Multi-objective optimization of a bow thruster based on URANS numerical simulations
Yukun Feng,Zuogang Chen,Yi Dai,Lianzheng Cui,Zheng Zhang,Ping Wang
DOI: https://doi.org/10.1016/j.oceaneng.2022.110784
IF: 5
2022-03-01
Ocean Engineering
Abstract:In this study, a multi-objective design optimization of a bow thruster, which involved both experimental and numerical analyses, was performed. By changing the chamfer form at the tunnel opening of the bow thruster and the position of the anti-suction tunnel, the optimization procedure aimed to improve the propulsion efficiency, reduce the amplitude of the fluctuating pressure on the tunnel wall, and balance the pressure difference on the hull caused by the speed. First, model tests were conducted to measure the thrust and torque of the bow thruster as well as the fluctuating pressure acting on the tunnel. Numerical simulations were also performed under the same conditions based on the unsteady Reynolds-averaged Navier-Stokes (URANS) method. The accuracy of the numerical results was verified by comparison with the experimental data. Second, the chamfer depth and angle of the tunnel opening and the distance between the anti-suction tunnel and the propeller axis in the length and height directions were taken as the design variables. A sample set that contained 30 sampling points was selected using the optimal Latin hypercube sampling (opt-LHS) method. A surrogate model was built based on the Support Vector Regression (SVR) algorithm, and the multi-island genetic algorithm was utilized during the tuning process of the surrogate model parameters. On this basis, the non-dominated sorting genetic algorithm II (NSGAII) was used for the three-objective optimization and the Pareto solution set was obtained. Due to the inverse relationship between the propulsion efficiency and the fluctuating pressure amplitude, four optimal solutions were selected from the Pareto optimal solution set, and direct numerical computations were conducted to verify the optimization results. The results showed that, compared with the original shape, the fluctuating pressure amplitudes for the four optimal solutions were reduced significantly after the optimization, with a maximum reduction of 18.56%, but their propeller propulsion efficiencies were slightly lower than the original hull. Increasing the chamfer angle and depth significantly reduced the flow-separation area around the tunnel entrance, but it correspondingly increased the flow velocity in the tunnel, which caused the propeller to not work at its highest efficiency point; thus, the propulsion efficiency was reduced to a certain extent. Moreover, the position of the anti-suction tunnel could be determined using the SBD method, but with an increase in the chamfer angle, the location where the jet of the bow thruster re-attached to the hull after mixing with the external flow moved towards the stern. This led to an expansion of the low-pressure area near the rear of the tunnel that correspondingly increased the pressure difference on the hull. If the chamfer angle was large enough, although the pressure difference could be reduced to a fairly low value through the anti-suction tunnel, it could not be balanced completely.
engineering, civil, ocean, marine,oceanography