Design optimization of the waterjet-propelled Delft catamaran in calm water using URANS, design of experiments, metamodels and swarm intelligence

A URANS-based design optimization for the waterjet-propelled Delft catamaran is presented. Iterative Latin hypercube and Hammersley sequence are used to generate initial designs; multiple metamodels are applied to reduce the computational cost, whereas a particle swarm method (PSO) is used to solve the optimization problem. The optimization is conducted for the catamaran advancing in calm water at Fr = 0.5, with the model free to sink and trim, and selfpropelled. The waterjet is appended to the optimal barehull obtained by dimensionality reduction of a free-form deformation space through Karhunen-Loève expansion, multiple metamodels and PSOs, which showed a 10% reduction in calm-water resistance (Chen et al., 2013). Two design variables are used for the waterjet design modifications. Optimization achieves a 17.2% reduction in effective pump power, compared to the original barehull with the original waterjet. Compared to earlier optimization, current studies provide an additional reduction in powering requirements by 6%.