Multi-Objective Optimization Design of Assembled Wheel Lightweight Based on Implicit Parametric Method and Modified NSGA-II

To improve the fatigue life of wheels, the level of lightweight design, and the efficiency and accuracy of optimization analysis, a wheel finite element analysis model combining shell-body elements is established in this paper. A design and optimization method for the assembled wheel with a magnesium alloy rim and an aluminum alloy disc structure is proposed. The conceptual design model of the wheel is established, and the topology of the discs is optimized by combining the bending and radial fatigue test conditions. An implicit parametric model of the rim was built using SFE-Concept software and connected to the disc structure by bolt elements to form an assembled wheel. The performance parameters such as stress, displacement, life and life safety factor of assembled wheels under bending and radial fatigue conditions, and modal frequency are analyzed to study the influence law of wheel structural parameters on fatigue performance. A parametric model for fatigue analysis of assembled wheels is created based on advanced mesh deformation technique and implicit parameterization technique, Design of experiment (DOE) sampling is performed, and 21 design variables are screened in combination with the contribution analysis method. A hybrid method of entropy weighted grey relation analysis (EGRA) combined with modified non-dominated sorting genetic algorithm-II (MNSGA-II) is proposed for multi-objective optimization of the assembled wheel in combination with an approximate model approach to obtain the Pareto frontier solution set and its grey relation order to filter the preferred compromise solution. The performance indexes of assembled wheels before and after optimization were compared by simulation analysis and verified by bending and radial fatigue tests. The results show: With the preferred compromise solution screened out, all indicators of the fatigue performance of the assembled wheel have been reasonably improved. The wheel’s weight is reduced by 10.17%, and the weight reduction effect is remarkable. In this paper, we propose a hybrid method with entropy-weighted grey relation analysis (EGRA) combined with modified non-dominated sorting genetic algorithm-II (MNSGA-II), which can sufficiently synthesize the performance indexes of the wheel. Under the condition of guaranteed computational accuracy, the combined wheel shell-body cell model proposed in this paper saves at least 46.44% computational time compared with the body element model. This paper presents the implicit parametric shell-body element combination model scheme and entropy-weighted grey relation analysis ranking multi-objective optimization solution set method, which provides a reference for lightweight wheel design.