The design development process of athletic footwear is a long and iterative process.
Currently the shoe goes through a design, prototype and mass production phase before the
shoe reaches the consumer. The basis of iteration is the performance of the shoe in the
laboratory and wear tests, which are mainly carried out in the prototype phase. To improve
the performance of the shoe in these tests either better tests are needed to allow a better
understanding of how the shoe performs or the shoes should be modelled and laboratory
tests simulated prior to prototype to eliminate the need for early iteration. This thesis
investigates the possibility and practical applicability of simulating test conditions using
the finite element method in order to predict the performance of modelled shoes.
The investigation is split into three objectives, the replication of existing laboratory
mechanical tests into virtual tests; the development of a technique to develop more realistic
and complex footwear tests and the integration of the technique into a virtual test.
A process was created which identified how the finite element method was applied into a
virtua1 footwear test highlighting how the method will be validated. This process was used
to confirm it is possible to replicate the loading scenarios of simple, repeatable laboratory
mechanical tests with confidence.
Realism and complexity of footwear tests can be improved either by modifying existing
tests or creating a method that develops new loading scenarios. A prosthetic foot was
modelled with success as a method to improve the realism of existing mechanical tests and
increase complexity of future tests. Kinematic and kinetic data taken from the stance
phase of the gait cycle formed a method that can generate boundary conditions for new
virtua1 footwear tests.
The positional measurement method to generate boundary condition was integrated into
virtua1 tests using a transformation mathematical technique. The kinematic data is used to
drive the virtua1 test and the kinetic data is used to compare the results with reality. Initial
results found virtual and reality were visually similar and small differences in the
quantifiable parameters could be accounted for. Modifications to improve efficiency in the
solving of the virtual test and therefore increase the usability of this method in the design
development process were investigated. This was successful and showed it was possible to
predict the performance of the shoe in a realistic and complex test loading situation.
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University