sensors Article
Mode Conversion Behavior of Guided Wave in a Pipe Inspection System Based on a Long Waveguide Feiran Sun 1 , Zhenguo Sun 1,2, *, Qiang Chen 1,2 , Riichi Murayama 3 and Hideo Nishino 4 1 2 3 4
*
Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China;
[email protected] (F.S.);
[email protected] (Q.C.) Yangtze Delta Region Institute of Tsinghua University, Jiaxing 314006, China Department of Intelligent Mechanical Engineering, Fukuoka Institute of Technology, Fukuoka 811-0295, Japan;
[email protected] Graduate School of Advanced Technology and Science, Tokushima University, Minami-Josanjima, Tokushima 770-8506, Japan;
[email protected] Correspondence:
[email protected]; Tel.: +86-10-6277-3860
Academic Editors: Dipen N. Sinha and Cristian Pantea Received: 22 August 2016; Accepted: 27 September 2016; Published: 19 October 2016
Abstract: To make clear the mode conversion behavior of S0-mode lamb wave and SH0-plate wave converting to the longitudinal mode guided wave and torsional mode guided wave in a pipe, respectively, the experiments were performed based on a previous built pipe inspection system. The pipe was wound with an L-shaped plate or a T-shaped plate as the waveguide, and the S0-wave and SH0-wave were excited separately in the waveguide. To carry out the objective, a meander-line coil electromagnetic acoustic transducer (EMAT) for S0-wave and a periodic permanent magnet (PPM) EMAT for SH0-wave were developed and optimized. Then, several comparison experiments were conducted to compare the efficiency of mode conversion. Experimental results showed that the T(0,1) mode, L(0,1) mode, and L(0,2) mode guided waves can be successfully detected when converted from the S0-wave or SH0-wave with different shaped waveguides. It can also be inferred that the S0-wave has a better ability to convert to the T(0,1) mode, while the SH0-wave is easier to convert to the L(0,1) mode and L(0,2) mode, and the L-shaped waveguide has a better efficiency than T-shaped waveguide. Keywords: mode conversion; waveguide; guided wave; EMAT; pipe inspection
1. Introduction Online monitoring of pipes at high temperature for wall thickness and defects is a key to many industrial non-destructive evaluation (NDE) demands, such as the power generation and the nuclear industry. It is quite cost effective to avoid plant shutdowns during monitoring of its structures; whereas “high-temperature monitoring” allows us to conduct the plant structure monitoring without shutting down the plant [1]. In recent years, the ultrasonic guided wave testing method has been widely applied for pipe inspections as it carries major advantages such as low attenuation, long distance propagation, and high detection efficiency [2–6]. There are two techniques which are commonly employed for generating guided waves: the piezoelectric transducer and electromagnetic acoustic transducer (EMAT) [2]. However, the piezoelectric ultrasonic testing requires a good sonic contact with the test piece, which restricts its inspection ability for certain applications, especially under high-temperature conditions [7,8]. EMAT has the ability to generate and detect ultrasonic waves without making a contact with the test object which makes it suitable to inspect high-temperature pipes with a large
Sensors 2016, 16, 1737; doi:10.3390/s16101737
www.mdpi.com/journal/sensors
Sensors 2016, 16, 1737
2 of 14
Sensors 2016, 16, 1737
2 of 14
lift-off. Even though the EMAT still requires a cooling system which causes many issues [9–11], in addition, the lift-off effect also influences the results [12–15]. Sensors 2016, 16, 1737 2 of 14 Therefore, a sensor system for high-temperature pipe inspection, which can transmit and receive Therefore, a sensor system for high-temperature pipe inspection, which can transmit and receive ultrasonic guideda sensor waves using afor long waveguide, has been developed by Murayama [16]. As shown Therefore, high-temperature can transmit ultrasonic guided wavessystem using a long waveguide, haspipe beeninspection, developedwhich by Murayama [16].and As receive shown inin Figure 1, the tested pipe was wound with a long plate as the waveguide, which could keep the ultrasonic guided wavespipe using a long waveguide, has been by Murayama [16].could As shown Figure 1, the tested was wound with a long platedeveloped as the waveguide, which keep ultrasonic sensor awayaway from the high-temperature The S0-mode lamb wave was excited in the in Figure 1, thesensor tested pipe from was wound with a longpipe. plate as the waveguide, which could keep the the ultrasonic the high-temperature pipe. The S0-mode lamb wave was excited plate-waveguide as the source, and it has been proved that the S0-wave can convert to ultrasonic sensor away from the high-temperature pipe.proved The S0-mode wavecan wasconvert excited to in the thethe in the plate-waveguide as the source, and it has been that thelamb S0-wave longitudinal mode guided wave the pipe pipe with theproved T-shaped waveguide and to torsional mode plate-waveguide the source, it has been that the S0-wave convert to the longitudinal modeas guided wave in inand the with the T-shaped waveguide and tocan thethe torsional mode guided wave with the L-shaped waveguide [16], as shown in Figure 2. However, the longitudinal longitudinal guided wave in the pipe with and to thethe torsional mode guided wavemode with the L-shaped waveguide [16],the as T-shaped shown in waveguide Figure 2. However, longitudinal and torsional waves have different different detection abilities for different types of defects guided wavemode with guided the L-shaped [16],detection as shownabilities in Figure 2. However, theoflongitudinal and torsional mode guided waveswaveguide have for different types defects duedue toto their different oscillation patterns, thus both of them are always employed for a comprehensive and torsional mode guided waves have different differentfor types of defects due their different oscillation thus both ofdetection them areabilities alwaysfor employed a comprehensive to their different oscillation patterns, in thus both of them are always employed a comprehensive pipe evaluation [6,16–18]. Therefore, system, ififthe ofoffor testing wave pipe evaluation [6,16–18]. Therefore, inthis this system, thetransformation transformation testing wavefrom fromthe pipe evaluation [6,16–18]. Therefore, in this system, if the transformation of testing wave from the the longitudinal mode to the torsional mode is required, the waveguide will be converted from longitudinal mode to the torsional mode is required, the waveguide will be converted from the T-shaped longitudinal mode the mode is required, thewhich waveguide will converted from the T-shaped tofixed thetoL-shaped fixed on the high-temperature pipe, which cannot be easily executed tothe theT-shaped L-shaped on torsional the high-temperature pipe, cannot bebeeasily executed in practice. to practice. the L-shaped fixed on the high-temperature pipe, which cannot be easily executed in practice. in
Figure systemusing usingaaalong longwaveguide waveguide [16]. Figure 1. Pipe inspection [16]. Figure 1. 1. Pipe Pipe inspection inspection system system using long waveguide [16].
(a) (a)
(b)(b)
Figure 2. Pipeinspection inspection system with with different shaped waveguides based on S0-mode lamb wave (a) Figure Figure2.2.Pipe Pipe inspectionsystem system withdifferent differentshaped shapedwaveguides waveguides based based on on S0-mode S0-mode lamb lamb wave wave (a) longitudinal mode with T-shaped waveguide; (b) torsional mode with L-shaped waveguide [16]. longitudinal mode with T-shaped waveguide; (b)(b) torsional [16]. (a) longitudinal mode with T-shaped waveguide; torsionalmode modewith withL-shaped L-shaped waveguide waveguide [16].
As a consequence, based on the already established pipe inspection system, a method to change As a consequence, based on the already established pipe inspection system, a method to change As a consequence, basedin onthe thepipe already established inspection a methodEMAT to change the mode of the testing wave by replacing thepipe S0-wave EMATsystem, with SH0-wave was the mode of the testing wave in the pipe by replacing the S0-wave EMAT with SH0-wave EMAT the mode of wave in the pipe by easier replacing S0-wavethe EMAT SH0-wave waswas proposed in the thistesting research, which was much thanthe removing fixedwith waveguide. AnEMAT SH0-wave proposed ininthis which was much mucheasier easierthan thanremoving removingthe the fixed waveguide. SH0-wave proposed thisresearch, research, which was fixed waveguide. AnAn SH0-wave periodic permanent magnet (PPM) EMAT and an S0-wave meander-line coil EMAT were developed, periodic permanent magnet (PPM) andan anS0-wave S0-wave meander-linecoil coil EMAT were developed, periodic (PPM) EMAT EMAT and EMAT were developed, then the permanent experimentsmagnet were conducted involving the mode meander-line conversion from SH0 mode and S0 mode then the experiments were conducted involving the mode conversion from SH0 mode and S0 mode then thewaves experiments were conducted mode conversion SH0 mode S0 mode guided to the longitudinal andinvolving torsional the mode guided waves, from respectively. In and addition, the guided waves to the longitudinal and torsional mode guided waves, respectively. In addition, guided waves to the longitudinal and torsional mode guided waves, respectively. In addition, the mode conversion behavior for each mode of guided wave was studied by calculating and comparingthe mode conversion behavior for mode of guided wavewas wasstudied studied calculating and comparing mode conversion behavior for each mode guided byby calculating and comparing the mode conversion efficiency, from SH0of and S0 to wave longitudinal and torsional, respectively. the fromSH0 SH0and andS0S0totolongitudinal longitudinal and torsional, respectively. themode modeconversion conversionefficiency, efficiency, from and torsional, respectively. 2. Method
2. Method In order to test a pipe with longitudinal and torsional mode guided waves, based on the previous In pipe orderinspection to test a pipe with longitudinal torsional modeofguided waves, based previous built system, a method to and change the mode the testing wave in on thethe pipe by built pipe inspection system, a method to change the mode of the testing was wave in the pipe changing the EMAT on the waveguide without removing the fixed waveguide proposed. That by is, the T-shaped waveguide and L-shaped waveguide would be switched with each other, and changing the EMAT on the waveguide without removing thenot fixed waveguide was proposed. That EMAT needed and to beL-shaped changed waveguide to the other would mode guided sensor to each complete theand is,the theS0-mode T-shaped waveguide not be wave switched with other, switch between the longitudinal and torsional mode, which was much to implement the S0-mode EMAT needed to bemode changed to the other mode guided waveeasier sensor to completeinthe
Sensors 2016, 16, 1737
3 of 14
Sensors 2016, 16, 1737 an industrial NDE.
3 of 14 Thus, the key was to find another mode guided wave in the plate-waveguide which could separately convert to the longitudinal and torsional mode guided waves using the opposite waveguide of the S0-mode. 2. Method
In order to test a pipe 2.1. Guided Wave into a Pipewith longitudinal and torsional mode guided waves, based on the previous built pipe inspection system, a method to change the mode of the testing wave in the pipe by changing The axially symmetric L(0,2) and T(0,1) modes are the most widely-used modes for pipelines [19]. the EMAT on the waveguide without removing the fixed waveguide was proposed. That is, the The oscillation patterns of the L(0,2) and T(0,1) mode guided waves are shown in Figure 3, in which T-shaped waveguide and L-shaped waveguide would not be switched with each other, and the the oscillation direction of L(0,2) is parallel to its propagation direction, but for T(0,1) mode, Sensors 2016, 16, 1737 3 ofthe 14 S0-mode EMAT needed to be changed to the other mode guided wave sensor to complete the switch oscillation direction is perpendicular to the propagation [16]. between the longitudinal mode and torsional mode, which was much easier to implement in an an industrial NDE. Thus, the key was to find another mode guided wave in the plate-waveguide industrial NDE. Thus, the key was to find another mode guided wave in the plate-waveguide which which could separately convert to the longitudinal and torsional mode guided waves using the could separately convert to the longitudinal and torsional mode guided waves using the opposite opposite waveguide of the S0-mode. waveguide of the S0-mode.
2.1. 2.1. Guided Guided Wave Wave into into aa Pipe Pipe The The axially axially symmetric symmetric L(0,2) L(0,2)and andT(0,1) T(0,1)modes modesare arethe themost mostwidely-used widely-usedmodes modesfor forpipelines pipelines[19]. [19]. The oscillation patterns of the L(0,2) and T(0,1) mode guided waves are shown in Figure 3, which in which The oscillation patterns of the L(0,2) and T(0,1) mode guided waves are shown in Figure 3, in the the oscillation direction of L(0,2) is parallel to its propagation direction, but for T(0,1) the oscillation direction of L(0,2) is parallel to its propagation direction, but for T(0,1) mode, themode, oscillation oscillation is perpendicular to the propagation [16]. direction isdirection perpendicular to the propagation [16].
Figure 3. Oscillation patterns of the L(0,2) and T(0,1) mode guided waves in pipes [16].
Then, the dispersion curves of the pipe and the plate-waveguide were built using a numerical method with MATLAB based on an analysis of the dispersion equations [20,21], as shown in Figures 4 and 5. The relevant parameters of the pipe and waveguide can be obtained from Table 1. Table 1. The parameters of the pipe and waveguide.
Parameter
Pipe
Waveguide (Plate)
Inner radius (width of plate)/mm 18.195 10 Figure 3. Oscillation patterns of the (a) L(0,2) and (b) T(0,1) mode guided waves in pipes Figure 3. Oscillation patterns of the L(0,2) and T(0,1) mode guided waves in pipes [16].[16]. Thickness/mm 3.22 0.6 Velocity of shear wave/(m/s) 3200 3200 Then, the curves of the pipe and the plate-waveguide were built using Then,Velocity the dispersion dispersion curves of the pipe and the plate-waveguide were built using aa numerical numerical of longitudinal wave/(m/s) 5940 equations [20,21],5200 method with MATLAB based on an analysis of the dispersion as shown in Figures in 4 method with MATLAB based on an analysis of the dispersion equations [20,21], as shown and 5. The relevant parameters of the pipe and waveguide can be obtained from Table 1. Figures 4 and 5. The relevant parameters of the pipe and waveguide can be obtained from Table 1. Table 1. The parameters of the pipe and waveguide.
Parameter
Pipe
Waveguide (Plate)
Inner radius (width of plate)/mm Thickness/mm Velocity of shear wave/(m/s) Velocity of longitudinal wave/(m/s)
18.195 3.22 3200 5940
10 0.6 3200 5200
(a)
(b)
Figure 4.4.TheThe dispersion curvescurves of group the pipe:in (a)the longitudinal mode and (b) torsional Figure dispersion ofvelocity group invelocity pipe: (a) longitudinal modemode. and (b) torsional mode.
Sensors 2016, 16, 1737 Sensors 2016, 16, 1737
4 of 14 4 of 14
(a)
(b)
Figure 5. 5. The The dispersion dispersion curves curves of of group group velocity velocity in in the the waveguide: waveguide: (a) lamb wave wave and and (b) (b) shear shear Figure (a) lamb horizontaL(SH) wave. wave. horizontaL(SH)
If the T(0,1) mode is one of the objective guided waves, the exciting frequency should be below Table 1. The parameters of the pipe and waveguide. the cut-off frequency of 500 kHz to eliminate the other interference torsional modes in Figure 4b. Based on Figure 4a, the L(0,1) mode guided wave may exist the test even if we try to avoid it Parameter Pipe during Waveguide (Plate) through the design of the EMAT, for it has the same frequency range as the L(0,2) mode. However, Inner radius (width of plate)/mm 18.195 10 at the frequency of around 100 kHz, the group velocity of L(0,2) is much higher Thickness/mm 3.22 0.6 than that of L(0,1) so that we can remove the L(0,1) partwave/(m/s) by reception signal selection [22]. Therefore, Velocity of shear 3200 3200 the exciting frequency Velocity longitudinal wave/(m/s) 5940could be computed 5200 as 50 mm and 32 mm, was chosen as 100 kHzofand the wavelengths of S0 and SH0 respectively, from Figure 5, which can help in the design of the EMAT. If the T(0,1) mode is one of the objective guided waves, the exciting frequency should be below 2.2. Method and Assumption the cut-off frequency of 500 kHz to eliminate the other interference torsional modes in Figure 4b. BasedBased on Figure 4a, the L(0,1) mode guided wave exist during theremains test evenunchanged if we try towhen avoidthe it on Figure 2a,b, it can be noted that the may oscillation direction through design of the EMAT, forthe it has the sameto frequency as the L(0,2) mode.waveguide However, at S0-modethe guided wave travels from waveguide the pipe range no matter which shaped is the frequency of around 100 kHz, the group velocity of L(0,2) is much higher than that of L(0,1) so used. The reason for generation of different modes of guided waves in the pipe is the different that we can remove by reception selection that [22]. is Therefore, the exciting connection patternstheofL(0,1) the part different shapedsignal waveguides, the T-shaped and frequency L-shaped was chosen asThe 100 kHz and the wavelengths of S0 andtoSH0 be computed as that 50 mm 32 mm, waveguides. T-shaped waveguide is parallel thecould axis of the pipe so theand consistent respectively, from Figure 5, which can help in the design of the EMAT. oscillation direction is parallel to the axis and the propagation direction is always along the axis, thus
forming the L(0,2) mode. This is the same for the L-shaped waveguide and T(0,1) mode. 2.2. Method and Assumption Consequently, based on this phenomenon, it can be assumed that a guided wave whose oscillation direction perpendicular to its in adirection plate can convert to the T(0,1) mode Based on Figure is 2a,b, it can be noted thatpropagation the oscillation remains unchanged when the guided wave using the travels T-shaped waveguide or L(0,2) using the L-shaped That is, S0-mode guided wave from the waveguide to mode the pipe no matter which waveguide. shaped waveguide theused. shearThe horizontaL(SH) wave is aofpolarized wave in-plane with respect is reason for generation different guided modes of guided waves in the pipeto is the the reference different interface [23]. Thereofare and anti-symmetric SH waveand modes. Thiswaveguides. study used connection patterns themultiple different symmetric shaped waveguides, that is the T-shaped L-shaped onlyT-shaped the lowest order SH is mode, SH0, is of a symmetric modeoscillation [24] just like the S0 The waveguide parallel towhich the axis the pipe sodispersionless that the consistent direction mode. Based on this experiments were performed involving the mode conversions from is parallel to the axisassumption, and the propagation direction is always along the axis, thus forming the L(0,2) S0-mode and T-mode and waveguide L-mode, respectively, shown in Figure 6. mode. This is SH0-mode the same fortothe L-shaped and T(0,1) as mode. Consequently, based on this phenomenon, it can be assumed that a guided wave whose oscillation direction is perpendicular to its propagation in a plate can convert to the T(0,1) mode guided wave using the T-shaped waveguide or L(0,2) mode using the L-shaped waveguide. That is, the shear horizontaL(SH) wave is a polarized guided wave in-plane with respect to the reference interface [23]. There are multiple symmetric and anti-symmetric SH wave modes. This study used only the lowest order SH mode, SH0, which is a symmetric dispersionless mode [24] just like the S0 mode. Based on this assumption, experiments were performed involving the mode conversions from S0-mode and SH0-mode to T-mode and L-mode, respectively, as shown in Figure 6.
2016, 16, 1737 1737 Sensors 2016, Sensors 2016, 16, 1737
5 of 14 5 of 14
Figure 6. The mode conversion experiments to prove the assumption. Figure6.6.The Themode modeconversion conversionexperiments experimentstotoprove provethe theassumption. assumption. Figure
3. Experimental System 3.3.Experimental ExperimentalSystem System As shown in Figure 7, the experimental system consists of a function generator and a power As Asshown shownin inFigure Figure7,7,the theexperimental experimentalsystem systemconsists consistsofofaafunction functiongenerator generatorand andaapower power amplifier, which generates the 100 kHz frequency and six volt amplitude electric power with four amplifier, amplifier,which whichgenerates generatesthe the100 100kHz kHzfrequency frequencyand andsix sixvolt voltamplitude amplitudeelectric electricpower powerwith withfour four cycles of burst shape, and is amplified 20 times, then input to an electromagnetic induced coil, a precycles of burst shape, and is amplified 20 times, then input to an electromagnetic induced coil, a precycles of burst shape, and is amplified 20 times, then input to an electromagnetic induced coil, a amplifier and a frequency filter, which amplifies the received signal by 50 dB, and selects the signal amplifier and and a frequency filter, which amplifies thethe received signal byby 5050 dB, pre-amplifier a frequency filter, which amplifies received signal dB,and andselects selectsthe thesignal signal between 50 kHz and 200 kHz, and an oscilloscope and central processing unit (CPU), which collects between 50 kHz and 200 kHz, and an oscilloscope and central processing unit (CPU), which collects between 50 kHz and 200 kHz, and an oscilloscope and central processing unit (CPU), which collects and evaluates the received signal. The two waveguides of 1 m long, 10 mm wide and 0.6 mm thick and andevaluates evaluatesthe thereceived receivedsignal. signal.The Thetwo twowaveguides waveguidesofof11mmlong, long,10 10mm mmwide wideand and0.6 0.6mm mmthick thick plate were separately connected around the pipe, and the transmit EMAT and receive EMAT were plate platewere wereseparately separatelyconnected connectedaround aroundthe thepipe, pipe,and andthe thetransmit transmitEMAT EMATand andreceive receiveEMAT EMATwere were placed on the terminal of the waveguides. Figure 8 shows the part of the experimental system which placed on the terminal of the waveguides. Figure 8 shows the part of the experimental system which placed on the terminal of the waveguides. Figure 8 shows the part of the experimental system which consists of the EMAT and the waveguide winding around the test pipe. consists consistsof ofthe theEMAT EMATand andthe thewaveguide waveguide winding winding around around the the test test pipe. pipe.
Figure Block diagram diagram of Figure 7. 7. Block of experimental experimental system. system. Figure 7. Block diagram of experimental system.
Sensors 2016, 16, 1737
6 of 14
Sensors 2016, 16, 1737
6 of 14
Sensors 2016, 16, 1737
6 of 14
Sensors 2016, 16, 1737
6 of 14
Figure 8. The photo of the pipe inspection system. Figure 8. The photo of the pipe inspection system.
4. Development and Optimization of EMATs Figure 8. The photo of the pipe inspection system. 4. Development and Optimization of EMATs Figure 8. The photo of the pipe inspection system.
4.1. for S0 Mode Wave 4. Meander-Line Development Coil and EMAT Optimization of EMATs 4.1. Meander-Line Coil EMAT for S0 Mode Wave
4. Development and Optimization of EMATs
meander-line coil EMAT been widely used to excite lamb in a plate-waveguide [25]. The meander-line coil has been widely used to excite a lambawave in wave a plate-waveguide [25]. 4.1. The Meander-Line Coil EMAT forEMAT S0has Mode Wave In this study, Coil permanent magnets (PM)were were used provide a static field. The PM was The 10 PM was 10 In this4.1. study, permanent magnets (PM) usedtoto provide a magnetic static magnetic field. Meander-Line EMAT for S0 Mode Wave The meander-line coil EMAT has been widely used and to excite a lambfield wave in a at plate-waveguide mm in10 length, 10 mm in width and 4 mm in thickness, the magnetic density the surfaceat the [25]. mm in length, mm in width and 4 mm in thickness, and the magnetic field density surface The coil EMAT has been widely used toprovide excite a lamb wave inwire a plate-waveguide wasmeander-line 326.4 mT. As shown in Figure 9, the meander-line was made of copper withfield. a diameter In this study, permanent magnets (PM) were used tocoil a static magnetic The [25]. PM was was 326.4 mT. Asmm, shown in magnets Figure 9, thethe meander-line coilequal was made copper wire with 10 a diameter In this permanent (PM) were to provide static magnetic field. The ofstudy, 0.315 and in the pitch between meander lines was to half theofwavelength of25PM mm 10 mm in length, 10 mm width and 4 mm in used thickness, and athe magnetic field density at was the surface of 0.315 and 10 the pitch between lines equal tofield halfdensity the wavelength mmmm, inaccording length, mm in width and 4the mmmeander in thickness, andwas the magnetic at the surfaceof25 mm to the drive principle. was 326.4 mT. As shown in Figure 9, the meander-line coil was made of copper wire with a diameter was 326.4 mT. As shown in Figure 9, the meander-line coil was made of copper wire with a diameter according to the drive principle. of 0.315 mm, and the pitch between the meander lines was equal to half the wavelength of 25 mm of 0.315 mm, and the pitch between the meander lines was equal to half the wavelength of25 mm according to the principle. according to drive the drive principle.
Figure 9. The structure of meander-line coil EMAT for S0-mode guided wave.
In order to excite a pure S0-mode lamb wave in the waveguide, the meander-line coil EMAT was optimized. First of all, to produce the best static magnetic field, the configuration of the permanent magnets above the coil was tested in the free waveguide whose dimensions were presented 9. The structure of meander-line meander-line coil EMAT for S0-mode guided wave.side in Table 1.Figure The transmit EMAT of and receive EMAT were separately placed onguided each of the Figure coil EMAT S0-mode wave. Figure 9. 9. The The structure structure of meander-line coil EMAT forfor S0-mode guided wave. waveguide, and there were three reasonable configurations of the PMs to be tested as shown in Figure 10.order The densities static magneticlamb field wave were measured in these three a In to exciteofathe pure S0-mode in the waveguide, theconfigurations meander-lineusing coil EMAT In order tometer, excite ashown pureinS0-mode lamb wave in the waveguide, the meander-line coil EMAT gauss also Figure 10. In order to excite a pure S0-mode lamb wave in the waveguide, the meander-line coil was optimized. First of all, to produce the best static magnetic field, the configuration EMAT of the was
was optimized. First of all,the tocoil produce the best static magnetic field, the configuration of the permanent above wasbest tested in the free waveguide whose dimensions were presented optimized. Firstmagnets of all, to produce the static magnetic field, the configuration of the permanent permanent magnets above coilin was inEMAT the free waveguide whose dimensions presented in Table 1. the Thecoil transmit EMAT and receive were separately placedwere on each sidewere ofinthe magnets above wasthe tested thetested free waveguide whose dimensions presented Table 1. inThe Table 1. The transmit EMAT and receive EMAT were separately placed on each side of waveguide, and there were three reasonable configurations of the PMs to be tested as shown in Figure transmit EMAT and receive EMAT were separately placed on each side of the waveguide, and the Theand densities of the static fieldconfigurations in the these three configurations using ain Figure waveguide, there were threemagnetic reasonable of to in beFigure tested10. as shown there10. were three reasonable configurations of thewere PMsmeasured to be tested as PMs shown The densities gauss meter, also shown in Figure 10. 10. densities of the static magnetic field were three measured in these three configurations using a of The the static magnetic field were measured in these configurations using a gauss meter, also (a) (b) (c) gauss meter, also shown in Figure 10. shown in Figure 10. Figure 10. Three configurations of permanent magnets (PMs) of EMAT: (a) two magnets; (b) four magnets; and (c) eight magnets.
(a)
(b)
(c)
Figure 10. Three configurations of permanent magnets (PMs) of EMAT: (a) two magnets; (b) four magnets; (a)and (c) eight magnets. (b) (c)
Figure of permanent permanentmagnets magnets(PMs) (PMs)ofofEMAT: EMAT: magnets; Figure 10. 10. Three Three configurations configurations of (a)(a) twotwo magnets; (b) (b) fourfour magnets; and (c) eight magnets. magnets; and (c) eight magnets.
Sensors 2016, 16, 1737
7 of 14
Sensors 2016, 16, 1737
7 of 14
The amplitude and signal to noise rate (SNR) value of the S0-wave were computed from the The signal, amplitude signal noise rate (SNR) of the to S0-wave were fromasthe received then,and these threetoconfigurations werevalue compared determine thecomputed best one, just in received signal, then, these three configurations were compared to determine the best one, just as in Figure 11. It indicated that the amplitude and SNR could reach the highest values when the 4-PM Figure 11. It indicated that the amplitude and SNR could reach the highest values when the 4-PM configuration was selected from these three candidates. configuration was selected from these three candidates.
Figure 11. The The amplitude amplitude and and signal signal to to noise noise rate rate (SNR) (SNR) value value of of S0 S0wave wavewith withdifferent different configurations configurations of PMs shown in Figure 10.
Next, the the number of turns coil was was optimized. Next, number of turns of of the the meander-line meander-line coil optimized. The The test test was was conducted conducted while while changing the number of turns from four to eleven, and the capacitance, inductance, resistance, and and changing the number of turns from four to eleven, and the capacitance, inductance, resistance, impedance of each coil coil were were measured measured using using an an LCR LCR (inductance, (inductance, capacitance, capacitance, resistance) resistance) meter meter as as impedance of each shown in Table 2. Figure 12 shows the relationship of how the amplitude and SNR of the S0 wave shown in Table 2. Figure 12 shows the relationship of how the amplitude and SNR of the S0 wave varied while while the the number number of of turns turns increased. increased. It amplitude increased increased with with the the varied It also also shows shows that that the the amplitude increasing number numberof ofturns, turns,and andthe theSNR SNR obtained maximum value when number of turns increasing obtained thethe maximum value when the the number of turns was was equal to ten. equal to ten. Table 2. The inductance, resistance, resistance, and and impedance impedance of of the the meander-line meander-line coil. coil. Table 2. The capacitance, capacitance, inductance,
Number of Turns
Capacitance/ Capacitance/nF nF
Inductance/ Inductance/µH H
Resistance/ Resistance/mΩ m
9 10 11
822.77 652.10 462.17 652.10 371.51 462.17 280.44 371.51 238.35 200.12 280.44 176.87 238.35 200.12 176.87
2.7921 3.5233 4.9716 3.5233 6.1839 4.9716 8.1950 6.1839 9.6395 11.481 8.1950 12.989 9.6395 11.481 12.989
289.45 318.75 386.84 318.75 439.51 386.84 506.31 439.51 533.51 626.32 506.31 656.48 533.51 626.32 656.48
Number of Turns 4 54 65 7 6 8 97 108 11
822.77
2.7921
289.45
Impedance/
Impedance/Ω 1.7815 1.7815 2.3468 3.3027 2.3468 4.1030 3.3027 5.4305 4.1030 6.3814 7.6005 5.4305 8.5942
6.3814 7.6005 8.5942
Sensors 2016, 16, 1737 Sensors 2016, 2016, 16, 16, 1737 1737 Sensors
8 of 14 of 14 14 88 of
Figure 12. and SNR SNR of of S0 12. The The amplitude amplitude and S0 wave wave with with aa different different number number of of turns turns of of meander-line meander-line coil. coil. Figure
As aaaresult, result, an EMAT with 10-turn meander-line coil and 4-PM configuration configuration has been been As result, EMAT aa 10-turn meander-line and 4-PM has As anan EMAT withwith a 10-turn meander-line coil andcoil 4-PM configuration has been developed developed asFigure shown in Figure 9,the which has theconditions best drive driveaccording conditionstoaccording according toand Figures 11 and and 12. 12. developed shown Figure 9, which the best conditions to Figures as shown inas 9,in which has besthas drive Figures 11 12. 11 4.2. PPM PPM EMAT EMAT for for SH0-Mode SH0-Mode Wave Wave 4.2. To generate generate an an SH0-mode SH0-mode guided guided wave, wave, the the periodic periodic permanent permanent magnet magnet (PPM) EMAT [26–28] [26–28] To (PPM) EMAT was employed employed in in this study. The PPM EMAT consisted of aaof racetrack-shaped coil with with magnet array PPM EMAT consisted of racetrack-shaped coil aa magnet array was inthis thisstudy. study.The The PPM EMAT consisted a racetrack-shaped coil with a magnet on top topon (two columns and several several rows rows of alternating alternating polarity magnets) as shown shown in Figure Figure 13. The The on (two and rows of polarity magnets) as in 13. array top columns (two columns and several of alternating polarity magnets) as shown in Figure 13. racetrack coilcoil hadhad mm width with eight turns ofof 0.2 mm diameter wire. The permanent racetrack coil had aa 55 amm width with eight turns of 0.2 The racetrack 5 mm width with eight turns 0.2mm mmdiameter diameterwire. wire.The Thepermanent permanent magnet magnet was as as long long as as half half the wavelength ofofthe the SH0 wave, and thethe magnetic field density at the the surface was was wave, and the magnetic field density at was half the thewavelength wavelengthof theSH0 SH0 wave, and magnetic field density at surface the surface 281.1281.1 mT. mT. 281.1 mT. was
(a) (a)
(b) (b) Figure 13. 13. (a) The The layout of of PPM EMAT; EMAT; (b) The The outlook of of PPM EMAT EMAT for for SH0 SH0 wave. wave. Figure Figure 13. (a) (a) The layout layout of PPM PPM EMAT; (b) (b) The outlook outlook of PPM PPM EMAT for SH0 wave.
Similar to to the the optimization optimization tests tests for for the the meander-line meander-line coil coil EMAT, EMAT, the the number number of of PPM PPM rows rows was was Similar optimized to to obtain obtain aa good good SNR SNR signal. signal. The The test test was was conducted conducted while while changing changing the the number number of of PPM PPM optimized rows from from two two to to eight, eight, and and accordingly, accordingly, the the length length of of the the racetrack racetrack coil coil had had to to be be changed changed to to fit fit the the rows
Sensors 2016, 16, 1737
9 of 14
Sensors 2016, 16, 1737
9 of 14
Similar to the optimization tests for the meander-line coil EMAT, the number of PPM rows was length of the array, andSNR the capacitance, inductance, resistance, and impedance each coil were optimized to PPM obtain a good signal. The test was conducted while changing the of number of PPM measured using an LCR meter as shown in Table 3. Figure 14 shows the result of the amplitude rows from two to eight, and accordingly, the length of the racetrack coil had to be changed to fit and the SNR ofofthe wave while of PPM rows wasresistance, set from two eight. Andofiteach is easy find length theSH0 PPM array, andthe thenumber capacitance, inductance, andtoimpedance coiltowere that when the number of PPM rows is equalintoTable five, the PPM EMAT can generate SH0 wave. measured using an LCR meter as shown 3. Figure 14 shows the resultthe ofbest the SNR amplitude and SNR of the SH0 wave while the number of PPM rows was set from two to eight. And it is easy to Table 3. The capacitance, inductance, resistance, and impedance of the racetrack coil. find that when the number of PPM rows is equal to five, the PPM EMAT can generate the best SNR SH0 wave. Number of the Capacitance/ nF Inductance/ H Resistance/ m Impedance/ PPMTable Rows 3. The capacitance, inductance, resistance, and impedance of the racetrack coil.
2
1305.3
Number of Capacitance/nF 3 the PPM Rows 859.75
4 5 6 7 8
2 3 4 5 6 7 8
734.08 584.11 503.73 429.12 370.45
1305.3 859.75 734.08 584.11 503.73 429.12 370.45
1.9406 Inductance/µH 2.9478 1.9406 3.4506 2.9478 4.3366 3.4506 5.0169 4.3366 5.0169 5.9027 5.9027 6.8380 6.8380
494.69 Resistance/mΩ 730.01 494.69 831.21 730.01 955.75 831.21 1156.0 955.75 1156.0 1271.4 1271.4 1445.7 1445.7
1.3159 Impedance/Ω 1.9891 1.3159 2.3216 1.9891 2.8874 2.3216 3.3643 2.8874 3.3643 3.9403 3.9403 4.5329 4.5329
different number number of of PM PM rows. rows. Figure 14. The amplitude and SNR of SH0 wave with different
5. Experiment 5. Experimentto toDetect DetectaaGuided Guided Wave Wave into into the the Pipe Pipe Using Using the the S0 S0 and and SH0 SH0 Waves Waves In order order to to select select the the appropriate appropriate parameters parameters for for the the device, device, such such as as the the function function generator generator and and In amplifier, as as well well as as to to compute compute the the mode mode conversion conversion efficiency efficiency later, later, we we conducted conducted an an experiment experiment amplifier, for the wave in in thethe freefree waveguide before traveling to thetopipe. The transmit EMAT for the S0 S0wave waveand andSH0 SH0 wave waveguide before traveling the pipe. The transmit and receive EMAT were separately placed on each side of the waveguide so that it could receive ideal EMAT and receive EMAT were separately placed on each side of the waveguide so that it could signals ideal of the S0 wave and through selectionselection as shown Figure 15. From receive signals of the S0 SH0 wavewave and SH0 waveparameter through parameter as in shown in Figure 15. Figure 15a,b, the velocities of the S0 wave and SH0 wave were separately computed to be about 5000 m/s From Figure 15a,b, the velocities of the S0 wave and SH0 wave were separately computed to be about and 3200 respectively, that agreedthat well with the velocity calculated from the dispersion 5000 m/s m/s, and 3200 m/s, respectively, agreed wellgroup with the group velocity calculated from the curve. Thecurve. amplitude of the firstofreceived signal ofsignal the S0 SH0 as 12.10 V and dispersion The amplitude the first received of and the S0 andwas SH0measured was measured as 12.10 V 13.75 V, respectively. and 13.75 V, respectively.
Sensors Sensors 2016, 2016, 16, 16, 1737 1737 Sensors 2016, 16, 1737
10 10 of of 14 14 10 of 14
(a) (a)
(b) (b)
Figure 15. The received signal in the free waveguide: (a) S0-wave; (b) SH0-wave. Figure 15. 15. The The received received signal signal in in the the free free waveguide: waveguide: (a) (a) S0-wave; S0-wave; (b) (b) SH0-wave. SH0-wave. Figure
The The experiments experiments were were then then processed processed in in the the mode mode conversion conversion from from S0 S0 and and SH0 SH0 to to the the The experiments were then processed in the mode conversion from S0 and SH0 to the longitudinal longitudinal mode and torsional mode guided waves, respectively, using the opposite waveguide longitudinal mode and torsional mode guided waves, respectively, using the opposite waveguide to to mode and torsional mode guided waves, respectively, using the opposite waveguide to confirm the confirm confirm the the previous previous assumption. assumption. The The distance distance between between the the two two waveguides waveguides was was set set at at 0.5 0.5 m, m, 11 m, m, previous assumption. The distance between the16 two waveguides wasexample set at 0.5 m, 1 m, and 1.5 m to and and 1.5 1.5 m m to to verify verify the the received received results. results. Figures Figures 16 and and 17 17 present present an an example of of the the received received signals signals verify the received results. Figures 16 and 17 present an example of the received signals converted converted converted from from the the S0 S0 wave wave and and SH0 SH0 wave, wave, respectively, respectively, when when the the distance distance between between two two waveguides waveguides from the S0 wave and SH0 wave, respectively, when the distance between two waveguides is 0.5 m is is 0.5 0.5 m m and and the the length length of of the the waveguide waveguide is is 11 m. m. The The velocities velocities of of the the L(0,2) L(0,2) mode, mode, L(0,1) L(0,1) mode, mode, and and and the length of the waveguide is 1m/s, m. The velocities of the L(0,2) mode, L(0,1) mode, dispersion and T(0,1) T(0,1) mode are about 5200 m/s, 2300 and 3200 m/s, respectively, from the developed T(0,1) mode are about 5200 m/s, 2300 m/s, and 3200 m/s, respectively, from the developed dispersion mode are about 5200 m/s, 2300 m/s, and 3200 m/s, respectively, from the developed dispersion curves. curves. We We could could then then compute compute the the transmit transmit time time for for each each mode mode of of guided guided wave wave and and speculate speculate which which curves. We could then compute the transmit time for each mode of guided wave and speculate which waveform waveform was was the the objective objective mode mode as as labeled labeled in in the the Figures Figures 16 16 and and 17. 17. This This was was the the same same with with the the waveform wasthe the objective mode as labeled in thewas Figures 16 and 17. This was the samesignals, with the signals when distance between the waveguides 1 m and 1.5 m. From the received signals when the distance between the waveguides was 1 m and 1.5 m. From the received signals, it it signals when the distance between the waveguides was 1 m and convert 1.5 m. From the received signals, can can be be proven proven that that the the SH0-mode SH0-mode guided guided wave wave can can successfully successfully convert to to the the longitudinal longitudinal mode mode it can be proven that the SH0-mode guided wave can successfullyofconvert to the longitudinal mode and and torsional torsional mode mode guided guided waves waves with with the the opposite opposite waveguide waveguide of the the S0-wave, S0-wave, and and the the method method of of and torsional mode guided wavesmode with thethe opposite waveguide of the can S0-wave, and the method of shifting shifting the the EMATs EMATs to to change change the the mode of of the guided guided wave wave in in the the pipe pipe can be be implemented. implemented. shifting the EMATs to change the mode of the guided wave in the pipe can be implemented.
(a) (a)
(b) (b)
Figure The received signal converted from wave when the between the is Figure 16. 16. received signalsignal converted from S0 S0from wave S0 when the distance distance between the waveguides waveguides is Figure 16.TheThe received converted wave when the distance between the 0.5 m: (a) Longitudinal wave using T-shaped waveguide; (b) Torsional wave using L-shaped waveguide. 0.5 m: (a) Longitudinal wave using T-shaped waveguide; (b) Torsional wave using L-shaped waveguide. waveguides is 0.5 m: (a) Longitudinal wave using T-shaped waveguide; (b) Torsional wave using L-shaped waveguide.
Sensors2016, 2016,16, 16,1737 1737 Sensors Sensors 2016, 16, 1737
1111 of of 1414
11 of 14
(a) (a)
(b) (b)
Figure The The received signalsignal converted from SH0 wave when the when distance between thebetween waveguides Figure17.17. received converted from SH0 wave the distance the is Figure 17. Longitudinal The received wave signalusing converted fromwaveguide; SH0 wave(b) when the distance between the waveguides is 0.5 m: (a) L-shaped Torsional wave using T-shaped waveguide. waveguides is 0.5 m: (a) Longitudinal wave using L-shaped waveguide; (b) Torsional wave using 0.5 m: (a) Longitudinal wave using L-shaped waveguide; (b) Torsional wave using T-shaped waveguide. T-shaped waveguide.
Furthermore, the amplitude of each mode of guided wave could be measured from the received Furthermore, the amplitude each mode of guided wave could beismeasured from received signal, then the mode conversionofefficiency could be calculated, which defined as the the ratio of the Furthermore, the amplitude of each mode of guided wave could be measured from the received signal, then the mode conversion efficiency could be calculated, which is defined as the ratio the amplitude of the guided wave in the pipe to the amplitude of the first received signal in theoffree signal, then the mode conversion efficiency could be calculated, which is defined as the ratio of amplitude of the guided wave in the pipe to the amplitude of the first received signal in the free waveguide. The average amplitude of pipe the to guided wave for thefirst three distances the the amplitude of the guided wave in the the amplitude of the received signalbetween in the free waveguide. The average amplitude of the guided wave for the three distances between the waveguides used. The results of tothree the different guided waves in the waveguide. was The average amplitude ofthe the mode guidedconversion wave for the distancesmode between the waveguides waveguides was used. The results of the mode conversion to the different mode guided waves in the pipe SH0-wave are shown Figure mode 18. It guided can be waves observed thatpipe thefrom mode was from used. the TheS0-wave results ofand the mode conversion to theindifferent in the pipe from the S0-waveS0-wave and SH0-wave are shown in Figurewave 18. Ithas canthe be best observed that the conversion from the T(0,1) mode andmode the the S0-wave and the SH0-wave aretoshown in Figure 18.guided It can be observed that the efficiency, mode conversion conversion from the S0-wave to the T(0,1) mode guided wave has the best efficiency, and the efficiency the S0-wave the L(0,2) mode efficiency. However, is quite from the of S0-wave to the converting T(0,1) modetoguided wave has has the the bestlowest efficiency, and the efficiencyit of the efficiency of the converting to the L(0,2)from mode has thewave, lowest However, is for quite the opposite forS0-wave the efficiency result converted the SH0 in efficiency. other theopposite SH0itwave is S0-wave converting to the L(0,2) mode has the lowest efficiency. However, it iswords, quite the the opposite for the efficiency result converted from the SH0 wave, in other words, the SH0 wave most likely to convert to the L(0,2) mode rather thanin other T(0,1) mode.the SH0 wave is most likely to is the efficiency result converted from the SH0 wave, words, most likely to therather L(0,2)than mode the T(0,1) mode. convert toto theconvert L(0,2) mode therather T(0,1)than mode.
(a)
(b)
(a) of mode conversion to different mode guided waves in (b)the pipe (a) converted Figure 18. The efficiency from S0 and (b) converted from SH0. Figure 18.18. The efficiency modeguided guidedwaves waves the pipe converted Figure The efficiencyofofmode modeconversion conversion to to different different mode inin the pipe (a)(a) converted from S0 and (b) converted from SH0. from S0 and (b) converted from SH0.
It is also necessary to compare the mode conversion behaviors between the S0-wave and SH0wave converttotocompare the samethe mode guided wave in the pipe between shown inthe Figure 19. Therefore, It when is alsothey necessary mode conversion behaviors S0-wave and SH0it can be concluded that the S0-wave has a better efficiency to convert to the T(0,1) mode guided wave wave when they convert to the same mode guided wave in the pipe shown in Figure 19. Therefore, theconcluded SH0-wave,that while SH0-wave more likely to convert to thetoL(0,1) and L(0,2) guided itthan can be thethe S0-wave has is a better efficiency to convert the T(0,1) modemode guided wave waves than the S0-wave, which is a very important characteristic of the mode conversion behavior. than the SH0-wave, while the SH0-wave is more likely to convert to the L(0,1) and L(0,2) mode guided
waves than the S0-wave, which is a very important characteristic of the mode conversion behavior.
Sensors 2016, 16, 1737
12 of 14
It is also necessary to compare the mode conversion behaviors between the S0-wave and SH0-wave when they convert to the same mode guided wave in the pipe shown in Figure 19. Therefore, it can be concluded that the S0-wave has a better efficiency to convert to the T(0,1) mode guided wave than the SH0-wave, while the SH0-wave is more likely to convert to the L(0,1) and L(0,2) mode guided waves Sensors 16, 1737 which is a very important characteristic of the mode conversion behavior. 12 of 14 than2016, the S0-wave,
(a)
(b)
(c) Figure betweenS0 S0wave waveand andSH0 SH0 wave when they convert to the same Figure19. 19.The Theefficiency efficiencycomparison comparison between wave when they convert to the same mode T(0,1) mode; mode;(b) (b)L(0,1) L(0,1)mode; mode; L(0,2) mode. modeguided guidedwave wavein inthe the pipe: pipe: (a) T(0,1) (c)(c) L(0,2) mode.
InInaddition, efficiencyofof different mode conversions be calculated when addition,the the average average efficiency thethe different mode conversions couldcould be calculated when the the same waveguide was used. For example, the average efficiency of mode conversions including same waveguide was used. For example, the average efficiency of mode conversions including S0 to S0T(0,1), to T(0,1), SH0 to L(0,1), and SH0 to can beasregarded as the forwaveguide, the L-shaped SH0 to L(0,1), and SH0 to L(0,2) canL(0,2) be regarded the efficiency for efficiency the L-shaped waveguide, and also the average efficiency for the T-shaped waveguide. The comparison between and also the average efficiency for the T-shaped waveguide. The comparison between the efficiency of efficiency the two waveguides shown in Figure 20, which that the L-shaped had a the of the twois waveguides is shown in indicates Figure 20, which indicateswaveguide that the L-shaped better efficiency than the T-shaped waveguide no matter which mode guided wave was excited waveguide had a better efficiency than the T-shaped waveguide no matter which mode guided in wave theexcited waveguide. was in the waveguide.
Sensors 2016, 16, 1737 Sensors 2016, 16, 1737
13 of 14 13 of 14
Figure Figure20. 20.The Theefficiency efficiencycomparison comparisonbetween betweentwo twowaveguides. waveguides.
6.6. Conclusions Conclusions InInthis thisstudy, study,based basedononthe theprevious previouspipe pipeinspection inspectionsystem, system,an anSH0-wave SH0-wavePPM PPMEMAT EMATand andan an S0-wave meander-line coil EMAT were developed and optimized based on the dispersion curves of S0-wave meander-line coil EMAT were developed and optimized based on the dispersion curves of the theguided guidedwaves. waves.The Theexperiments experimentswere weresuccessfully successfullyperformed performedon onthe themode modeconversion conversionfrom fromthe the SH0-mode SH0-modetotothe thelongitudinal longitudinalmodes modeswith withthe theL-shaped L-shapedwaveguide waveguideand andtotothe thetorsional torsionalmode modewith with the theT-shaped T-shapedwaveguide. waveguide.The Theresults resultsshowed showedthat thatthe themode modeswitch switchbetween betweenthe thelongitudinal longitudinaland and torsional torsionalcan canbebeaccomplished accomplishedby bythe theshift shiftbetween betweenthe theS0-EMAT S0-EMATand andSH0-EMAT SH0-EMATwithout withoutshifting shifting the thefixed fixedwaveguide, waveguide,which whichwill willprovide providesignificant significantconvenience convenienceininthe theprocess processofofpipe pipeevaluations. evaluations. Moreover, Moreover,the themode modeconversion conversionbehaviors behaviorswere werestudied studiedbybycalculating calculatingand andcomparing comparingthe theamplitude amplitude efficiency. efficiency.ItItcan canbebeinferred inferredthat thatthe theS0S0wave wavehas hasa abetter betterability abilitytotoconvert converttotothe theT(0,1) T(0,1)mode, mode,while while the theSH0 SH0wave wavehas hasa abetter betterability abilitytotoconvert converttotothe theL(0,1) L(0,1)mode modeand andL(0,2) L(0,2)mode, mode,and andthe theL-shaped L-shaped waveguide waveguide. waveguidehas hasbetter betterefficiency efficiencythan thanthe theT-shaped T-shaped waveguide. Author AuthorContributions: Contributions:Z.S., Z.S.,H.N. H.N.and andR.M. R.M.proposed proposedthe theidea; idea;F.S. F.S.and andR.M. R.M.conceived conceivedand anddesigned designedthe the experiments; F.S. performed the experiments; F.S. also contributed to writing and editing the manuscript. Z.S., Q.C. experiments; F.S. performed the experiments; F.S. also contributed to writing and editing the manuscript. Z.S., and R.M. guided the research and were charged with the critical revisions of the manuscript; all of the authors Q.C. andthe R.M. guided the research and were charged with the critical revisions of the manuscript; all of the analyzed results. authors analyzed the results. Conflicts of Interest: The authors declare no conflict of interest. Conflicts of Interest: The authors declare no conflict of interest.
References References 1. 1. 2. 2. 3. 3. 4. 4. 5. 5. 6. 6. 7.7.
8.
Burrows, S.E.; Fan, Y.; Dixon, S. High temperature thickness measurements of stainless steel and low carbon Burrows, S.E.; Fan, Y.; Dixon, S. High temperature of stainless steel and low steel using electromagnetic acoustic transducers. NDT Ethickness Int. 2014,measurements 68, 73–77. [CrossRef] carbon steel acoustic transducers.patch NDTtransducer E Int. 2014,array 68, 73–77. Liu, Z.; Hu, Y.;using Fan, J.electromagnetic Longitudinal mode magnetostrictive employing a multi-splitting Liu, Z.; coil Hu,for Y.;pipe Fan,inspection. J. Longitudinal mode magnetostrictive patch transducer array employing a multimeander NDT E Int. 2016, 79, 30–37. [CrossRef] splitting meander coil for pipe inspection. NDT E Int. 2016, 79, Duan, W.; Kirby, R. A numerical model for the scattering of elastic30–37. waves from a non-axisymmetric defect in W.; Kirby, A numerical for the scattering of elastic waves from a non-axisymmetric defect a Duan, pipe. Finite Elem. R. Anal. Des. 2015,model 100, 28–40. [CrossRef] in a pipe. Elem. Anal. Des. 2015, 100, 28–40. Willey, C.L.;Finite Simonetti, F.; Nagy, P.B. Guided wave tomography of pipes with high-order helical modes. Willey, C.L.; Simonetti, Nagy, P.B. Guided wave tomography of pipes with high-order helical modes. NDT E Int. 2014, 65, 8–21. F.; [CrossRef] NDT P.B.; E Int.Simonetti, 2014, 65, 8–21. Nagy, F.; Instanes, G. Corrosion and erosion monitoring in plates and pipes using constant Nagy,velocity P.B.; Simonetti, F.; Instanes, G. Corrosion erosion monitoring in plates[PubMed] and pipes using constant group Lamb wave inspection. Ultrasonicsand 2014, 54, 1832–1841. [CrossRef] group E.; velocity wave inspection. Ultrasonics 2014, of 54,buried 1832–1841. Leinov, Lowe,Lamb M.J.S.; Cawley, P. Ultrasonic isolation pipes. J. Sound Vib. 2016, 363, 225–239. Leinov, E.; Lowe, M.J. S.; Cawley, P. Ultrasonic isolation of buried pipes. J. Sound Vib. 2016, 363, 225–239. [CrossRef] Cegla, Davies, J.O. J.O.High-temperature High-temperatureultrasonic ultrasonic crack monitoring using SH waves. Cegla,F.B.; F.B.;Jarvis, Jarvis, A.J.C.; A.J. C.; Davies, crack monitoring using SH waves. NDT NDT E 2011, Int. 2011, 44, 669–679. [CrossRef] E Int. 44, 669–679. Cegla, F.B.; Cawley, P.; Allin, J. High-temperature (>500 °C) wall thickness monitoring using dry-coupled ultrasonic waveguide transducers. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2011, 58, 156–167.
Sensors 2016, 16, 1737
8.
9. 10. 11. 12. 13. 14. 15.
16. 17. 18. 19.
20. 21. 22. 23.
24. 25. 26. 27. 28.
14 of 14
Cegla, F.B.; Cawley, P.; Allin, J. High-temperature (>500 ◦ C) wall thickness monitoring using dry-coupled ultrasonic waveguide transducers. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2011, 58, 156–167. [CrossRef] [PubMed] Murayama, R.; Kobayashi, M.; Matsumoto, K. Ultrasonic Inspection System Using a Long Waveguide with an Acoustic Horn for High-Temperature Structure. J. Sens. Technol. 2014, 4, 177–185. [CrossRef] Hernandez-Valle, F.; Dixon, S. Preliminary tests to design an EMAT with pulsed electromagnet for high temperature. Am. Inst. Phys. 2009, 1096, 936–941. Hernandez-Valle, F.; Dixon, S. Initial tests for designing a high temperature EMAT with pulsed electromagnet. NDT E Int. 2010, 43, 171–175. [CrossRef] Tian, G.Y.; Sophian, A. Reduction of lift-off effects for pulsed eddy current NDT. NDT E Int. 2005, 38, 319–324. [CrossRef] Huang, S.; Zhao, W.; Zhang, Y. Study on the lift-off effect of EMAT. Sens. Actuators A Phys. 2009, 153, 218–221. [CrossRef] Yu, Y.; Yan, Y.; Wang, F. An approach to reduce lift-off noise in pulsed eddy current nondestructive technology. NDT E Int. 2014, 63, 1–6. [CrossRef] Morrison, J.P.; Dixon, S.; Potter, M.D.G. Lift-off compensation for improved accuracy in ultrasonic lamb wave velocity measurements using electromagnetic acoustic transducers (EMATs). Ultrasonics 2006, 44, e1401–e1404. [CrossRef] [PubMed] Murayama, R.; Matsumoto, K.; Ushitani, K. Pipe Inspection System by Guide Wave Using a Long Distance Waveguide. Mod. Mech. Eng. 2015, 5, 139–149. [CrossRef] Lee, H.; Park, H.W.; Sohn, H. Pipe Defect Visualization and Quantification Using Longitudinal Ultrasonic Modes. Int. J. Struct. Stab. Dyn. 2014, 14, 1440008. [CrossRef] Murayama, R.; Weng, J.; Kobayashi, M. Pipe inspection system of a pipe by three-modes guide wave using polarized-transverse wave EMATs. Proc. SPIE 2014, 9302, 93022T. [CrossRef] Xu, J.; Wu, X.; Kong, D. A Guided Wave Sensor Based on the Inverse Magnetostrictive Effect for Distinguishing Symmetric from Asymmetric Features in Pipes. Sensors 2015, 15, 5151–5162. [CrossRef] [PubMed] Zheng, X.M.; Zhao, Y.Z.; Shi, Y.W. Calculation for lamb wave dispersion curves. Nondestr. Test. 2003, 25, 66–68. Xu, X.N.; Yu, Z.J.; Zhu, L.Q. A graphical method to solve a dispersion equation of Lamb wave. J. Electron. Meas. Instrum. 2012, 11, 9. [CrossRef] Lowe, M.J.S.; Alleyne, D.N.; Cawley, P. Defect detection in pipes using guided waves. Ultrasonics 1998, 36, 147–154. [CrossRef] Murayama, R.; Makiyama, S.; Kodama, M. Development of an ultrasonic inspection robot using an electromagnetic acoustic transducer for a Lamb wave and an SH-plate wave. Ultrasonics 2004, 42, 825–829. [CrossRef] [PubMed] Petcher, P.A.; Burrows, S.E.; Dixon, S. Shear horizontaL(SH) ultrasound wave propagation around smooth corners. Ultrasonics 2014, 54, 997–1004. [CrossRef] [PubMed] Murayama, R.; Mizutani, K. Conventional electromagnetic acoustic transducer development for optimum Lamb wave modes. Ultrasonics 2002, 40, 491–495. [CrossRef] Andruschak, N.; Saletes, I.; Filleter, T. An NDT guided wave technique for the identification of corrosion defects at support locations. NDT E Int. 2015, 75, 72–79. [CrossRef] Petcher, P.A.; Dixon, S. Weld defect detection using PPM EMAT generated shear horizontal ultrasound. NDT E Int. 2015, 74, 58–65. [CrossRef] Hill, S.; Dixon, S. Frequency dependent directivity of periodic permanent magnet electromagnetic acoustic transducers. NDT E Int. 2014, 62, 137–143. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
