基于部分电磁屏蔽的航空航天材料边缘缺陷检测

Titanium alloy and carbon fiber reinforced polymer (CFRP) are widely used in the area of aerospace and aviation. In terms of preventing accidents and reducing economic losses, effective detection methods for such materials are important. Compared with non-edge defects, the stress at edge defects is more concentrated and hence the potential safety hazard is higher. Due to edge effects, edge defect detection is challenging. Therefore, it is urgent to develop a non-destructive testing method that is able to detect edge defects. This project proposes a method based on partially electromagnetic shielding to improve the sensitivity of edge defect detection. Firstly, the edge effect theoretical framework of partially electromagnetic shielding method will be established. The quantitative relationship between sensor parameters and edge effect will be investigated. Secondly, the edge defect model will be added into the established theoretical model via electric field source equivalent for the defect model, and quantitative analysis of the relationship between edge defects and edge effects will be carried out. In addition, this project will investigate irregular edges, including corner and wavy edges, via finite element method. Finally, based on theoretical analysis, numerical caculation and experimental verification, the sensor and detection system will be optimized, well-designed experimental measurements will be carried out, and the feature information of the defect response will be extracted and a database will be established. By means of deep learning, this project aims to realise the prediction and hazard classification for edge defects within titanium alloy and CFRP. The proposed theories and methods can be extended to the edge defects analysis of other conductive materials.

钛合金及碳纤维增强聚合物(CFRP)是航空航天常用材料，有效的检测手段对预防安全事故及减少经济损失均有重要意义。相较于非边缘缺陷，边缘缺陷处应力更集中，安全隐患更高。由于边缘效应的影响，边缘缺陷检测很具挑战性，因此迫切需要研发边缘缺陷的无损检测手段。本项目提出部分电磁屏蔽技术提高边缘缺陷检测的灵敏性。首先，建立部分电磁屏蔽的边缘效应理论框架，探索传感器参数与边缘效应的定量关系。其次，在建立的理论模型中引入边缘缺陷模型-将缺陷等效为电场源，定量分析边缘缺陷与边缘效应的关系。然后，通过有限元探索非规则边缘-边缘角及波浪形边缘的边缘缺陷行为。最后，在理论分析、数值计算及实验研究的基础上，优化传感器及检测系统，开展系统性实验测量，提取缺陷响应特征信息并建立数据库，基于深度学习，拟实现对钛合金及CFRP边缘缺陷的预估及危险定级。本项目提出的研究理论及方法可扩展到其他导电类材料的边缘缺陷检测与分析。

钛合金具有密度低、比强度高、导热率低、耐蚀性好、无毒无磁及可焊接等特性，广泛用于航空航天结构件。针对钛合金制品，有效的检测手段对降低缺陷的安全隐患、减少经济损失、预防安全事故等均有着重要的社会意义和现实意义。已经有大量学者对远离边缘处的缺陷进行调研，并研发了相应的检测设备。但对边缘缺陷的研究还较少，且缺乏系统的检测理论及高效、精确而且便宜的检测手段。相较于非边缘缺陷，边缘缺陷处应力更集中，更易造成裂纹的扩散，安全隐患更高。因此，迫切需要研发一种有效的边缘缺陷检测手段。由于边缘效应的影响，边缘缺陷检测很具挑战性。.本申请围绕航空航天材料-钛合金的边缘缺陷检测难题，从边缘缺陷检测理论模型、传感器设计与优化、缺陷分类与评估方法等方面展开研究。具体研究内容及成果包括：1）建立了边缘效应理论框架，阐明了传感器参数与边缘效应的定量关系；2）在建立的理论模型中引入边缘缺陷模型-将缺陷等效为电场源，定量分析了边缘缺陷与边缘效应的关系；3）提出金属构件多参数解耦测量方法，可快速准确的反演电导率和厚度（相对误差小于3.3%）；4）在理论分析、数值计算及实验研究的基础上，优化了传感器及检测系统，开展了系统性实验测量，提取缺陷响应特征信息并建立数据库，建立深度学习网络，可实现钛合金边缘缺陷（长1mm、宽0.1mm）的检测与分类。本项目提出的研究理论及方法可扩展到其他导电类材料的边缘缺陷检测与分析，对预防安全事故及减少经济损失均有重要意义。