Abstract:

The paper addressed a practical problem of the non-uniform clearance between the outer ring and the inner ring after the nosing process in the assembly of spherical plain bearings, the extrusion process of GEW12 spherical plain bearings was simulated by using ABAQUS software. BP neutral network was applied to identify the relationship between the die profile and the peak difference between the maximum clearance and minimum clearance. A genetic algorithm was used to optimize the die profile, which yielded more uniform clearance distribution. Thus, the FEM, neutral network, and genetic algorithm were combined to develop a method for the design of the optimal shape of a extrusion die. The experimental results show that the clearance distribution between outer ring and inner ring is improved greatly and the metal flow of the bearing becomes more uniform.

Key words: spherical plain beating, numerical simulation, BP neural network, genetic algorithm(GA)

摘要：

针对实际生产中关节轴承内外圈间隙分布不均问题，以GEW12关节轴承为例，应用ABAQUS软件对关节轴承挤压装配过程进行了数值模拟。采用BP神经网络建立了挤压模具形状与内外圈最大间隙与最小间隙之差的映射关系。以关节轴承内外圈间隙均匀分布为目标，结合遗传算法，提出了一种集数值仿真、人工神经网络和遗传算法为一体的关节轴承挤压模具型腔优化设计方法。实验结果表明，模具型腔经过优化后，轴承内外圈间隙均匀性得到了很大的改善，轴承金属流动速度更加均匀。

关键词: 关节轴承, 数值模拟, BP神经网络, 遗传算法