Study On Multi-objective Optimization Of Liquid Mechanical Seal With Spiral Grooves And Micro-cavities Coupling

The optimization of the surface textures and performance is always one of the hot topics in mechanical seal research.A non-contact mechanical seal with spiral grooves and micro-cavities has good seal and lubricity performance based on both the pumping and dynamic pressure effects.In the optimization of mechanical seal,there are few researches involves the combination of surface textures,multi-parameters and multi-objects.In this paper,the internal flow field performance analysis of mechanical seal with spiral grooves and micro-cavities was carried out under the support of the National Natural Science Foundation of China?51279067?.Based on the multiple regression analysis,the neural network and the intelligent algorithm,the multi-objective optimization design of the mechanical seal under the modeling was carried out,so as to provide some references for the optimization and design of mechanical seal.The main research work and conclusions are as follows:1.Considering the cavitation model,the mechanical seal with four different groove combinations were numerically simulated with the dynamic grid technique.By comparing the static pressure distribution,liquid film stiffness,friction torque and leakage of the flow field in different schemes,the optimal combination of the spiral grooves and micro-cavities was obtained,which the spiral grooves were in the inner diameter of moving ring and the micro-cavities were in the seal dam.The optimal combination was the basis for the textures coupling optimization.2.By comparison and previous research,the geometrical parameters of groove depth,helix angle,cavities depth and area ratio of cavities were selected to be the design variables,and the sealing performance of the liquid film K,the leakage Q were the optimization targets.The uniform test tableU₅₀?50⁴? was established,then the approximate equation between variables and target was obtained to establish the approximate model,and the response graph was made by using MATLAB.At last,the influence rule of the geometric parameters on the sealing performance and the optimal combination value range were obtained.3.Combined neural network and intelligent algorithm to establish the multi-objective optimization strategy for mechanical seal with spiral grooves and micro-cavities.Firstly,using MATLAB neural network toolbox to construct the BP neural network and training the sample pairs.Then the fitness function between the variable and the target was obtained,and the optimization toolbox was used for multi-objective optimization on fitness function.And in order to ensure prediction accuracy,genetic algorithm inner loop and prediction accuracy check calibration cycle are taken into the optimization process.At last,set the suboptimal results as new samples to training,and the Pareto Frontier solution was obtained by adjusting the final geometric parameters constantly.Two groups of the Pareto results were selected to compare before and after optimization.The results show that the liquid film stiffness of opt1 and opt2 is 12.58% and 15.47% higher than that of the original,and the leakage is reduced by 26.56% and 27.91%.4.The surface texture of the sealing rings before and after the optimization was machined by laser processing technology.And the optimization results were verified by sealing performance test which carried out on MSTS-IV mechanical sealing test bench.