19 th World Conference on Non-Destructive Testing 21 Detection and Imaging of Internal Cracks b Tangential Magnetic Field Component Analsis using Low-Frequenc Edd Current Testing Takua YASUGI, Yatsuse MAJIMA, Kenji SAKAI, Toshihiko KIWA, Keiji TSUKADA Okaama Universit, Okaama-Shi, Japan Contact e-mail: tsukada@cc.okaama-u.ac.jp Abstract. Edd current testing using a pickup coil is usuall applied to the surface detection of cracks because of its high frequenc. However, for metal infrastructure, detection of not onl the surface but also internal cracks is desired. Therefore, we have developed a low-frequenc edd current testing sstem to inspect interior structures using a magneto-resistive (MR) sensor with a wide measuring frequenc range, ranging from DC to high frequencies. For imaging of the crack shape, especiall the front edge of the crack, a tangential component detection and analsis method that can directl detect the edd current distribution was developed. This distribution reflects the shape of the crack because the current distribution is changed b the eistence of the crack and the current is concentrated at the front edge of the crack. Edd current testing is difficult to appl to steel constructions because of magnetic noise because of changes in permeabilit and magnetization of ferromagnetism. In order to investigate the influence of magnetism, we compared two different tpes of materials: aluminum plates as a non-magnetic material and steel plates as a magnetic material. We found that, regardless of the material, internal cracks were imaged b the highl sensitive MR sensor at a low frequenc and with tangential magnetic field measurements. 1 Introduction Metal is a material used in a variet of products such as industrial parts. Because defects in metals can cause accidents in societ, guaranteeing the reliabilit of structures made of metallic plates is important. Therefore, a non-destructive evaluation method for detecting a defect and estimating its shape in a metallic plate is required. However, the surface of the metal is usuall painted to prevent corrosion, resulting in an obstacle during the inspection process. For this reason, a low-cost and fast evaluation method is desirable. Among non-destructive testing methods, edd current testing (ECT) has the advantage that it can find a flaw in the metal from the painted surface. However, owing to skin effects, the detectable range of flaws using ECT is the surface or near the surface. We have previousl reported on a low-frequenc ECT sstem that can be applied to the detection of deep metal cracks [1]. Although a low-frequenc magnetic field has a large skin depth, it is difficult for a pick-up coil to measure a low-frequenc magnetic field. To measure a lowfrequenc magnetic field, a magnetoresistive (MR) sensor, having a sensitivit ranging from DC to high frequenc, has been emploed in our research. In contrast, the magnetic field License: http://creativecommons.org/licenses/b-nd/3./ 1 More info about this article: http://ndt.net/?id=19258
strength that an MR sensor can measure is limited because of its high sensitivit. If the ecitation and detection directions are the same, the MR sensor will be saturated. To overcome this problem, the MR sensor should be placed to reduce the applied magnetic field at the sensor position [2-4]. In this stud, a tangential magnetic field detection sstem was developed using an MR sensor. Obtaining a current distribution pattern is possible b detecting a tangential magnetic field [1], and the edd current induced b a low-frequenc magnetic field enables the detection of internal defects from the surface. Compared to the conventional pick-up coil used in ECT, an MR sensor has a high spatial resolution. This implies that the specific magnetic field produced b the partial current can also be detected. Thus, the detected magnetic field reflects the partial current. To create an image that reflects the edd current in the sample, line scans can be repeated using an automatic scanning stage. In addition, we also investigated the influence of magnetism b comparing the magnetic responses of aluminum and steel plates. This was performed since it is difficult for ECT to evaluate magnetic materials. 2 Eperimental 2.1 Measuring Sstem The measuring sstem consisted of a measurement probe with an MR sensor, an ecitation coil, an oscillator, a current source, a lock-in amplifier, an - stage, and a personal computer, as shown in Fig. 1. The ecitation coil was 3 mm in diameter, had 1 turns, and operated using a sinusoidal current of.5 A. The frequenc of the ecitation coil was 5 Hz and 1 khz. The MR sensor attached to the bottom of the probe detected and magnetic components parallel to the metal surface. An ecitation coil induces a uniform edd current in the sample. A crack in the sample disturbs the uniform edd current distribution and generates a difference in the current densit. The edd current abnormalit in the sample can be detected using measurements of the tangential magnetic component that is perpendicular to the applied magnetic field on the surface. The magnetic field reaches a maimum just above the position of the current. Thus, the magnetic field distribution is represented as a vector b detecting the and direction magnetic fields. The lock-in amplifier detects the output signal from the MR sensor and calculates the signal intensit R and the phase θ. In the current sstem, the MR sensor detected onl one magnetic field direction at a time. Thus, after the sample was measured from one sensitive direction, the sample was rotated 9 degrees and scanned at the same measurement range. To obtain the defect signal with particular phase components, the phase shift method was used. The imaginar part of the signal with the phase-shift value was calculated using the following equation [5]: = sin +, (1) where and are the measured signal intensit and phase, respectivel, and is a phaseshift value. This value depends on the measurement environment and sample. The phaseshifted value of reflects the component of electric current flow. This value is used as an evaluation. Furthermore, the two-dimensional edd current intensit distribution pattern is given b the following formula: = 2 + 2, (2) where is a constant. 2
12 mm Current source Function generator z Induction Coil MR sensor Sample Lock-in amplifier PC - stage 2.2 Specimens and conditions Fig. 1. A schematic diagram of the edd current sstem. Two tpes of material, aluminum and iron plates, were used in the eperiments as samples. Figure 2 shows the configuration of two tpes of aluminum plates. One is overlapped aluminum plates, and the other is a continuous plate. The overlapped aluminum plates were composed of 1 laers. Each plate measured 12 15 1 mm 3, and onl one plate had a rectangular through-hole in the center. The shape of the hole was 15 mm in length and 1 mm in width, and the depth was 1 mm (through-hole). The aluminum plate with a through-hole was placed either at the first laer or the third laer from the top. The continuous aluminum plates with a hole measured 12 15 1 mm 3. The shape of the hole was 15 mm in length and 1 mm in width, and the depth was 1 mm (through-hole) or 8 mm (non-through hole). Edd current testing is difficult to appl to a magnetized material due to the change in permeabilit and the magnetization of ferromagnetism. Steel plates (SS4) were used to investigate the influence of magnetism. The configurations of the steel plates were the same as for the continuous aluminum plates. Details regarding the measurement samples are summarized in Table 1. A sample was placed on the - stage and measured in the range of 4 4 mm 2 around the defect with an interval of 2 mm and 21 21 steps. The signal intensit and phase were obtained at each point using a lock-in amplifier. 2 mm 8 mm 3rd laer Defect Laered strucuture 15 mm Hole 1 mm Continuous strucuture Fig. 2. A schematic of the test samples. 3
Table 1. Sample number and material. Sample no. #1 #2 #3 #4 #5 # Material Al Al Al Al Steel Steel Defect position from surface Top laer 3rd laer Through 2 mm Through 2 mm Structure Piled Piled Continuous Continuous Continuous Continuous 3 Results and Discussion 3.1 Measurements of aluminum plates The multi-laered and continuous aluminum samples were measured at 1 khz. Although the MR sensor was attached to the central ais of the coil, the influence of an applied magnetic field and edd current that flows through the no-defect part was included as an offset. Therefore, the difference between the test and no-defect parts was calculated. This process clearl shows the change derived from a defect. Figure 3 shows the phase-shifted tangential magnetic field imaging of multi-laer aluminum with a surface laer defect (#1). The black lines in the images represent the slit defect location. The phase-shift values varied from 8 to, and the peak value changed depending on the phase-shift value. The magnetic signal change was large at the edge of the defect in the -direction and at the short side of the slit defect in the -direction. Figure 4 shows the results for multi-laered aluminum with a defect in the third laer (#2). The phase-shift values ranged from 2 to. Although the magnetic field signal was almost the same as that of the multi-laered sample, a change in the upper and lower area appeared in the direction, as shown in Fig. 4. Figures 5 and show the magnetic field images of continuous aluminum samples with a through-hole and a non-through-hole located 2 mm from the surface (#3 and #4, respectivel) when the phase-shift values were 2 and, respectivel. Both samples #3 and #4 ehibited similar imaging results compared with the multi-laered structure. This result indicates that a defect located less than 2 mm from the surface can be also detected. The difference between these two samples is that the signal variation for the sample with a non-through-hole (#4) was one order smaller than that with a through-hole (#3). 4
4 4 4 3 3 3 3 2 2 2 1 1 1-3 1 2 3 4 1 2 3 4 1 2 3 4-4 3 2 1 1 2 3 4 4 3 2 1 1 2 3 4 1 2 3 4-8 -4 4 3 2 1 3-3 - Fig. 3. Magnetic field images of sample #1 in the direction and direction. 4 4 4 1. 3 3 3.8 2 2 2. 1 1 1 -.8 1 2 3 4 1 2 3 4 1 2 3 4-1. 4 4 4 1. 3 3 3.8 2 2 2. 1 1 2 3 4 1 1 2 3 4 1 2 3 4-2 2 1 -.8-1. Fig. 4. Magnetic field images of sample #2 in the direction and direction. 4 12 4 1 3 3 8 2 2-1 - 1-8 1 2 3 4-12 1 2 3 4-1 Fig. 5. Magnetic field images of sample #3 in the direction and direction. 5
4 3 2 1 1 2 3 4 1.5.5 -.5-1.5 4 3 2 1 1 2 3 4 1.5.5 -.5-1.5 Fig.. Magnetic field images of sample #4 in the direction and direction. 3.2 Measurements of steel plates Steel plates (a magnetic material) were measured under the same conditions as aluminum plates. Figure 7 shows the phase-shifted tangential magnetic field of a steel plate with a through-hole (#5). Although the signal changed around the slit defect located in the center, the magnetic field distribution is different compared to that of an aluminum plate (Fig. 5). From phase analsis, it is presumed that the edd current is disturbed in a wide area compared with the case of an aluminum plate. Net, the sample with a non-through-hole (#) was measured. To avoid the skin effect, the applied magnetic field frequenc was selected to be 5 Hz. Figure 8 shows the magnetic field images for sample #. Although imaging in the direction showed that the phase-shift was 13 and was correlated to the defect shape, it was found that the imaging was not clear. The cause of the unclear image is that the signal strength is two orders smaller than that of the through-hole sample (Fig. 7 ). Figure 8 (c) shows the profile values in the direction at 13 as illustrated in Fig. 8 b a black dashed line. Although the signal variation was small, a signal change related to the defect was observed. On the other hand, the defect signal in the direction increased according to the -direction position. Figure 8 (d) shows the profile values in the direction at 13, as illustrated b a black dashed line in Fig. 8. If the red line shown in Fig. 8 (d) is regarded as the offset value, the change in the signal is considered to include the background slope caused b the offset value. This means that a small signal change related to the defect can be obtained if the influence of the offset values is eliminated. 4 3 4 3 4 3 5 2 2 2 1 1 2 3 4 1 1 2 3 4 1 1 2 3 4-5 4 4 4 5 3 3 3 2 2 2 1 1 2 3 4 1 1 2 3 4-14 -1 - Fig. 7. Magnetic field images of sample #5 in the direction and direction. 1 1 2 3 4-5
4 3.4.2 4 3.5 4 3.4.2 2. 2. 2. 1 1 2 3 4 -.2 -.4 1 1 2 3 4 -.5 1 1 2 3 4 -.2 -.4 4 3 2 1 1 2 3 4 (c).5. -.5 1..5. -.5-1. 4 3 2 1 1 2 3 4-17 -13-9 1 2 3 4 direction position (mm) Fig. 8. Magnetic field images of sample # in the direction and direction. (c) Profile of direction signal ( = 8), and (d) profile of direction signal ( = 2). (d).2.1. -.1 -.2.2. -.2 4 3 2 1 1 2 3 4 1 2 3 4 direction position (mm) 1..5. -.5-1. 3.3 Edd current imaging Figure 9 shows the edd current intensit distribution of the aluminum samples with a through-hole and a defect located at 2 mm from the surface. In both cases, the current densit was high at the corner of the defect, whose location is illustrated using a black line in the figure. When the defect was located inside the aluminum plate, the current densit distribution spread out. This is because the distance between the defect and the MR sensor increased. From the edd current images, the defect position and the approimate shape can be inferred. 4 3 2 1 1 2 3 4 8 4 2 4 1 2 3 4 Fig. 9. Edd current images of aluminum samples: sample #3 and sample #4. 3 2 1 1.8 1.2.. 7
4 Conclusion A tangential magnetic field detection sstem using the edd current testing (ECT) method was developed for imaging crack shapes. The magnetoresistive (MR) sensor detected the tangential magnetic field, which is perpendicular to the applied magnetic field. Because defect detection for a magnetized material is difficult using ECT, the influence of magnetism was investigated b comparing the measurements from aluminum plates as a non-magnetic material and steel plates as a magnetic material. Two tpes of aluminum plates (multi-laered and continuous) and one tpe of steel plate (continuous) were used. The measurement results for a steel plate with a through-hole ehibited a clear magnetic field image that reflected the defect. For an internal defect in a steel plate, a small signal change related to the defect was detected. Therefore, the developed sstem can detect both surface and internal defects, even in a magnetic material. References [1] K. Tsukada, T. Kiwa, T. Kawata, and Y. Ishihara, Low-Frequenc Edd Current Imaging Using MR Sensor Detecting Tangential Magnetic Field Components for Nondestructive Evaluation, IEEE Trans. Magn., vol. 42, no. 1, pp. 3315-3317, Oct. 2. [2] H. Yamada, T. Hasegawa, Y. Ishihara, T. Kiwa, and K. Tsukada, Difference in the detection limits of flaws in the depths of multi-laered and continuous aluminum plates using low-frequenc edd current testing, NDT E Int., Vol. 41, pp. 18-111, Mar. 28. [3] L. Ferrigno, and M. Laracca, GMR-Based ECT Instrument for Detection and Characterization of Crack on a Planar Specimen: A Hand-Held Solution, IEEE Trans INSTRUM MEAS., vol. 1, no. 2, pp. 55-512, Feb. 212 [4] T.Dogaru and Stuart T. Smith, Giant Magnetoresistance-Based Edd-Current Sensor, IEEE Trans. Magn., vol. 37, no. 5, pp. 3831-3838, Sept. 21 [5] K. Tsukada, M. Yoshioka, Y. Kawasaki, and T. Kiwa, Detection of back-side pit on a ferrous plate b magnetic flu leakage method with analzing magnetic field vector, NDT E Int., vol. 43, no. 4, pp. 323-328 Jan. 21. 8