engleski

Genetic algorithm in material model calibration for biomaterials behavior simulation

Modeling of biomaterials behavior has anchored its place in ever more extensive field of material research. In order to simulate behavior of such materials, an appropriate constitutive material model must be calibrated to follow real stress stretch relationship of material under consideration. The basic material models utilized for modeling behavior of soft tissues, in this case human cervical spine ligaments, were the ones initially developed for rubber-like materials showing hyperelastic characteristics: well-known Mooney's model and enhanced rendition, a Mooney-Rivlin's material model. Although a portion of these material models are moderately straightforward, commonly placed additions to them comprise of complex articulations with increase in number of material parameters, which are unknowns that have to be identified. The identification of the unknowns of these models requires the utilization of a numerical methodology which can result with the as accurate as possible parameters' set and thus calibrate chosen material model. Genetic algorithm, which is based on evolutionary principles, presents such a method, but in order for it to be efficient in discontinuous, stochastic and highly nonlinear systems, its operators and properties must be specifically developed and tested during material model calibration process.

Genetic algorithm ; Inverse modeling ; Biomaterials

nije evidentirano