焊缝典型缺陷磁记忆特征提取与反演的定量化基础研究

项目名称： 焊缝典型缺陷磁记忆特征提取与反演的定量化基础研究

项目编号： No.11272084

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 数理科学和化学

项目作者： 邢海燕

作者单位： 东北石油大学

项目金额： 72万元

中文摘要： 焊缝结构广泛应用于石油化工、能源电力、航空航天等关系国计民生的重要领域，通常在高温、高压、腐蚀、易燃易爆的环境下运行，一旦发生破坏，往往造成灾难性后果。磁记忆检测被称为21世纪最具应用前景的绿色无损检测技术，不仅可以发现宏观缺陷，还能检测早期损伤和应力集中，在焊缝检测上有着传统无损检测技术不可比拟的优势。但是定量评定一直是其难点问题，本课题以参数分散性与不确定性条件下焊缝缺陷磁记忆特征提取、缺陷等级量化及磁记忆缺陷反演的逆问题为落脚点，通过实验与理论研究，为焊缝缺陷的定量评定提供基础性理论研究和方法依据。主要内容包括：（1）多场耦合下焊接残余应力的磁记忆表征及有限元仿真；（2）焊缝典型缺陷演化的磁记忆定量化实验研究与缺陷磁荷模型建立；（3）考虑参数分散性、不确定性的焊缝缺陷特征提取与缺陷分级量化；（4）建立基于优化算法的磁记忆逆向缺陷反演定量化方法，进行焊缝定量评定的基础性研究。

中文关键词： 焊缝缺陷；磁记忆检测；特征提取；反演；定量化

英文摘要： Welded joints are widely used in petrochemical, energy，power, aerospace and other important areas. Welded joints usually work in the environment of high temperature, high pressure, corrosive, flammable and explosive. Once they are destroyed, catastrophic accidents will be caused. Metal magnetic memory (MMM) testing is known as one of the most promising green nondestructive testing(NDT) in the twenty-first century, which has more advantage on welded joint testing than the conventional NDTs. However, the welded joint quantitative assessment is still the MMM bottleneck. This subject majors in the quantitative weld defect feature extraction, grade and reversion based on the experimental and theoretical investigates considering disperse and uncertain parameters, in order to provide the theory and method foundation for weld defects quantitative assessment. The main contents include: (1) Welding residual stress MMM characteristic and welded joint multi-field coupling finite element simulation; (2) MMM quantitative experiment study of typical weld defects evolution and defect magnetic charge model; (3) Weld defect MMM feature quantitative extraction and defect grade by considering dispersion and uncertainty parameters; (4) MMM defect reversion quantitative method based on optimization algorithm, which provides the bas

英文关键词： Weld defect；metal magnetic memory test；feature extraction；inversion；quantification