电磁热无损检测缺陷特征建模与表征算法研究

发布时间：2018-06-10 01:32

本文选题：无损检测 + 脉冲涡流热成像　；参考：《电子科技大学》2015年硕士论文

【摘要】：无损检测是保障产品制造质量及重大工程设备可靠运行的核心技术。脉冲涡流热成像技术(Eddy Current Plused Thermography,ECPT)作为电磁无损检测的一种方法,具有非接触、安全快速、单次测量范围大等优势,目前已经广泛应用于导体材料缺陷的识别和评估,并获得了丰硕的成果。然而,现有的基于脉冲涡流热成像技术的缺陷检测模型和特征提取方法对微小缺陷,特别是针对尚未形成宏观微裂纹的隐性缺陷,难以进行有效、完整的检测和评估。并且,仅仅利用简单图像处理技术获取ECPT缺陷特征而不做深入的物理与数学关系建模将难以真实描述材料损伤程度变化而导致各类微缺陷的错误量化。因此,针对复杂构件缺陷区域的电磁热信号特征物理数学建模与提取算法的研究具有重大意义。本文旨在利用脉冲涡流热成像技术的电、磁、热等物理效应,通过理论与实验相结合的方法,建立针对复杂构件微弱缺陷检测的模型(即为热光流模型);研究并分析所建模型对复合材料的低冲击损伤以及齿轮早期疲劳隐形缺陷检测的通适性。研究内容主要有:1)研究缺陷脉冲涡流热成像检测机理,研究热像仪采样频率对缺陷检测以及材料纹理分析精度的影响,优化脉冲涡流热成像实验参数;2)分析脉冲涡流热成像热传导机理,建立热光流场模型,提出(热光流)散度以及(热光流)熵特征提取算法;3)分析热光流场模型和散度、熵算法对于复合材料的低冲击损伤以及齿轮早期疲劳隐形缺陷检测的灵敏度、准确性,建立(热光流)散度/熵特增值与不同缺陷损伤之间的映射关系,实现缺陷/损伤的定位分析和定量检测。研究结果表明,针对缺陷检测和辨识,热像仪存在临界采样频率,当采样频率控制在临界数值时,可实现对脉冲涡流热成像无损检测装置的参数优化。基于脉冲涡流热成像机理建立的热光流模型具有检测灵敏度高,物理含义和信息量丰富等优点,能有效的用于目前存在挑战的复合材料低冲击损伤和金属材料早期疲劳隐形缺陷的定位检测和定量分析。本文建立的基于脉冲涡流热成像机理的热光流模型、热光流-散度算法及热光流-熵算法,为复杂构件的无损检测以及全寿命安全评估研究提供了一种新的思想和方法。
[Abstract]:Nondestructive testing (NDT) is the core technology to guarantee the manufacturing quality of products and the reliable operation of major engineering equipment. As a method of electromagnetic nondestructive testing (ENDT), Eddy current used Thermography (ECPT) has the advantages of non-contact, fast safety and large measurement range, so it has been widely used in the identification and evaluation of defects in conductor materials. And has obtained the rich achievement. However, the existing defect detection models and feature extraction methods based on pulsed eddy current thermal imaging are difficult to detect and evaluate microdefects effectively and completely, especially for the hidden defects which have not yet formed macroscopic microcracks. Moreover, it is difficult to describe the damage degree change of materials by using simple image processing technology to obtain ECPT defect features without deep modeling of physical and mathematical relationship, which leads to the error quantization of various microdefects. Therefore, it is of great significance to study the physical mathematical modeling and extraction algorithm of electromagnetic thermal signal features in the defect region of complex components. The purpose of this paper is to utilize the physical effects of electrical, magnetic and thermal effects of pulsed eddy current thermal imaging, and to combine theory with experiment. A model for weak defect detection of complex components (i.e. thermo-optical flow model) is established, and the suitability of the model to the low impact damage of composite materials and the detection of early fatigue invisible defects of gears is studied and analyzed. The main research contents include: (1) study the mechanism of defect pulse eddy current thermal imaging, study the influence of sampling frequency of thermal imager on defect detection and material texture analysis precision. Optimizing the experimental parameters of pulsed eddy current thermal imaging (2) analyzing the thermal conduction mechanism of pulse eddy current thermal imaging, establishing a thermo-optical flow field model, and proposing an algorithm for extracting (thermal optical flow) divergence and entropy feature (3) to analyze the thermal optical flow field model and divergence. The sensitivity and accuracy of entropy algorithm for the detection of low impact damage of composite materials and early fatigue stealth defects of gears are studied. The mapping relationship between the divergence / entropy increment and the damage of different defects is established. The localization analysis and quantitative detection of defect / damage are realized. The results show that the critical sampling frequency exists in the thermal imager for defect detection and identification. When the sampling frequency is controlled at the critical value, the parameters of the pulsed eddy current thermal imaging nondestructive testing device can be optimized. The thermo-optical flow model based on the mechanism of pulsed eddy current thermal imaging has the advantages of high detection sensitivity, rich physical meaning and abundant information. It can be used in the localization and quantitative analysis of low impact damage of composite materials and early fatigue stealth defects of metal materials. The thermo-optical flow model based on the mechanism of pulsed eddy current thermal imaging, the thermo optical flow-divergence algorithm and the thermo-optical flow-entropy algorithm are established in this paper, which provide a new idea and method for the study of nondestructive testing and life safety assessment of complex components.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41;TB302.5

【参考文献】