Numerical analysis of propeller-induced higher-order pressure fluctuations on the ship hull

Abstract This thesis documents and explains the development and validation of a hybrid simulation method for investigating higher-order hull pressure fluctuations induced by cavitating propellers. Two forms of propeller cavitation are considered in this work: coherent structures of sheet cavitation on the propeller blades and tip vortex cavitation. The interaction between sheet cavitation and developed tip vortex cavitation can be responsible for notable higher-order pressure fluctuations. The essential element of this novel simulation method is panMARE, the in-house panel code used to calculate the propeller flow including effects of sheet cavitation. Furthermore, relevant parts of the hull surface above the propeller are incorporated in the panel model in order to evaluate fluctuations of pressure in the aft ship region. The propeller operates in the effective wakefield of the ship which results from the viscous interaction between hull and propeller flow. It is calculated by the RANSE solver ANSYS CFX in combination with panMARE. A body force coupling approach is used to couple both methods. Here in, the viscous hull flow is determined by ANSYS CFX and the impact of the propeller is approximated by a corresponding distribution of bodyforces applied to the viscous flow which in return is calculated by means of panMARE. In order to model tip vortex cavitation, the vortex cavity is divided into a large number of cylindrical segments, where each of them are treated separately. This breaks down the formerly three- dimensional problem into a two-dimensional one, which is much easier to handle. For each segment, the momentum equations in cylindrical coordinates, leading to a Rayleigh-Plesset-like equation for the dynamical behaviour of the cavitating core, are solved by means of the newly developed code VoCav2D. Interaction with sheet cavitation is taken into account by correlating the initial cavitation radius with the cavity thickness at the trailing edge of the blade in the tip region. This and other tip vortex parameters are extracted from detailed RANS simulations of the blade tip flow made in advance for a number of representative loading conditions. For validation purposes, three vessels are investigated. The numerical results are compared to those obtained from experiments and–if available–from full-scale measurements. Furthermore, two types of scale effects due to the Reynolds number are investigated by the method: the wake scale effect on sheetcavitation and the influence of the viscous core radius on moderate and tip vortex cavitation in the stage of formation in an idealised manner.