View article

The solution of two optimal shape design problems related to industrial CFD using genetic algorithms are considered. The first one is a single objective optimization problem, where the geometry of a flow divider of a paper machine headbox is designed subject to prescribed goals and restrictions. The second problem is a two-dimensional airfoil design problem, where the objectives are to minimize the drag and the electromagnetic backscatter while the lift is larger than a given value. The flow and backscatter are modeled by the thin-layer Navier-Stokes equations and the time-harmonic Maxwell equations, respectively. Hence, this is a multiobjective and multidisciplinary shape optimization problem. Numerical experiments demonstrate capability of the genetic algorithms to solve these two industrial optimal shape design problems.