A compact guided-wave EMAT with pulsed electromagnet for ferromagnetic tube inspection

Abstract

Small-diameter tubes that are widely used in petroleum industries and power plants experience corrosion during long-term services. In this paper, a compact inserted guided-wave EMAT with a pulsed electromagnet is proposed for small-diameter tube inspection. The proposed transducer is noncontact, compact with high signal-to-noise ratio and unattractive to ferromagnetic tubes. The proposed EMAT is designed with coils-only configuration, which consists of a pulsed electromagnet and a meander pulser/receiver coil. Both the numerical simulation and experimental results validate its feasibility on generating and receiving L(0,2) mode guided wave. The parameters for driving the proposed EMAT are optimized by performance testing. Finally, feasibility on quantification evaluation for corrosion defects was verified by experiments.

Keywords

Guided wave EMAT pulsed electromagnet tube inspection

1. Introduction

Small diameter tubes are widely used for fluid transportation and heat exchange in petroleum industries and power plants. These tubes experience various types of corrosion due to high temperature, high pressure, fluid flush and radiation effect during long-term services [1]. Small diameter tubes are tested presently by ultrasonic testing (UT) method and eddy current testing (ECT) method using an inserted rotating probe to scan through the tubes [2]. Both of them suffer from the fact that extensive cleaning has to precede any inspection process. They require a probe-head to travel through the entire length of the tube, either by pushing and subsequent pulling by hand or through the use of a propelling system, so that obstructed and bent segments cannot be tested.

Ultrasonic guided wave (GW) testing method is efficient for structures with thin wall such as pipelines and plates. Guided wave is ultrasound propagating through the tube wall, so signals of guided wave involve defect information of the whole tube [3]. An ultrasonic GW probe only needs to be inserted within on end of the tube, and remains in the fixed position during the test. No push-pulling unit is required. Moreover, ultrasonic guided wave has advantages such as low attenuation, mode variety and accessible for bent tubes [4,5]. Among the modes of guided wave in tubes, the longitudinal wave L(0,2) is sensitive to corrosion defects. It is of low leakage in tubes full of liquid, because the particle displacement is in axial direction involving in-plane component.

At present, conventional GW probes including piezoelectric transducer (PZT), magnetostrictive sensor (MsS) and electromagnetic acoustic transducer (EMAT) are complex in configuration with large size. This feature may limit the application on some extreme objects such as small diameter tubes in narrow space. For the contact probes like PZT and MsS, couplant or surface treatment is needed for a good coupling efficiency, which can make the inspection process complicated. The permanent magnets of EMATs attract and move difficultly from ferromagnetic objects. Such behavior may disturb the inspection, and make the installation and removal of the transducer become difficult and even cause mechanical damage to the transducers or test objects [6].

In this paper, a compact inserted guided-wave EMAT with a pulsed electromagnet is proposed for corrosion inspection in small diameter tubes. It is a noncontact and coupling-free transducer based on electromagnetic mechanism, which can simplify inspection process. Owing to the pulsed electromagnet, the proposed EMAT has more compact configuration and higher signal-to-noise ratio (SNR). The pulsed electromagnet can induce a momentary strong magnetic field so the transducer can be easily placed inside ferromagnetic tubes. The inserting design is suitable for tubes with open edges such as small diameter tubes in steam generator or boiler.

2. Basic design and numerical simulation

2.1. Configuration and working principles

The diagram of the inspection system and configuration of the proposed inserted guided-wave EMAT is shown in Fig. 1. The EMAT is composed with a pulsed electromagnet in form of a solenoid and a circumferential meander coil as shown in Fig. 1(b). The electromagnet is wrapped in gap of the bobbin with enameled copper wire. Energized by a rectangular long-pulse current lasting 2 ms with more than 100 A of magnitude, the electromagnet can induce a momentary strong magnetic field with a large flux density B. The pulser/receiver coil (meander coil) is wrapped outside the electromagnet. During the period of long-pulse current driven into the pulsed electromagnet, the pulser/receiver coil is energized by a radio frequency (RF) pulse current to induce dynamic magnetic field. Guided wave can be generated and received via magnetostrictive mechanism and its inverse effect respectively. The magnetostrictive force for guided wave generation is in axial direction, which is valid for L(0,2) mode. In the inspection system, the long-pulse generator and RF pulser are triggered by a signal generator successively with an adjustable time delay.

Fig. 1.

Sketch diagram of the proposed guided-wave EMAT and inspection system.

2.2. Parameter determination

The dispersion character is determined by the size and material of the tubes. For a carbon steel tube with 28 mm of outer diameter, 2 mm of thickness, the dispersion curve for both group velocity c_g(f) and phase velocity c_p(f) can be calculated theoretically. To avoid the effect of multi-mode excited, frequency lower than the cut-off frequency of L(0,3) mode is alternative for excitation. Then according to differential curve of dc_g(f)∕df calculated as shown in Fig. 2(a), the excitation frequency is determined while dc_g(f)∕df = 0, for avoiding the wave pack expansion during long distance propagation. The wavelength can be determined as c_p(f)∕f in Fig. 2(b) according to the dispersion curve of phase velocity. The distance between meanders of the pulser/receiver coil is set as half of wavelength. In this case of the carbon steel tube, the excitation frequency is set as 0.21 MHz and the distance between meanders as 13 mm.

Fig. 2.

Curves of dispersion character of guided wave.

2.3. Simulation method and results

The principle of EMATs for ultrasonic wave generation is sufficiently studied based on magnetostrictive mechanism. The governing equation derived from the Ampere’s law and Faraday’s law due to the pulsed electromagnet and pulser/receiver coil can be expressed as [7]: $\begin{eqnarray}\displaystyle {\nabla}\times \frac{1}{{\mu}}\times \mathbf{A}+{\sigma}_{c}\frac{\partial \mathbf{A}}{\partial t}=\mathbf{J} & & \displaystyle\end{eqnarray}$ (1) where μ is the magnetic permeability for the ferromagnetic tubes; σ_c is electrical conductivity; $\mathbf{A}$ is the induced magnetic vector potential and $\mathbf{J}$ is the driven current for both coils. The magnetostrictive force $f_{i}^{ms}$ in the subsurface of the tube inner wall can be expressed as [8]: $\begin{eqnarray} \displaystyle f_{i}^{ms}=-{\displaystyle \frac{1}{2}}(3\stackrel{\frown}{{\lambda}}+2\stackrel{\frown}{{\mu}})(1-2{\nu}){\displaystyle \frac{\partial {\lambda}_{s}}{\partial M_{oi}}}{\displaystyle \frac{\partial m_{i}}{\partial x_{i}}} & & \displaystyle \end{eqnarray}$ (2) where $\stackrel{\frown}{\lambda}$ and $\stackrel{\frown}{\mu}$ are Lamé constants; 𝜈 is Poisson’s ratio; 𝜆_s is line magnetostriction ratio; M_oi is the magnetization intensity affected by the pulsed electromagnet and m_i is dynamic magnetization intensity affected by the pulser/receiver coil. Then inducing the magnetostrictive force to the wave equation as follow, the ultrasonic field and signals of velocity can be calculated. $\begin{eqnarray} \displaystyle {\mu}{\nabla}^{2}\mathbf{u}+(\stackrel{\frown}{{\lambda}}+\stackrel{\frown}{{\mu}}){\nabla}({\nabla}\bullet \mathbf{u})+\mathbf{f}_{\mathbf{i}}^{\mathbf{ms}}={\rho}{\displaystyle \frac{\partial ^{2}\mathbf{u}}{\partial t^{2}}} & & \displaystyle \end{eqnarray}$ (3) The biased magnetic field with correlated vectors on the inner surface of tube induced by the electromagnetic is shown in Fig. 3 The length of the electromagnet is 52 mm, with 11 mm of the radius. Both radial and axial component distribution along axial position of the biased magnetic field is extracted as shown in Fig. 4. The induced magnetic flux near middle of the electromagnet is in axial direction, which is conductive to exciting and receiving L(0,2) mode guided wave. The peak value of the axial component is larger than 0.25 T, which is larger than magnetic flux density of permanent magnet with the same size. The signal-to-noise ratio of EMAT is proportional to square of magnetic flux density, so the signal-to-noise ratio of the proposed EMAT is enhanced compared with conventional EMAT [9].

The EMAT with 4 meanders is placed at 125 mm from the tube edge. The vibration velocity of guided-wave signal in the induced area for intact tube is extracted as shown in Fig. 4(b). All the echoes are reflected by the edge of the tube. By measuring the group velocity of the wave pack, these echoes can be distinguished as L(0,2) mode guided wave.

Fig. 3.

Simulation results: biased magnetic field.

3. Experimental validation

3.1. Performance of the transducer

The proposed guided-wave EMAT used in experiment is shown in Fig. 5. The electromagnet is wrapped in gap of the bobbin with enameled copper wire as 0.5 mm of the diameter. The meander coil is wrapped outside the electromagnet with enameled copper wire as 0.1 mm of the diameter. Carbon steel tubes with 28 mm of outer diameter, 2 mm of thickness and 1 m of length are used as specimens in experiment. The transducer is placed at 125 mm from the edge of the intact tubes. Both the received signal and driven long-pulse current are acquired as shown in Fig. 6. All the echoes are signals of guided wave reflected from the tube edges. Both the L(0,1) and L(0,2) mode guided wave echoes are received, while the amplitude of the L(0,2) mode is much larger than the L(0,1) mode. The echoes number ① and ② are echoes from the near and far edges respectively; number ③ is overlapped by echoes in the opposite propagation directions.

Fig. 4.

Simulation results of the proposed EMAT.

Fig. 5.

The proposed transducer.

Fig. 6.

Received guided wave signal.

To maximize the signal-to-noise ratio of received signal using the proposed EMAT, parameters in experiment are discussed in performance testing. The factors which can mainly affect the signal amplitude are considered as variable parameters in the experiment of performance testing. These parameters are: driving voltage of the electromagnet, time delay between the long-pulse current and the RF pulses, excitation frequency and burst duration of the RF pulses. The amplitudes of the number 3 echo in Fig. 6 are measured among different parameter settings in performance testing as shown in Fig. 5.

The magnitude reaches a maximum value while the excitation frequency is set as 0.22 MHz, which is approximate to 0.21 MHz as expectation. The rising tendency of magnitude decreases while the driving voltage is larger than 250 V due to the nonlinearity of line magnetostriction ratio 𝜆_s and B-H curve of carbon steel. The coefficient effect of magnetostriction and magnetization has an optimum with a proper voltage. The delay of RF pulse can affect the magnitude due to the nonlinear performance during magnetization with rectangular pulse. According to the result of performance testing, 0.22 MHz of frequency, 276 V of driving voltage, 3 cycles of burst and 100 μs of delay are determined in the following experiments.

Fig. 7.

Results of the performance testing.

Fig. 8.

Sketch diagram of the artificial wall thinning.

Fig. 9.

Time domain signals of the proposed EMAT.

Fig. 10.

Sketch diagram and results of the thinning defects.

3.2. Defects inspection

Artificial wall thinning defects are fabricated on the outer surface of tubes to imitate corrosion. Circumferential thinning defects with different depth and location, local thinning defects with different proportion and location are fabricated, for each tube with one defect, as shown in Fig. 8. The proposed EMAT is placed near the edge of the tubes. Figure 9 shows typical time domain signals of guided wave for intact tube and tube with number 2 of the local thinning defect. It can be seen that the echoes from defect appear while the magnitudes of echoes from edges decrease.

The size of defects can be quantified by measuring magnitude of the echoes from defects. The magnitude of first echo is plotted with the size of defects, as shown in Fig. 10. It reveals an approximately linear correlation between the sectional area of defects and the magnitude of the echoes. The results validate the feasibility on quantification evaluation of defects with the proposed EMAT.

4. Conclusions

A new compact inserted guided-wave EMAT with pulsed electromagnetic is proposed in form of a coil-only design. The proposed EMAT consists of an electromagnet driven by a rectangular long-pulse current and a meander coil driven by RF pulses. It has the advantages such as compact, noncontact and high SNR. The inspection process can be accomplished by placing the transducer near the edge of the tube without pushing into the tubes. The feasibility of the proposed EMATs for L(0,2) mode guided wave generating and receiving was validated by both numerical simulation and experiment. The results of performance testing based on measurement of the echo magnitudes determined the optimized parameters for the proposed EMAT. The results of defect inspection show a linear correlation between echo magnitude and defect area. And such results demonstrate the feasibility on quantitative evaluation of defects using the proposed EMAT.

Footnotes

Acknowledgements

The authors would like to thank the National Natural Science Foundation (No. 11927801) and the State Key Lab of Digital Manufacturing Equipment & Technology (No. DMETKF2018012) for funding.

References

Shim

H.S.

Choi

M.S.

Lee

D.H.

and Hur

D.H.

, A prediction method for the general corrosion behavior of Alloy 690 steam generator tube using eddy current testing, Nuclear Engineering and Design297 (2016), 26–31.

Gori

Giamboni

DAlessio

Ghia

Cernuschi

and Piana

G.M.

, Guided waves by EMAT transducers for rapid defect location on heat exchanger and boiler tubes, Ultrasonics34(2–5) (1996), 311–314.

Nurmalia, N. Nakamura, H. Ogi and M. Hirao, EMAT pipe inspection technique using higher mode torsional guided wave T(0,2), NDT & E International87 (2017), 78–84.

Ashigwuike

E.C.

Ushie

O.J.

Mackay

and Balachandran

, A study of the transduction mechanisms of electromagnetic acoustic transducers (EMATs) on pipe steel materials, Sensors and Actuators A: Physical229 (2015), 154–165.

Hao

K.S.

Huang

S.L.

Zhao

Wei

Wang

and Huang

Z.J.

, A new frequency-tuned longitudinal wave transducer for nondestructive inspection of pipes based on magnetostrictive effect, Sensors Applications Symposium IEEE (2010), 64–68.

Pei

C.X.

Xiao

Zhao

S.Q.

and Chen

Z.M.

, Development of a flexible film electromagnetic acoustic transducer for nondestructive testing, Sensors and Actuators A: Physical258 (2017), 68–73.

Xie

S.J.

Chen

Z.M.

Takagi

and Uchimoto

, Efficient numerical solver for simulation of pulsed eddy-current testing signals, IEEE Transactions on Magnetics47(11) (2011), 4582–4591.

Wang

Y.M.

Kang

Y.H.

and Wu

X.J.

, Theoretical and experimental study on generation of longitudinal guided waves in circular pipes based on magnetostrictive effect, Chinese Journal of Mechanical Engineering41(10) (2005), 174–179.

Wang

S.J.

Xin

P.H.

Lei

and Zhai

G.F.

, Research on influence of Lorentz force mechanism on EMAT’s transduction efficiency in steel plate, Industrial Electronics & Applications (2010).