Design And Optimization Of Duct Propeller Blade For Underwater Vehicle

Underwater robots with extensive application in detection,salvage,sampling,mining and other aspects are particularly of great significance in deep-sea mining operation.Propeller is the important core component of underwater robots.The overall performance and reliability of working robots are directly influenced by the propeller's efficiency in underwater propulsion.Therefore,it is of great engineering value and important significance to conduct optimization design of the structure of the propeller for its propulsion efficiency and reliability improvement.In this paper,parametrization design of the propeller blade was performed by using the graph method according to the actual working conditions of underwater robots,and a geometric model of the propeller was established.At the same time,the standard k-? turbulence model was used to calculate the pressure distribution of the propeller blade under different water pressure conditions.MRF watershed analysis model was used to calculate the corresponding open water coefficient values of the model for different feed rate coefficients.Numerical simulation of the propeller blade's open water performance was conducted.The calculation results indicated that the simulation results with high accuracy could be used for subsequent simulation analysis.Due to the great impact of different types of ducts on the propeller blade's open water performance,analogue simulation of the propeller blade's three motion modes was performed,i.e.,underwater direct route,oblique route and retreat.The duct type was determined as JD7704 through comparative analysis,thus confirming the structure scheme of the ducted propeller blade.To improve the open water efficiency of the ducted propeller and the propeller thrust,a performance optimization model based on BP neural network was proposed in this paper.Correlation analysis of the dimensional parameters of the ducted propeller was performed by using Latin Latin Hypercube test method and the structure parameters with a greater impact on the ducted propeller's performance were determined.According to the simulation data obtained by Fluent,a nonlinear model reflecting the performance indexes and design variables of the ducted propeller was established by using BP neural network,which provided theoretical basis for the multi-objective optimization of the ducted propeller.Therefore,a multi-objective optimization model with propeller thrust and open water efficiency as the objective functions,and blade strength as the constraint condition was solved by using genetic algorithm.The optimization results showed that the optimized thrust of the ducted propeller blade increased by 27.3% and open water efficiency increased by 5.7%,indicating the reliability of the optimization model.To verify the reliability of the ducted propeller,the propeller indoor thrust test and the complete machine lake test were conducted in this paper.The laboratory test results indicated that the thrust decreased with the increase of the advance coefficients.The smaller the advance coefficients,the greater the propeller thrust;and the greater the advance coefficients,the smaller the propeller thrust.The test curve was close to the simulated value curve with small deviation,indicating the optimization scheme was correct.In addition,the lake test showed that the underwater robot could work reliably in different conditions at different diving depths without any abnormalities.The test results indicated that the ducted propeller designed in this project could output thrust reliably and operate stably.It could meet the design need while verifying the reliability of the optimization scheme in this paper.