Zhongyi Wang

Multi-stage optimization method for air-intake system of hovercraft based on autonomous optimization

Abstract In this study, an optimal structural design program was designed and developed for Computational Fluid Dynamics based on self-optimization, effectively reducing the time required for structural optimization. Through experimental design using this program, the effects of various design variables on the optimization objectives were evaluated, and an adaptive simulated annealing algorithm was used for global optimization. Furthermore, response surface methodology and a nonlinear quadratic programming algorithm were utilized to obtain a global optimum solution after repeated iterations. Moreover, using a hovercraft air-intake system as the optimized object, the total pressure loss of the system was completely optimized by using a porous medium model and Matlab analysis program, and the accuracy of the structural design optimization program was validated. After the global optimization, the total pressure loss of the air-intake system was reduced by 20.5% compared to the original model. An average nonuniformity of 4.36% of engine inlet speed and 5% local nonuniformity of 11.19% satisfy the design requirements of the hovercraft engine. This method can be directly applied to engineering optimization problems as well as multiobjective optimization tasks after improving the relevant methodologies.

Investigation on the effects of blade corrosion on compressor performance

Abstract Corrosion is a common reason for marine gas turbine blade failure, and salt spray is largely responsible for corrosion of metallic objects near the coastline. It will directly lead to geometrical deviations of the compressor blade, forcing an increase in roughness height of the surface of the blade. By using a three-dimensional numerical method, this paper deals with the corrosion of a 1.5 stage axial compressor. Two different schemes, respectively based on blade geometrical model modification and turbulent near wall functions, were employed in numerical simulations for corrosion rate of the blades. The compressor map, derived from the numerical simulations by adjusting the back pressure, shows that there is a decline of compressor efficiency and pressure ratio when the blade is corroded, which will result in performance degradation of the compressor. The corrosion properties of stainless steel were studied, using the static salt spray corrosion test under laboratory conditions. Additionally, combined with numerical methods, the compressor performance parameters along the extension of operating time were simulated and a binomial relationship between performance degradation and corrosion time was established. This research provides a technical guide for compressor performance prediction under less than ideal conditions.