Optimizing The Pipe Network Of Lubrication System To Achieve Flow Distribution Based On Isight

Gearing is an important part of the tank.Precise flow distribution of pipe network is directly related to the normal running of components and the reliability of the entire tank.The shortage of oil supply will lead to the lack oflubrication,the excessive supply of oil will cause too much power loss. Existing flow distribution of pipe network lubrication studies generally focus on artificial change in the design parameters for better result,Itlacks automatic optimization processes, the flow distribution precision can not meet the requirements. Therefore, research on theautomatic optimization of flow distribution for pipe network is very important. This article combine the fluid finite element method and optimization method.Thendevelop Isightsoftware platform for the automatic flow distribution of pipe network to achieve precise flow distribution.Provide important reference and technical support for future work.To explore the relationship between oil flow distribution and pipeline network length and diameter dimension, separate the flow parameters and the other parameters of head loss, thus theoretically infers themonotone invariance of the flow and size.We get the result that flowrate is more sensitive to the diameter diamention than rhe length diamention.The simulation verify the above conclusion. Considering the sensitivity and space constraints we choose the diameter diamention as the design variables in the optimization. Simulation results show that the partial derivative of flow variation width size is not large, it was decided to select a gradient-based optimization algorithm to achieve flow distribution pipe network optimization.By writing code of programs, respectively achieve the three-dimensional solid model of pipe network automatic update, automatic mesh division, automatic calculation and the automatic outputof the calculation results. Use the directional automatic segmentation method to solve the scheduling pronlem of faces. integrationeach automated process, developeoptimization software platform of flow distribution based on Isight.Analyze the features of several optimization algorithms, choose gradient-based Quasi-Newton algorithm to achieve automatic flow distribution optimization of pipe network. Construct uniformized objective function. Realize the algorithm according to the relationship between the objective function value- flow- size.Use the developed of Isight automatic flow distribution optimization software platform for optimization of simple lubrication system pipe network. Compares several optimization algorithms, quasi-Newton method converges the objective function at high-speed and high precision. Choose best partial derivatives segment step size of the Quasi-Newton method through simulation experiment.Improve the accuaracy,stability,speed of the automatic optimization process. Explores the flow distribution space of simple pipe network.Therelative error of flow distribution for each outlet is less than 5%in a wide range of flow distribution space.Propose the flow distribution method of layering and grouping according to tree pipe network flow pressure characteristics,analysethis method theoreticaly.Optimize complex lubrication system pipe network in this method. The result of flow distribution is good,therelative error of flow rate for each outlet is less than 10%. The method solves the problem of small flow distribution space and non-convergence objective function in direct optimization method.Explores the influence of lubricating oil viscosity on flow distribution accuracy, the results show the influence is small in a large range of viscosity, so we can achieve accurate flow distribution of pipe network in a variety of climatic conditions andworking conditions.