CFD-based hull optimization in calm water using adaptive grid deformation method

Abstract

This paper presents an adaptive grid deformation technique for optimizing ship hull forms using computational fluid dynamics (CFD). The proposed method enables accurate and smooth updates of the hull surface and 3-D CFD grids in response to design variables. This technique incorporates a two-level point-transformation approach to move the grid points by a few design points. Initially, generic B-splines are utilized to transform grid points according to the displacements of the control points within a defined control box. This ensures surface modification accuracy and smoothness, similar to those provided by non-uniform rational B-splines. Subsequently, radial basis functions are used to interpolate the movements of the control points with a limited set of design points. The developed method effectively maintains the mesh quality and simulation efficiency. By applying this method to surface and grid adaptation, a regression model is proposed in the form of a second-order polynomial to represent the relationship between the geometric parameters and design variables. This polynomial is then used to introduce geometric constraints. Furthermore, a radial basis function surrogate model for the calm-water resistance is constructed to approximate the objective function. An enhanced optimization framework is proposed for CFD–based hull optimization and applied to KVLCC2 to validate its feasibility and efficiency.