Structural Optimization Of Drive Axle Housing Of Wheeled Loader Based On Surrogate Model

Drive axle housing of wheeled loader is a very important component in thewhole vehicle system although it is a shell. It not only protects the internal drivesystem, but also supports the weight of entire vehicle.The drive axle housing often appears deformation,cracks and fracture due to theexcessive load, which could reduce the useful life and lead to uneven force of internaltransmission systems in drive axle housing.Therefore, it is very important to study thedeformation and failures of drive axle housing, especially the structure optimization.Thispaper optimizes the structure of DA1170axle based on Surrogate Models and geneticalgorithm to improve the performance of the drive axle housing.Firstly,the research status of drive axle housing, surrogate models and geneticalgorithm of structural optimization were introduced.The working condition of shovelwas taken as typical working condition after summing the working conditions ofwheel loader.The three-dimensional models of the drive axle housing and mainreducer shell were established.The stress of drive axle housing was calculated in orderto determine the loads and constraints of drive axle housing, and Finite elementanalysis (FEA) was carried out on the drive axle housing.Secondly,the "shuttle-type" structure can be instead of original bridge shell byanalyzing the impact of the main parameters of hollow Chary flexural modulus ofsection and referencing the latest driver axle housing structure.The feasibility ofimprovement was confirmed after FEA.The safety factor of improved drive axlehousing was increased by simulant experiment of drive axle housing in finite elementssoftware.Six parameters of drive axle housing were identified to create Surrogate Modelsof RBF based on analysis of improved drive axle housing.30sets of test samples and15sets of error analysis samples were obtained using Latin Hypercube to create the Surrogate Models of quality, maximum deformation and maximum stress of the driveaxle housing.The accuracy of Surrogate Models was verified to meet the engineeringrequirements using Multiple Correlation Coefficient and Root Mean Square Error.Finally,the minimum weight of the drive axle housing was taken as the objectivefunction,the amount of maximum deformation of the drive axle housing does notexceed2.8mm and the maximum stress value not exceeding300MPa were taken asthe boundary conditions,six parameters of the drive axle housing were optimizedusing the multi-island genetic algorithm based on the surrogate models.Thedeformation and maximum stress of the drive axle housing were reduced afteroptimization, and achieve the goal of light weight.