Austenitic welds are extensively used in nuclear, petrochemical and process industries. Due to the strong material anisotropy and coarse grain size in the dendritic weld zone, they are difficult to inspect with ultrasound. In this regard, the shear horizontal (SH) wave mode is far superior to the more conventional shear vertical (SV) and longitudinal wave modes. In this paper, an electromagnetic acoustic transducer (EMAT) is designed and used for the inspection of two austenitic weld samples. Despite the low efficiency of EMAT generation due to low conductivity of austenitic stainless steel material and strong attenuation in the weld zone, good signal to noise ratio is achieved with optimized EMAT probes and state-of-the-art instrumentation. The angle beam EMAT probe successfully detected all defects in the samples with good signal to noise ratio including a 2% defect. The capability of detection a defect across a 2’’ inch thick and 2’’ wide austenitic weld zone is also demonstrated in the paper.

1. INSPECTION OF AUSTENITIC WELD WITH EMATS
H. Gao, S. M. Ali, and B. Lopez
Innerspec Technologies, Inc. Lynchburg, VA 24501
ABSTRACT. Austenitic welds are extensively used in nuclear, petrochemical and process industries.
Due to the strong material anisotropy and coarse grain size in the dendritic weld zone, they are
difficult to inspect with ultrasound. In this regard, the shear horizontal (SH) wave mode is far
superior to the more conventional shear vertical (SV) and longitudinal wave modes. In this paper, an
electromagnetic acoustic transducer (EMAT) is designed and used for the inspection of two austenitic
weld samples. Despite the low efficiency of EMAT generation due to low conductivity of austenitic
stainless steel material and strong attenuation in the weld zone, good signal to noise ratio is achieved
with optimized EMAT probes and state-of-the-art instrumentation. The angle beam EMAT probe
successfully detected all defects in the samples with good signal to noise ratio including a 2% defect.
The capability of detection a defect across a 2’’ inch thick and 2’’ wide austenitic weld zone is also
demonstrated in the paper.
Keywords: Ultrasonic, Austenitic Weld, EMAT
PACS: 43.35.Cg, 43.38.Dv
INTRODUCTION
Austenitic stainless steel is widely used due to its superior resistance to corrosion. There
are many types of austenitic weld joints used in the nuclear and petrochemical industries as
well as other high-temperature processing piping environments. An austenitic weld can be
a weld joining two stainless steel plates, a dissimilar metal weld between stainless steel and
regular carbon steel, or a weld joining stainless steel and INCONEL, etc. Ultrasonic non-
destructive evaluation of structures with austenitic welds has been a difficult task for more
than three decades. This is primarily due to significant wave skew and scattering from
dentritic shaped coarse grain structure in the austenitic welds [1,2]. Shear vertical wave
commonly used in ultrasonic testing suffers most from the skew effect due to the
anisotropy of austenitic crystal structures. Longitudinal wave experiences mode conversion
at structural and weld boundaries. In order to avoid mode conversion problems and
facilitate interpretation, access to both sides of weld is needed when longitudinal wave is
used to inspect austenitic welds. Shear horizontal does not show mode conversion at
structure boundaries and also has much smaller skew effect, so it has been recognized as
the most promising technique for inspection of austenitic welds since the 1980s’.

2. Ultrasonic waves can be generated in a conductive material or magnetic material
using electromagnetic acoustic transducer (EMATs) via Lorentz force or magnetostriction
[3, 4]. Since the sound is generated in the test material, EMAT transducers can be operated
without contact and without couplants making EMAT especially suitable for high speed
inspection applications. EMATs can be also used to generate almost all types of ultrasonic
bulk wave and guided wave modes including SH waves, which are difficult to generate
with piezoelectric transducers. Although EMAT excitation of SH waves can be very
efficient in many engineering materials such as steel, aluminum and copper, austenitic
stainless steel has low conductivity and low or no magnetism which makes SH wave
excitation with EMAT very challenging. As a result, although some research has been done
in the past [5], there is not practical equipment available, and it is not an accepted practice
for thick austenitic weld inspection.
The purpose of our work is to develop an EMAT SH wave system that can be used in the
field for austenitic weld inspection. In the last 3 years, Innerspec Technologies has
developed a new line of EMAT instruments under the temate®
PowerBox name that
combines extremely powerful tone-burst generators capable of providing up to 20KW,
efficient signal amplifiers, and advanced digital signal processing techniques designed for
the most demanding applications. Using this instrumentation, we have been able to
overcome the challenge of weak signal strength and obtain good signal to noise ratio in
austenitic samples. In this paper, we presented our development of an angle beam SH wave
EMAT using a permanent magnet array and some inspection results on two austenitic weld
samples.
SH EMAT PROBE VALIDATION ON AUSTENITIC STAINLESS STEEL
Principle of Single Channel EMAT Beam Steering
Among all the possible EMAT configurations, the permanent magnet array shown in
Figure 1 is used to generate angle beam SH waves in austenitic material using Lorentz
Force. The alternating magnetic poles create a periodic pattern of Lorentz forces on the
material surface. The waves generated from all the magnet poles constructively interfere
with each other at angle  , when
fD
v
D


)sin( (1)
Since  is dependent on the excitation frequency ( f ), the SH wave beam can be steered
to different directions using a specified frequency.
FIGURE 1. A sketch of the principle of SH wave EMAT beam steering using permanent magnet array.

3. A numerical program is developed to simulate the transmission beam profile from the
EMAT using distributed point source method. This provides more details of the beam
profile than Equation 1. Figure 2 shows four situations of beam steering when the period
of the magnet pole is 0.25 inch. Figure 2(a), (b) are for frequency equal to 750kHz and 650
kHz. The incident angle reduces while the frequency increases. However, there is a
limitation of the minimum beam steering angle, below which a grating lobe at larger angle
becomes stronger than the desired angle. The angle for the first grating lobe happens when
fD
v
D
33
)sin( 1 

 . (2)
Therefore, the minimum practical incident angle without significant grating lobe is when
D
v
f
3
 , and o
20 . Figure 2(c) shows an example of beam profile with side lobe
when frequency equals to 1600kHz.
The lower critical frequency for the beam steering happens when o
90 , at which
D
v
f  . If the frequency is lower than that, no major lobe occurs in the field of interest.
Figure 2 (d) shows the subsurface SH wave beam profile when the frequency is 500kHz.
(a) (b)
(c) (d)
FIGURE 2. Examples of beam steering angle and beam profiles, (a) 750 kHz , (b) 650 kHz, (c) 1600kHz, (d)
500 kHz.

4. Beam Steering Verification
The performance of beam steering of the angle beam SH EMAT was tested on a 1.5 inch
thick austenitic stainless steel sample. The test was carried out by fixing the distance
between the transmitter and the receiver probes at 3inch, 3.5inch 4 inch, 4.5 inch, and 5
inch, and sweeping the excitation frequency through the range of 500kHz and 800kHz.
The results shown in Figure 3 indicates that when the distance increases, a smaller
frequency is needed to achieve the full hop from the transmitter to the bottom surface and
then to the receiver. Two peaks are shown when the separation is 5 inch, because a double
hop signal is detected at high frequency. The frequency that produces maximum amplitude
is recorded, and the beam steering angle is calculated by the equations of trigonometry.
Figure 4 shows a good match between steering angle form the theoretical calculation and
the experiment.
FIGURE 3. Results of frequency sweep when the transmitter and the receiver are separated at several
distances.
FIGURE 4. A comparison between theoretical beam steering angle and experimental result.

5. DEFECT DETECTION IN AUSTENITIC WELDS
Figure 5 shows a picture and a sketch of a pseudo pulse echo inspection scheme. The wave
path forms a full skip from the transmitter to the receiver. Reflection signals from defects
are then detected by the receiver as shown in Figure 5 (b) with the dashed line. By
adjusting the excitation frequency and distance from the transmitter and receiver, optimal
reflection signal can be detected at the receiver. The pseudo pulse echo technique works
similar to the situation of a pulse echo with a delay line. Reflection signals close to the
transducer can be isolated from the main bang signal.
Figure 6 shows the test result on a 1.5 inch thick sample with six notch defects. Three of
them are in the weld zone and three of them are in the heat affected zone (HAZ).
Inspecting from side 1, the three defects on the bottom are in the same side as the
transducers; inspection from side 2 the three defects on the bottom are in the opposite side
of the transducers. The area of interest in the stacked A-scan image is enclosed with a
rectangle. All six defects are detected either when the transducers are on the same 1 or
side 2. The location of the defect can also be indicated by the arrival time of the reflection
signal. When the defect is from the opposite side of the weld, the reflected signal travels a
longer distance. Therefore, the reflected wave package is later for the lower three defects.
(a) (b)
FIGURE 5. Concept of a pseudo pulse echo test scheme (a) picture (b) sketch.
(a) Inspection from side 1 (b) Inspection from side 2
FIGURE 6. Inspecting results of the sample with six notch defects.

6. Figure 7 shows the test result on a 2 inch thick sample with 6 thermal fatigue type defects
all in the HAZ on one side of the weld. Figure 7(a) is a sketch of the defect profiles, in
which defect B and D are only 10% and 2% deep respectively. These defects also have
different orientation with respect to the surface. Figure 7 (b) and (c) are the stacked A-scan
image obtained by scanning the transducers from side 1 and side 2 respectively. Although
the reflection amplitude varies due to defect geometry and orientation, all six defects
produce measurable indication on both sides.
(a) Sketch of defect profile
(b) Inspection from side 1 (c) Inspection from side 2
FIGURE 7. Inspecting results of the sample with six thermal fatigue defects.
CONCLUSIONS AND DISCUSSION
Thick austenitic stainless steel weld samples were tested with an SH wave EMAT probe. Using
a temate®
PowerBox 2 instrument the one channel EMAT is capable of detecting defects as
small as 2% in depth of a two inch thick sample. The capability of detecting defects from the
opposite side of the weld makes it possible to use this technology where there is limited
accessibility.
The results presented in this paper are from a linear scan along the weld direction from a one
channel system. Therefore, the reflection signal amplitude is also affected by the orientation
and location of the defects. In order to achieve quantitative defect sizing, the scanning results
perpendicular to the weld direction or a sector scan by beam steering is needed. A portable
phased-array system using the same technology will be launched at the end of 2009 to provide
better resolution and defect sizing capability.

7. REFERENCES
1. R.J. Huggell and B.S. Gray, “The ultrasonic inspection of austenitic materials-
state of the art report”, 1985.
2. K. J. Langenberg, et.al, “Application of modeling techniques for ultrasonic
austenitic weld inspection”, NDT&E International, Vol. 33, 465-480, (2000).
3. M. Hirao and H. Ogi, “EMATs for Science and industry-NonContacting
Ultrasonic Ultrasonic Measurements”, Kluwer Academic Publishers, 2003.
4. H. Gao, S. Ali, B. Lopez, “Delamination Detection in Composite Clad Products
Using Ultrasonic Guided Wave EMATs”, in Review of Progress in Quantitative
Nondestructive Evaluation, Edited by D. O. Thompson and D.E. Chimenti, AIP
Conference Proceedings, vol. 1096, American Institute of Physics, 2009, pp. 1121-
1126.
5. G. Hubschen, H.J. Salzburger, M. Kroning, et.al, “Results and experiences of ISI
of Austenitic and dissimilar welds using SH-waves and EMUS-Probes”, Elsevier
Science Publishers, K. Kussmaul, Editor, 1993.