摘要
漆层下钢材腐蚀的无损检测和评估是钢结构健康监测领域的一大挑战。脉冲涡流热成像(ECPT)技术可以通过钢材表面红外辐射信号反映漆层下的钢材物理性质。本文提出基于红外信息主成分分析(PCA)的特征提取、图像重构,以及腐蚀时间表征方法,用于防腐漆层下钢材表面腐蚀的无损检测与评估。分析试样表面温度变化发现,电导率对早期腐蚀的焦耳热产生和分布起主导作用。结合特征提取和重构方法,对于6个月以内的早期腐蚀,第二主成分(2ndPC)能够准确检测腐蚀轮廓;归一化的2ndPC均值具有良好的检测灵敏度,并能单调表征10个月以内的钢材表面腐蚀,从而证明了脉冲涡流热成像技术和主成分分析法是防腐漆层下钢材表面腐蚀无损检测和评估的有效方法。
Paint coated corrosion non-destructive detection and evaluation is a big challenge to steel structure health monitoring. Eddy current pulsed thermography (ECPT) technique can reflect the steel prop-erties through the Infrared signal from the paint coated steel surface. This paper provides a princi-pal component analysis (PCA) based approach for Infrared information feature extraction, image reconstruction, and corrosion time characterization, and utilizes this approach to nondestructively test and evaluate the steel corrosion under coating. By surface temperature variation analysis, the electrical conductivity is considered to dominate the Joule heat generation and distribution. With feature extraction and reconstruction method, the 2nd principal component (PC) can accurately de-tect the profile for early stage corrosion, within 6 months. The normalized mean of 2nd PC can test the corrosion with good sensitivity and characterize the steel surface corrosion within 10 months monotonously. Therefore, the proposed approach which combines ECPT technique and PCA method is appropriate for nondestructive testing and evaluation for paint coated steel surface corrosion.
出处
《传感器技术与应用》
2022年第4期538-549,共12页
Journal of Sensor Technology and Application
