Ship design optimization with Variable Fidelity Models

THe use of High Fidelity, CPU time expensive solvers is now quite common in the trial-and-error ship design process. Their use in Simulation Based Design (SBD) is however still reduced by the large amount of computational effort needed in the optimization cycles, and only few examples are available in optimization of ship design literature. In this paper, to reduce the computational cost of running complex engineering simulations, a metamodeling technique is applied in order to produce a global approximation of the high fidelity solvers, solving the optimization problem by means of a combined use of analytical models, less accurate but also incomparably less expensive, and high-fiedlity models. The use of a consistency condition ensures the accuracy of the final solution.