论文部分内容阅读
采用搅拌摩擦工艺以A356合金为基体金属制备B_4C/A356复合材料。利用人工神经网络(ANN)和非支配排序遗传算法-Ⅱ研究复合材料的显微组织和力学性能。首先,研究不同加工条件下制得的复合材料的显微组织。结果表明,搅拌摩擦工艺参数如搅拌头的旋转速度、横向移动速度和形状显著影响基体中初始Si颗粒的尺寸、复合材料层中B_4C增强剂的分散效果及体积分数。采用高旋转/移动速度比和螺纹销形状搅拌头能获得较好的颗粒分布、较细的Si颗粒和较少的B_4C团聚体。其次,通过硬度和拉伸试验研究复合材料的力学性能。结果显示,经搅拌摩擦工艺处理后样品的断裂机理由脆性断裂转变为延性断裂。最后,利用人工神经网络技术建立了搅拌摩擦工艺参数与复合材料显微组织和力学性能的关系。采用结合多样性保护机制的NSGA-Ⅱ法,即ε消除算法得到搅拌摩擦工艺参数的Pareto最优解集。
The friction stir process was used to prepare B_4C / A356 composite with A356 alloy as base metal. Using Artificial Neural Network (ANN) and Non-dominated Sequencing Genetic Algorithm-Ⅱ to Study the Microstructure and Mechanical Properties of Composites. First, the study of the microstructure of the composite material obtained under different processing conditions. The results show that the parameters of the friction stir process, such as the rotating speed, lateral moving speed and shape of the stirrer head, significantly affect the size of the initial Si particles in the matrix, the dispersion effect and the volume fraction of the B_4C reinforcing agent in the composite layer. Good particle distribution, finer Si particles and fewer B_4C agglomerates were obtained with the high rotation / moving speed ratio and the shape of the screw pin. Secondly, the mechanical properties of the composites were studied by hardness and tensile tests. The results show that the fracture mechanism of the sample after the friction stir process changes from brittle fracture to ductile fracture. Finally, the relationship between the friction stir process parameters and the microstructure and mechanical properties of the composites was established by artificial neural network technology. The Pareto optimal solution set of friction stir process parameters was obtained by NSGA-Ⅱ method, which is a combination of diversity protection mechanism, that is, ε elimination algorithm.