Ultrasonics 51 (2011) 675682

Contents lists available at ScienceDirect

Ultrasonics
journal homepage: www.elsevier.com/locate/ultras

Improving the signal amplitude of meandering coil EMATs by using ribbon

soft magnetic ux concentrators (MFC)
R. Dhayalan, V. Satya Narayana Murthy, C.V. Krishnamurthy, Krishnan Balasubramaniam
Centre for Nondestructive Evaluation and Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai 600 036, India

a r t i c l e

i n f o

Article history:
Received 30 May 2010
Received in revised form 20 January 2011
Accepted 25 January 2011
Available online 2 February 2011
Keywords:
Meander coil EMAT
Lorentz force
Magnetic ux concentrator (MFC)
Soft magnetic ribbons

a b s t r a c t
This paper presents a new method of improving the ultrasonic signal amplitude from a meander line
EMAT by using soft magnetic alloy ribbon (Fe60Ni10V10B20) as a magnetic ux concentrator (MFC). The
ux concentrator is a thin soft amorphous magnetic material (Fe60Ni10V10B20) which is very sensitive
to a small ux change. The MFC is used with the EMAT to improve the signal amplitude and it was
observed that the peak signal amplitude increases by a factor of two compared to the signal without
MFC. Two dimensional numerical models have been developed for the EMAT with MFC to quantify the
improvement of the received signal amplitudes. Model calculations and experiments have been carried
out for a wide range of ultrasonic frequencies (500 kHz1 MHz) in different materials.
2011 Elsevier B.V. All rights reserved.

1. Introduction
Electromagnetic acoustic transducers (EMATs) [15] are advanced Ultrasonic transducers that can generate and detect sound
waves in an electrically conducting material without making physical contact with it. Such noncontact sensors are attractive in applications in the area of Nondestructive testing (NDT) of electrically
conducting materials. Their operation is based on the Lorentz force
principle and/or by magnetoelastic effects, depending on the type
of the material either ferromagnetic or not, bias eld strength
and the EMAT conguration employed [68,37]. It can generate
different ultrasonic wave including modes that are difcult to generate with conventional piezoelectric crystals, e.g., SH (horizontally
polarized shear) waves [911]. Since the coupling between the
transducer and the component is magnetic, it has many advantages over the conventional ultrasonic techniques that require a coupling media. Hence, it is possible to eliminate any inconsistency
arising from the use of couplant during the inspection. This type
of transducer can be easily designed and fabricated based on specic application required.
The two main important components of an EMAT are a coil carrying a high current alternating pulse when in transmission mode
and a permanent magnet to induce a strong static magnetic ux
within the skin depth of the test specimen directly below the
EMAT. The pulsed alternating current fed to the transmitter EMAT
coil induces eddy currents, ~
J e , within the skin depth of the test
piece. In the presence of a large bias magnetic ux, B~S , these eddy
Corresponding author.
E-mail address: balas@iitm.ac.in (K. Balasubramaniam).
0041-624X/$ - see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ultras.2011.01.009

currents lead to body forces, F~L , at the surface layer of the specimen
[1214]. Fig. 1(a) shows a single coil and magnet leading to the
generation of Lorentz force inside the material. The Lorentz forces,
F~L , on the eddy currents are transmitted to the solid by the electrons exchanging momentum with the atoms in the metal. The
direction of the transient force in the solid below each element
of the ML coil alternates at the frequency of the driving current
and acts as the source of ultrasonic waves.

~
FL ~
Je  ~
BS

EMATs can be used as generators as well as detectors of ultrasound.

EMATs are better as detectors than generators of ultrasonic waves
[15] and in the receiving mode, the magnetic ux density B~S from
the EMATs permanent magnet interacts as a result of ultrasonic
motion with particles having a velocity v within the material. This
produces eddy currents ~
Je within the sample according to Lorentz
force mechanism, expressed as [16],

~
J e r~
v  ~BS

where r is the electrical conductivity of the sample. The current

density generated inside the sample in turn induces an e.m.f. in
the pick-up coil of the receiver EMAT which is normally placed
above the sample surface as shown in Fig. 2. This occurs as a result
of the interaction between an acoustic wave and a static magnetic
eld in the test sample. In conducting ferromagnetic materials, this
interaction can produce Lorentz and magnetostrictive current densities. Since the amplitude of the received signal is very small (in the
range of 10 lV100 lV) high gain (above 60 dB), high input

676

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

tromagnetic model has been utilized as the inputs for wave

propagation in the elastodynamic model. The numerical modeling
of meander coil EMATs have been explained in more detail elsewhere [23,24] and here only the inclusion of the MFC in the model
will be discussed. In this paper, 2D numerical models have been
developed to validate the experimental results obtained with
EMAT with MFC.
2. Numerical simulation

Fig. 1. Schematic of (a) a single coil and magnet leading to the generation of Lorentz
force and (b) the meander coil conguration used.

In simulation part, the soft magnetic ribbon (MFC) is modeled

with a permanent magnet to examine the variation of the magnetic
ux density. Then the performance of MFC with EMAT has been
analyzed and compared with the experimental results.
2.1. Modeling of MFC

Fig. 2. EMAT receiving mechanism.

impedance (50 O) and low noise electronic ampliers (less than

2 dB) are used to amplify the received signals [17].
In this paper, the improvement of the EMAT signal amplitudes
has been reported with the introduction of a new soft magnetic
ribbon (Fe60Ni10V10B20) as magnetic ux concentrator (MFC). One
of the most signicant properties of this amorphous soft magnetic
materials is magnetoimpedance (MI) [1820]. Such an effect involves a very large and sensitive change of the impedance of some
soft magnetic materials under the action of a static magnetic eld
and a small alternating current. The skin effect and domain wall
motion, depending on the frequency range, are the main physical
origins of giant magnetoimpedance (GMI) [21]. MI is useful for
magnetic-eld detecting sensor applications [22]. These devices
are very sensitive and have quick responses to magnetic elds.
The soft magnetic ribbons (MFC) were used with the EMAT to
improve the signal amplitudes and it has been observed that the
amplitude has increased for different wave modes in different
materials. Additionally, a 2D nite element model has been developed for the EMAT to quantify the improvement of the signal
amplitudes (peak to peak) of different wave modes with MFC
and are compared to the experimentally obtained peak signal
amplitudes with and without MFC for a wide range of frequency
from 500 kHz to 1 MHz.
The numerical simulation of an ML EMAT has been done in two
steps. First, an electromagnetic model has been developed for the
Lorentz force density calculation. The commercial nite element
software COMSOL 3.5 was used for the electromagnetic modeling
of the ML EMAT. This software package has predened applications
called modes with built-in mathematical solutions to facilitate
modeling. The Lorentz force density calculation has been done
for a wide range of frequency varying from 500 kHz to 1 MHz. Second, the wave propagation simulation has been done by using another commercial explicit nite element package ABAQUS 6.6.8. It
uses an explicit integration scheme for solving the transient dynamic and quasi-static analyses. The output of the COMSOL elec-

A square permanent magnet of size 25  25  12.5 mm has

been used for this study and a thin soft magnetic material of
30 lm thick and 10 mm length has been introduced with the magnet. The permanent NdFeB (N38H) magnets with residual magnetic
strength (Br = |Br|) of 1.254 T were used for both simulation studies
and experiments. This work involved two steps: rst the permanent magnet was modeled alone and second, the magnet with
the MFC was modeled for the magnetic ux density (B) measurements. In the second case, the MFC was placed 0.5 mm below the
magnet surface and the size of the MFC (10  0.02 mm) covers exactly the active area of the EMAT coil. Two dimensional magnetostatic models have been developed for the spatial distributions of
the magnetic ux density (B) arising from both the permanent
magnet and the permanent magnet with MFC. In such a model,
the magnetic scalar potential Um for static magnetic eld can be
dened as [25],

~ rUm
H

where H is the magnetic eld intensity. By using the Gauss law

of magneto statics, the divergence of magnetic ux density B at any
point is described by [26],

r ~
B0

The constitutive relation of magneto statics for both B and H can be

written as,

~
~ M
~
B l0 H

where l0 is the permeability of free space and M is the magnetization which can also be written as,

~ M
~ 0 l  1H
~
M
r

where M0 is the pre-magnetization vector of the magnet and lr is

the relative permeability. Combining all the equations, it can be
shown that,

~ 0 0
r  l0 lr rUm  l0 M

In order to solve for the magnetic scalar potential (Um) based on

Eq. (7), the commercial nite element software COMSOL (FEMLAB
3.2) has been used. The permanent magnet and the magnet with
MFC have been modeled using 2D static analysis with predened
coefcients of l = l0lr and c = l0M0 [27]. The pre-magnetization
constant (M0) value for the NdFeB magnet (N38H) can be calculated based on the magnetization hysteresis curve when H = 0
and B = Br. Therefore Eqs. (6) and (7) reduce to M0 = Br/l0 [28]
and the pre-magnetization value is calculated to be
|M0| = M0 = 998 kA/m. The value of relative permeability for the
soft magnetic material (MFC) is lr = 2  106 as provided by the
manufacture of the soft magnetic ribbons. The saturation magnetic

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

eld for this material is 100 Oe and the electrical resistivity is

134 lX cm. Additional details about the soft magnetic ribbon
(MFC) are explained in the experimental part. The active region
of the EMAT coil has covered by the ribbon and the arrangement
of the ribbon with the coil and magnet is shown in Fig. 3a.
After dening the material geometries, the appropriate components of M0 are applied for both the models. For the rst case, the
magnetic eld is applied along the positive Y axis for the magnet
and the Y axis component of M0 (M0Y) is set equal to M0 and the
other component M0X is set to zero. For the second case, the magnetic elds are applied along the positive Y axes like the previous
case the appropriate lr values are applied along the positive Y axes
other axes set to zero for the MFC. With boundary conditions, the
2D mappings of the magnetic ux densities have been modeled for
both cases. Fig. 4a and b shows spatial distribution of the magnetic
ux densities concentrated around the edges of the magnet and the
magnet with MFC. Fig. 4c and d shows the observed magnetic ux
densities below the surface (0.6 mm) of the permanent magnet and
the magnet with MFC respectively. The black lines just below the
magnet and the magnet with MFC are the led observation lines.

677

From Fig. 4c and d, it has been noted that the magnetic ux density is increased for second case, the magnet with MFC. From the
numerical model, it has been observed that the maximum ux
density is about 0.3 T for the permanent magnet and 0.41 T for
the permanent magnet with MFC. In the rst case, the magnetic
ux lines from the magnet diverges away from the edges and cover
a larger spatial area but in the second case, the soft magnetic ribbon (MFC) concentrates the ux lines from the end face of the magnet and cover the small active region of the EMAT coils.
3. Experimental details
For experimental studies, a set of meander coil EMATs have
been developed which consists of a normal biasing magnet and a
copper coil with periodicity of 3 mm. The coils were fabricated
by printed circuit board (PCB) technique which allows to fabricate
any type of at coil of arbitrary pattern. The printed circuits boards
are made by 150 lm polyester based exible laminate with 30 lm
thick copper clad. The permanent NeodymiumIronBoron (Nd
FeB) sintered magnets were used for bias magnetic eld.

Fig. 3. Arrange of EMAT coil and the magnet (a) with soft magnetic ribbon and (b) without soft magnetic ribbon.

Fig. 4. Magnetic ux density distributions of (a) the magnet alone and (b) the magnet with MFC. The cross-sectional eld proles across both the images represents the
variation of the magnetic ux density with (c) the magnet alone and (d) the magnet with MFC.

678

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

3.1. Meander coil EMAT

The transmitter coil is driven by an alternating current which
induces eddy current inside the test specimen. The interaction of
eddy current and the perpendicular magnetic eld induces Lorentz
force parallel to the surface, whose directions change alternatively
with the meandering period. Such a force distribution generates
Rayleigh waves along surface and oblique bulk waves (longitudinal
and transverse waves) into the specimen [29,30]. Fig. 5a shows the
cross sectional view of the meander coil EMAT and the wave
modes generation inside a thick specimen. The meander coil EMAT
can also be used to generate Lamb waves in thin plates. Two
different sizes of aluminum (400  200  2 mm and 300  755 
50 mm) and stainless steel plates (300  200  2 mm and
250  125  30 mm) were used in the experiments.
Fig. 5b shows the schematic diagram of the instrumentation set
up for the experimental studies. In order to generate a high current
RF tone burst, a RITEC RPR4000 gated amplier was used and the
high power output was applied to the transmitter excitation coil.
An impedance matching network system was used between the
high impedance of the pulser to the low impedance of the transmitter coil. By, adjusting the control setting present in the gated
amplier, the output power to the coil was varied. The inbuilt
broadband receiver was used to amplify the received signal from
the receiver EMAT. A 60 dB gain setting at the receiver has been
used for all the measurements. The receiver EMAT is connected
to another impedance matching network to match the low impedance receiver coil to the high impedance receiver. The output from
the broadband receiver is fed to the Agilent DSO6032A Digital storage oscilloscope which is interfaced with PC for data acquisition.
Before starting the actual experiment, all the instruments and
the power level of the RITEC with the control setting are tested
and calibrated.
3.2. Magnetic ux concentrator (MFC)
The magnetic ribbon (Fe60Ni10V10B20) was prepared by the melt
spinning technique in argon atmosphere. The ribbon (Fe60Ni10V10B20) has high content Fe with negligible or zero magnetostriction and it can be used for the design of low magnetic eld
sensors such as magneto impedance (DZ/Z%) sensors [31,32]. One
of the most signicant properties of amorphous soft magnetic
material is magnetoimpedance (MI). Such an effect involves a very
large and sensitive change of the impedance of some soft magnetic
materials under the action of a static magnetic eld and a small
alternating current. The impedance of the magnetic material is
very sensitive to a small change in the magnetic eld. Fig. 6b shows
the magnetoimpedance variation of this material with respect to
the magnetic eld. A large change in magneto impedance can be

observed under the action of static magnetic eld. Fig. 6c shows

the variation of magnetoimpedance with respect to the frequency.
At low frequency the variation of magnetoimpedance is around 5
10. The Fe60Ni10V10B20 alloy was prepared by arc melting the high
pure constituent elements in argon atmosphere. The ingot was
melted several times to obtain a homogenized mixture. The weight
loss after the melting was less than 0.5%. Refs. [33,34] explains in
more detail the process used here for the fabrication of the amorphous ribbons by using melt spinning method. The ribbons that
were obtained are 3040 lm thick, 12 mm wide and approximately 2 m long. It is very easy to make the soft magnetic material
with less cost. It has been used as the magnetic ux concentrator to
increase the ux lines from the bias magnet to the active part of
the EMAT coils. A layer of ribbons were placed between the pickup coil and the permanent magnet of the EMAT.
4. Results and discussion
Experiments were conducted to evaluate the inuence of the
magnetic ribbon (Fe60Ni10V10B20) as a MFC for a meander coil
EMAT. The measurements were performed using EMATs, with
MFC and without MFC in a through transmission (pitchcatch)
method. A tone burst RF signal with a peak current value of
approximately 50 A and a 6 ls pulse duration was applied to the
excitation coil. Different ultrasonic modes were observed as described below.
4.1. Generation of Lamb waves
First the Lamb waves generation have been done on 2 mm thick
aluminum and stainless steel plates at 500 kHz. Fig. 7 shows the
measurement set up for Lamb wave generation for both experiment and simulation. At 500 kHz, the EMAT transmitter generates
the fundamental Lamb wave modes (S0 and A0) and the receiver
EMAT was placed 100 mm away from the source EMAT.
In both the simulation and experimental outputs, it was observed that the peak-to-peak value of amplitudes of the Lamb
wave modes (S0 and A0) showed a 100% increase in the case where
MFC was used when compared to the case where MFC was not
used. The results are quantied in Table 1. Fig. 8a and b shows
the comparison of experimental and simulation outputs of Lamb
wave signals with and without MFCs on the 2 mm thick aluminum
and stainless steel plates.
4.2. Generation of Rayleigh waves
Fig. 9 shows the measurement setup for generating Rayleigh
waves using meander coil EMAT with MFC and without MFC for
both simulation and experiment. A 50 mm thick aluminum and a

Ritec Pulser Receiver

RPR-4000

EMAT
Transmitter

Oscilloscope

EMAT
Receiver

Test Specimen

(a)

(b)

Fig. 5. (a) Meander line EMAT (b) Schematic diagram of the experimental set-up.

679

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

(a)

(b)
35
30

Z/Z%

25
20
15
10
5
0

10

15

Frequency (MHz)

(c)
Fig. 6. (a) Photograph of EMAT coil and the magnetic ribbon and variation of magneto impedance with (b) magnetic eld and (c) frequency.

Fig. 7. Measurement set-up for Lamb waves generation on 2 mm thin plate.

Table 1
Comparison of the amplitudes of different wave modes in different materials with and without MFC.
Material

Method

Condition with MFC

R-wave
(1 MHz)

Lamb wave (S0 mode)

(500 kHz)

Lamb wave (A0 mode)

(500 kHz)

Aluminium

Experimental output Amplitude

(mV)

Without MFC
With MFC
Amplitude ratio
(AR)
Without MFC
With MFC
Amplitude ratio
(AR)

2.92
6.27
2.15

2.91
5.93
2.04

2.76
5.65
2.05

2.24e16
4.52e16
2.02

3.2e17
6.42e17
2.01

1.81e16
3.64e16
2.01

Without MFC
With MFC
Amplitude ratio
(AR)
Without MFC
With MFC
Amplitude ratio
(AR)

1.12
2.37
2.12

0.76
1.48
1.95

1.69
3.62
2.14

8.32e17
1.66e16
2.00

1.08e17
2.18e17
2.02

7.04e17
1.44e16
2.04

Simulation output Displacement (m)

Stainless
steel

Experimental output Amplitude

(mV)

Simulation output Displacement (m)

30 mm thick stainless steel blocks are used for this measurement.

The transmitter EMAT was located on the middle of the at surface
of the specimen and the receiver was placed 100 mm away from
the centre of the transmitter, which was oriented along the axis

of the wave front generated by the source EMAT. The Rayleigh

wave generation was examined at the excitation frequency of
1 MHz, Fig. 10a and b shows the comparison of Experimental and
simulation results of Rayleigh waves on two different materials

680

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

Experimental output

Simulation output

(a) Aluminum

(b) Stainless steel

Fig. 8. Comparison of time histories of simulation and experimental results of Lamb wave modes (a) on an aluminum plate with and without out MFC and (b) on a stainless
steel plate with and without MFC.

Fig. 9. Schematic diagram for Rayleigh wave measurement set up.

Experimental output
(a) Aluminum

with MFC and without MFC. It has been observed that the amplitudes of Rayleigh wave signals have been increased twice for EMAT
with MFC and is quantied in Table 1.
Table 1 summaries the performance of meander coil EMAT with
MFC and without MFC for both experimental and simulation results. The various sound modes have been generated on different
material for a wide range of frequencies and the signal amplitudes
have been measured for both bulk and guided wave modes. The
increasing signal amplitudes with MFC indicate that the soft magnetic material (Fe60Ni10V10B20) acts as a ux concentrator and increases the normal bias magnetic eld. To quantify the

Simulation output

(b) Stainless steel

Fig. 10. Comparison of time histories of simulation and experimental results of Rayleigh waves (a) on an aluminum block with and without MFC and (b) on a stainless steel
block with and without MFC.

681

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

performance of MFC, the signal amplitude (peak to peak) of different wave modes were calculated and compared with MFC and
without MFC. For example, comparing the peak signal amplitudes
of Rayleigh (R-waves) waves on aluminum plate with MFC and
without MFC

Amplitude ratioAR

PeakAmplitudewithMFC
6:27 mV

PeakAmplitudewithoutMFC 2:92 mV

 2:15:

In nonmagnetic materials, the mechanisms for both generation

and detection of elastic waves with EMATs are linear and obey
standard reciprocity relations. There are excellent treatments of
these in the literature and little would be served by reproducing
them here [6,35]. For a totally isotropic elastic body described by
elastic constant C and mass density qM, the wave equation is

u
u ~
@ 2~
@ 2~

C
Fx
@x2
@t 2

where F(x) is the Lorentz force driving force per unit volume.
Therefore,

~
Bs y
Fx ~
J e z  ~

10

The induced current density can be obtained by solving

p the
Maxwells equations. This current density, given by j 1,



~0
1  j~
nH
z1  j
~
exp
J e z
dc
dc

11

and results from the magnetic eld H0 produced by the EMAT RF

coil in close proximity to the surface. In Eq. (10), dc = (2q/l0x)1/2,
q is the metal resistivity, l0 = 4p  107 H/m, and x is the angular
frequency [37,38].
When an elastic wave is reected from a metal surface, it also
generates a current density and electric eld inside the metal
and, of course, a corresponding electric and magnetic eld outside
the surface. The magnetic eld (B) is sensed by the received coil. It
can be shown that the surface electric eld is given by ~
v  ~B. Consequently neglecting some complexities due to a nite skin depth,
the receiver coil output voltage is given by [36]

~
x~
u ~
Bc  dl

12

where integration is over the entire length of wire in the receiver

coil. only that wire is very close to the metal surface is effective
in receiving the elastic wave. When the magnetic eld is perpendicular to the surface, the displacement is given by [6,7]

~
u

BR BT NR NT AR AT 1=2
ZA

14

where BR, BT, NR, NT, AR and AT are the magnetic elds, number of
turns and effective areas of coils of the transmitter and receiver
EMATs respectively. The parameter bM is dened by

bM

5f q

15

v2

where f is the operating frequency. If the transmitter and receiver

are identical, the Eq. (13) becomes

4.3. Transfer impedance ratio (TIR)

qM

ZT

~0 B  j 
H
xZ A 1  jbM

13

where bM is the analytical function of thickness and ZA is the acoustic impedance of the material. In the limit of small bM, the transfer
impedance for free surface boundary conditions is given by,

ZT

B2 N 2 A
ZA

16

So, the efciency of an EMAT transmitter and receiver system can

be measured by the Transfer impedance ratio ZT (TIR) which is
the ratio of the received voltage by the receiver to the input driving
current. For an ideal case without acoustic attenuation and the effect of coil lift-off, the transfer impedance ratio ZT (TIR) of an EMAT
can be expressed as [12],

Z T TIR

V Received N2 B2 A 4pG=D

e
ZA
IDriving

17

where N is the number of turns in each coil (three turns) and A is the
area covered by the coil (9  30 mm). The magnetic ux density B is
0.3 T for EMAT without MFC and 0.41 T for EMAT with MFC and G is
the magnetic lift-off (1 mm). D is the coil spacing (1.5 mm) and IDriving is the input driving current (50 ).
To verify the performance of the MFC, the transfer impedance
ratio ZT (TIR) of the meander coil EMAT with MFC and without
MFC have been calculated for both the bulk wave on aluminum
and stainless steel samples. Table 2 summarizes the calculated
physical parameters such as the transfer impedance ratio ZT (TIR)
and the received voltage (VReceived) for different wave modes on different materials with and without MFC. and guided waves and
summarized in Table 2. The magnetic ux density of the EMAT
without MFC is 0.3 T and the EMAT with MFC is 0.41 T.
The actual improvement for sole generation (or detection) of
EMAT with MFC is 1.3, and only when two EMAT transducers
(through transmission method) are used, or pulseecho test is employed, then the amplitude increment will be a factor of two.
Moreover these kind of improvement occurs only in non-ferromagnetic materials.
From the transfer impedance ratio ZT (TIR) calculations, it has
been observed that the peak signal amplitudes of all the received
wave modes with MFC has been increased almost by a factor of
two for both aluminum and stainless steel. Hence by using the soft
magnetic alloy ribbon (Fe60Ni10V10B20) as a MFC with EMAT improves the amplitudes of the wave modes. The magnetic material
improves the signal strength and it can be used for defect detection
to increase the sensitivity of the defect.

Table 2
Comparison of the transfer impedance ratio ZT (TIR) and received voltage for different wave modes in different materials with and without MFC.
Material

Physical parameter

Condition with MFC

R-wave (1 MHz)

Lamb wave (S0 mode)

(500 kHz)

Lamb wave (A0 mode)

(500 kHz)

Aluminum

ZT (TIR) (X)

Without MFC
With MFC
Without MFC
With MFC

62.37e8
119.35e8
31.18e6
59.68e6

33.41e8
63.93e8
16.71e6
31.97e6

58.47e8
111.89e8
29.24e6
55.95e6

Without MFC
With MFC
Without MFC
With MFC

22.33e8
42.74e8
11.17e6
21.37e6

12.46e8
23.84e8
6.24e6
11.92e6

21.23e8
40.63e8
10.62e6
20.32e6

VReceived (V)
Stainless steel

ZT (TIR) (X)
VReceived (V)

682

R. Dhayalan et al. / Ultrasonics 51 (2011) 675682

5. Summary and conclusions

In this paper a novel method for improving the EMAT signal
amplitude has been presented by introducing a soft magnetic
material (Fe60Ni10V10B20) as a magnetic ux concentrator (MFC).
This soft magnetic material (MFC) has been used along with the
meander coil EMAT to increase the bias magnetic eld. The
improvements of the magnetic eld with MFC have been calculated by numerical simulation models and validated by experimental observations. The permanent magnet and the magnet with MFC
have been modeled by using the commercial nite element software COMSOL (FEMLAB 3.2). The magnetic ux density of the
magnet with MFC increased from 0.3 T to 0.41 T which is applied
to the active region of the EMAT coil. A coupled 2D numerical model has been designed for the meander coil EMAT which is working
under the principle of Lorentz force mechanism. For this rst an
electromagnetic model has been developed for calculating the Lorentz force density. The force density calculation has been done for
a wide range of frequency (500 kHz1 MHz) for different sound
wave generation. Then, the elastodynamic model has been developed for wave propagation and the calculated force densities have
been utilized as the sources for the ultrasonic wave propagation
modeling.
A set of meander coil EMATs have been developed by using
3 mm periodicity copper coils and the permanent magnets (Nd
FeB) of 0.3 T ux density for the experimental studies. The experimental measurements have been done for various sound modes
on two different materials for wide range of frequency. It has been
observed that the strength of the received signals have been increased with MFC for all the wave modes. The magnetic material
acts as a ux concentrator which allows more ux lines for the active part of the meander coil. The improvement of the signal amplitudes have been quantied by comparing the signal amplitudes
(peak to peak) of the wave modes with MFC and without MFC. It
has been observed that the peak signal amplitudes of all the received wave modes with MFC have been increased almost by a factor of two. The numerical simulation results show a good
agreement with the experimental measurements. The performance
of the EMAT with MFC is also veried by calculating the Transfer
impedance ratio ZT (TIR) for the shear wave generation on aluminum and stainless steel samples. From the Transfer impedance ratio ZT (TIR) calculation, it has been shown that the received signal
amplitude is increased twice by the new magnetic eld.
For the future, we noted that the similar magnetic ribbon (Fe60Ni10V10B20) with different composition of materials may improve
the signal amplitude signicantly.
Acknowledgement
The authors of this paper would like to acknowledge Dr. G.
Markandeyulu, Department of Physics and the Magnetics laboratory of the Indian Institute of Technology Madras for providing
the experimental facilities for this work.
References
[1] B.W. Maxeld, C.M. Fortunko, The design and use of electromagnetic acoustic
wave transducers (EMATs), Mater. Eval. 41 (1983) 13991408.
[2] M.G. Silk, Electromagnetic Acoustic Transducers, Ultrasonic Transducers for
Nondestructive Testing, Adam Hilger Ltd., Bristol, 1984 (pp. 111119).
[3] J.R. Houck, H.V. Bohm, B.W. Maxeld, J.W. Wilkins, Direct electromagnetic
generation of acoustic waves, Phys. Rev. Lett. 19 (1967) 224227.
[4] X. Jian, S. Dixon, S.R. Edwards, Ultrasonic generation and optimization for
EMAT, Rev. Prog. QNDE 24A (2004) 10411046.

[5] M. Hirao, H. Ogi, EMATs for Science and Industry Noncontacting Ultrasonic
measurements, Kluwer Academic Publishers, 2003.
[6] M.R. Gaerttner, W.D. Wallace, B.W. Maxeld, Experiments relating to the
theory of magnetic direct generation of ultrasound in metals, Phys. Rev. 184
(1969) 702.
[7] B.W. Maxeld, Electromagnetic Generation and Defection of Acoustic Waves,
Cornel1 Materials Science Center Report No. 1573, June 1971.
[8] B.W. Maxeld, A. Kuramoto, J.K. Hulbert, Evaluating EMAT designs for selected
applications, Mater. Eval. 45 (10) (1987) 11661183.
[9] H. Ogi, Field dependence of coupling efciency between electromagnetic eld
and ultrasonic bulk waves, J. Appl. Phys. 82 (1997) 39403949.
[10] R.B. Thompson, New EMAT conguration for generating SH-waves in
ferromagnetic materials, IEEE Trans. Son. Ultrason. 26 (2) (1979) 149.
[11] M. Hirao, H. Ogi, An SH-wave EMAT technique for gas pipeline inspection,
NDT&E Int. 32 (3) (1999) 127132.
[12] G.A. Alers, L.R. Bums, EMAT designs for special applications, Mater. Eval. 45
(1987) 11841189.
[13] H.M. Frost, Electromagnetic ultrasound transducers: principles, practice, and
applications, in: R.N. Thurston, A.D. Pierce (Eds.), Physical Acoustics, vol. 14,
Academic Press, San Diego, 1979.
[14] X. Jian, S. Dixon, S.R. Edwards, Optimal ultrasonic wave generation of EMAT for
NDE, Non-Destruct. Eval. (2005) 4262.
[15] J. Green, E. Robert, Non-contact ultrasonic techniques, Ultrasonics 42 (2004)
916.
[16] V.M. Ristic, Principles of Acoustic Devices, John Wiley and Sons, Toronto, 1983.
[17] S. Boonsang, R.J. Dewhurst, A sensitive electromagnetic acoustic transducer for
picometer-scale ultrasonic displacement measurements, Sensor Actuator A:
Phys. 127 (2006) 345354.
[18] V. Satya Narayana Murthy, G. Markandeyulu, Magnetoimpedance studies on
Fe60 Ni10 V10 B20 ribbons, in: Proc. International Symposium for Research
Scholars on Metallurgy Materials Science and Engineering (ISRS), Chennai,
India, 2006.
[19] S.O. Volchkov, V.A. Lukshina, A.P. Potapov, G.V. Kurlyandskaya, A sensitive
element based on the giant magnetoimpedance effect for detecting stray
elds, Russ. J. Nondestruct. Test. 45 (9) (2009) 595603.
[20] K. Mohri, K. Kawashima, T. Kohzawa, Y. Yoshida, L.V. Panina, MagnetoInductive effect (MI effect) in amorphous wires, IEEE Trans. Magn. 28 (5)
(1992).
[21] M. Vkquez, G.V. Kurlyandskaya, J.M. Garcia-Beneytez, J.P. Sinnecker, Frequency
dependence of the magnetoimpedance FeCuNbSiB with high transverse stressinduced magnetic anisotropy, IEEE Trans. Magn. 35 (5) (1999).
[22] M. Yamaguchi, Y. Miyazawa, K. Kaminishi, H. Kikuchi, S. Yabukami, K.I. Arai, T.
Suzuki, Soft magnetic applications in the RF range, J. Magn. Magn. Mater. 268
(2004) 170177.
[23] R. Dhayalan, K. Balasubramaniam, A hybrid nite element model for
simulation of Electro-magnetic Acoustic transducer (EMAT) based plate
waves, NDT&E Int. 43 (2010) 519526.
[24] R. Dhayalan, K. Balasubramaniam, A two stage nite element model for a
Meander
coil
Electromagnetic
Acoustic
Transducer
Transmitter,
Nondestructive Testing and Evaluation (Taylor and Francis), 2010, in press.
[25] J.D. Jackson, Classical Electrodynamics, Wiley, New York, 1963.
[26] J.D. Grifth, Introduction to Electrodynamics, Prentice-Hall Inc., 1999.
[27] COMSOL, Users Guide and Introduction, Version 3.2 byCOMSOL AB 2006,
<http://www.comsol.com>.
[28] P.P. Silvester, R.L. Ferrari, Finite Elements for Electrical Engineers, Cambridge
University Press, Cambridge, UK, 1983.
[29] H. Ogi, M. Hirao, Line-focusing of ultrasonic SV wave by electromagnetic
acoustic transducer, J. Acoust. Soc. Am. 103 (5) (1998). Pt.1, May.
[30] H. Ogi, M. Hirao, T. Ohtani, Line-focusing electromagnetic acoustic transducers
for the detection of slit defects, IEEE Trans. Ultason., Ferroelec. Freq. Contr. 46
(2) (1999).
[31] W. Klement Jun, R.H. Willens, P. Duwez, Non-crystalline structure in solidied
goldsilicon alloys, Nature 187 (1960) 869.
[32] P. Duwez, R.H. Willens, W. Klement Jr., Continuous series of metastable solid
solutions in silvercopper alloys, J. Appl. Phys. 31 (1960) 1136.
[33] V. Satya Narayana Murthy, G. Markandeyulu, Effect of annealing on
magnetoimpedance of Co68 Fe4 Zr10 Cu2 B16 ribbons, J. Magn. Magn. Mater.
303 (2006) 178.
[34] Emile Josso, Iron-cobalt-vanadium alloys: a critical study of the phase
diagrams in relation to magnetic properties, IEEE. Trans. Magn. 10 (1974) 161.
[35] K. Kawashima, Quantitative calculation and measurement of longitudinal and
transverse ultrasonic pulses in solids, IEEE Trans. Son. Ultrason. Su-31 (2)
(1984) 83.
[36] R.J. Seran, D. Fryberger, M. Epstein, Measurement of strain and velocity of
ultrasonic surface waves, Rev. Sci. Instrum. 38 (1967) 1395.
[37] R.B. Thompson, A model for the electromagnetic generation of ultrasonic
guided waves in ferromagnetic metal polycrystals, IEEE Trans. Son. Ultrason.
SU-25 (1) (1978) 715.
[38] K. Kawashima, Electromagnetic acoustic wave source measurement and
calculation of vertical and horizontal displacements of surface waves, IEEE
Trans. Son. Ultrason. SU-32 (4) (1985) 514522.