Structural Damage Evaluation of Carbon Steel Based on Dynamic Scanning Eddy Current Infrared Thermography
Tu Yanxin1, Mei Hongwei1, Liu Lishuai2, Shen Zekai1, Wang Liming1
1. Guangdong Engineering Technology Research Centre of Power Equipment Reliability in Complicated Coastal Environments Shenzhen International Graduate School Tsinghua University Shenzhen 518055 China; 2. Key Laboratory of Pressure Systems and Safety of Ministry of Education East China University of Science and Technology Shanghai 200237 China
Abstract：As a common metal, carbon steel has been utilized widely in the power industry. Nevertheless, faults such as fatigue, cracks, and delamination are unavoidable in carbon steel during the process of manufacture and use, and these defects can even lead to the fracture failure of equipment. Detecting structural defects in carbon steel in a timely manner is crucial for ensuring the safety of power industrial systems. Compared to traditional non-destructive testing techniques, dynamic scanning eddy current infrared thermography (DSECT) has application potential for the inspection of carbon steel materials in the power industry. For the practical needs of structural damage detection, the principle of DSECT and related algorithms are explored, as well as the applicability of the technology for detecting faults in carbon steel. Firstly, the thermal wave propagation model of DSECT is derived based on Fourier heat conduction. Secondly, an algorithm for thermographic signal reconstruction (TSR) is proposed, in which the thermal wave signal is fitted in the logarithmic coordinate system to eliminate the noise of the thermal wave signal. Thirdly, an experimental study is carried out to verify the effectiveness of the TSR algorithm by detecting artificial defective carbon steel specimens using DSECT. Finally, the frequency of the minimum phase is extracted by the fast Fourier transform for the purpose of quantitatively detecting the defect depth, and the defect depth is estimated quantitatively by using the linear relationship between the defect depth and the inverse of the square root of the frequency of the minimum phase. In a word, DSCET overcomes the drawbacks of traditional non-destructive testing, and the related algorithm enhances imaging and defect depth measurement. Experimental test results on artificial defect samples show that the coil arrival order matches the peak temperature arrival order. The temperature at the detection point remains largely constant while the coil does not reach the detection point, increases rapidly as the coil leaves the detection point, and peaks as the coil leaves the effective heating region. The TSR-based thermogram sequence algorithm is used to reconstruct the original experimental data and the results show that, the thermophysical change process in the reconstructed thermogram sequence conforms to the thermal wave propagation law; the reconstruction algorithm can eliminate coils from the field of view and achieve uniform heating of the tested specimen. In addition, the minimum phase in the reconstructed thermogram is linearly proportional to the frequency, and the frequency domain properties are correlated with defect depth. The defect depth can be fitted with the square root inverse of the frequency of the minimum phase as a linear relationship, and the goodness of fit is 0.991 2. Results show that the DSECT approach and related algorithms proposed can depict faults in two-dimensional imaging more simply and clearly, and also offer quantitative measures of defect depth. The key conclusions are as follows. (1)DSECT's dynamic coil scanning allows for a single, wide-range specimen detection, overcoming the single-coil excitation method's limitations. (2)By using TSR algorithm, a smooth reconstruction of the thermal wave signal can be accomplished and the thermal wave signal noise is eliminated. (3)The TSR-based reconstruction method of the thermogram sequence eliminates the coils in the thermogram field of view and achieves uniform heating of the specimen in the thermogram. (4)Minimum phase is the effective feature for quantitative detection, and the defect depth can be quantitatively estimated using the linear relationship between the defect depth and the inverse of the square root of the minimum phase frequency.
2023, Vol. 38 
