철재 구조물의 표면 결함 진전 방향 평가를 위한 자기광학 영상화 시스템 개발

Abstract

Large equipment structures such as power generation facilities, railway vehicles, bridges, and high-rise buildings are built with a huge investment of budget and time. Therefore, in order to secure social and economic benefits from the investment, they should be operated safely while maintaining the target performance until the design lifespan. On the other hand, according to the damage tolerance design concept, if the target performance, safety, and economic feasibility are ensured, the structure can be continuously operated even after the design lifespan has elapsed. In such damage tolerance engineering, determining the allowable dimensions of damage, detecting and evaluating flaws through non-destructive testing, and implementing appropriate maintenance and replacement are crucial factors. Various types of cracks occur, such as fatigue cracks caused by voids, corrosion, or fatigue. Particularly, fatigue cracks can rapidly propagate and lead to shell failure or delamination when the direction of the flaw is perpendicular or inclined. This not only poses safety issues but also results in significant social and economic disadvantages. Therefore, this study proposes a non-destructive testing method that can quickly inspect and quantitatively evaluate the propagation direction of fatigue cracks. Prior to the development of the inspection method, the relationship between fatigue cracks and magnetic flux density was theoretically approached. Three magnetic domain models, namely RDM (rotation domain model), CDM (concentration domain model), and VDM (vertical domain model), were proposed and simulated by theoretical equations. As a result, it was found that the magnetic flux density distribution of RDM, CDM, and VDM was changed according to the depth direction of the fatigue crack even without external magnetic field applied. Based on the above results, a 3-axis TMR sensor experiment was conducted. Test specimens made of SS400 with depth directions of fatigue cracks at 0°, 30°, and 60° from the surface were fabricated. The magnetic flux density of each manufactured test piece was measured at intervals of 0.004 mm by a 3-axis TMR sensor. Through the measured results, the presence or absence of a defect was determined, the depth direction of the defect was estimated, and quantitative evaluation was conducted. In addition, when compared with the simulation results of the magnetic domain model presented in the theory, the X-direction and Z-direction components are very similar to RDM and the Y-direction components to VDM, proving the simulation results. However, although the 3-axis TMR test has the advantage of being able to measure finely, it has a major disadvantage of requiring a lot of test time. Therefore, this study proposes the development of a magneto-optical imaging system that can detect defects with fast inspection time and high spatial resolution and evaluate the surface length and depth direction. Based on the magneto-optical effect theory, this MONDI system composed of polarizer, analyzer, beam splitter, magneto-optical device (A-type and D-type MO sensor), CMOS camera, and computer was constructed. Furthermore, magnetic flux density was measured in the same test piece as in the 3-axis TMR experiment in the state of residual magnetization, horizontal magnetization, and vertical magnetization after spontaneous magnetization and vertical magnetization, and then horizontal magnetization. Although the presence or absence of defects could be estimated from the result of subtracting the background signal from the original, it was difficult to evaluate the depth direction of the defects due to the spatial resolution of the camera, noise caused by the magnetic domain width of the MO sensor, lens distortion, and noise caused by the light source. Therefore, an algorithm was developed to remove the defect through moving average and detrend and evaluate the depth direction of the defect. According to the developed algorithm, a primary equation for obtaining 0°, 30°, and 60° was derived from the test specimen results under four conditions measured with A-type and D-type MO sensors, and the error was obtained and quantitatively evaluated. As a result, the lowest error was obtained when measuring the horizontally magnetized specimen with the A-type MO sensor. Therefore, through the magneto-optical system proposed in this paper, it is possible to estimate the depth of the defect, the length of the defect, and the depth direction, and the condition for measuring the horizontally magnetized specimen with the A-type MO sensor is the most suitable.