Using computational fluid dynamic and artificial neural networks to predict the performance and cavitation volume of a propeller under different geometrical and physical characteristics

作者：Shora Mohammad Mahdi; Ghassemi Hassan^*; Nowruzi Hashem

来源：Journal of Marine Engineering and Technology, 2018, 17(2): 59-84.

DOI：10.1080/20464177.2017.1300983

In the current study, simulations of hydrodynamic characteristics of a propeller under different geometrical and physical characteristics are conducted by the computational fluid dynamic (CFD). Then, by designing appropriate artificial neural networks (ANNs), the hydrodynamic performance and cavitation volume of propellers are predicted under intended conditions. For this accomplishment, finite volume-based Navier-Stokes equations associated with incompressible large eddy simulation turbulence model are used. In order to verify the computational procedure, mesh sensitivity analysis and validation study are conducted and appropriate accuracy is observed. In the CFD simulations, propeller thrust, torque and cavitation volume are computed under different pitch ratio (P/D), rake angle (RA) and skew angle (SA), advance velocity ratio (J) and cavitation number (sigma). By the CFD results, a significant increase in propeller thrust and torque is observable by enhancement of P/D and positive value of RA. Moreover, maximum mean square errors of ANNs output in the prediction of propeller thrust, torque and cavitation volume achieved are 0.000111, 7.4206E-5 and 0.000667, respectively. Also, related to ANNs' weights and bias, four set of equations are proposed to predict the performance and cavitation volume of propellers.

• 出版日期2018