Robust Design Optimization for Operational Profile

Abstract

The work deals with the actual necessity of creating optimal ship designs under uncertainty, also referred as Robust Design Optimization (RDO). The consideration of uncertainties in the optimization process is the future trend of ship design as a more realistic scenario is taken into account leading to the development of more efficient vessels and thus more environmental friendly. Several optimization methods were tested for a hull form improvement concerning the total resistance of the vessel at forward speed in calm water condition. The operational profile has been taken into account as a single-objective problem in which several operating conditions are suitably weighted, as a single-objective problem in which one operating point is dominant while others are considered as constraints and as a multi-objective optimization task in which the best design is selected from the Pareto-set. Furthermore, a hybrid method via Gauss-Markov estimation was also tested. The robustness was considered via sensitivity check on slight variation of uncertain operational characteristics. All the processes are compared among them and, in addition, with a single objective optimization process for one specific operational condition. Each approach merits and drawbacks are pointed out in an investigation of which approach is particularly suited for hull form development. The criteria to judge pros and cons are quality of the achieved results (e.g. energy efficiency), computational effort (e.g. number of Computer Fluid Dynamics (CFD) runs needed) and sensitivity to changes (e.g. influence of slight changes on the operational conditions). The study case is an existing vessel that had a change in the operation leading to higher fuel consumption. In addition, the optimization is focus on changes of the forebody form of the vessel (e.g. bulbous bow). Finally, the analysis is suitable to assist ship designers regarding the necessity of performing a robust optimization or a typical optimization process plus sensitivity analysis and thus create in short time optimal ship designs even under uncertainties.