density). It is shown that a realistic detailed prediction of long-term bone-morphology changes can be obtained with such an empirical theory.

In the chapters III and Г a similar model is applied to human femoral hip components, in two dimensional finite element models. It is shown that most stems lead to drastic bone atrophy, probably threatening the long-term stability of the THA. In a parametric analysis it is shown that either a low modulus material (chapter III) or a smooth stem with a proximal press fit (chapter IV) can overcome this problem. Both solutions do have negative aspects as well. Low modulus materials lead to high interface stresses (chapter III), so the geometry and interface strength must be optimized to guarantee a good initial stability. Press-fit or debonded implants provoke relatively large motions at the bone implant interface (chapter IV), which is highly dependent on the precise geometry, in particular the taper, of the implant. Chapter V presents a three dimensional model of a proximal femur with an uncemented stem in one particular configuration, constructed from a CT-scan. Again the amount of bone atrophy is studied with adaptive bone remodeling theory, in particular with respect to the size of the threshold levels in the remodeling description.

Chapter VI studied more fundamentally the behavior of the adaptive bone remodeling simulation models. The remodeling rule applied is, in fact, a self-optimization process, with a local objective. It is speculated that the existence of cancellous bone as a porous structure can be considered as a 'fractal' (a mathematical figure with self-similarity). The adaptive bone-remodeling process can be considered as a multi-variable dynamical system with a positive feedback:​ denser bone attracts more load and becomes even denser. This mechanism results in a discontinuous patchwork of dense bone with a maximum possible density alternated with complete loss off bone (a void). Such an irregular structure as a result of a positive feedback system is characteristic for self-organizing systems. It is hypothesized that trabecular bone is a chaotically ordered structure, whereby the rather irregular (coherent) morphology with its (rares, plates and struts can then be determined completely from parameters affecting this process and characterized by the mathematical formulations, and the external loading conditions. This hypothesis is in fact not so different from the original 'T^aw of Bone Transformation" (Wolff, 1892) and can be regarded as its modern interpretation. This hypothesis is not only of great importance for the fixation of orthopaedic implants, but also for its proposed role in the development of diseased bone with an adverse morphology (osteoporosis).

Part II of the thesis describes implant interface reactions. The implant/bone interface has a complex irregular shape with properties which are difficult to describe. In a global model, whereby the whole implant is modeled, one can characterize the most important aspects only. It is shown in chapter VII that the effects of debonding and an intermediate soft tissue layer between implant and bone results in an unfavorable effect on the stability (stress transfer and motion). Chapter VIII studies the secondary stability of an implant. Interface debonding and subsequent fibrous tissue formation at the interface were mathematically described and integrated with the finite element models. It seems better not to rely on the stability after debonding of the interface. The results of chapter VIII are very preliminary in this respect, and it is certainly not clear yet why certain debonded implants with an initially growing soft intermediate layer can stabilize on the longer term, while others do not. Well defined experiments, guided with accurate simulation models could probably improve our knowledge relative to this question.