Computational Fluid Dynamics-Based Optimization of a Surface Combatant

Computational fluid dynamics (CFD)-based optimization of a surface combatant is presented with the following main objectives: (1) development of a high-performance optimization module for a Reynolds averaged Navier-Stokes (RANS) solver for with-free-surface condition; and (2) demonstration of the capability of the optimization method for flow- and wave-field optimization of the Model 5415 hull form. The optimization module is based on extension of successive quadratic programming (SQP) for higher-performance optimization method by introduction of parallel computing architecture, that is, message passing interface (MPI) protocol. It is shown that the present parallel SQP module is nearly m (= 2k + 1; k is number of design parameters) times faster than conventional SQP, and the computational speed does not depend on the number of design parameters. The RANS solver is CFDSHIP-IOWA, a general-purpose parallel multiblock RANS code based on higher-order upwind finite difference and a projection method for velocity-pressure coupling; it offers the capability of free-surface flow calculation. The focus of the present study is on code development and demonstration of capability, which justifies use of a relatively simple turbulence model, a free-surface model without breaking model, static sinkage and trim, and simplified design constraints and geometry modeling. An overview is given of the high-performance optimization method and CFDSHIP-IOWA, and results are presented for stern optimization for minimization of transom wave field disturbance; sonar dome optimization for minimization of sonar-dome vortices; and bow optimization for minimization of bow wave. In conclusion, the present work has successfully demonstrated the capability of the CFD-based optimization method for flow- and wave-field optimization of the Model 5415 hull form. The present method is very promising and warrants further investigations for computer-aided design (CAD)-based hull form modification methods and more appropriate design constraints.