Nonlinear System Identification for the Prediction of Unsteady Vertical Plane Hydrodynamic Forces on a Planing Hull

Abstract

A method to derive a nonlinear model of the vertical plane hydrodynamic forces and moments on a planing hull using forced motion Unsteady RANSE CFD simulations and nonlinear system identification is investigated. The use of forced motion simulations on the fully three-dimensional hull form allows three-dimensional effects to be included in the nonlinear force model which have previously been neglected by strip theory or approximated by 2D+T theory. Nonlinear system identification allows the most significant dynamic terms in the hydrodynamic model to be included in the model without over regressing the response. The process of identifying the nonlinear model is investigated in detail to determine the best collection of forced motion simulations to derive an adequate model. This methodology is intended to provide higher fidelity predictions of vertical plane motions of planning hulls than previous methods while requiring a lower computational expense than a comprehensive CFD evaluation in waves. The final nonlinear model, developed for the Generic Prismatic Planing Hull (GPPH), is validated against static and dynamic simulations not included in the system identification and is found to produce a promisingly accurate model of the hydrodynamic forces and moments When used to predict the heave force and trimming moment of a forced coupled heave and trim motion, the selected nonlinear model is found to predict the heave force and trimming moment with an average absolute error of 3.9% of the maximum force and 5.5% of the maximum moment respectively.