Abstract
Eddy current testing is widely used for nondestructive evaluation of metallic structures in characterizing numerous types of defects occurring in various locations. It offers remarkable advantages over other nondestructive techniques because of its ease of implementation. This paper presents a technical review of Eddy current technique in various scope of defect detection. The first part presents Eddy current evaluation on various defects location and orientation such as steam generator tubes, stress crack corrosion, and fatigue cracks. The next section analyzes the use of pulsed Eddy current and pulsed Eddy current thermography as an alternative method for monitoring the growth of cracks with the aid of computational techniques for postsignal analysis.
1. Introduction
Preserving structural reliability is the most vital issues faced by the industries especially in aircraft, civil, and oil and gas sectors. Thus, detection of defects must be carried out in a nondestructive approach to prevent severe problems. Nevertheless, not all of the defects will essentially lead to a malfunction of the component. For instance, a very tiny crack would not affect the function of the component at the time of inspection. Thus, an immediate repair or a replacement of the component would lead to an unnecessary financial burden if the evaluation assures that the crack remains sufficiently small and will not affect the integrity of structures until the next scheduled inspection. Therefore, it is not enough to simply detect the presence of a defect, whilst the most crucial part is the ability to assess the consequence of the defects on the operation of the component. As a result, the quantitative characterization of defects using a nondestructive technique is crucial. This implies the important roles played by the nondestructive techniques which are not only to detect cracks but also to evaluate them.
Eddy current inspection is one of the nondestructive techniques which applied to conductive materials. The main advantage of Eddy current is its range of inspections and measurements which covers the crack detection [1, 2], material and coating thickness measurement [3–5], conductivity measurement for material identification [6, 7], heat damage detection, case depth determination, and heat treatment monitoring [8]. It offers numerous advantages such as high sensitivity, rapid scanning, and flexibility which contributed to its broad utilization. In particular, Eddy current is known for its contactless inspection technique with no physical contact between the test probe and the test piece.
According to the state of the art, Eddy current technique has a strong application in defect detection. The sensitivity to characterize defects and other parameters can be improved by an optimal choice of probes and operation frequency. One of the considerable solutions for this problem is to observe the growth of a crack by positioning a probe at the detected crack and continuously collecting the signals. Ultrasonic testing is superior in evaluating the profile of cracks compared to numerous nondestructive methods. However, in general, it is not suitable for monitoring crack growth due to the need for couplant. For that reason, Eddy current technique is regarded as promising technique compared to other nondestructive methods.
2. Defect Location and Orientation
Since the 1950s, the role of Eddy current technique has found prevalent use in aircraft safety inspection procedure, piping and vessels inspection, and aerospace as well as nuclear and automotive industries. The extensive research and development in highly sensitive Eddy current sensors and instruments over the last sixty years indicate that Eddy current technique is currently a widely used inspection technique. The Eddy current technology lends itself well to the detection of near-surface or surface breaking defects such as surface scratches and corrosion and distinguishes types of conductive materials. Hellier [9] compared several nondestructive methods with the Eddy current technique for different types of shapes and locations as in Table 1.
Summary of nondestructive methods for various shapes and locations.
U: unsatisfactory; P: possible; A(1): first order of preference; A(2): second order of preference; A(3): third order of preference; and A(4): fourth order of preference.
Inspection of billet possible.
Possible using surface waves.
Assuming it is completely internal.
may be seen on ends by penetrant or magnetic particle.
If open to surface.
Thermal tests possible.
Note. Acoustic emission testing (AET) has not been included in this guide because this method applies only to those discontinuities that propagate under applied load.
Commonly employed Eddy current procedures are capable of reliably detecting cracks as small as 0.050 inches while maintaining false calls below 1%. However, to achieve such detection rates requires careful settings of threshold levels and appropriate setup standards [10]. Spencer [10] summarized the potential of advanced Eddy current technology using several advanced systems which involved inspections of several rivet skin splices, representative of actual aircraft structure containing cracks ranging from 0.040-, 0.060-, and 0.080-inch standards with thresholds set to the lowest reasonable level for the particular system. The results demonstrated that some of the systems were able to detect cracks as small as 0.040 inches with false call rates which remained less than 1%. Meanwhile, defect detection of magnitude crack length and depth of 0.4 mm × 0.12 mm was detected on the heated metallic components [11].
2.1. Steam Generator Tubes
Steam generator (SG) tube plays a critical safety role because it composes one of the primary blockades between the radioactive and nonradioactive sides of the plant. Therefore, the reliability of the tubing is essential in lessening the leakage of water between the two sides of the plant. Steam generator tubes in nuclear power plants have a long history of cracking due to stress corrosion cracking (SCC). Periodic inspection using Eddy current technique is commonly employed to detect various forms of tube degradation including SCC. As a result, it is crucial to assess and validate the reliability of the Eddy current technique used for evaluating the structural dependability of components [12].
Remote field Eddy current (RFEC) technique had been used for in-service inspection of the ferromagnetic SG tubes made of modified 9Cr-1Mo steel [13], in nuclear power plants [14] and on nonmagnetic SG tubes [15]. Expansion bends were provided in the SG to accommodate differential thermal expansion [13]. Meanwhile, Kobayashi et al. [16] employed the RFEC technique to inspect the helical-coil-type double wall tube steam generator with the wire mesh layer. The use of multiphase rotating magnetic fields RFEC was proposed for detection of SCC in gas transmission pipelines [17].
Eddy current probes were used extensively to evaluate the integrity of SG tubes in nuclear power plants and to detect cracks in tube walls. Standard practice of Eddy current probes employed for SG tube inspection is bobbin coil probe, rotating probe with two pancake coils and one plus coil, and array probe. However, the probe design has evolved from simple bobbin probes to mechanically rotating multicoil probes providing high resolution images of tube integrity. A rotating probe with a higher sensitivity and spatial resolution was developed to compensate for bobbin coil limitations [18]. The sizing accuracy and the probability of detection for Eddy current probes were dependent on the location and orientation of the defects and the artifacts such as corrosion deposits and tube support structures [19]. A rotating field Eddy current probe with bobbin pickup coil was developed by Xin et al. [20]. Meanwhile, Li et al. [21] proposed the array probe which consists of multiple coils arranged around the circumference of the probe in addition to a standard bobbin coil probe. The research by Kim and Lee [22] also reported on Eddy current probe with structure designed to be sensitive to circumferential cracks and to axial defects. In addition to the various Eddy current probe designs, Todorov [23] optimized and designed two encircling Eddy current magnetic sensors with uniform field.
Tian et al. [24] focused on the signal extraction and classification from mixed RFECT signals of signal processing and defects using a neural network and a statistic. Wavelet transform (WT) was used in the probe wobble denoising by Lopez et al. [25]. Jo and Lee [26] proposed a feature extraction from Eddy current signals, as an input vector and multilayer perceptron (MLP) neural networks were then used to classify defect types and to predict defect size.
A surface profiling technique for monitoring local deformation and identification of surface wear in the pressure tubes was presented using an Eddy current probe mounted in a small (50 mm × 25 mm) planar probe [27]. Optimized Eddy current array technique was used to find material degradation, especially intergranular SCC of SG tubes during the periodical in-service inspection [28].
2.2. Fatigue Crack
In recent years, several studies have focused on small fatigue cracks due to 70–80% of the total fatigue life spent in the crack initiation and small crack growth stages. Previous researches showed that cracks initiated at the very beginning of fatigue life and the propagation of small crack were influenced by the inherent microstructure resulting in the abnormal propagation behavior as compared with the long crack behavior [29–32]. Therefore, the precise measurement of small crack growth rate and the good understanding of small crack growth mechanism are very important in the reliable prediction of the fatigue life of materials [33–46].
The numerical modeling of fatigue and SCC was studied in Eddy current simulations using a 316 stainless steel to assess the sizing of cracks which appeared in general structure [47]. An online inspection for imaging multiple cracks caused by rolling contact fatigue of rail transportation was proposed using pulsed Eddy current thermography (PECT) [48]. Meanwhile, the Eddy current inspection of small fatigue cracks caused by different levels of static load in Ti-6AL-4V [49] and a modelling of thermal fatigue [50] were evaluated using the finite element (FE) model.
Bohacova [51] proposed the Eddy current technique in an operating frequency range between 200 Hz and 100 kHz with a single-value interpretation of the individual EC signals for the detection of short fatigue cracks hidden under a rivet head on the aircraft wing. On a different study, a uniform Eddy current probe with 23 arrayed detectors was designed to measure signals from six mechanical fatigue cracks initiated into type 316L austenitic stainless steel plates [52]. Weekes et al. [53] used the induction PECT for detection of fatigue cracks in steel, titanium, and Waspaloy.
As reported by many researchers, fatigue cracks are also developed around fasteners found in multilayer aluminum structures on aging aircraft or originating near bolt holes in the inner layers of aircraft lap joints. Pulsed Eddy current was employed for detection of holes and EDM notches beneath rivet heads in subsurface layers of stratified samples [54] and for detection of deep-lying cracks originating near bolt holes in the inner layers of aircraft lap joints [55]. Meanwhile, Joubert et al. [56] reported the usage of an array probe for imaging of submillimetric surface breaking defects in bore holes of metallic parts. Underhill and Krause [2] employed Eddy current techniques within the bolt holes for risk assessments used in evaluating the serviceability of the aircraft. In contrast, Yang et al. [57] developed an Eddy current-giant magnetoresistive (EC-GMR) sensor system used in detecting subsurface cracks at fastener sites with aluminum fasteners present.
2.3. Corrosion
Corrosion of ferromagnetic components is a widespread damage in oil and gas, chemical, electric power, metallurgy, and other related industries. As an example, ferromagnetic pipes and containers are used to transport and store liquid or gaseous corrosive media with most components which usually work under the conditions of high temperature and high pressure. As a result, corrosion of ferromagnetic components is unavoidable. On the other hand, climate conditions are the most significant cause of corrosion between the layers of aircraft fuselage.
A few researches had been described on the practice of quantitative evaluation of stress corrosion cracks [58, 59]. However, current studies had revealed that the accuracy of evaluation strongly depends on the crack modeling [47, 60–62]. In contrast, Hosseini and Lakis [63] used pulsed Eddy current to detect corrosion and cracks in multilayer aluminum structures in aircraft applications.
Wall-thinning defects usually caused by corrosion are potential hazards to safety and could lead to pipeline leakage, explosion, or other accidents to the ferromagnetic objects such as carbon steel pipes and vessels which are commonly used in petrochemical and power generation industries [64, 65]. Therefore, regular in-service inspection and evaluation for remaining wall thickness of ferromagnetic components are crucial to ensure safe operation. Due to the presence of thick thermal insulation and metallic cladding wrapped around the outside of the insulation, the evaluation of the pipe wall-thinning without removal of the insulation and cladding is a challenging task. An electromagnetic acoustic transducer-Eddy current (EMAT-EC) dual probe was developed to assess wall-thinning [66].
2.4. Other Applications
The material microstructure at the heat affected zones near welds was characterized using Eddy current technique [67]. The estimation of pearlite percentages in low to high plain carbon steels was studied using Eddy current technique in [68, 69] and Ghanei et al. [70] estimated the pearlite percentage in ductile cast irons for prediction of mechanical properties in cast irons.
The measurement of conductive material thickness was performed using Eddy current technique [71]. On the other hand, Chen and Lei [72] used the PEC for measuring the thickness of a ferromagnetic plate. Table 2 classifies the Eddy current techniques employed to defect detection in major areas of crack occurrence.
Eddy current used for various defect locations and orientations.
3. Pulsed Eddy Current (PEC)
In recent years, several studies have focused on the pulsed Eddy current (PEC) technique as an effective method of quantifying defects in multilayer structures due to its richer information in time and frequency domain. It has been used in inspection of aircrafts, oil/gas pipelines, nuclear steam pipes, and high-speed rails.
A study by Abidin et al. [54] demonstrated the advantage of PEC to detect holes and EDM notches beneath rivet heads in subsurface layers of stratified samples without the need for reference samples through the varied pulse width feature. Babbar et al. [55] employed the finite element modeling with the PEC in detecting deep-lying cracks originating near bolt holes in the inner layers of aircraft lap joints with ferrous fasteners present. Differential signals from both the top layer and the bottom layer cracks in different orientations and with different probe displacements were analyzed using a modified principal components analysis (PCA) to differentiate cracks from blanks. Meanwhile, Horan et al. [73] analyzed the PEC generated by a probe designed to utilize the ferrous fastener as a flux conduit to detect simulated cracks within the spar with the wing skin present. The PCA was also used to overcome variability in PEC signal response due to variability in magnetic coupling to the fastener.
In a different study, Yu et al. [74] modeled the lift-off effect introduced by various coating thicknesses from irregular sample surface or movement of transducers. However, simplification of the measurement should be further improved and widen the proposed application on the sample with the ferrous material and on the subsurface defect. Tian et al. [75] investigated the spectral response of PEC under varying probe liftoff, material properties, and directional tensile stress by using normalization in frequency domain. However, the application of the proposed method can be enhanced in corrosion characterization, displacement measurement, and profile inspection using the time-domain features. On the contrary, He et al. [76] developed the automated defect classification using PCA and SVM under different interlayer gaps and lift-off effects.
The fast crack profile reconstruction methods using transient time and spectral components of PEC signals were carried out on EDM slots of 3 mm, 4 mm, 6 mm, and 8 mm depth [77]. The method provided initial approximate profiles for crack shape reconstruction using different PEC signals such as amplitude, phase, real and imaginary values of spectral components, and transient slices using reconstruction mean square errors (MSE) which reduces the computing time. In another study, He et al. [78] classified surface and subsurface defects using features-based rectangular PEC sensor. Nevertheless, the proposed method should be improved by performing real-time defect identification. The defect classification in the con-casting slabs (CCS) was investigated through PEC with the help of PCA-LDA and PCA-Bayes [79]. In a different study, He et al. [80] examined the magnetic field intensity and conductivity to characterize the low-energy impacts in carbon fibre reinforced plastics (CFRP) laminates and internal inserted defects in honeycomb sandwich panels. Although it offers higher reliability and better detection ability for deep defect, the proposed method was time-consuming.
Xiao and Li [81] focused on solving PEC at a high speed while maintaining the accuracy using the combined analytical-numerical approach. Hosseini and Lakis [63] converted the PEC data from time domain to time-frequency domain and analyzed the data using maximum variances of PCA. Theodoulidis et al. [82] evaluated the PEC interaction with a crack in a planar conductor using the Fourier superposition with the current pulses decomposed into distinct frequencies.
A signal denoising method for PEC signal was established for ferromagnetic material which transformed the averaged PEC signal from Cartesian domain to double logarithmic domain [83]. However, the method was not applicable for real-time denoising.
4. Pulsed Eddy Current Thermography (PECT)
In recent years, several studies have focused on the combination of pulsed Eddy current and thermography techniques known as pulsed Eddy current thermography (PECT) for area imaging of defects without scanning. The PECT technique can be divided into two categories which are the induction and heat diffusion. The induction PECT is carried out using a short burst of electromagnetic excitation applied to the material under inspection which induces Eddy current flowing in the material. If the induced Eddy current encountered a discontinuity, it will be forced to divert which leads to area of increased and decreased current density. The area with increasing current density experiences higher levels of Joule heating (ohmic); thus the defect can be identified from the IR image sequence during the heating and cooling period.
In contrast, the heat diffusion PECT applies the flash thermography. The interactions between the heating mechanism and the defect occur in two ways through diversion of induced Eddy current and through heat dissipation. This results in greater change in heating around defects especially for vertical and surface breaking defects. However, as with traditional Eddy current inspection, the orientation of a particular defect with respect to induced currents has a strong impact. The sensitivity decreases with defect depth and the technique is only applicable for samples with a minimum level of conductivity such as ferromagnetic, paramagnetic, and conductive nonmetals as the heating is directly proportional to the Eddy current density and diffused heat.
The relationship between divergence and transient thermal patterns as well as between divergence and impact energy was analyzed within CFRP using the induction PECT [84]. In contrast, Pan et al. [85] employed the heat diffusion PECT to investigate the carbon fibre structure, delamination, and impact in CFRP. Nevertheless, the study only focused on the qualitative information on delamination and impact of CRFP which can be improvised on the quantification information on the impact and delamination of the CFRP. Research by Liu et al. [86] utilized the heat diffusion PECT in investigating the impact of thermal image sampling rates versus feature extraction for defect characterization. Meanwhile, Wilson et al. [48] conducted feasibility study using heat diffusion PECT for imaging multiple cracks caused by rolling contact fatigue. The report which illustrated the method was viable solution for detecting real defects and measurement for multiple natural cracks on a complex shape.
Another studies by Xu et al. [87] modeled the peak value and the time-to-peak features of wall-thinning insulated ferromagnetic pipes using the induction PECT. In addition, Xie et al. [88] proposed a numerical solver for simulation of induction PECT signals based on the Fourier series and interpolation strategy. Table 3 provides the summary of the signal analysis tools used with the PEC and PECT.
Signal analysis tools used in the studies of PEC and PECT.
5. Development of Special Eddy Current Probes Design
A special attention had been focused on designing Eddy current probes for specific applications in the recent years. The design of Eddy current probes was optimized in order to increase the sensitivity and resolution. The study of the magnetic fields in the vicinity of a probe was suggested by Zergoug et al. [67] to characterize the field activity and optimize the controlled measuring process and the relevant sensitive probe. Zergoug et al. [67] outlined the important parameters in designing a probe which were the reduction of the reluctance, the optimal energy exchange between the probe and the material, the use of a high magnetic permeability and a low electric conductivity core material, and the choice of the excitation current appropriate for the ferromagnetic material under testing.
The use of PEC and a specific probe design was performed for detection of defects near rivets [54]. He et al. [89] employed the two-stage differential coil probe to detect defects between third layer and fourth layer in riveted structures and Underhill and Krause [2] used a multiple frequencies probe for defect detection around bolt holes of aircraft lap joints.
The planar design and a differential operation probe were proposed for inspection of imperfections along friction stir welding (FSW) joints at root and top zones of FSW beads [90] and the ionic probe for defects detection on aluminum solid state processed alloys as FSW and Friction Spot Welding [91].
6. Signal Processing Using Special Software Program
With the advancement of computer systems and numerical methods, the interest in Eddy current technique had resolved the complex problem concerning the postsignal analysis. Besides the defect characterization, actual studies dealt with the metallurgical evaluation of materials. Surface assessment allows the prediction of the material strength and consequently its life span. This ability has made Eddy current technique a sophisticated method capable of fulfilling the specific demands connected with inspections. The precise and rapid forward numerical simulations were performed for different Eddy current applications [92]. Meanwhile, Yusa et al. and Badics et al. [59, 93] solved the inverse problem for defect characterization and evaluation for SCC using computational physics. In many circumstances, the design of Eddy current probes for different applications and testing conditions was supported by simulations with finite element modeling [56, 94], [95]. The closed form expressions were obtained to evaluate the probe responses [96]. The probe wobble denoising method was used in conjunction with the Eddy current evaluation [25]. A pseudomonitoring tests were designed to measure signals from six mechanical fatigue cracks introduced into 316L austenitic stainless steel plate [97].
7. Conclusion
Eddy current technique has played its vital role as part of the nondestructive technique chosen by the industry. Reliability of the technique during an in-service inspection was evaluated by a destructive examination and proven to show a better capability in defect detection. In spite of its very useful advantages, the Eddy current technique has a particular drawback that indeed originates from its underlying principles. However, the developments of pulsed Eddy current technique and pulsed Eddy current thermography as well as specific Eddy current probes design have overcome chronic problems and limitations in Eddy current inspection. The designed probes significantly increase the scope of inspection using standard commercially available instruments. Furthermore, the evolution of computer technology has led to various analytical solutions and thus eliminates unwanted signals, promising a more certain and unambiguous resolution of deep defects in thick structures.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Footnotes
Acknowledgment
This research was funded by the Ministry of Higher Education (MOHE) through the Fundamental Research Grant Scheme (FRGS/1/2013/SG02/TATI/03/1).