From the earliest records on skeletal research, there has never been a doubt about the interrelation
between structure and function of bone. Today, there is a consensus on bone functional
adaptation and optimisation to its mechanical environment. However, it is rare for bone to be
modelled using a structural mechanics as opposed to a continuum mechanics approach. Yet,
there is a need for a considered compromise between the high resolution of microscale continuum
models and the computational e ciency associated with macroscale continuum models.
This thesis aims at developing novel translational and cost-e ective computational methods
for the prediction of bone architecture and fracture, based on a structural representation of
bone where cortex and trabecular bone tissue are represented using idealised shell and truss
or beam elements, respectively. The PhD project approaches this topic from four di erent
yet complementary viewpoints, including mesoscale predictive modelling of bone structure using
phenomenological rules for bone remodelling, simulations of bone fracture, manufacture of
physical mesoscale bone models, and implementation of a surrogate model for bone remodelling
at the microscale poroelastic scale.
The methods implemented as part of this thesis allow for the prediction of bio delic inner
structural architecture in an entire long bone such as the femur or the tibia based on loading
associated with daily living activities, rigorously computed using musculoskeletal simulations.
Reorientation of trabecular elements is e ciently predicted using metamodelling or poroelastic
mechanisms. The approaches presented here also allow for prediction of bone fracture onset
and progression until complete structural failure, as well as assessment of the in
uence of
subject-speci c activity regimes and bone outer geometry on structural organisation and failure
behaviour. Physical models are produced using selective laser sintering based on the modelling
results, for applications in testing of protective equipment mitigating trauma.
Informed recommendations are drawn from this work. The importance of multiple physical
activities, speci cally sit-to-stand, to direct femoral architecture is established. The superiority
of elasto-plastic over purely elastic bone material formulation for fracture prediction is assessed
and the addition of a separate shear strain criterion is shown ine ectual. The strong in
uence
of the manner of load application in side fall simulations is also highlighted.
These methods provide a number of tools to support investigations in various domains of
skeletal biomechanics including rehabilitation, implant and prosthetic design, study of clinical
conditions such as osteoporosis and osteoarthritis, design of sca olds for tissue engineering, and
development of injury mitigation measures.