Prediction Method for Wall Shear Stress of Water Jet Based on KGWO-BP Neural Network
The wall shear stress of ultra-high-pressure water jet impinging on the ship hull can be regarded as an important indicator to reflect the working efficiency of rust removal operation,and it also acts as a crucial criterion to assess the reasonability of the internal structure parameters related of the nozzles.Several parameters should be considered when the hydrodynamic performance of nozzles is optimized.If only relying on CFD simulations,the computational consumption is too high.In this paper,the straight-cone nozzle is taken as the research object,a prediction model for water jet performance is established,by using CFD simulation and neural network modelling simultaneously.By means of adopting the orthogonal test rule to set simulation cases and regarding the wall shear stress of water jet on target surface as a core indicator,the primary and secondary effect sequence of the internal structure parameters of nozzle on the hydrodynamic performance of water jetting is obtained.It provides a reference for the shape lines design of nozzle,as well as for the optimization of the structure parameters.The results show that the length of the exit section has the greatest effect on the water-jet performance among the four internal structure parameters,and the optimal combination of those four parameters corresponding to the larger wall shear stress is also provided.The proposed KGWO-BP model can provide reliable technical support for the optimization of the straight-cone nozzle,because of its high prediction accuracy and high generalization performance.
Ultra-high-pressure water jetStraight-cone nozzleStructure parameterGrey wolf algorithmBP neural network
万方数据
维普
