基于改进遗传算法的C/SiC拉伸损伤声发射模式识别

采用层次聚类及基于改进遗传算法的无监督模式识别方法,对2D-C/SiC复合材料常温拉伸试验过程的声发射数据进行分析,结合试样断口的扫描电镜(SEM)照片,得到拉伸过程中5类损伤模式及其典型声发射特征参数。通过对各类损伤的能量分布、累计事件数和累计能量的分析,研究C/SiC复合材料的损伤演化过程,发现其过程可分为基体微裂纹和界面失效为主的初始损伤阶段、基体微裂纹停滞导致层间剥离及纤维失效占主导地位的裂纹饱和阶段、基体长裂纹和界面失效为主的损伤积累发展阶段和纤维束大量失效的宏观断裂阶段。

Acoustic Emission Pattern Recognition on Tensile Damage Process of C/SiC Composites Using an Improved Genetic Algorithm

The acoustic emission data collected during room temperature tensile test of 2D-C/SiC composites were analyzed by hierarchical clustering and unsupervised pattern recognition method based on an improved genetic algorithm. Combined with the SEM observation on the fracture surface, five damage modes were identified and their typical acoustic emission characteristics were obtained. According to the analysis of energy distribution, cumulative event number and cumulative energy of different damage modes, the damage evolution process of C/SiC composites can be divided into four stages. The first stage (damage initiation stage) shows mainly matrix microcracks and interface debonding. In the second stage, matrix crack reaches saturated and then causes a considerable quantity of interlaminar delamination and fiber failure. The third stage is a gradual damage development stage and all kinds of damage keep occurring except the breakage of fiber bundles. In the last stage, a large amount of fiber bundles break and the sample eventually fails.