Multi-objective Bayesian hull form optimisation for high-speed craft

Abstract

Simulation-based design plays an important role in many engineering fields enabled by advancements in computing resources, numerical algorithms and computer aided engineering software. In naval architecture, it opens new opportunities for improving hull form designs by reducing calm water resistance and hence minimising required propulsive energy. In this paper, we propose a low dimensional shape morphing tool suitable for designing novel and realistic hull forms for high-speed craft. We provide an efficient simulation-based design framework for optimising hull forms across multiple operating conditions by taking advantage of a multi-objective Bayesian optimisation method. We demonstrate the proposed framework for optimising the hull form of a conceptual high-speed catamaran at two speeds, corresponding to Froude numbers of Fr=0.35 and Fr=0.65. The results highlight the benefit of multi-objective optimisation for high-speed craft along with a nonlinear CFD (computational fluid dynamics) solver to evaluate design solutions. We show the proposed framework can design highly efficient hull forms along the multi-objective Pareto front that have a reduced total resistance when compared to a generic baseline hull by 31%–43% and 5.7%–8.2% at low and high Froude numbers, respectively.