Research on valve ball motion characteristics and multi-objective optimization of reciprocating pump inlet valve
In order to study the motion characteristics of the valve ball in the reciprocating pump valve and optimize its performance,the dynamic grid and UDF technology were used to conduct the calculation of coupling between the valve ball motion and the flow field of the pump valve and experimental verification during the liquid inlet process of the reciprocating pump,so as to obtain the motion law of the valve ball,and analyze the influences of the valve ball mass,the guide hole structure of the valve guide sleeve and the limit height of the valve sleeve on the valve movement and performance.In order to further optimize the model and find the optimal combination scheme of pump and valve structure,the radial basis neural network(RBFNN)proxy model was proposed and the optimization framework of its Pareto optimal solution set was obtained by multi-objective particle swarm(MOPSO)algorithm and was verified.The results show that the motion lift of valve ball increases first and then decreases,which is greatly affected by the flow rate of valve gap and hydrodynamic force.The mass of valve ball,the number of valve guide holes and the limit height of valve sleeve have nonlinear relationship with the maximum lift and seating speed of valve ball.Based on the RBFNN proxy model and MOPSO algorithm,the optimal combination model was found.After optimization,the maximum lift of the valve ball was increased by 8.12 mm,and the seating speed of the valve ball was reduced by 31.4%.The optimization effect was remarkable.The research results can provide reference for the optimization design of reciprocating pumps.
万方数据
维普
