Sensors and Actuators A 248 (2016) 162172

Contents lists available at ScienceDirect

Sensors and Actuators A: Physical

journal homepage: www.elsevier.com/locate/sna

Inuence of metallic shields on pulsed eddy current sensor for

ferromagnetic materials defect detection
Deqiang Zhou a,b, , Jun Wang a,b , Yunze He c, , Dewen Chen d , Ke Li a
a
The key laboratory for advanced food manufacturing equipment technology of Jiangsu province, Wuxi, 214122, China
b
School of Mechanical Engineering, Jiangnan University, Wuxi, 214122, China
c
College of Electrical and Information Engineering, Hunan University, Changsha, 410082, China
d
National Center of Quality Supervision and Inspection for Commodity, Yiwu, 322000, China

a r t i c l e i n f o a b s t r a c t

Article history: Pulsed eddy current (PEC) with transient analysis is a signicant advance over conventional eddy current
Received 1 June 2016 testing using impedance analysis at a particular frequency. The fabrication and optimization of PEC sensor
Received in revised form 27 July 2016 are important issues in PEC applications in nondestructive testing (NDT) eld especially for ferromagnetic
Accepted 27 July 2016
materials. In this work, the inuence of metallic shields on circular PEC sensor for defect detectability
Available online 29 July 2016
in ferromagnetic metals has been investigated through nite element models (FEM) and experimental
studies by comparing iron and aluminum shields with no-shield. Both simulations and experiments
Keywords:
indicate that PEC probe with an aluminium shield can effectively enhance the amplitude of PEC response
Pulsed eddy current (PEC)
Sensor optimization
in both surface and subsurface defect detection. However, the sensitivity is decreased. In contrast, iron
Nondestructive testing (NDT) shield has better shielded performance and can effectively improve the sensitivity of both surface and
Ferromagnetic materials subsurface defects detection in ferromagnetic materials.
Metallic shield 2016 Elsevier B.V. All rights reserved.
Sensitivity

1. Introduction ment of wall thinning in insulated ferromagnetic pipes [10,11];

As one of the advanced electromagnetic nondestructive testing
(NDT) technologies, pulsed eddy current (PEC) has a lot of advan-
tages over eddy current testing (ECT) [1,2] and other NDT methods.
Comparing to conventional ECT, PECs frequency spectra are wider
and information is richer, thus it can be used to detect defects at
different depths at the same time. And, various features extracted
from time and frequency domains (such as peak time, rising time
and frequency to zero) can be used to estimate sizes, locations and
types of defects [3]. Unlike ultrasonic, PEC does not require cou-
plant and is totally non-contact. Contrary to optical methods which
only detect surface damages, PEC can inspect inside of the material
due to electromagnetic led diffusion. Lastly, PEC does not have
any radiation and it is environment friendly. Due to these advan-
tages, PEC has been used in material characterization, geometry
measurement and defect detection for these kinds of materials:
(1) non-ferromagnetic metal and its alloy, such as aluminum and
copper slabs and multi-ply structure in aircraft; (2) ferromagnetic
materials [49], such as steel corrosion characterization and assess-
Corresponding authors.
E-mail addresses: zhoudeqiang@jiangnan.edu.cn (D. Zhou), hejicker@163.com,
hejicker@gmail.com (Y. He).
(3) conductive composites, such as carbon ber reinforced plastic
(CFRP) and sandwich honeycomb structures [12]. In these applica-
tions, the fabrication and optimization of PEC sensor, the analysis
and feature extraction of PEC sensor signal are the important issues.
By now, a lot of works are focused on sensor design of PEC and other
electromagnetic NDT.
The PEC probe usually comprises a driving unit for magnetic eld
excitation and sensors for magnetic eld measurement. The driving
unit is normally an induction coil in either cylindrical or rect-
angular shape [13]. The sensors for magnetic eld measurement
are usually bobbin coil or magnetic sensor. For non-ferromagnetic
metal and conductive composites, more attention has been paid
on structural design and feature extraction [14]. However, ferro-
magnetic material is special due to the great permeability. Due
to electric conductivity and the iron magnetic eld resulting from
magnetic permeability, the induced magnetic eld in the ferromag-
netic material is the superposition of eddy current (EC) induced
magnetic eld [15]. There is a signicant difference in PEC sensor
signals between ferromagnetic and non-ferromagnetic materials.
Therefore, not only PEC sensor but also other EM NDT such as
magnetic ux leakage (MFL) sensor should be optimized for fer-
romagnetic material detection. On the basis of the analysis of
magnetic refraction, Sun et al. [1620] put forward a new princi-
ple of MFL based on magnetic vacuum leakage, in which the strong

http://dx.doi.org/10.1016/j.sna.2016.07.029
0924-4247/ 2016 Elsevier B.V. All rights reserved.
 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172
background magnetic eld is eliminated. Song [21] reported that
detection signals of EC probe are integrating MFL and EC signals
on the saturation magnetization state. For the ferromagnetic met-
als, the PEC will magnetize the ferromagnetic sample at the same
time. Therefore, PEC probe can detect defects based on the leakage
of magnetic ux and eddy current. On the other hand, the remote
eld eddy current (RFEC) has been applied for the inspection of
the conducting tubular products such as oil and gas pipes [2224].
Shijn [25] reported the progress in designing a shielded encircling
RFEC probe to inspect nuclear fuel rods and investigated resulting
signal characteristics. These results show that the totally shielded
probe dramatically improved signal characteristic. H. WANG [26]
designed a new RFEC transducer for inspection of at conductive
plates with shielding structures. The results of numerical simula-
tions and experimental study indicate that the new transducer has
the capability of detecting deep buried defects. In order to eliminate
the disturbance of the background magnetic eld, the technology
of shielding was put forward to obtain the disturbed eld prefer-
ably above the defect. The above studies indicate that the technique
shields play important roles in MFL and RFEC.
The motivation of this work is explained as follows. According to
the state-of-the-art review of sensor of PEC and other Electromag-
netic NDT including EC, MFL, and RFEC for ferromagnetic materials,
it can be found that detection signals of PEC are complex and irreg-
ular because of eddy current magnetization of ferromagnetic metal,
and the shielding technology can remarkably improve the sensor
performances. However, the inuence of the metallic shield on the
PEC sensor for defects detectability in ferromagnetic metals has not
been investigated. Furthermore, as explained in [27], non-ferrite
metallic material (shields) will reduce the quality factors of eddy
current sensor and ferrite metallic material (shields) will increase
the quality factors, high permeability material in particular. There-
fore, different metallic shields will have different behaviors of PEC.
This work is devoted to solve this problem. The lift-off is commonly
known to be one of the main obstacles for effective eddy current
NDT testing, which can disturb the response signals and lead to the
erroneous results, as the same in PEC testing. However, this paper
is focused on investigating the inuences with the different shield
structure of PEC probes without considering the inuences of vari-
able lift-off. And, this work has not considered the defect detection
with big lift-off in the 25 cm range, because this is the preliminary
study. The big lift-off will be investigated in the next work. Finite
element models (FEMs) and experimental studies are conducted
to investigate the PEC sensor with aluminum and iron shields and
Fig. 1. (a) The model of PEC sensor without shield and (b) with shield.
163
no-shield for defect detection in ferromagnetic materials. The rest
of the paper is organized as follows. Firstly, simulation studies are
conducted in Section 2. Then, the PEC experimental system is built
in Section 3, which is followed by experimental studies and dis-
cussion in Section 4. Finally, the conclusion and future work are
outlined.
2. Simulations with FEM
2.1. Simulation setup
A 2D axis-symmetric model of the cylindrical PEC probe has
been established using COMSOL Multiphysics 4.3b [28]. The phys-
ical eld was selected as magnetic eld under AC/DC module and
then solved using the transient solver. In an effort to investigate the
inuence of shield on defect detectability in ferromagnetic metals,
three models were established, which are: 1) without a shield (with
no-shield); 2) with iron shielding; and 3) aluminum shielding. The
PEC sensor model without shield is shown in Fig. 1(a), while the
models with iron and aluminum shields are presented in Fig. 1(b).
The simulation parameters were set as follows: the excited current
(in square waveform) was set with maximum amplitude of 0.5A,
excitation frequency of 100 Hz. The turns of coil are 667 and the
cross-sectional area of enameled wire is 1 mm2 . The inner diam-
eter, outside diameter and height of the PEC probe are 20 mm,
30 mm and 20 mm, respectively. The iron plate was selected as a
sample with electrical conductivity of 1.12 107 S/m and relative
permeability of 4000, thickness of 10 mm. Material settings of the
shielding are iron and aluminum, respectively. The relative per-
meability and conductivity of aluminum are 1 and 3.77 107 S/m,
respectively. The relative permeability and conductivity of iron are
4000 and 1.12 107 S/m, respectively. The shield is 1 mm thick,
10 mm high and 5 mm wide. In the actual situation, the gap exists
between the specimen and shield, and the distance was set to
0.05 mm in the simulation. Because of the dimension of the probe
and the built-in Hall sensor in the shield, 1 mm of thickness is more
suitable for the investigation.
2.2. Distribution of eddy currents and magnetic eld
The distributions of eddy currents and magnetic eld induced
by PEC sensor were observed under no-shield, iron and aluminum
shields. The current density and magnetic induction lines were
acquired through the solution. Fig. 2(a)(c) show the distribution of
164 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172

Fig. 2. The distribution of EC and magnetic induced lines of PEC sensor (a) without shield, (b) with iron shield and (c) with aluminum shield.

EC and magnetic induced lines without shield, with iron shield and 1.05 107 , 5.11 106 and 1.07 107 A/m2 , respectively. The value
aluminum shield, respectively. It can be observed in Fig. 2 that the of induced current density the under iron shield is the smallest due
maximums of induced current density under three conditions are to the inuence of high permeability shield on magnetic eld. The
 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172 165

(a)
100
non-shield
aluminum shield
80 iron shield

60

Amplitude
40

20

0
0 500 1000 1500 2000 2500 3000
Frequency/Hz
(b)
non-shield
1.2 aluminum shield
iron shield

Normalized amplitude
1

0.8

0.6

0.4

0.2

0
0 500 1000 1500 2000 2500 3000
Frequency/Hz
Fig. 3. (a) Response of PEC sensor without shield and with aluminum shield; (b) Fig. 4. (a) Spectrum under different shields; (b) Spectrum after normalization under
Response of PEC sensor with iron shield. different shields.

induced current density with aluminum shield is the highest. This
indicates that the induced current density with aluminum shield
can be enhanced. However, it can be seen from Fig. 2(b) that the
distribution of induced current density with iron shield is more
centralized near the shield. In addition, the magnetic induced lines
could be deviated by the metallic shield with high permeability.
The simulation results demonstrate that magnetic induced lines
pass through the prole of the shield and the magnetic lines can be
shielded in the shield space. Iron shield signicantly weakens the
original excitation, and EC on both sides of the shield will spread
into the defect nearby. The disturbance caused by defects can be
captured by Hall sensors in the shield. In addition, aluminum can
impede high frequency magnetic eld and let low frequency pass
because the excited frequency is 100 Hz, which belongs to lower
frequency. As is well known to all, aluminum shield works well
in high frequency electromagnetic elds [29,30]. Therefore, the
signal will be inuenced by strong excited elds (primary elds),
and magnetic elds caused by defects (second elds) are relatively
obtained less. However, PEC with an iron shield can turn around
this condition. Therefore, the iron shield is effective to shield the
excited magnetic, and weaken the EC. On the other hand, it is very
noticeable that a max of magnetic ux density with the iron shield
in Fig. 2(b) is 3.18 104 wb/m, which is bigger than the other two
cases shown in Fig. 2(a) and (c). That is to say, the iron shield will
magnetize the iron plate better than the other two cases. The PEC
probe can also obtain MFL signal caused by defects in ferromag-
netic sample. Therefore, the signal of PEC response is complex and
irregular.
2.3. PEC responses under different shields
To further analyze the inuence of metallic shielding on the
transient response of PEC sensor, the magnetic ux density in Z
component at 1 mm height was extracted under no-shield, iron
and aluminum shields. The extracted point of data can be seen
in Fig. 1(a). This height 1 mm is suitable to simulate the lift-off of
Hall sensor. Fig. 3(a) shows the response of the PEC sensor with-
out shield and with aluminum shield, respectively. It is very clear
that the PEC response can be divided into two phases: the rising
edge of the transient response and the stable phase. Furthermore,
the permeability effects are prominent in the stable phase of the
transient response, which shows a good agreement with reference
[15]. As shown in Fig. 3(a), peak value of the PEC response with
aluminum shield is obviously larger than that with no-shield. The
Aluminum shield does not weaken the original signal. Instead, the
peak value of the signal is improved. It is generally known that the
PEC signal of ferromagnetic materials is relatively weak because of
its non-homogeneity while non-ferromagnetic materials are more
stable [31]. When aluminum is invoked as shield magnetic eld,
the excited magnetic eld can exist in iron plate and aluminum
shield at the same time. In this case, the whole EC contains por-
tions in the iron plate and some of the aluminum. Thus, the signal
with aluminum shield is preferable to that with no-shield and this
enhancement occurs. Fig. 3(b) shows the response the of PEC sensor
with iron shield. It is clear that the PEC response signal reaches the
max value rapidly and then drops to a stable value. Furthermore,
the value of the signal under the iron shield is lower than the other
166 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172

Fig. 5. (a) Surface defect and (b) sub-surface defect in the simulation models.

Fig. 6. Surface defect signals of the PEC sensor (a) with no-shield, (b) iron shield, and (c) aluminum shield in FEM.

two cases. It is very different from the PEC response signals with
the other two cases. It is well known that ferromagnetic materials
perform well for shielding of low frequency signals [32]. At the ris-
ing edge of the transient signals, it can be concluded that the iron
shielding shields the part excitation magnetic eld and weakens
the induced eddy current. The reason is that low frequency com-
ponents exist more in the stable phase of PEC signals because of
eddy current magnetization. Therefore, the amplitude of response
signal is lower than the other two cases. Furthermore, a sudden
decrease occurs in the stable phase of PEC signals. In order to verify
the effectiveness of shielding materials, spectrums of PEC signals
with different shields were obtained and ten times of harmonics
 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172 167

were extracted as low frequency components, which are shown in

Fig. 4(a). It can be seen that the amplitude under the aluminum
shield is signicantly larger than the others and the amplitude of
iron shield is the smallest. In other words, iron shield can weaken
the original signal extensively. Fig. 4(b) shows the spectrum after
normalization under different shields. It is clear that the shielding
ability of iron to high frequency is decreasing. The conclusions can
be drawn that iron shield has a great inuence on the PEC signal and
can distort the original signal in an ideal condition in the simulation.
In the following, the sensitivity of signals with different shielding
will be introduced to adequately illustrate their detection effects.

2.4. PEC signals of defects with different depths

In previous subsections, the mechanism of shielding in PEC sen-

sor has been analyzed and response signals have been achieved
in time domain. In this paper, the aim of sensor optimization
with shield is to enhance defect detectability. Therefore, the PEC
response of surface and sub-surface defects with different depth
were compared and the feature was extracted. The surface and
sub-surface defect was simulated as shown in Fig. 5(a) and (b).
The diameter of defects is 2 mm and the depth varies from 2 mm
to 8 mm. Fig. 6(a)(c) represent surface defect signals of the sen-
sor with non-shield, iron shield and aluminum shield, respectively.
From Fig. 6, it can be seen that non-shield and aluminum shield
have similarities and iron shield shows obvious difference, which
leads to sharp peaks on the rising edge of response. It is well known
that PEC signals contain two stages. One is the rising edge (transient
signal) and the other is the stable stage. The rising edge means high
Fig. 7. The detection sensitivity of different shields for surface defects in FEM.
frequency components of PEC signals and stable stage owns low
frequency components. It has been investigated that ferromagnetic
materials are effective with low frequency signals, but not sensitive
to high frequency ones. Therefore, the high frequency components
of the rising edge cant be damped adequately while low frequency
components at constant stage can be substantially attenuated. The
signal of 0mm defect is regarded as the reference signal with-
out defect. Because the reference signal is different in three cases
of shields, the comparison should be conducted with individual
reference signal. The formula (1) was chosen to calculate the rela-

Fig. 8. Subsurface defects signals of PEC sensor with (a) no-shield, (b) iron shield, and (c) aluminum shield in FEM.
168 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172

tive sensitivity and estimate detectability of defects under different

shields, where Bref means the peak value of reference signal [9],
Bdefect is the peak value of defect signal, and S was used to reect
sensitivity of defect testing. The values of S for different conditions
are presented in Fig. 7, which reveals that the curves of sensitivity
for non-shield and aluminum shield nearly overlap and that of iron
shield is distinctive. In detail, the curve of iron shield is above the
other curves and the slope of the curve of iron shield is the largest.
Based on this case, it indicates that the iron shield can improve sur-
face defect sensitivity and the defects with different depths can be
distinguished easily.
The above research is about surface defects and the following
is for the sub-surface defects. The time-domain signals of defects
with varied depths are extracted, as shown in Fig. 8(a)(c). It can
be seen that the testing effect of sub-surface defects is poor com-
pared to surface defects. In addition, the discrimination between
defects is not sufciently obvious. Relative sensitivity under dif-
ferent cases was also calculated according to the formula (1). The Fig. 9. The detection sensitivity of different shields for sub-surface defects in FEM.
result of testing sensitivity is illustrated in Fig. 9, which reveals that
aluminum shield can not enhance testing sensitivity of PEC. How-
ever, the iron shield can improve testing sensitivity and the trend 3. Experimental set-up
of enhancement is not the same prominent as that of the surface.
The experimental system consists of a pulse signal generator, a
PEC sensor with iron and aluminum shields shown in Fig. 10(a),
Bdefect Bref a power amplier, a signal conditioning circuit, a data acquisi-
S= (1)
Bref tion module, and a specimen. The experimental system is shown

Fig. 10. (a) The probe with metallic shield; (b) The experiment system; (c) The photo of iron plate and the size (mm).
 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172
Fig. 11. Surface defect signals of the PEC sensor with (a) no-shield, (b) iron shielding and (c) aluminum shielding.
in Fig. 10(b). The YUANLONG VD1641 function generator was
adopted as the pulse signal generator. The power amplier module
is LPA05 B of the Newton Newtons4th by Science and Technology
Company. The cylindrical probe with metallic shield is shown in
Fig. 10(a). The parameters of the probe are as follows: the number
of turns of the excited coil is 800 with 0.3 mm-diameter wire; the
external diameter and inner diameter of the drive coils are 30 and
20 mm, respectively; the height is 20 mm. The shield is 1 mm thick,
10 mm high and 5 mm wide. The type of Hall sensor is SS95A226.
The dimension of Hall sensor is 4 2.8 1.5 mm3 . The signal con-
ditioning circuit includes the low-pass lter circuit that composed
of the core chip of UA741 and the signal amplication circuit with
INA111 of ANALOGDEVICES as the core chip. The data acquisition
module is Adlink DAQ2010. The sampling frequency was set at
100 kHz. The toolbox in MATLAB was applied to collect data. The
photo and size of the iron plate are shown in Fig. 10(c). The speci-
men is the iron plate with several defects. The surface defects were
simulated using rectangular slots with the same width 2 mm but
the different depths (the distance from slot bottom to surface) from
2, 4, 6, and 8 mm. In experiments, these defects were named with
the slots depth. Subsurface defects were simulated by reversing the
specimen. The thickness of specimen is 10 mm. So the depths for
subsurface defects (the distance from slot bottom to subsurface)
are also 2, 4, 6 and 8 mm, respectively. The experimental defects
are not consistent with the defects of simulation. However, it is
still effective to investigate the inuences of shield on PEC sen-
sor. In the experiment, the excitation is the square wave of 1 V and
100 Hz, which passes through the power amplier. The Hall sen-
sor was used to capture signals of defects. And then, signals were
processed by the signal conditioning circuit. The cut-off frequency
of the low pass lter is 1000 Hz. The data processing ow refers to
169
the reference [33]. The defects were tested using PEC sensor with
no-shield, aluminum and iron shields, respectively.
4. Experiment results and discussion
4.1. Surface defects detection
Surface defects with different depths from 0 to 8 mm in step
of 2 mm were tested. Fig. 11(a)(c) show the surface defect sig-
nals of PEC sensor without shield, with iron shield and aluminum
shield, respectively. In experiments, the noise affects the PEC sig-
nal and ideal shielding is difcult to be achieved. So its difcult to
obtain the totally same results with FEM. However, the experimen-
tal results show the same qualitative results. It can be concluded
that the discrimination among the signals of different defects under
the iron shield is more discriminating than that under no-shield
and aluminum shield. The amplitude of the signal under iron shield
drops due to weakening the preliminary magnetic eld. According
to reference [21], the MFL signal existed in PEC signals because
of the max of magnetic ux density shown in Fig. 2(b). The mag-
netic ux density will play an important role to magnetize and
identify the defects of ferromagnetic metal. On the basis of the
magnetic refraction, diffusion and magnetic compression effect, the
leakage mechanism of magnetic ux of defects can be obtained
[16,17]. Therefore, the iron shield will be effective to suppress
the background noise and enhance MFL signals from the defect
of ferromagnetic metal. The recognition of defects is improved. In
addition, Fig. 11(c) reveals that aluminum shield will enhance the
peak value of testing signal and aluminum shield also cannot bring
a clear discrimination for the varied defects. In experiments, the
ideal shielding is inapplicable because of the built-in Hall sensor.
170 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172

Considering this point, it will cause a gap between the shield and
specimen. So the decline of the signal is relatively less than that of
the simulation.
To further analyze the inuence of different shields for surface
defects, the same calculation was made based on formula (1). The
relative sensitivity S was calculated for different defects, as shown
in Fig. 12. Obviously, there is a linear relationship between sensi-
tivity and defect depth and the sensitivity increases with the defect
depth increases. The sensitivity of the PEC sensor with iron shield
is considerably larger than that of no-shield and aluminum shield.
The aluminum shield, causing larger amplitude does not increase
sensitivity for surface defect detection.

4.2. Subsurface defects detection

Reverse the specimen with slots and put the sensor on the spec-
imens subsurface to simulate the detection of subsurface aws.
These sub-surface defects have different depths (the depth of slot)
from 0 to 8 mm in step of 2 mm 0 means no subsurface defect. The
Fig. 13(a)(c) show the sub-surface defect signals of PEC sensor
without shield, with iron shield and aluminum shield, respectively.
From Fig. 13(a)(c), the conclusions can be drawn that 1) the PEC
signals from the sub-surface are relatively weaker than that of sur-
face and 2) the recognition of defects in the sub-surface is also lower
than that of surface defects as shown in Fig. 11. According to the
formula (1), the sensitivity S of the sub-surface defect signals is cal-
culated as shown in Fig. 14. It can be concluded that although the
curves of sensitivity for three conditions are close to each other,
the iron shield still has the best detectability for sub-surface defect.
However, the enhancement is not as remarkable as surface defect.
Fig. 12. The detection sensitivity of different shields for surface defects.
Furthermore, the sensitivity cannot be improved by an aluminum
shield, although the signal is more enhanced than that of no-shield.
5. Conclusions
This paper extensively investigates the inuence of metallic
shields on PEC sensor for defect detection in ferromagnetic plates.
Both FEM and experiments where PEC probes with iron and alu-
minum shields and without shield used to detect the surface and
subsurface defects in ferromagnetic material were conducted to
analyze the PEC signals and sensitivity. The relative sensitivity,

Fig. 13. The sub-surface defect signals of PEC sensor with (a) no-shield, (b) iron shielding, and (c) aluminum shielding.
 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172
Fig. 14. The detection sensitivity of different shields for sub-surface defects.
rather than absolute sensitivity, was introduced as a specic fea-
ture to estimate performance of the PEC sensor for defect detection.
The conclusions are drawn as follows: (1) For PEC sensor, the iron
shield could effectively reduce the inuence of excited eld (the
rst eld) and weaken the induced eddy currents in ferromagnetic
materials; (2) The iron shield could effectively improve the defect
detection performances on the surface and subsurface in iron plate,
although this tendency of enhancement fades for the sub-surface
defects; (3) The aluminum shield could enhance peak values of sig-
nals. However, it cannot contribute to the increase of sensitivity of
defect detection. The characteristics of an eddy current sensor are
dependent on the electrical and magnetic properties of the target
material. It can be concluded that the inuence of an iron shield
on PEC sensor depends mainly on the equalization and the quan-
tity of permeability of the shield [27]. The conclusions of this work
have some potential applications including but not limited to: (1)
it can provide a guide to design and optimize a PEC sensor for fer-
romagnetic materials; (2) it can eliminate the noise and improve
the detection sensitivity of defect and stress; and (3) it can give
some interesting acknowledge for magnetic NDT methods such as
MFL, magnetic Barkhausen noise and magnetic acoustic emission.
However, there are some limitations. The rst is the disconnecting
between FEM and simulation; the second is the lift-off effect has
not been considered. Further research involves 1) the consistency
between FEM and experiments; 2) the detection and evaluation of
natural cracks and corrosion in ferromagnetic metals with big lift-
off distance using PEC sensor and (3) EC sensors combination with
eddy current pulsed thermography [34] for ferromagnetic materi-
als.
Acknowledgements
The work was supported by the National Natural Science Foun-
dation of China under Grants 51107053 and 61501483, China
Postdoctoral Science Foundation under Grant 2012M520994, and
Science Technology Department of Zhejiang Province under Grant
2015C32108. The authors would like to thank the China Scholar-
ship Council, China, for sponsoring Dr. Deqiang Zhous study visit to
Tohoku University, Japan and Dr. Yunze He to Newcastle University,
UK.
References
[1] R.A. Smith, Hugo GR: transient eddy current NDE for ageing aircraft
capabilities and limitations, Insight 43 (1) (2001) 1425.
[2] Yunze He, Mengchun Pan, Feilu Luo, et al., Reduction of lift-off effects in
pulsed eddy current for defect classication, IEEE Trans. Magn. 47 (12) (2011)
47534760.
171
[3] Yunze He, Feilu Luo, Mengchun Pan, et al., Defect classication based on
rectangular pulsed eddy current sensor in different directions, Sensors
Actuators A: Phys. 157 (2010) 2631.
[4] Mengchun Pan, Yunze He, Gui Yun Tian, et al., PEC frequency band selection
for locating defects in two-layer aircraft structures with air gap variations,
IEEE Trans. Instrum. Meas. 62 (2013) 28492856.
[5] Young-Kil Shin, Dong-Myung Choi, Young-Joo Kim, et al., Signal
characteristics of differential pulsed eddy current sensors in the evaluation of
plate thickness, NDT&E Int. 42 (3) (2009) 215221.
[6] Dengfeng Chen, Qichun Ji, Hongcai Zhang, et al., Application of pulsed eddy
current in plate thickness evaluation, ICIEA (2009) 32863288.
[7] Maxim Morozov, Gui Yun Tian, Philip J. Withers, The pulsed eddy current
response to applied loading of various aluminium alloys, NDT&E Int. 43
(2010) 493500.
[8] D.G. Park, C.S. Angani, G.D. Kim, et al., Evaluation of pulsed eddy current
response and detection of the thickness variation in the stainless, IEEE Trans.
Magn. 45 (10) (2009) 38933896.
[9] Gui Yun Tian, Yunze He, Ibukun Adewale, et al., Research on spectral response
of pulsed eddy current and NDE applications, Sensor Actuat. A: Phys. 189
(2013) 313320.
[10] Yunze He, Gui Yun Tian, Hong Zhang, Steel corrosion characterization using
pulsed eddy current systems, IEEE Sensors J. 12 (6) (2012)
21132120.
[11] Zhiyuan Xu, Xinjun Wu, Jian Li, et al., Assessment of wall thinning in insulated
ferromagnetic pipes using the time-to-peak of differential pulsed
eddy-current testing signals, NDT&E Int. 51 (2012) 2429.
[12] Yunze He, GuiYun Tian, Mengchun Pan, et al., Non-destructive testing of
low-energy impact in CFRP laminates and interior defects in honeycomb
sandwich using scanning pulsed eddy current, Compos: Part B 59 (2014)
196203.
[13] Noritaka Yusa, Eiji Machida, Ladislav Janousek, et al., Application of eddy
current inversion technique to the sizing of defects in Inconel welds with
rough surface, Nucl. Eng. Des. 235 (2005) 14691480.
[14] Yunze He, Feilu Luo, Menchun Pan, et al., Pulsed eddy current technique for
defect detection in aircraft riveted structures, NDT&E Int. 43 (2010) 176181.
[15] Ibukun Dapo Adewale, Gui Yun Tian, Decoupling the inuence of permeability
and conductivity in pulsed eddy current measurements, IEEE Trans. Magn. 49
(3) (2013) 11191127.
[16] Yanhua Sun, Yihua Kang, New MFL nondestructive testing method based on
magnetic vacuum leakage, J. Mech. Eng. 46 (2010) 1823.
[17] Yanhua Sun, Yihua Kang, Magnetic compression effect in present MFL testing
sensor, Sensors Actuators A: Phys. 160 (2010) 5459.
[18] Yanhua Sun, Yihua Kang, Chen Qiu, A permanent magnetic perturbation
testing sensor, Sensors Actuators A: Phys. 155 (2009) 226232.
[19] Yanhua Sun, Yihua Kang, Chen Qiu, A new NDT method based on permanent
magnetic eld perturbation, NDT&E Int. 44 (2011) 17.
[20] Yanhua Sun, Yihua Kang, A new MFL principle and method based on
near-zero background magnetic eld, NDT&E Int. 43 (2010) 348353.
[21] Song Kai, Yihua Kang, Yanhua Sun, et al., Study of signal mechanism based on
combining magnetic ux leakage and eddy current testing[J], J. Mech. Eng. 45
(7) (2009) 233237.
[22] Noriyasu Kobayashi, Souichi Ueno, Satoshi Nagai, et al., Remote eld eddy
current testing for steam generator inspection of fast reactor [C], The 18th
International Conference on Nuclear Engineering (ICONE-18) 24 (12) (2011)
46434648.
[23] H. Fukutomi, T. Takagi, M. Nishikawa, Remote eld eddy current technique
applied to non-magnetic steam generator tubes, NDT&E Int. 34 (2001) 1723.
[24] D. Kim, L. Udpa, S. Udpa, Remote eld eddy current testing for detection of
stress corrosion cracks in gas transmission pipelines, Mater. Lett. 58 (2004)
21022104.
[25] Young-Kil Shin, Design of encircling remote eld eddy current probe, IEEE
Trans. Magn. 38 (2) (2002) 12731276.
[26] H.T. Wang, Q.F. Luo, X. Wang, et al., Simulation and experimental study of
remote eld eddy current testing on at conductive plate, Int. J. Appl.
Eectromagn. Mech. 33 (2010) 12611266.
[27] G.Y. Tian, Z.X. Zhao, R.W. Baines, The research of inhomogeneity in eddy
current sensors, Sens. Actuators A 69 (1998) 148151.
[28] https://cn.comsol.com.
[29] Jincheng Wang, Mechanism and development of electromagnetic shielding
materials, Chin. New Chem. Mater. 30 (7) (2002) 1618.
[30] Jianhua Li, Zhong da, Ganghua Mei, et al., Research on electromagnetic screen
of electronic equipment, Chin. Electron. Compon. Device Appl. 6 (5) (2004)
5154.
[31] Huang Cheng, Theory and Instrument for Pulsed Eddy Current Wall-thinning
Measurement of Ferromagnetic Objects, Wuhan: Hua zhong University of
Science and Technology, 2010.
[32] Jean-Michel Thomassin, Christine Jerome, Thomas Pardoen, et al.,
Polymer/carbon based composites as electromagnetic interference (EMI)
shielding materials, Mater. Sci. Eng. R 74 (2013) 211232.
[33] Deqiang Zhou, Yong Li, Xiangyang Yan, et al., The investigation on the optimal
design of rectangular PECT probes for evaluation of defects in conductive
structure, Int. J. Appl. Electromagn. Mech. 42 (2) (2013) 319332.
[34] Ruizhen Yang, Yunze He, Bin Gao, et al., Lateral heat conduction based eddy
current thermography for detection of parallel cracks and rail tread oblique
cracks, Measurement 66 (2015) 5461.
172 D. Zhou et al. / Sensors and Actuators A 248 (2016) 162172

Biographies
Deqiang Zhou is an associate professor in Jiangnan Uni-
versity, China. He nished the Joint PhD study at School of
Electrical, Electronic Engineering, Newcastle University,
UK and Nanjing University of Aeronautics and Astronau-
tics, China in September 2009. He got the PhD degree at
College of Automation Engineering, Nanjing University of
Aeronautics and Astronautics in June 2010. His research
is focused on electromagnetic testing, thermography, and
composite testing in the eld of Non-destructive testing
(NDT) and Structural health monitoring (SHM). In recent
5 years, he has chaired or participated in more than 10
projects including NSFC, China Postdoctoral Science Foun-
dation etc. And he has published more than 30 academic
papers in journals and conferences. He has applied or holds about 10 patents in
China. He is a reviewer for over 10 academic journals.
Dewen Chen is a senior engineering in National Cen-
ter of Quality Supervision and Inspection for Commodity,
China. He received PhD degree at college of Food sci-
ence, Huazhong Agricultural University in June 2009. His
research is focused on the development and research of
consumer goods safety technology. In recent 10 years,
he has chaired more than 10 projects including Ministry
of Science and Technology etc. And he made or partici-
pated in 12 the formulation of national standards. He has
published about 30 academic papers in journals and con-
ferences.
Ke LI received his Ph.D. from Mie University, Japan in
2012. He is currently an associate professor in the School
of Mechanical Engineering, Jiangnan University, China. His
research interests include fault diagnosis of plant machin-
ery and signal processing.He has published more than 40
academic papers in journals and conferences.

Jun Wang is a M.Sc. candidate in Jiangnan University.

His research direction is electromagnetic non-destructive
testing. He has published several papers in Chinese jour-
nals. He has applied or holds about several patents in
China.

Yunze He received the B.Sc. degree in measurement

and control technology and instrument from Xian Jiao-
tong University (20022006), China, received MSc degree
and PhD degree in instrumentation science and tech-
nology from College of Mechatronics Engineering and
Automation, National University of Defense Technology
(20062008, 20092012). He nished the Joint PhD study
at School of Electrical, Electronic Engineering, Newcas-
tle University, UK and National University of Defense
Technology, China in February 2012. From 2013 to 2016
20132016, he is a Lecturer in National University of
Defense Technology, China. Currently, he is an Associate
Professor with the College of Electrical and Information
Engineering, Hunan University, Changsha, 410082, China. His research is focused on
Non-destructive testing (NDT), Sensing and imaging, and Structural health monitor-
ing (SHM). In recent 5 years, he has chaired or participated in more than 20 projects
including NSFC, EPSRC etc. And he has published more than 50 academic papers
in journals and conferences, which have been cited more than 900 times in Google
scholar and the h-index is 17. He has published a book titled Eddy current thermog-
raphy non-destructive testing in Chinese, which is the rst monograph focusing on
eddy current thermography non-destructive testing research in China and is col-
lected by many university libraries in China. He is a reviewer for over 25 academic
journals.