基于相位突变的异种材料扩散焊缺陷超声检测和识别方法研究

项目名称： 基于相位突变的异种材料扩散焊缺陷超声检测和识别方法研究

项目编号： No.51205083

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 机械工程学科

项目作者： 栾亦琳

作者单位： 黑龙江科技大学

项目金额： 25万元

中文摘要： 针对常规超声波检测由信号幅度高低无法辨别异种材料扩散焊缺陷的问题，提出一种基于相位突变的缺陷检测和识别方法，以相位突变和幅频缓升特征自动识别异种材料扩散焊界面弱接合、微孔和未焊合缺陷，解决了传统超声波检测不出微米弱接合和微孔缺陷的难题。以异种材料扩散焊界面超声波信号为研究对象，采用连续小波变换研究缺陷信号相位突变和幅频缓升特性，提取与缺陷类别相关的特征，建立表达函数，采用支持向量机构建缺陷识别模型，实现缺陷智能识别。根据缺陷识别结果获得焊合率与抗剪强度对应关系，实现接头强度的预测。通过本项目的研究，阐明声波在异质界面的传播行为、与缺陷的相互作用机制，揭示缺陷信号相位突变和幅频缓升的规律，在研究声波与连接界面的相互作用等方面具有较高的科学意义，同时可为异种材料连接界面缺陷的无损检测与质量评价提供思路。

中文关键词： 扩散焊；异种材料；超声波；相位突变；支持向量机

英文摘要： Aiming at the difficulty of ultrasonic amplitude in the evaluation of diffusion bonding of dissimilar materials, a defects testing and recognition method is proposed based on phase mutation. Kissing bond, micropore and unbond in dissimilar diffusion bonding can be automatic recognized by phase mutation and amplitude-frequency-slight-creasing characteristics, which solve the problem of convention ultrasonic testing in the detection of micron kissing bond and micropore. Phase mutation and amplitude-frequency-slight-creasing characteristics of ultrasonic interface signals of dissimilar diffusion bonding are analyzed by continuous wavelet transformation. Correlative features are extracted to establish expression function. Defects recognition models based on support vector machine are estabilished to realize defects intelligent recognition. Correlation between bonding radio and shear strength is obtained to predict strength of diffusion bonding joints. Propagation of ultrasonic in the interface of dissimilar materials and mechanism of ultrasonic interacting with the defects are clarified by this study. Regularity of phase mutation and amplitude-frequency-slight-creasing characteristics is disclosed. It has scientific significance in the study of interaction between ultrasonic wave and diffusion bonding interface and

英文关键词： diffusion bonding；dissimilar materials；ultrasonic；phase mutation；support vector machine