Research On Multi Objective Optimization Of Mixer Blade Performance Based On Genetic Algorithm

The twenty-first Century is an era of energy shortage and environmental pollution,so it is urgent to improve energy efficiency and reduce energy waste caused by unreasonable mechanical design.The mixing machine is widely used in the field of industrial engineering,and the design of high efficiency and uniform mixer becomes the focus of researchers.The blade is the key of the mixer,and the blade profile directly affects the performance of the blade.This paper used a based hypervolume multi-objective optimization genetic algorithm to optimize the blade curve of the blade of a mixer,and then apply it to the whole blade to improve the efficiency of the mixer blade.In this paper,the B spline curve was used to fit the blade profile of the mixer,and the smooth curve was obtained.Based on the surface panel method,the performance of the B spline curve was calculated and the pressure characteristics were analyzed.In this paper,the genetic algorithm was studied,and the principle of super volume was added to the original genetic algorithm to improve the performance of multi-objective optimization.At the same time,the surface element method was introduced into the genetic algorithm,and the flow field calculation was coupled to the optimization design of the blade.According to the actual requirements,the three optimization objectives of the blade optimization are listed as follows:increasing the lift,reducing the leaf area and reducing the pressure variance.This paper optimizes the design of the 8 leaf sections using genetic algorithm,optimization results show:the blade profiles of the front-end and back-end part from 0.2R to 0.6R changes greatly,but the ones from 0.7R to 0.9R changes hardly,but the overall results shift upward.The results show that the other two optimization objectives of the optimized blade have obvious improvement under the condition of the smaller area difference change.For the optimization of the shape of the blade profile,the two-dimensional coordinate of the leaf profile is transformed into the Descartes coordinates of the three-dimensional space by the software.By virtue of the powerful surface modeling function of Pro/E,the leaf profile of each region is mixed and the three-dimensional model of the blade is established.In this paper,the CFD method was used to predict the performance of the mixer blade,and the three-dimensional model of the blade was calculated by ICEM.According to the characteristics of each part of the blade,the 3D numerical simulation model was established by the method of partition and meshing.And fluid field of the blade models before and after the optimization were calculated by Fluent,the results show that:In the case of the all feed coefficient known,the optimized thrust coefficient and torque coefficient of the blade can be improved,when the thrust coefficient increases,leaf efficiency can be also improved.The results show that According to the speed of cloud images show that scope of vortex flow field in the optimized model is smaller,which proves that the mixing is more uniform.The dynamic performance of the blade is up to the expected goal,and it is proved that the hyper volume criterion combined with the multi-objective genetic algorithm is effective for the blade profile optimization.