ABSTRACT
A statistical analysis is performed on the key geometric parameters that influence hemodynamics in the human carotid artery bifurcation. A parametric CAD model of the bifurcation is used to automatically construct a range of different geometries. A design of experiments (DoE) approach is employed to generate a set of candidate geometries for flow analysis using an unsteady three dimensional Navier–Stokes solver. The values of the integral of negative average shear stress (INASS) and the maximal average shear stress (MASS) corresponding to these geometries are then used to construct a Bayesian surrogate model. A shape optimization problem is proposed using the surrogate model which predicts arterial geometries having minimal INASS and MASS.