ASNT NDT Handbook Volume 7 ultrasonic Testing

PART 4. Phased Arrays

Introduction circular array or some more complex form
(see Fig. 37). Most applications use linear
Ultrasonic phased arrays use multiple arrays, because these are the easiest to
ultrasonic elements and electronic time program and are significantly cheaper
delays to generate and receive ultrasound, than more complex arrays because of
creating beams by constructive and fewer elements. As costs decline and
destructive interference. As such, phased experience increases, greater use of the
arrays offer significant technical more complex arrays can be predicted.
advantages over conventional
single-probe ultrasonic testing: the phased The elements are ultrasonically isolated
array beams can be steered, scanned, from each other and packaged in normal
swept and focused electronically. probe housings. The cabling usually
consists of a bundle of well shielded
Electronic scanning permits very rapid micro coaxial cables. Wireless systems
coverage of the components, typically an have increased since 2005. Commercial
order of magnitude faster than a multiple-channel connectors are used
single-probe mechanical system. with the instrument cabling.

Beam forming permits the selected Elements are typically pulsed in groups
beam angles to be optimized from 4 to 32, typically 16 elements for
ultrasonically by orienting them welds. With a user friendly system, the
perpendicular to the discontinuities of computer and software calculate the time
interest — for example, lack of fusion in delays for a setup by using either operator
welds. input on interrogation angle, focal

Beam steering (usually called sectorial FIGURE 37. Array types: (a) one-dimensional linear array of
scanning) can be used for mapping 16 sensors; (b) two-dimensional matrix array of 32 sensors;
components at appropriate angles to (c) sectorial annular array of 61 sensors.
optimize probability of detection.
Sectorial scanning is also useful for (a) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
inspections where only a minimal
footprint is possible. (b) 4 8 12 16 20 24 28 32

Electronic focusing permits optimizing 3 7 11 15 19 23 27 31
the beam shape and size at the expected
discontinuity location, as well as 2 6 10 14 18 22 26 30
optimizing probability of detection.
Focusing improves signal-to-noise ratio 1 5 9 13 17 21 25 29
significantly, which also permits operating
at lower pulser voltages.

Overall, phased arrays optimize
discontinuity detection while minimizing
test time.

Operation (c) 47 46
30 45
Ultrasonic phased arrays are similar in 48 17
principle to phased array radar, sonar and 49 31 8 29
other wave physics applications. However, 50 32 18 28 44
ultrasonic development is behind the
other applications because of a smaller 51 33 19 9 3 16
market, shorter wavelengths, mode 7 15 27 43
conversions and more complex
components. Industrial applications of 2 6 14 26 42
ultrasonic phased arrays have increased in
the twenty-first century.31-36 52 34 20 10 4 5 13 25 41 61

Phased arrays use an array of elements, 53 35 21 11 12 24 40 60
all individually wired, pulsed and time 54 36 22 23
shifted. These elements can be a linear 55 37
array, a two-dimensional matrix array, a 56 39 59
38

58
57

90 Ultrasonic Testing

distance, scan pattern and other test testing, then phased arrays would
circumstances or by using a predefined typically start by using the same
file (see Fig. 38). The time delays are back frequency, aperture size, focal length and
calculated using time-of-flight from the incident angle.
focal spot, and the scan assembled from
individual focal laws. Time delay circuits While phased arrays require well
must be accurate to around 2 ns to developed instrumentation, one of the
provide the phasing accuracy required. key requirements is good, user-friendly
software. Besides calculating the focal
Each element generates a beam when laws, the software saves and displays the
pulsed; these beams constructively and results, so good data manipulation is
destructively interfere to form a wave essential. As phased arrays offer
front. (This interference can be seen, for considerable application flexibility,
example, with photoelastic imaging.37 The software versatility is highly desirable.
phased array instrumentation pulses the Phased array inspections can be manual,
individual channels with time delays as semiautomated (that is, encoded) or fully
specified to form a pre-calculated wave automated, depending on the application,
front. For receiving, the instrumentation speed, budget and other considerations.
effectively performs the reverse, i.e. it Encoder capability and full data storage
receives with precalculated time delays, are usually required.
then sums the time shifted signal and
displays it. This is shown in Fig. 39. Although it can be time consuming to
prepare the first setup, the information is
The summed waveform is effectively recorded in a file and only takes seconds
identical to a single-channel discontinuity
detector using a probe with the same FIGURE 39. Beam: (a) emitting; (b) receiving.
angle, frequency, focusing, aperture and
other settings. Figure 39 shows typical (a) Pulses Incident
time delays for a focused normal beam Sensors
and transverse wave. Sample scan patterns Trigger wave front
are shown in Fig. 40 and are discussed Acquisition
below. Phased
unit array
Implementation unit
Discontinuity
From a practical viewpoint, ultrasonic
phased arrays are merely a means of (b) Echo signals Reflected
generating and receiving ultrasound; once Sensors wave front
the ultrasound is in the material, it is Acquisition
independent of generation method, unit Phased Discontinuity
whether generated by piezoelectric, array
electromagnetic, laser or phased arrays. unit
Consequently, many of the details of
ultrasonic testing remain unchanged; for
example, if 5 MHz is the optimum test
frequency with conventional ultrasonic

FIGURE 38. Generation of scans using phased arrays: (a) linear focusing; (b) sectorial focusing;
(c) depth focusing.

(a) (b) (c)

Delay Focal law 1
(relative scale)
Focal law 5
Acoustic field 1 Acoustic field 5 αN Linear array
αl Beam spot N
Beam spot 1

Generation and Detection of Ultrasound 91

to reload. Also, modifying a prepared elements but alter the time delays to
setup is quick in comparison with sweep the beam through a series of angles
physically adjusting conventional probes. (see Fig. 42). Again, this is a
straightforward scan to program.
Scan Types Applications for sectorial scanning
typically involve a stationary array,
Electronic pulsing and receiving provide sweeping across a relatively inaccessible
significant opportunities for a variety of component like a turbine blade root,38 to
scan patterns, as shown in Fig. 40 and map out the features and discontinuities.
below. Depending primarily on the array
frequency and element spacing, the sweep
Electronic Scans angles can vary from ±0.3 rad (±20 deg)
up to ±1.4 rad (±80 deg).
Electronic scans are performed by
multiplexing the same focal law (time Combined Scans
delays) along an array (see Fig. 41).
Typical arrays have up to 128 elements. Combining linear scanning, sectorial
Electronic scanning permits rapid scanning and precision focusing leads to a
coverage with a tight focal spot. If the practical combination of displays (Fig. 43).
array is flat and linear, then the scan Optimum angles can be selected for welds
pattern is a simple B-scan. If the array is and other components whereas electronic
curved, then the scan pattern will be scanning permits fast and functional tests.
curved. Electronic scans are For example, combining linear and
straightforward to program. For example, longitudinal wave sectorial scanning
a phased array can be readily programmed permits full ultrasonic testing of
to perform corrosion mapping, or to test a components over a given angle range,
weld using 0.8 rad (45 deg) and 1 rad such as ±0.3 rad (±20 deg). This type of
(60 deg) transverse waves, which mimics test is useful when simple normal beam
conventional manual inspections. tests are inadequate, such as titanium
castings in aerospace where
Sectorial Scans (S Scans) discontinuities can have random
orientations. A related approach applies to
Sectorial scanning is unique to phased weld inspections, where specific angles are
arrays. Sectorial scans use the same set of often required for weld geometries; for

FIGURE 40. Schematic time delays FIGURE 41. Electronic scanning.
(histograms): (a) focused normal beam;
(b) focused transverse wave. Active group
16
(a)
1
Delay

Sensors

Resulting Scanning direction
wave
FIGURE 42. Sectorial scanning on turbine rotor for sequence
surface of N scans.

(b) 1 2 3N
Scan sequence
Applied
delay

Sensors

Angle steering

Resulting wave surface

92 Ultrasonic Testing

these applications, specific beam angles probe pan is scanned linearly round or
are programmed for specific weld bevel along the weld, while each probe sweeps
angles at specific locations. out a specific area of the weld. The
simplest approach to linear scanning is
Linear Scanning of Welds found in pipe mills, where a limited
number of probes test electric resistance
Manual ultrasonic weld inspections are welded pipe.39
performed using a single probe, which the
operator rasters back and forth to cover Phased arrays for linear weld tests
the weld area. Many automated weld test operate on the same principle as the
systems use a similar approach (see multiprobe approach; however, phased
Fig. 44a), with a single probe scanned back arrays offer considerably greater flexibility
and forth over the weld area. Rastering is than conventional automated ultrasonic
time consuming because the system has testing. Typically, it is much easier to
dead zones at the start and finish of the change the setup electronically, either by
raster. modifying the setup or reloading another;
often it is possible to use many more
In contrast, most multiple-probe beams (equivalent to individual
systems and phased arrays use a linear conventional probes) with phased arrays;
scanning approach (see Fig. 44b). Here the special inspections can be implemented
simply by loading a setup file.
FIGURE 43. Phased array imaging patterns: (a) scanning
pattern using sectorial and linear scanning; (b) image using Applications
all data merged together.
Ultrasonic phased arrays are flexible and
(a) can address many types of problems.
Consequently, they are used in a wide
(b) variety of industries where the technology
has inherent advantages. These industries
FIGURE 44. Scanning: (a) conventional include aerospace, nuclear power, steel
raster; (b) linear. mills, pipe mills, petrochemical plants,
(a) pipeline construction, general
manufacturing and construction, plus a
Welded pipe selection of special applications. All these
applications take advantage of one or
Sensor more of the dominant features of phased
arrays:
(b) Welded pipe
1. Speed — scanning with phased arrays
Scan direction is much faster than single probe
Legend conventional mechanical systems,
with better coverage.
= Data collection step
= Raster step 2. Flexibility — setups can be changed in
a few minutes, and typically a lot
more component dimensional
flexibility is available.

3. Test angles — a wide variety of test
angles can be used, depending on the
requirements and the array.

4. Small footprint — small matrix arrays
can give significantly more flexibility
for testing restricted areas than
conventional probes.

5. Imaging — an image (enhanced to
simulate three dimensions) of
discontinuities is much easier to
interpret than a waveform. The data
can be saved and redisplayed as
needed.

Each feature generates its own
applications. For example, speed is
important for pipe mills and pipelines,
plus some high volume applications.
Flexibility is important in pressure vessels
and pipeline welds due to geometry
changes. Test angle is key for pipelines,
some pressure vessel and nuclear
applications. Small footprint is applicable

Generation and Detection of Ultrasound 93

to some turbine applications. Imaging is
useful for weld tests.

Phased array nondestructive testing is
relatively new and still requires some
setup effort, especially for complex
three-dimensional applications.
Two-dimensional setups are generally
straightforward, provided the software is
user friendly. For example, automated
setup procedures have been developed for
weld tests. Phased array systems have
sometimes been more costly than
single-channel systems; however, the
higher speed, data storage and display,
smaller footprint and greater flexibility
can often offset the higher costs,
especially with the newer portable
instruments.

94 Ultrasonic Testing

PART 5. Focused Beam Immersion Techniques40

Focused Transducers FIGURE 45. Improvement with focused
transducer on curved test surface:
Sound can be focused by lenses in a (a) distorted A-scan image with flat
manner analogous to focusing light. The transducer; (b) elimination of distortion with
basic difference between the two is the focused transducer.
ratio of the lens thickness to the (a)
wavelength. In optics, the lens thickness
is 104 to 105 times the wavelength. In Flat
ultrasonic testing, the lens thickness is transducer
about 10 times the wavelength. As a
result, sound waves with opposite phases Tubing
are emitted from the surface of an
acoustic lens in concentric rings several (b)
millimeters apart and interfere at the focal
plane. Contoured
transducer
Acoustic lenses can improve testing
reliability by reducing and controlling Tubing
certain energy losses. They are usually an
integral part of the transducer assembly.
In most applications, the lens
concentrates the energy into a long and
narrow beam, increasing its intensity.
Special sharp focused transducers can be
made with a usable test range less than
6.4 mm (0.25 in.). Such a focused
transducer can resolve a 0.4 mm
(0.015 in.) flat bottom hole located 1 mm
(0.04 in.) beneath the surface of a steel
block. These transducers are particularly
useful for tests of thin materials, high
resolution C-scan imaging and the
determination of bond quality in
sandwich structures.

Generally, focused transducers allow
the highest possible resolving power with
standard equipment because the front
surface is not in the focal zone and the
concentration of energy at the
discontinuity makes its reflection very
high, providing a ratio up 104 to 1
between the front surface and the
discontinuity echo.

Using a cylindrical lens, an
improvement in resolving power can be
achieved to a lesser degree without
decreasing the horizontal width of the
beam. With this construction, a 75 mm
(3 in.) wide beam can provide clear
resolution of discontinuities from about
2.0 mm (0.08 in.) below the surface to a
depth of 13 mm (0.5 in.). For standard
ultrasonic equipment operating at
10 MHz, such transducers are produced for
tests of thin airframe aluminum
extrusions.

Focused beams reduce the effect of
surface roughness and the effect of
multiple minute discontinuities such as

Generation and Detection of Ultrasound 95

grain boundaries and porosity. In where V is acoustic velocity (meter per
addition, these transducers produce plane second) and ρ is density. The index 1
wave behavior at the focal spot and are values are for the water medium. Index 2
used in appropriate experimental values are for the lens medium. A high
studies.41 index of refraction (V2·V1–1) is required to
allow a small radius of curvature. The
Focused transducers are more energy absorption and scattering must be
commonly used in the immersion as small as possible.
environment where they are not exposed
to wear or erosive conditions. Such It is difficult to find materials that
transducers are also preferable for tests of completely meet lens design
curved surfaces. When an ultrasonic beam requirements. Lenses undergo specific
encounters divergence at a curved surface, and, in some ways, unique fabrication
a flat transducer is used. A focused beam processes, demanding particular material
maintains a circular or cylindrical wave characteristics within limited tolerances.
front and is not distorted by a curved In addition, when a fluid lens is used, the
surface (see Fig. 45). fluid must not be harmful to personnel
and must be inert to the materials it
Ultrasonic Lenses contacts.

Ultrasonic beams are focused using three The reflection coefficient for most host
basic techniques: (1) a curved, ground and lens combinations is significantly
piezoelectric material, (2) a plano concave high unless a fluid or a lithium lens is
lens cemented to a flat piezoelectric used in a fluid host. Note that air has a
crystal and (3) a biconcave lens placed in very low acoustic impedance and cannot
front of the transducer. Curved be present in any form in an ultrasonic
transducers provide a well defined focusing system.
acoustic field with limited noise and
energy losses. The need to fix a damper Biconcave Lenses43
on the back of the crystal (to obtain broad
band characteristics) makes the use of When a lens is immersed in front of a
curved crystals less practical. Lenses glued transducer, the lens must be concave on
to crystals are more common. both surfaces to accommodate beam
divergence from the transducer. The half
Spherical lenses are used in most angle of divergence r is:
nondestructive testing applications. When
a line focus is needed, cylindrical (28) sin θ = 1.22 λ
transducers are used. In some special d
applications, an external lens is attached
in front of a flat transducer to focus its FIGURE 46. Attachment of acoustic lens to
beam and improve test sensitivity. Such transducer: (a) transducer separated from
lenses produce a certain amount of lens induces transverse wave in lens because
disturbance and absorption because some of angular incidence; (b) bonded lens has
materials, such as vulcanized rubber, near zone longer than focal length of lens.
contain fillers.
(a) θ F
Lens Design Criteria42
d D
Lenses can be made of solids or liquids. R1
For most nondestructive test applications,
the lens is a solid and the host or the (b)
conducting medium is fluid. If so, the
acoustic velocity is higher in the lens than F
in the host medium. This produces a R1
converging lens that is concave (the
reverse of optic lenses). Legend

When designing an ultrasonic lens, D = lens diameter
there are several necessary considerations, d = sensor diameter
including acoustic impedance, acoustic F = focal length
velocity, attenuation, fabrication and R = refractive concavity
nonhazardous materials. θ = angle of diffraction

Matching acoustic impedances is
required to minimize energy loss from
reflections in the lens. This condition is
expressed in Eq. 27.

(27) ρ1V1 = ρ2V2

96 Ultrasonic Testing

where d is the diameter (millimeter) of the Focused Beam Profile and
transducer and λ is the ultrasonic Intensity
wavelength (millimeter) in the lens.
At the focal plane, the normalized
As shown in Fig. 46a, the acoustic amplitude distribution is expressed as
beam impinges on the solid lens at an follows:
angle. This angular incidence on the
fluid-to-solid interface of the lens (32) P = 2J1 (kamr)
produces a transverse wave in addition to Pmax kamr
the longitudinal wave. These two waves
have different velocities and generate a where k is 2π·λ–1 (wave number), am is
double focus. arc sin (R·F–1), F is focal length

The radius of curvature R1 of the first
surface of the lens is:

(29) R1 = Dd (millimeter), J1 is the bessel function of
2.44 λ the first kind and first order, R is radius of

the lens aperture (millimeter), r is

where D is the diameter of the lens cylindrical coordinate in the focal plane
aperture in millimeters. The focal length
of a biconcave lens can be determined and λ is wavelength (millimeter);
from the following expressions.
For small angles, where am < 0.5 rad
(30 deg), am = R·F –1. Figure 47 shows the
–1
distribution of P·P max as a function of the
parameter kamr. lobe of the
The first

(30) F = R2 × R1 described distribution also defines the
1− V1
⎛ V1 − ⎞ + − diffraction limit for a given frequency,
V2 d ⎜ 1⎟ R1 R2
⎝ V2 ⎠ namely the smallest focal spot diameter.

This diffraction limit DL is also known as
an airy disk of the first order.

where d is the thickness of the lens FIGURE 47. Normalized acoustic pressure distribution on
focal plane for radius R = 10 mm and R = 8.5 mm:
(millimeter), R1 and R2 are the radii of (a) frequency f = 5 MHz and focal length F = 60 mm;
curvature (millimeter), V1 is the acoustic (b) f = 2.5 MHz and F = 60 mm; (c) f = 2.5 MHz and
velocity in the host (meter per second), V2 F = 40 mm.
is acoustic velocity in the lens (meter per
second) and V1·V2–1 is the index of (a) 1.0
refraction of the lens material relative to

the host medium.

Plano Concave Lenses Amplitude
(relative scale)
When an ultrasonic lens is glued directly
to a transducer (see Fig. 46b), then the Distance (relative scale)
near field is usually longer than the focal (b)
length of the lens and no clear focus can
be produced.44 To determine the focal 1.0
length F for a plano concave lens, Eq. 31
can be used. The equation contains a Distance (relative scale)
correction term for large angles of (c)
aperture.
1.0
⎛ V1 ⎞ Amplitude
d⎜ ⎟ (relative scale)
R ⎝ V2 ⎠
(31) F = 1 − V1 ×
⎛ V2 ⎞
V2 2 ⎜ − 1⎟
⎝ V1 ⎠

where R is the radius of curvature in Amplitude
millimeters. For small apertures, d is (relative scale)
effectively 0. With the increase of the
angle of aperture, the focus moves toward
the lens and the focal area is widened.
This reduces the concentration of energy
and the accuracy of the focal position.

543210 1 234 5
Distance (relative scale)

Generation and Detection of Ultrasound 97

(33) DL = 1.22 λ F add the radius of the ball to the water
r path. This path is measured by a pulse
echo time-of-flight test. The transducer is
where F is the lens focal length moved back and forth with the sphere
(millimeter) and r is the radius of aperture along its axis, until maximum amplitude
of the lens (millimeter). is measured.

The theory is discussed in further detail The focal distance becomes shorter
elsewhere.45 when the ultrasonic beam propagates
from a fluid to a solid material. The
For a 10 MHz wave in water, the reduction of the focal distance can be
wavelength is 0.148 mm. With a lens of determined from a geometrical analysis of
13 mm diameter and 25 mm focus, the the position of the front surface of the
diffraction limit is 0.35 mm. Effectively, material along the beam path (Fig. 48):
this is also the smallest focus that can be
obtained. This limit value can be a ( )(36)Xw + Vw
constraint on the effective use of an FL – Xw Vtm
acoustic lens at low frequencies. At
1 MHz, the diffraction limit becomes R = F
3.5 mm.
where R is the ratio of the focal length
The gain of lenses has been derived as
follows:43 change, Vtm is acoustic velocity in the test
object (meter per second), Vw is acoustic
(34) Gp = πD2 T velocity in water (1485 m·s–1) and Xw is
2λF 2 the one-way water path between the

and: transducer and the front surface of the

solid material (millimeter).

The focal depth inside the solid is:

4Z1Z2
Z1 + Z2 2
( )(35) T = ( )(37) Vw
Xtm = FL – FL – Xw Vtm

where T is the transmission coefficient, Z1 Because of the difference in velocities
is ρ1V1 (acoustic impedance of the host for metal and water, changes of the water
medium) and Z2 is ρ2V2 (acoustic path have a relatively small effect on the
impedance of the lens material). focal depth in the metal. The metal
surface forms a second lens much more
The main disadvantages of acoustic powerful than the acoustical lens itself.
lenses are aberrations and the energy loss This effect pulls the focal spot very close
from reflections and attenuation. Most to the metal surface, compared to the
lenses are made from plastics for a low focal length of a transducer in water.
reflection coefficient. Unfortunately,
plastics are highly attenuative. To reduce FIGURE 48. Ultrasonic focus effect in metals,
attenuation, ultrasonic applications use so demonstrating effect of second lens as result
called zone lenses, the acoustic equivalent of immersion in water.
of fresnel lenses in optics. In a zone lens,
rings are scribed on a plate so that every Focused
second ring is in phase, generating transducer
constructive interference at a
predetermined point. This point is Lens
dependent on the frequency and is
regarded as the focal spot. The rings that Beam
are out of phase are covered to eliminate
their contribution. Zone lenses have
found very limited application in
nondestructive testing.

Focal Distance Water Beam refracted
with greater
The focal distance of a transducer is Metal convergence
measured experimentally with a small ball
target from which its reflection is New point of Divergence
examined. It is assumed that a spherical focus in metal beyond focus
wave front is produced and the surface of
the ball behaves as an equal phase Focal distance
reflector. The focal point is inferred to be if in water
at the geometric center of the ball. When
the ball diameter is larger than the
diffraction limit, then it is common to

98 Ultrasonic Testing

The second lens has three other microscopes are available commercially
important effects: it sharpens the beam, for tests of thin structures such as
increases the sensitivity to objects in the integrated circuits.47 The acoustic
focal zone and makes the transducer act microscope can also be used to study
as a very directional and distance sensitive bond integrity, microstructure formation
receiver. Large increases in sensitivity are and material stress effects.
produced by these complex interactions.
This makes it possible to locate minute In scanning acoustic microscopes
discontinuities and to study areas that (SAM), a piezoelectric crystal is bonded to
produce very low amplitude reflections, a sapphire or quartz substrate to which it
including the bond juncture between transmits plane waves. At the back of the
stainless steel and electroformed copper, substrate is a spherical lens machined
for example. with a low reflection coating. The lens
produces a highly focused beam in water
Pencil Shape Focus allowing the pair — transducer and lens
— to operate as a focused transducer in a
In many ultrasonic applications, a long pulse echo mode. A shallow region close
focused beam is required to test a large to the surface of the test material is
depth range. Two lens configurations examined.
produce such a focused beam,21 as shown
in Fig. 49. The long axial test ranges are The scanning procedure is similar to an
limited mainly by the rate at which signal ultrasonic C-scan and produces an image
amplitude falls off in the far field of the of the test area on the monitor. The
beam. contrast of the image is determined by the
surface reflectivity of the test object and
the phase of the reflected wave.

Acoustic Microscopy Closing

An acoustic microscope uses acoustic Ultrasonic techniques are implemented
waves and a set of one or more lenses to primarily in the pulse echo and
obtain information about the elastic through-transmission modes, with contact
microstructural properties of test objects. or immersion coupling. Pulse echo
The finest detail resolvable by an acoustic immersion is a coupling technique not
microscope is determined by the commonly used in field applications
diffraction limit of the system. Such because of the complexity associated with
maintaining the couplant. Several means
FIGURE 49. Acoustic field where are commercially available to overcome
dF·dN–1 ≅ F·N –1 for: (a) concave lens; the limitations of immersion coupling,
(b) angled concave lens.46 including wheels and boots.

(a) Pulse echo techniques provide very
detailed information about test objects.
dN This capability is attributed to the various
parameters that can be analyzed,
dF including time of flight, amplitude of
back surface reflection and amplitude of
F extraneous reflections. The 1990s saw
Focal length significant improvements in the capability
of the technique with the introduction of
N microprocessor controlled
Near field length pulser/receivers, signal analyzers and
computerized C-scan controllers. Tests
(b) became more reliable; data, easier to
interpret; systems, capable of testing
dN complex object shapes.

dF Pulse echo immersion may be used
with materials made of metal, plastic,
F composites and ceramics — the raw
Focal length materials for most engineering structures.
A wide variety of discontinuities can be
N detected and characterized with pulse
Near field length echo techniques: the location, depth, size
and discontinuity type can be determined.

Generation and Detection of Ultrasound 99

PART 6. Lamb Waves

The theory of lamb waves was originally The phase velocity of the incident
developed by Horace Lamb in 1916 to longitudinal wave is then given by:
describe the characteristics of waves
propagating in plates.48 Frequently, they (38) Vp = VL
are also referred to as plate waves. Lamb sin ϕ
waves can be generated in a plate with
free boundaries with an infinite number where VL is the group velocity of the
of modes for both symmetric and incident longitudinal wave, Vp is the
antisymmetric displacements within the phase velocity of the incident
layer. The symmetric modes are also called longitudinal wave and ϕ is the angle of
longitudinal modes because the average incidence of the incident longitudinal
displacement over the thickness of the wave.
plate or layer is in the longitudinal
direction. The antisymmetric modes are Lamb waves are extremely useful for
observed to exhibit average displacement detection of cracks in thin sheet materials
in the transverse direction and these and tubular products. Extensive
modes are also called flexural modes.49,50 developments in the applications of lamb
The infinite number of modes exists for a waves provides a foundation for the
specific plate thickness and acoustic inspection of many industrial products in
frequency which are identified by their aerospace, pipe and transportation. The
respective phase velocities. Figure 50 generation of lamb waves can be
shows a typical example of generating performed using contact transducers,
lamb waves in a solid plate using an angle optical, electromagnetic, magnetostrictive,
wedge. The normal way to describe the and air coupled transducers.
propagation characteristics is by the use of
dispersion curves based on the plate mode Magnetostrictive transducers operate by
phase velocity as a function of the producing a small change in the physical
product of frequency times thickness. The dimensions of ferromagnetic materials,
dispersion curves are normally labeled as resulting in a deformation of crystalline
S0, A0, S1, A1 and so forth, depending on parameters.51 Applying high frequency
whether the mode is symmetric or power to the transducers then produces
antisymmetric. ultrasonic waves in the material. Lamb
waves are produced when thin or tubular
Although the dispersion diagrams are materials are excited by high frequency
very complex, they can be simplified by oscillations. The technique also works
using the incidence angle of the exciting with nonferromagnetic samples: a
wave to determine which mode is to be ferromagnetic sheet, such as nickel, is
dominant. A particular lamb wave can be bonded to the nonferromagnetic sample
excited if the phase velocity of the being tested. The lamb waves generated in
incident longitudinal wave is equal to this manner can be used to detect cracks
phase velocity for the particular mode. or other material characteristics in areas
away from the excitation source because
FIGURE 50. Lamb wave propagating in plate: the waves propagate along the sample for
(a) symmetric; (b) antisymmetric. long distances. When used this way, the
ultrasonic waves are also called guided
(a) waves. Technology using guided waves
developed into a very useful ultrasonic
(b) test technique in the 1990s and is
expected to continue to develop in the
twenty-first century.52 For example, new
sensor development introduces smaller
transducers, such as capacitive
micromachined devices,53 air coupling54
and new characterization studies on the
lamb wave modes propagating in plate
structures.55

100 Ultrasonic Testing

PART 7. Ultrasonic Guided Waves

Introduction results are actually points that end up on
the phase velocity dispersion curve for the
Ultrasonic guided waves are well structure. Elsewhere is strong cancellation.
documented in the technical To solve for the points on the dispersion
literature.56-62 Compared to ultrasonic curve, either a partial wave summation
bulk waves that travel in infinite media process could account for all reflections
with no boundary influence, guided and mode conversions or an appropriate
waves require a structural boundary for boundary value problem in wave
propagation. As an example, some guided propagation can be solved.
wave possibilities are illustrated in Fig. 51
for a rayleigh surface wave, a lamb wave The use of ultrasonic guided waves
and a stonely wave at an interface increased tremendously after 1990 for
between two materials. There are many various reasons, especially because of
other guided wave possibilities, of course, improved analytic techniques. The
as long as a boundary on either one or principal benefits of guided waves can be
two sides of the wave is considered. summarized as follows.
Natural waveguides include plates (such as
aircraft skin), rods (rails, cylinders, square 1. Inspection over long distances from a
rods), hollow cylinder (pipes, tubing), single probe position is possible,
multilayer structures, curved or flat giving complete volumetric coverage
surfaces on a half space and one or more of the test object.
layers on a half space.
2. There is no need for scanning: all of
Most structures are natural wave guides the data are acquired from a single
provided the wavelengths are large probe position. Often, greater
enough with respect to dimensions in the sensitivity than that obtained in
wave guide. If the wavelengths are very conventional normal beam ultrasonic
small, then bulk wave propagation can be testing or other nondestructive
considered, those waves used traditionally techniques can be obtained, even with
for many years in ultrasonic low frequency ultrasonic guided wave
nondestructive pulse echo and test techniques.
through-transmission testing. One very
interesting difference, of many, associated FIGURE 51. Guided wave types: (a) rayleigh
with guided waves, is that many different (surface) waves; (b) lamb waves; (c) stonely
wave velocity values can be obtained as a waves.
function of frequency whereas, for most
practical bulk wave propagation, wave (a)
velocity is independent of frequency. In
fact, tables of wave velocities are available (b)
from most manufacturers of ultrasonic
equipment — tables applicable to bulk (c)
wave propagation in materials, showing
just a single wave velocity value for
longitudinal waves and one additional
value for transverse waves.

To get some idea of how guided waves
are developed in a wave guide, imagine
many bulk waves bouncing back and
forth inside a wave guide with mode
conversions between longitudinal and
transverse constantly taking place at each
boundary. The resulting superimposed
wave form traveling along the wave guide
is just a sum of all of these waves
including amplitude and phase
information. The outcome can strongly
depend on frequency and introductory
wave angles of propagation inside the
structure. The strongly superimposed

Generation and Detection of Ultrasound 101

3. There is also an ability to inspect an assumed harmonic solution for
hidden structures, structures under displacement, elasticity permits derivation
water, coatings, insulations and of equations to satisfy the boundary
concrete because of the ability to test conditions. This leads to a transcendental
from a single probe position via wave equation, or a characteristic equation. In
structure change and controlled mode extracting the roots from the
sensitivity along with an ability to characteristic equation, associated with a
propagate over long distances. system of homogeneous equations, the
determinant of the coefficient matrix
4. There is also a tremendous cost must be set equal to zero. In this case, the
effectiveness associated with guided roots extracted determine the values of
waves because of the test simplicity phase velocity versus frequency that can
and speed. be plotted. Figure 53 shows the phase
velocity dispersion curves and group
Dispersion velocity dispersion curves for a particular
traction free aluminum plate. The modes
The subject of dispersion and the are labeled as antisymmetric A0, A1 and
propagation of either dispersive or so on, or symmetric S0, S1 and so on. The
nondispersive modes is very critical to particular limits in the diagram as plate
understand when dealing with ultrasonic velocity, surface wave velocity, transverse
guided waves. Figure 52 shows an wave velocity and cutoff frequencies are
example of dispersive and nondispersive all shown in the figure. Details on the
guided wave propagation. For development and the nomenclature
nondispersive wave propagation, the considered here can be found in
pulse duration remains constant as the references.15 Derivable from the phase
wave travels through the structure. On velocity dispersion curves are sets of
the other hand, for dispersive wave group velocity dispersion curves. The
propagation, because wave velocity is a values of the group velocity dispersion
function of frequency, the pulse duration curves depend on the ordinate and slope
changes from point to point inside the values of the phase velocity dispersion
structure. The change is because each curves. The group velocity is defined as
harmonic of the particular input pulse the velocity measured in a wave guide of
packet travels at a different wave velocity. a packet of waves of similar frequency.
There’s a decrease in amplitude of the This group velocity is what you actually
waveform and an increase in pulse measure in an experiment.
duration but energy is still conserved —
unless of course lossy media are If an aluminum plate is under water,
considered. there will be energy leakage as the wave
travels along the plate, because of an
Consider the development of a phase out-of-plane displacement component
velocity computation in a wave guide, that would load the liquid. The in-plane
such as a plate having boundary displacement components would not
conditions with traction free upper and travel into the liquid medium because this
lower surfaces. If we now consider some would be like shear loading on the fluid.
form of a governing wave equation and If you solve this wave propagation
problem — or, as another example, the
FIGURE 52. A0 nondispersive and S0 wave propagation associated with
dispersive waves: (a) S0 dispersive, bitumen coating on a pipe — there would
time = 10.0 s; (b) S0 dispersive, also be leakage of ultrasonic energy as the
time = 20.0 s; (c) A0 non-dispersive, wave propagates along the plate.
time = 10.0 s; (d) A0 non-dispersive, Following the phase and group velocity
time = 14.0 s. dispersion curves, the complex roots from
the characteristic equation would then
(a) lead to a set of attenuation dispersion
curves.
(b)
A sample set of these attenuation
(c) dispersion curves for bitumen coating on
a pipe structure is illustrated in Fig. 54. A
(d) pipe sample problem is used here. For the
plate problem the modes would be labeled
as A0, A1, A2, S0, S1, S2 and so on
because of their symmetric and
antisymmetric character (see Fig. 53). In
the case of guided wave in pipes, the
axisymmetric longitudinal waves may be
labeled as L(0,1), L(0,2), L(0,3) and so on
and the axisymmetric torsional waves as
T(0,1), T(0,2), T(0,3) and so on. Flexural
modes are also possible because of partial

102 Ultrasonic Testing

loading around the circumference of a the outer, center or inner surface; with
pipe. See elsewhere63 for more details on only in-plane vibration on the surface to
flexural modes. Note in Fig. 54 that avoid leakage into a fluid; or with
attenuation does not always increase as minimum power at an interface between a
frequency is increased as in a usual bulk pipe and a coating. A sample set of wave
wave problem. Some modes attenuate structure curves are illustrated in Fig. 55
more quickly than others. Note the dotted to illustrate this point. In this case, the S0
curve for the L(0,3) mode in this case. mode propagation in an aluminum plate
One of the mode’s attenuations improved is considered. Notice the in-plane
significantly with higher frequency but vibration behavior across the thickness
this is the surface wave on the uncoated compared to the out-of-plane motion. The
side of the pipe. For other modes, for wave structure changes from point to
higher frequency, the wave amplitudes are point along every mode on a dispersion
significantly reduced. curve. The characteristics of every point
on each dispersion curve are different,
All guided wave problems have primarily with respect to wave structure, a
associated with them the development of critical feature for the development of an
appropriate dispersion curves and efficient test technique for a particular
corresponding wave structures. Of structure. Notice that for an f·d value
thousands of points on a dispersion curve, (value of frequency f times thickness d) of
only certain ones lead to a valid test — for 0.5, the in-plane displacement is totally
example, those with greatest penetration dominant across the thickness with
power; with maximum displacement on

FIGURE 53. Dispersion curves for a traction free aluminum plate: (a) phase velocity dispersion
curves; (b) group velocity dispersion curves.

(a)

Cutoff frequencies

20 A2 A4 A5 A7
A1 S2 A3 S6 S7
S1 S3
18 8 S5 A6
46 S4
16 14 16
10 12
Phase velocity 14
(mm·s–1)
12

10

8 CT

Cplate 6
S0

4

2 A0 2 CR 20
0 18

0

Frequency dispersion
(MHz·mm)

(b) 7

Group velocity (mm·s–1) 6 S4 S5 S6
S0
S3
5
S1 S2
4

3
2 A0

1 A3
A1 A2 A4 A5 A6

0
0 2 4 6 8 10 12 14 16 18 20

Frequency dispersion
(MHz·mm)

Generation and Detection of Ultrasound 103

almost no out-of-plane vibration. This ordinary frequency spectrum. These two
mode, as an example, would travel very spectral band widths, frequency and
far even if the aluminum plate were under phase velocity, make it difficult to excite a
water. If we now move forward and specific point on a dispersion curve. Some
consider the product of frequency times interesting discussion on the source
thickness equal to 2.0, the in-plane influence problem can be found
displacement on the outer surface of the elsewhere.15,64.65 Mode separation in the
plate is almost zero whereas the dispersion curve then becomes useful for
out-of-plane vibration is a maximum on single-mode excitation potential.
the upper and lower surfaces. If this S0
mode were to propagate at an f·d value of Guided wave energy can be induced
2.0, the leakage would be substantial and into a wave guide by various techniques.
waves under water would not propagate The challenge is to excite a particular
very far along the plate. mode at a specific frequency. Normal

In addition to the lamb waves FIGURE 55. Wave structure for various points
illustrated so far in Figs. 53 to 55, there on S0 mode of aluminum plate:
could be transverse horizontal guided (a) frequency f × thickness d = 0.5;
wave propagation in the plate as well, (b) f × d = 1.0; (c) f × d = 1.5; (d) f × d = 2.0.
depending on the sensor loading
situation. In this case, the transverse (a) 0.5 d
horizontal wave produces an in-plane
component perpendicular to the wave –0.2 u,w 1.4
propagation or wave vector direction but –0.5 d
still in the plane of the plate. For the
transverse horizontal mode, there is no (b) 0.5 d
out of plane displacement. The leakage
into a fluid media from an aluminum –0.4 u,w 1.2
plate would be nonexistent as far as wave –0.5 d
propagation is concerned. Keep in mind,
however, that mode conversion at a (c) 0.5 d
discontinuity could create some leaky
reflected waves.

Source Influence

The development of the dispersion curves
discussed so far uses harmonic plane wave
excitation in the wave guide. Because of a
bounded transducer problem, however, it
is necessary to study a source influence
problem for a particular size sensor. The
finite size of a transducer and various
vibration characteristics give rise to a
phase velocity spectrum in addition to the

FIGURE 54. Attenuation dispersion curves for a 100 mm
(4 in.) schedule 40 steel pipe with a 125 µm (0.005 in.)
bitumen coating.

–0.4 u,w
–0.5 d
0 L (0,3)
Attenuation (dB·m–1) –1 (d) 0.5 d
–2 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
–3 Attenuation (MHz) –0.3 u,w 0.3
–4
–5 –0.5 d
–6 Legend
–7
–8 = u, in-plane
–9 = w, out-of-plane
–10

0

104 Ultrasonic Testing

beam probes can be used. Angle beam phased array focusing techniques are
sensors can also be used to impart beams reported.68,69 Beam focusing is possible,
that lead to desired kinds of guided waves although a different computation
in a pipe or plate. A comb transducer (a technique to achieve focusing is necessary
number of different elements at a specific being different than the computations
spacing) can be used to pump ultrasonic required in phased arrays for bulk wave
energy into the plate, causing wave focusing.
propagation of a certain wavelength in
the wave guide. The excitation zones in A sample configuration of pipeline
the phase velocity dispersion curve can be testing with a typical wraparound guided
evaluated by the source being considered wave sensor arrangement is illustrated in
in the problem. Again, references15,64.65 Fig. 56. There are a number of
provide details in this exercise. A comb arrangements for this application via
transducer, as an example, could be normal beam sensors, angle beam sensors,
wrapped completely around a pipe or laid electromagnetic acoustic transducers,
out as fingers or an interdigital transducer magnetostrictive sensors and others at
design on a plate. frequencies ranging from 20 to 800 kHz,
depending on the distance of propagation
Pipeline Inspection and the discontinuities sought. Typically,
low frequency test systems find
Guided wave inspection of pipeline discontinuities that have a 5 percent cross
materials is particularly useful because a sectional area or more. Higher frequencies
large area can be tested from a single can go down to 1 percent cross sectional
transducer position.66,67,68 Some of the area or less.
initial work done in this area was for
steam generator tube testing. It was Aircraft Inspection
discovered that these waves could go far
and still be able to evaluate A variety of different problems can be
discontinuities at a long distance from the tackled in the aircraft industry.70 Aircraft
transducer position.66 In looking at skins are well suited to guided wave
pipeline inspection over long distances, testing. Note in Fig. 57 a possibility of
guided wave testing. In Fig. 57a, if
FIGURE 56. Representative wraparound ultrasonic energy can be passed from a
ultrasonic guided wave array for long range transmitter to a receiver across a lap splice
ultrasonic guided wave inspection of piping: joint, the integrity of the bond line can be
(a) attached; (b) detached. evaluated. Nevertheless, the problem is
not as simple as it looks because the wave
(a) structure has to be adjusted to have
sufficient energy at the interface to allow
propagation into medium 2. The wave
structure variation and the kind of energy
obtained can result from calculations of
wave structure for a particular mode and
frequency in a phase velocity dispersion
curve. Once the technique is developed,

FIGURE 57. Lap splice test sample problem:
(a) through-transmission; (b) double-spring
hopping probe.

(a)

Transmitter

(b) Receiver

1
2

(b)

Generation and Detection of Ultrasound 105

tools can be used as illustrated in Fig. 57b,
as an example, the double spring hopping
probe illustrated here can be placed on a
material quite easily and at the
appropriate mode and frequency can
evaluate the integrity of the lap splice
joint.

Closing

Because of tremendous advances in the
understanding of guided wave
propagation and the superb
computational ability by mathematical
and finite element analysis, guided wave
testing is a practical test option. The
technique can be used to solve many
problems using guided wave analysis in
nondestructive testing and structural
health monitoring is very bright.

106 Ultrasonic Testing

PART 8. Optical Generation and Detection of
Ultrasound71

Generation and detection of ultrasound in experimental systems. Because laser
by optical means, which together are beams can be easily scanned, array of
known as laser ultrasonics or laser based small generation sources or detection
ultrasound, represent an area of intensive spots can be realized. In addition, the
research and development which was spectral content of laser generated elastic
until the twenty-first century limited to waves may be extremely broad band.
the laboratory and pilot demonstrations.72 Using optical pulses of picosecond
The technology has matured and duration, acoustic signals of only 2 ns
applications have been transitioned to have been generated. This provides a
industry. It is also commercially available. potential for discontinuity detection in
Laser ultrasonic techniques are also useful materials as thin as a razor blade. Narrow
for material characterization and optical band signals can be generated as well by
detection can be used to characterize delaying in time or separating in space an
piezoelectric transducers. array of sources.

Advantages of Optical Optical generation and detection of
Ultrasound Generation ultrasound does have some limitations
and Detection that primarily affect detection sensitivity.
Elastic waves generated by optical sources
Noncontact Tests often have relatively low amplitude. Using
higher laser intensities to increase acoustic
A major advantage of optical techniques amplitude may damage to the test object
is that there is no mechanical contact surface. High intensity is often prohibited
made with the test object surface. In fact, but in many applications invasive
these techniques are not simply marking by the generation laser does not
noncontacting but may be used for matter, such as probing of hot metals
remote sensing — a clear distinction and during processing.
potential advantage over electromagnetic
acoustic transducer and capacitance Optical Generation of
transducer techniques. Elastic Waves

Because transduction of light energy to When light radiation is absorbed by the
acoustic energy is performed by the test irradiated portion of a test object, thermal
material, no intervening couplant is expansion results, producing elastic
needed. Likewise, material surface ultrasonic waves. A contribution might
vibrations are directly encoded onto a also come from the momentum transfer
light beam, also without couplant. These of the reflected light but these radiation
techniques make possible ultrasonic pressure effects are extremely small
testing in conditions difficult for other compared to those associated with light
techniques — probing hot materials, absorption.
testing in vacuum (in space) and probing
moving objects either transversely or With increasing incident optical
toward the transducer. Important for intensity, the temperature rise at the
industrial applications is the ease of object surface can be so great that
testing objects with contoured surfaces vaporization of the material may occur.
and nonplanar shapes. The momentum transfer of the ablated
material leaving the surface and the
With laser ultrasonic techniques, there pressure of the produced plasma result in
is no requirement for precise transducer a force normal to the surface that also
orientation, such as that found in gives rise to elastic waves. This subject has
conventional piezoelectric techniques. been reviewed in the literature and its
main features are outlined below.72,73
Other Advantages
Waves from Free Surface
Also notable is the optical resolution
obtained with laser ultrasonics. For Optical absorption produces two
example, detection spot diameters down mechanisms for elastic wave generation
to 10 µm (4 × 10–4 in.) are used routinely and these may be used to generate
ultrasound in at least three ways.

Generation and Detection of Ultrasound 107

FIGURE 58. Types of laser generated Figures 58a and 59a illustrate what
ultrasound: (a) thermoelastic or free surface; happens when light is absorbed near the
(b) constrained surface; (c) ablated surface. free surface of a material at a power
density below that which causes material
(a) ablation (about 105 W·mm–2 for
aluminum). Note that the thermoelastic
Air Steel expansion of the source volume is not
constrained by the surface. As a result,
Laser beam Heated zone thermoelastic generation on the free
Heated zone surface gives rise to little compressional
(b) Heated zone energy directed along the axis
perpendicular to the surface. Instead,
Laser beam longitudinal (compressional) waves are
directed in a pattern that takes the form
Transparent of a hollow cone with an apex half angle
bonded of about 1 rad (60 deg).
plate
This conical radiation pattern is
(c) observed for sources smaller than the
ultrasonic wavelength. The pattern was
Laser beam predicted theoretically by early models
Ablated that assumed point sources of radiating
material force vectors acting in the plane of the
surface.73,74 Experimental results using an
FIGURE 59. Radiation patterns of laser aluminum hemisphere as a test object
generated ultrasound for surface conditions substantially confirm these predictions. In
shown in Fig. 58 (longitudinal wave fact, epicentral measurements
directivity for a source smaller than the (measurements on axis) show a significant
ultrasonic wavelength): (a) thermoelastic retraction of the test object surface
surface; (b) constrained surface or ablated opposite the source following the arrival
surface. of the longitudinal wave (Fig. 60).

(a) Figure 60 shows a small outward
deflection of the object‘s back surface
1.5 rad immediately after arrival of the
(90 degrees) longitudinal wave. Early models for the
thermoelastic source did not predict this
leading spike because they assumed the
source had no thickness. More recently,
investigators have developed models
without heuristic assumptions that show
the spike.75,76 Other developments have
linked its origin to thermal-to-acoustic

FIGURE 60. Epicentral displacement caused
by point laser source in thermoelastic mode.

Laser Longitudinal
impinges
1 rad (60 degrees)
here 0 rad (0 degrees)

(b) Distance (relative scale)

1.5 rad
(90 degrees)

0 rad (0 degrees) Transverse

1 2345
Time (µs)

108 Ultrasonic Testing

mode conversion at the test object Although the ablative technique is
surface.77 destructive, it may be used when surface
finish is not critical during early stages of
In addition to longitudinal waves, both material processing. Also note that in this
transverse and rayleigh waves are case, the displacements are comparable to
generated in the free surface thermoelastic those produced by conventional
mode. Rayleigh wave directivity has been piezoelectric transducers excited by a few
demonstrated using focused line sources.78 hundred volts. The displacements are
An annular illumination pattern has also weaker in the thermoelastic and free
been demonstrated, giving rise to surface mode.
convergent rayleigh waves and very
strong amplitude at the center of the To generate ultrasound in these three
annulus.79 Other generation patterns can modes, a pulsed laser is used. Many
also be used with various advantages, reported studies have been performed
such as an array of lines that gives narrow with quantum based switched solid lasers
band emission and minimizes surface such as the neodymium
damage by distributing laser energy. yttrium-aluminum-garnet laser with pulse
Generation can also be enhanced by lengths from 5 to 30 ns. Pulsed gas lasers
sweeping the line or array of lines also may be used, such as the transversely
source.80 For samples of thickness smaller excited atmospheric pressure carbon
than the rayleigh wavelength, plates dioxide laser, used for inspecting polymer
(lamb) waves can be similarly generated. matrix composites.72

Waves from Constrained Surface Optical Detection of
Ultrasound
Acoustic waves also may be optically
generated as illustrated in Fig. 58b, where Optical techniques for ultrasonic wave
the surface of the test object is detection can be divided into two classes.
constrained by a transparent material. The first includes techniques that permit
Constraint may be obtained by bonding a real time detection of ultrasonic
glass plate to the surface or more simply disturbances at a single point or over a
by coating the surface with a thin layer of single zone on a test object surface. The
water, oil or grease. second category includes full field
techniques that provide maps of the
The displacement out of the surface is acoustic energy distribution over an entire
redirected into the bulk of the test object. field of view at one instant in time.84
As a result, the directivity pattern for
longitudinal waves is altered so that most Full field techniques have insufficient
of the longitudinal wave energy is sensitivity and are not used, so the
directed on axis (Fig. 59b).74 present discussion focuses on single-spot
Enhancement of the longitudinal and detection techniques. Such techniques
transverse waves is observed as high as have been widely reviewed in the
30 dB over free surface generation, literature85 except for the variants
depending on the test object material and described below. Generally, in the case of
the type of surface constraint. Similar laser generated ultrasound, the ultrasonic
effects occur when light penetrates deeper displacements (on the surface of
into the test object, producing a buried generation and any other surface) have a
ultrasonic source.72,81 This effect is found nonvanishing normal component.
particularly important for testing polymer Therefore, it is generally sufficient to
materials, in particular polymer matrix detect this component, although in-plane
materials.72 motion can be detected by optical
techniques.86 For detecting normal
Waves from Ablated Surface displacement, two interferometric
techniques and two variants of them have
A third way for generating acoustic waves found application.
optically occurs when the source energy
density is sufficiently high to cause Simple Interferometric Detection
ablation of material from the test object
surface (Fig. 38c).82,83 The ablated material A technique called optical heterodyning or
may be the base material or it may be a simple interferometric detection uses a wave
coating ablated after absorbing the light scattered by the surface to interfere with a
energy. The momentum transferred to the reference wave directly derived from the
surface during ablation produces a greatly laser (Fig. 61). Such a technique is
increased normal force on the surface. As sensitive to optical speckle and the best
is the case with a constrained surface, sensitivity is obtained when one speckle is
both the directivity and the amplitude of effectively detected. This means that the
the longitudinal wave energy are changed. mean speckle size on the focusing lens
Owing to the large normal force, has to be about the size of the incoming
longitudinal waves are directed most beam and that this beam should be
strongly along the axis normal to the
surface.

Generation and Detection of Ultrasound 109

focused onto the surface. This technique confocal fabry-perot interferometer). This
generally allows the measurement of interferometer is made simply with two
ultrasonic displacement over a small spot concave mirrors separated by a distance
and provides point detection, except at equal to their radius of curvature and is
high frequencies. widely used for industrial applications.85

Compensation for vibrations can be Interferometric Technique
performed by an electromechanical Variants
feedback loop that generally uses a
piezoelectric translator for path length Velocity interferometry does not have a
compensation. For more severe vibration flat detection response but generally has a
environments, the heterodyne large throughput corresponding to its
configuration is preferred. In this setup, large detection area. Simple
the optical frequency in one arm is interferometric detection, on the other
shifted by a radiofrequency and the hand, has a flat response (limited by the
detector receives a signal at this shift detector cutoff frequency or the shift
frequency, phase modulated by frequency) but also has a small
ultrasound and vibrations. Electronic throughput corresponding to its small
circuits can be devised to retrieve the detection area. Variants of these
ultrasonic displacement independently of techniques have however been developed
vibrations. that provide both broad detection band
width and large throughput.
Velocity Interferometry, or Time
Delay Interferometry A first variant consists in using the
confocal fabry-perot interferometer in
The second detection method, called reflection.87 In this case, there is
velocity interferometry or time delay interference between the light directly
interferometry,85 is based on the doppler reflected by the front mirror and light
frequency shift produced by surface reflected and leaked out of the fabry-perot
motion and its demodulation by an cavity, both having wavefronts
interferometer having a filter response substantially matched. The light leaked
(Fig. 62). This technique is primarily out in reflection is stripped from its
sensitive to the velocity of the surface and sidebands for frequencies above Δν (as
is insensitive to low frequencies. The filter defined in Fig. 62) and acts as a reference
response is obtained by giving a path wave. This system has a response
delay between the interfering waves
within the interferometer. Two-wave FIGURE 62. Ultrasound detection with time
interferometers (michelson, delay velocity interferometer: (a) test setup;
mach-zehnder) or multiple-wave (b) principle of detection.
interferometers (fabry-perot) can be used.
(a)
Unlike optical heterodyning, this
technique permits reception of many Laser
speckles. A large detecting spot (several
millimeters or more) is obtained when Ultrasound
known modifications are used to increase
throughput and field of view (field
widened michelson interferometer,

FIGURE 61. Configuration for optical Intedrfeetreocmtoerter and
heterodyning or simple interferometric
detection. A frequency shifter such as a (b)
bragg cell can be introduced in either arm
(heterodyne michelson interferometer). Response

Mirror

Δυ

Test object Frequency shifter

Lens Laser Optical
Beam splitter frequency

Ultrasound

Detector Laser frequency

110 Ultrasonic Testing

essentially flat above Δν (which means in ultrasonic transducers. Probes based on
practice a few megahertz). optical heterodyning are preferred for this
purpose because they can be easily
A second variant consists in using an calibrated and allow the measurement of
adaptive beam splitter/mixer. By two-wave the absolute value of ultrasonic
mixing in a photorefractive crystal, the displacements.
wave reflected by the surface beats with a
pump wave directly derived from the laser Optical probes can be used for
to produce a real time hologram in the detection of ultrasonic transducer
crystal.88,89 This hologram diffracts a malfunctions resulting from improper
reference wave matched to the manufacture or aging. Optical probes can
transmitted wave, as sketched in Fig. 63. also be used to clarify complicated
This system has a flat response from very ultrasonic testing procedures by mapping
low frequencies (1 kHz to 100 kHz the ultrasonic displacement field over the
depending on the pump wave power) to surfaces of a test object.90
the detector cutoff frequency and a large
throughput. Gaging

Note that the flexibility of such Laser ultrasonic testing, like any other
interferometric detection systems is ultrasonic technique, can be used for
greatly increased by including an optical measuring thickness. Because it operates
fiber link between the interferometer and at a distance, however, it allows gaging
the test surface. In the case of optical parts at elevated temperature. As with
heterodyning, single mode fibers must be conventional ultrasonic testing, this
used. Velocity interferometry and the two application is based on measuring the
variants mentioned above may use large time-of-flight between consecutive echoes
core fibers that make application simpler. and then, by using the value of the
acoustic velocity in the material, to relate
Finally, despite the diversity of time-of-flight to thickness. This velocity
interferometric detection system designs, has to be calibrated by independent
their sensitivities to ultrasonic surface measurements because it depends on
motion are generally of the same order of temperature.
magnitude. In the case of shot noise
limited detection, the sensitivity is on The technique has been particularly
order of 0.01 nm for an electronic band developed for wall thickness gaging of
width of 10 MHz and an intensity seamless tubes at elevated temperature on
received by the interferometer of about a production line and is now
1 mW. commercially available.91 Figure 64 shows
the measuring head of a system installed
Applications on line above a hot and rotating tube in a
seamless tube production plant. This head
Ultrasonic Metrology is linked by optical fibers to the lasers and
the interferometer, located remotely in an
Optical probes for detection of ultrasound environmentally controlled enclosure.
can be used to map ultrasonic fields at the This arrangement allows adequate
surface of test objects or at the surface of servicing of the lasers (for example,
change of the flash lamps) away from the
hot and dusty environment of the plant.

FIGURE 63. Ultrasound detection with two-wave mixing FIGURE 64. Wall thickness gaging and
photorefractive interferometer. austenite grain size determination on
seamless tubes: view of measuring head on
Ultrasound top of hot tube.

Photorefractive
crystal

Pump beam Diffracted
pump beam
Detector
(reference)

Laser

Generation and Detection of Ultrasound 111

The system includes also a pyrometer to Force.94 The technology has also been
measure the temperature from which the used at the validation and production
proper velocity can be found from the stages by aerospace companies in Europe95
previous calibration. Such a system allows This technology could be used not only
controlling the tube making process and for inspecting fabricated parts but also for
has demonstrated significant productivity inspecting an aircraft during
gain. It has also the advantage compared maintenance.96 Unlike conventional water
to gamma-ray tomography, another jet ultrasonics, laser ultrasonics allows
technology that could be used for scanning to the very edge of the part. For
determining the wall thickness of tubes, the C-scan of a horizontal stabilizer, a
to permit measurement with a piercing transversely excited atmospheric pressure
mandrel inside the tube. pulsed carbon dioxide laser operating at
10.6 µm has been used to generate
Discontinuity Detection ultrasound. For detection, a long pulse,
high stability neodymium
Laser ultrasonics produces an ultrasonic yttrium-aluminum-garnet laser specially
source at the surface of a test object and developed for laser ultrasonics is used and
allows detection from the object surface, coupled to a confocal fabry-perot
independently of shape and orientation. interferometer. Photorefractive
Curved and complex geometries such as interferometers with better low
pipes, rotor blades and the edges of frequencies sensitivity could be
aircraft wings can then be tested. advantageously used for thick part testing,
more than 12 mm (0.5 in.).
In particular, delaminations in flat or
curved graphite epoxy laminates can be On metallic parts, the detection of
easily detected.92 Figure 65 shows the surface breaking cracks has been
results (C-scan and B-scans) obtained by demonstrated by using laser generated
raster scanning a U shaped specimen with surface acoustic waves. When the
delaminations on the flat surfaces and generation laser beam passes over the
along a corner. These results are made crack, the signal detected at some distance
possible by using a generation laser that from generation is very different from the
provides adequate penetration and one encountered from a crack free region
absorption of light in the top epoxy layer and allows reliable detection of surface
and contributes to the constraining effect cracking.97 By using the filtering effect of
mentioned above. The generated the crack (low frequencies go through,
longitudinal wave is always essentially high frequencies are blocked), these
normal to the surface, independently of discontinuities can also be sized. In
the direction of the laser beams. metallic parts, because the acoustic source
Therefore, there is no need to know or to is at the surface and can be made very
follow the part contour as with small, the technique can easily generate
conventional ultrasonic testing. by scanning an array of ultrasonic A-scans
and is advantageously coupled to a
This application to polymer matrix numerical focusing technique such as the
composite inspection has been extensively synthetic aperture focusing technique
developed and is now routinely used for (SAFT).98 In combination with such a
part inspection by a major military technique, high resolution imaging can be
aircraft manufacturer.93 A turnkey system
has been built for the United States Air

FIGURE 65. Laser ultrasonic test of U shaped graphite epoxy composite showing time-of-flight
C-scan above B-scans at indicated locations.

Top
Corner

Laser beams Side

112 Ultrasonic Testing

obtained throughout the part volume. are essentially the same as the ones used
Fourier domain processing has been with conventional ultrasonics and are
developed to minimize computer based on the monitoring of ultrasonic
processing time.99 Figure 66 shows an velocity, ultrasonic attenuation and
example in which a test specimen with backscattered microstructure noise.
stress corrosion cracks was imaged from Optical detection is particularly
the surface opposite to cracking.98 The advantageous for measuring backscattered
C-scan at the crack opening surface is noise, which relates to grain size in
compared with an image obtained by metals, because a small detection spot is
liquid penetrants. The amplitude of the readily obtained.
crack indication obtained after numerical
reconstruction provides also some A first type of application consists in
information on crack depth. One will using the velocity information for the
note that in this combined technique, determination of elastic constants.
longitudinal or transverse waves can Because a laser generates longitudinal and
either be used for imaging because they transverse waves at the same time, both
are both generated by the laser. velocities can be deduced from the
measurement of the two propagation
Materials Characterization times. Assuming there is a suitable model
to link velocities and elastic constants,
The approaches for material these constants also can be determined.
characterization by laser ultrasonic testing Ultrasonic velocity measurements have
been reported in numerous research
FIGURE 66. Stress corrosion cracks on contributions for this purpose. The
stainless steel sample: (a) laser ultrasonic materials include metals (aluminum and
image; (b) liquid penetrant test. steel), ceramics and metal ceramic
(a) composites, at room temperature and at
elevated temperatures.100-103 These
(b) experiments were performed by
generating ultrasound with a short pulse
laser (generally a Q switched, neodymium
yttrium-aluminum-garnet laser) on one
side of the test object and detection from
the other side. Different approaches have
been considered to improve the accuracy
of the time interval determination,
including the use of ablation to produce
strong spike pulses,100,102 off epicenter
probing100 to enhance transverse features,
modeling of the source101 and cross
correlation between consecutive
echoes.103

Velocity measurements can also be
used to measure material anisotropy, or
texture. Two approaches can be used for
plate samples. A first one consists in
generating rayleigh or lamb waves
propagating in a given direction using
line source generation.104 Wave arrival is
detected at some distance from
generation. A second approach is based
on normally propagating bulk waves
(longitudinal and transverse) and the
frequency analysis of the multiple echoes.
If there is anisotropy, resonance will occur
at different frequencies for the transverse
waves according to their polarization.105
This has been demonstrated to be
applicable for monitoring annealing of
steel and aluminum alloys because during
this process the new grains grow usually
with a different orientation.105-107 Velocity
monitoring can also be used to monitor
phase change or to determine phase
composition.108 In particular, the retained
austenite fraction in steel can be
determined (Fig. 67).

It is known that materials can also be
characterized by measuring ultrasonic

Generation and Detection of Ultrasound 113

attenuation. Such a measurement made occurred will increase. One promising area
with conventional piezoelectric of application is the steel industry as an
techniques is generally difficult because it on-line sensing technology for
requires a test object with parallel microstructure and phase transformations.
surfaces, a precise orientation of the Other industries that use laser based
transducer (immersion technique) or a manufacturing techniques, such as the
uniform bond. Furthermore, to apply automotive industry, can benefit from
diffraction corrections, the transducer laser ultrasonic testing. The nuclear
should satisfy to a good approximation industry because of the radiation
the assumption of a piston source and of environment could also benefit from
a uniform baffled receiver.109 Laser remote sensing with lasers.
ultrasonic testing requires test objects of
sufficiently uniform thickness but is free Among the barriers preventing a wider
of orientation and bond problems because use of laser ultrasonics is the possible
the source and the receiver are on the damage to the material surface, laser
surface of the test object. ocular safety and beyond these aspects,
some complexity of the technology. This
It has been shown that sufficiently complexity leads to a relatively high cost,
accurate measurements can be performed usually higher than conventional means.
by operating in two cases: (1) the Therefore the hybrid approaches that use
spherical wave limit corresponding to laser generation, electromagnetic acoustic
point generation and detection and transducers or air coupled detection may
(2) the plane wave limit corresponding to find use in special cases. However,
large generation and detection spots.110 electromagnetic acoustic transducers are
On steel, good correlation was observed limited to metals and proximity sensing
between measured attenuation and grain and transmission through air suffers high
size.111 More recently this application has losses increasing with frequency. There
been extended to the measurement of have also been efforts to make the laser
austenite grain size on line.112,113 ultrasonic technology more affordable.116

Further Developments In conclusion, although the principles
of the technology are not expected to
Since the 1990s, laser ultrasonic testing change, its industrial application is
has made the transition from the expected to grow in the twenty-first
laboratory to industry in two important century.
areas: (1) polymer-matrix composite
testing and (2) gaging and microstructure
determination in the steel industry,
specifically for seamless tubes. The
technique is also used in microelectronics
for characterizing very thin layers or
measuring their thickness.114,115

It is reasonable to expect that the use
of the technology in the areas for which
the transition to industry had already

FIGURE 67. Austenite grain growth
monitoring by laser ultrasonic testing versus
metallography.

Grain size (µm) 120

70

20 300 600
0 Time (s)

Legend

= Metallography
= Laser ultrasonics

114 Ultrasonic Testing

PART 9. Electromagnetic Acoustic Transduction117

The electromagnetic acoustic transducer current and act as a source of ultrasonic
(EMAT) is a device for the excitation and waves. The process is in many ways
detection of ultrasonic waves in similar to that which creates motion in an
conductive or magnetic materials. No electrical motor. Reciprocal mechanisms
physical contact is required with the test also exist whereby waves can be detected,
object because the coupling occurs a process analogous to operation of an
through electromagnetic forces. The electrical generator.
working distances are typically less than a
millimeter and the probe often is allowed If the material is ferromagnetic,
to rest on the surface of the test object. additional coupling mechanisms are
This ability to provide reproducible found. Direct interactions occur between
signals with no couplant is often more the magnetization of the material and the
important than noncontact operation. dynamic magnetic fields associated with
the eddy currents. Magnetostrictive
Physical Principles processes are the tendency of a material to
change length when magnetized. These
The physical principles of electromagnetic processes can also play a major role in
acoustic transducer transduction118 are generating ultrasound. Again, reciprocal
shown in Fig. 68. Suppose that a wire is processes exist whereby these mechanisms
put next to a metal surface and driven by can contribute to detection.
a current at the desired ul_trasonic
frequency. Eddy currents Jw are induced Probe Configurations
within the met_al and, if a static magnetic
bias induction B0 is also present, the eddy Practical electromagnetic probes consist of
c_urrents experience periodic lorentz forces much more than a single wire. It is
FL given by: usually necessary to wind a coil and
design a bias magnet structure so that the
(39) FL = Jw × B0 distribution of forces predicted by Eq. 39
couples to a particular wave type. Figure
The lorentz forces on the eddy currents 69 shows the cross sections of five coil
are transmitted to the solid by collisions types.
with the lattice or other microscopic
processes. These forces on the solid are Included are probes that couple to
alternating at the frequency of the driving (1) radially polarized transverse beams,
(2) longitudinal or (3) transverse plane
FIGURE 68. Single element of polarized beams propagating normal to
electromagnetic acoustic transducer. the surface and (4) longitudinal or
vertically polarized transverse beams or
I (5) horizontally polarized transverse
B¯o horizontal beams propagating at oblique
angles. By virtue of the spatially periodic
F¯L stresses that it excites, the meander coil
electromagnetic acoustic transducer
shown in Fig. 69d can also excite rayleigh
waves on surfaces or lamb modes in
plates. The periodic permanent magnet
transducer in Fig. 69e can also excite
horizontally polarized transverse modes in
plates.

J¯ω Advantages of
Electromagnetic Acoustic
Legend Transducers
B¯o = magnetic bias induction
F¯L = body forces The major motivation for using
I = applied current electromagnetic acoustic transducers is
J¯ω = eddy current their ability to operate without couplant

Generation and Detection of Ultrasound 115

or contact. Important consequences of Modeling of
this include operation on moving objects, Electromagnetic Acoustic
in remote or hazardous locations, at Transducer Measurements
elevated temperatures, in vacuum and on
oily or rough surfaces. Moreover, Because the coupling of electromagnetic
alignment problems may be reduced acoustic transducers is very reproducible,
because the direction in which the wave is it is possible to model them with high
launched is primarily determined by the accuracy. This is important for three
orientation of the test object surface reasons.
rather than the probe. Finally,
electromagnetic acoustic transducers have 1. First, it allows the design of
the ability to conveniently excite measurement systems that fully
horizontally polarized transverse waves or exploit the ability of the transducers
other special wave types that provide test to excite special wave types and
advantages in certain applications. radiation patterns.

It must be noted that the cost of 2. It provides a tool for optimizing
realizing these advantages is a relatively system performance and counteracting
low operating efficiency. This inefficiency low sensitivity problems.
is overcome by high transmitter currents,
low noise receivers and careful electrical 3. In ferromagnetic materials, it allows
matching. In ferromagnetic materials, the full use of the magnetization and
magnetization or magnetostrictive magnetostriction mechanisms for
mechanisms of coupling can often be optimized transduction efficiency.
used to enhance signal levels.
As an example, a physical transducer
model based on the theory of
electromagnetic ultrasonic transduction in
ferromagnetic metals is described
below.119 Its predictions have been
analyzed to derive essential rules of
transducer behavior and these have been

FIGURE 69. Cross sectional view of practical electromagnetic acoustic transducer configurations: (a) spiral coil exciting radially
polarized transverse wave propagating normal to surface; (b) tangential field electromagnetic acoustic transducer for exciting
plane polarized longitudinal waves propagating normal to surface; (c) normal field transducer for exciting plane polarized
transverse waves propagating normal to surface; (d) meander coil transducer for exciting oblique longitudinal or vertically
polarized transverse waves, rayleigh waves or guided modes of plates; (e) periodic permanent magnet for exciting obliquely
propagating horizontally polarized transverse waves or guided horizontally polarized shear modes of plates.

(a) (b) (c)

S SN

N NS
NS

(d) (e)

S SN SN S N SN
NSNSN SNS
N

116 Ultrasonic Testing

systematically checked in a series of The numerical model is a strong
experiments. engineering tool for optimizing
transducers. In addition, careful
For generating ultrasound, the impedance matching and proper
transducer is modeled through a electronic design are necessary to
two-dimensional pattern of overcome insertion losses.
radiofrequency currents distributed in the
space above the material surface and a Thickness Gaging with
homogeneous magnetic bias field. For this Electromagnetic Acoustic
configuration, the displacements in the Transducers
ultrasonic wave field generated in the
material are calculated. For ultrasonic Ultrasonic techniques are widely used for
reception, the model gives the electrical thickness gaging and electromagnetic
voltage induced in the coil winding acoustic transducers expand the possible
pattern during the incidence of an range of applications. Because of the
ultrasonic wave. To obtain these results, electromagnetic acoustic transducer‘s
the coupled dynamic equations of the ability to operate at high speed and
electromagnetic field and of elastic stress elevated temperatures, thickness gaging is
and strain have been solved.119 well suited for online measurements
during materials processing. With these
Results of Theoretical Studies transducers, it is also particularly easy to
generate transverse waves. This has
Theoretical results have been checked advantages when measuring thin
systematically in experimental studies of materials because the transverse velocity is
transverse wave transduction using a roughly half the longitudinal wave
semicylindrical block of soft iron. As a velocity. For a given thickness, the echo
measure of transduction efficiency, the occurs later, is more easily resolved from
transfer impedance of a transducer pair is electrical leakage and the change in
used, a quantity that can readily be arrival time per unit change in thickness
measured and is also straightforwardly is greater.
obtained from the theoretical treatment.
The model gives directly the sensitivities Beams Propagating Normal to
of the transmitter and receiver. The Surfaces
multiplication of both yields the transfer
impedance, that is, the ratio of the Delay lines are commonly used with fluid
receiving voltage to the transmitting coupled transducers to measure thin
current. Losses from attenuation and material. They cannot be used with
diffraction must also be properly included electromagnetic acoustic transducers (later
in these calculations. reflections can often be used to
circumvent problems with system
The theoretical and experimental recovery time). Although electromagnetic
directivity pattern of vertically polarized acoustic transducers can operate as high
transverse waves can be generated by a as 12 MHz, a more practical limitation at
meander coil designed to enhance the present is the 5 to 7.5 MHz range (even at
radiated intensity at an angle of 5 MHz, cable lengths must be kept short).
θ = sin–1(λD) with respect to the surface This limits measuring thicknesses to those
normal (where D is coil period and λ is greater than about 1 mm (0.04 in.).
wavelength). The experiments were
conducted on a half cylinder of soft iron Measurements in thick materials are
with a magnetic bias field normal to the eventually limited by the ratio of signal to
surface. The ordinate is the transfer noise. In practice, beam spread losses (that
impedance (drawn to a logarithmic scale) decrease as the transducer coil gets larger)
occurring when the line probe receiver is must be balanced with available drive
moved along the cylindrical surface of the currents (that usually increase as the coil
test object. The transmitter coil has a gets smaller).
dolph-chebychev tapering to obtain
lowest side lobe levels. The transfer In most situations, any coil can operate
impedance measured at the main lobe reliably over a 3:1 thickness range and
maximum is about 20 percent lower and still accommodate a coil liftoff variation
the measured side lobe level is somewhat of at least 0.5 mm (0.02 in.). Smaller
higher than the calculated values. liftoff variation often allows a coil to
Through the dependence on wavelength, operate over a 10:1 thickness range.
the main lobe‘s angle of incidence
depends on the frequency. The largest The most efficient electromagnetic
discrepancy between experimental and acoustic transducer for thickness
theoretical values is about 20 percent. measurements with normally propagating
These results demonstrate the usefulness bulk transverse waves uses a planar coil. If
of the applied model in calculating separate transmit and receive coils are
absolute transducer efficiencies and used, then they should be of nearly
directivity patterns.

Generation and Detection of Ultrasound 117

identical geometry. Figure 70 shows a 13 m (42 ft) tube moving past the sensor
particularly simple configuration with a at 0.9 m·s–1 (3 ft·s–1) in an operating steel
single spiral coil (Fig. 69a). The coil mill.120 The transducer introduces
diameter, wire size and magnet size transverse waves into the pipe wall at an
depend on the application and the angle that allows the transmitter and
current source available for driving the receiver to be physically separated,
transducer. lengthening the transit time for thin
walled tubes.121
Angle Beams
An accuracy better than ±1 percent was
Online monitoring of wall thickness is obtained from the installed unit. The
desirable but the technique must be able system has eight electromagnetic acoustic
to operate on rough, scale covered transducers around the circumference to
surfaces that are not only moving rapidly yield eight thickness profile graphs like
past the testing station but may be at the one in Fig. 71. All these data are
temperatures above the boiling point of processed in real time by a dedicated
water. By using electromagnetic acoustic computer so that the operator is given a
transducers, the sensitivity to surface display of the test results immediately
speed, temperature and cleanliness can be after the pipe passes through the station.
minimized. Furthermore, the sensors can
be mounted on simple carriages that track This report provides the average wall
the surface as it bounces through the thickness, the length of the tube, its
testing station. weight per unit length, the minimum wall
thickness detected and the maximum
Seamless steel tubing is manufactured eccentricity of the center bore hole. In
at speeds that approach 1 m·s–1 addition, the display shows the standard
(200 ft·min–1). Abnormal variations in wall deviation in thickness and eccentricity as
thickness are the major cause for well as the location and thickness values
rejection. Figure 71 shows a plot of the of any points thinner than the limit set
thickness profile measured by an for rejection.
electromagnetic acoustic gage positioned
to record a thickness value every 3 mm A paint system is activated by the
(0.125 in.) along a longitudinal path on a computer to mark the location of
critically thin spots so that they can be
FIGURE 70. Permanent magnet transverse verified with subsequent manual
wave electromagnetic acoustic transducer scanning. All of the data, including the
about 25 mm (1 in.) in diameter and eight individual thickness profiles, are
38 mm (1.5 in.) high. Magnet pole diameter stored in computer memory for later
should be about twice coil diameter. analysis and for statistical summaries of
production runs.
Magnet
Weld and Cladding Tests
Radiofrequency
coil The testing of austenitic welds can be
strongly influenced by wave speeds in the
Connector weld metal that generally differ from
those in the base metal. Because of the
FIGURE 71. Plot of wall thickness profile produced by strong elastic anisotropy of weld
materials, the difference in base metal and
automatic thickness gage using electromagnetic acoustic weld metal acoustical impedance leads to
reflection and refraction of the ultrasonic
transducer on 350 mm (14 in.) diameter, 9.5 mm (0.375 in.) wave at the interface. These phenomena
wall seamless steel tube at production line speed of 0.9 m·s–1 have been studied in detail for vertically
(3 ft·s–1) in operating steel mill. polarized transverse and longitudinal
waves. For horizontally polarized
Thickness, 10.9 (0.431) transverse waves, most of the existing
mm (in.) 9.5 (0.375) work covers propagation in austenitic
8.0 (0.318) weld metal.122-126
0 1.5 3.0 4.5 6.0 7.5 9.0 10.5 12
(5) (10) (15) (20) (25) (30) (35) (40) Horizontally Polarized Waves

Distance along pipe, m (ft) Compared to vertically polarized
transverse and longitudinal waves,
horizontally polarized transverse waves
have a number of advantages for testing
welds. When reflecting from surfaces
parallel to the polarization direction,
transverse horizontal waves do not mode
convert to other types of waves and less
clutter is observed. When passing through

118 Ultrasonic Testing

interfaces with surfaces parallel to the investigated in detail.127 It has been
polarization direction, transmission demonstrated that such waves are almost
coefficients are often higher than for totally transmitted into the weld metal in
other wave types. a wide range of incidence angles between
0.8 and 1.5 rad (48 and 90 deg). Figure 72
Furthermore, these waves can be shows the computed coefficients of
excited over a wide range of angles. With reflection and transmission at the
electromagnetic acoustic transducers, an interface between base metal and weld
angle range from 0.3 to 1.5 rad (20 to metal as a function of incidence angle. It
90 deg) can be covered. In contrast, is assumed that the columnar grains are
transverse horizontal waves can only be oriented perpendicular to the material
excited with difficulty125 when using surface and parallel to the interface
conventional piezoelectric transducers between the base metal and the weld.
(transmission of a transverse wave with a
steel wedge and highly viscous coupling). These theoretical values have been
For practical applications, the transverse confirmed by measurement (Fig. 73). Two
horizontal method has not been electromagnetic acoustic transducers were
extensively used. Because transverse used to transmit through the base metal
horizontal waves can be excited and (Fig. 73b) and through the base metal
detected in electrically conductive with the weld metal (Fig. 73c) at different
materials without contact using angles of incidence. The received
electromagnetic acoustic transducers, ultrasonic amplitudes were measured and
more applications are possible.127,128 recorded for comparison. For incidence
angles between 0.9 and 1.5 rad (55 and
Tests of Austenitic Welds 90 deg), the amplitude values are the same
(Fig. 73a). For angles smaller than 0.9 rad
The potential of transverse horizontal (55 deg), an increasing part of the
waves for austenitic weld testing has been incident energy is reflected at the
interface between the base metal and the
FIGURE 72. Horizontally polarized transverse weld metal, and the transmitted signal
waves at base metal weld interface: amplitude diminishes in the test object.
(a) diagram of austenitic weld;
(b) computed reflection and transmission FIGURE 73. Experimental measurements of
coefficients. ultrasonic transmission of horizontally
polarized transverse waves: (a) plot of
(a) amplitude versus incidence angle;
(b) diagram of transmission through base
Austenitic weld metal; (c) diagram of transmission through
base metal, welds and interfaces.

(a)

Computed coefficient θS 40
ReceivedAustenitic base material30
20
(b) amplitude (dB) 10

2 0.70 0.87 1.05 1.22 1.40 1.57
(40) (50) (60) (70) (80) (90)

Angle of incidence, rad (deg)

⏐R⏐ T (b) Austenitic
1 base metal

⏐T⏐ T
0R 60 α

θS R
0.3 0.5 0.8 1.0 1.3 1.6
(15) (30) (45) (60) (75) (90) (c) T Austenitic weld metal
60 α
Angle of incidence, rad (deg)
R
Legend
R = reflection coefficient
T = transmission coefficient
θS = shear incidence angle

Generation and Detection of Ultrasound 119

Another important aspect of wave and transmission coefficient of the
propagation in austenitic materials is transverse horizontal wave at the interface
beam skewing. Beam skewing has been of the austenitic cladding and the ferritic
extensively studied in austenitic and base metal as a function of incident
dissimilar metal welds with a ray tracing transverse horizontal wave θs in the
model.126 According to these studies, the cladding.
smallest skewing is observed for transverse
horizontal waves and longitudinal waves. The transverse horizontal wave is
transmitted almost completely in an angle
Tests of Austenitic Cladding range of θs from 0 to 1.2 rad (0 to 70 deg).
Between 0 and 0.3 rad (0 and 20 deg), the
Reflection and transmission at the reflected amplitude is about 11 percent of
interface between a ferritic base metal and the incident amplitude. Between 0.3 and
austenitic cladding have been studied as 0.8 rad (20 and 47 deg), the value of the
further examples of the interactions of reflected amplitude decreases until it
transverse horizontal waves at an interface vanishes. It subsequently rises to an
between an anisotropic and isotropic amplitude of 2.5 percent of the incident
medium. amplitude before passing through
0 percent again. Above 1.1 rad (68 deg),
It is assumed that the columnar grains the reflected portion increases rapidly up
in the cladding are directed perpendicular to 1.2 rad (76 deg). For higher angles, the
to the surface. The incidence angle with reflection and transmission coefficients
respect to the material surface corresponds become complex. The reflection
to the incidence angle θs at the interface. coefficient |R| is equal to 1 (the incident
Figure 74 shows the computed reflection wave is totally reflected in this angle
range).
FIGURE 74. Horizontally polarized transverse waves at
interface of ferritic base metal and an austenitic weld: Because of the additional reflection of
(a) diagram of weld; (b) computed reflection and the transverse horizontal wave at the
transmission coefficients. stress free outer surface of the cladding, a
guided transverse horizontal wave can
(a) propagate at flat incidence angles in the
cladding by zigzag reflection Z (Fig. 75). At
Cladding the clad side of a test object, an
electromagnetic acoustic transducer with
Incident transverse wave a period of 5 mm (0.2 in.) is transmitting
a transverse horizontal wave with a
frequency of 572 kHz at grazing incidence
— a 6 dB drop of the main lobe is about
1.1 rad (65 deg).

At the front surface of the test object,
an ultrasonic signal is received with a
piezoelectric transducer (Y quartz, 1 MHz,

Base metal FIGURE 75. Guided horizontally polarized transverse waves in
austenitic cladding: (a) received signal in range of cladding
Computed coefficient(b) T (love wave); (b) received signal in base metal.
R
2 (a)

T Transverse horizontal
1 wave transmitter

R 1.6 Y quartz
(90)
0 240 mm
0.3 0.5 0.8 1.0 1.3
(15) (30) (45) (60) (75) (b)

Angle of incidence, rad (deg) Transverse horizontal
–1 wave transmitter

Legend Y quartz
R = reflection coefficient
T = transmission coefficient 240 mm

120 Ultrasonic Testing

polarization direction parallel to clad metals such as platinum and certain
surface). Figure 75a shows the received nickel-chromium alloys (these materials
signal propagating in the cladding. This are only available in small quantities and
sort of signal is called a love wave, with a are not yet available commercially). Many
group velocity lower than the phase materials used for high temperature
velocity of a bulk transverse horizontal insulation are thick compared to normal
wave in the base metal. Figure 75b shows insulation and surface adhesion tends to
the received signal propagating in the be poor. It is possible to produce
base metal. The velocity inferred from electromagnetic acoustic transducer drive
Fig. 75b corresponds nearly to the velocity and receive coils that operate up to
of a bulk transverse horizontal wave in 1500 °C (2800 °F).
the base metal. The amplitude of the
transverse horizontal wave guided in the Most magnetic steels cannot be used as
cladding is 3.4× the amplitude of the bulk magnet pole materials above 550 °C
transverse horizontal wave in the base (1000 °F) but some cobalt alloys function
metal. as magnet poles up to 820 °C (1500 °F).
Most permanent magnet materials cannot
This has important consequences for be used above 120 °C (250 °F) and some
testing austenitic cladding with transverse high field materials degrade rapidly above
horizontal waves from the clad side. 100 °C (212 °F). Consequently, these
Because of the guidance of the wave in magnets require some cooling for
the cladding, the beam divergence losses operation near surfaces such as hot
are small and large distances can be aluminum or steel.
tested. It is also possible to test curved
geometries like the corner range of a Cooling Procedures for High
pressure vessel nozzle with waves Temperature Tests
propagating in the circumferential
direction. Studies show the success of cooling both
the magnet and the radiofrequency coil
The general advantages of transverse for high temperature tests. One compact
horizontal waves include surface test, approach uses a pulsed bias magnetic field
corner reflection independent of the that is practical when a low duty factor is
incidence angle and no mode conversion. satisfactory.129
These features open new possibilities for
the nondestructive testing of austenitic Another effective system is designed for
and dissimilar welds in hazardous continuous wall thickness measurements
environments.128 on hot seamless steel pipe at 980 °C
(1800 °F) at six positions equally spaced
High Temperature Tests around the circumference. For this
application, both the radiofrequency coils
Electromagnetic acoustic transducers are and the magnet poles are cooled.130
well suited for high temperature Because an electromagnet is used, the
measurements because no fluid coupling system is massive but this is not typically
is required. In general, there are three a problem in a steel mill.
options available for designing these
transducers for high temperature A second alternative is to cool the
environments: (1) cool both the magnet (incidental cooling of the
radiofrequency coil and the magnet, radiofrequency coil may also occur).
(2) cool only the magnet or (3) cool When small uncooled or slightly cooled
neither the magnet nor radiofrequency radiofrequency coils are used, a smaller
coil. In practice, all three approaches have liftoff is of significant value for achieving
been used successfully. an acceptable signal-to-noise ratio.

High Temperature Material Figure 76 shows an electromagnetic
Considerations acoustic transducer made with high
temperature insulation on a high
High temperature interferes with the temperature wire. This unit can receive
operation of electromagnetic acoustic either longitudinal or transverse waves.131
transducers in several ways. The Using metal enameling or ceramic glazing
insulation on standard copper magnet techniques, the radiofrequency coil is
wire for electric motor windings is seldom bonded to a copper cooling plate. Care
rated above 220 °C (430 °F). Polyimide must be taken not to have the electrically
insulators can be used to higher conducting cooling plate too close to the
temperatures and in some cases, where radiofrequency coil. The cooling plate
low voltage insulation is adequate prevents the permanent magnet from
protection, long term operation to 350 °C heating above 40 °C (104 °F), even when
(660 °F) is possible. the radiofrequency coil is immediately
above a surface at 1090 °C (2000 °F). One
Ceramic insulation can be applied to transducer of this design had a noise
copper wire as well as high temperature threshold of 0.03 nm, with an output
sensitivity of 0.5 mV·nm–1 at 1 MHz. A
band width of 10 MHz and a spatial

Generation and Detection of Ultrasound 121

resolution of 4 × 6 mm (0.15 × 0.2 in.) In one application of such an
were also achieved. electromagnetic acoustic transducer
system, the transit time of the sound
Testing Metals in Fabrication through a steel billet is measured to
determine its internal temperature.120 This
One of the most important applications of measurement is made on billets as they
high temperature electromagnetic acoustic emerge from a continuous casting furnace
transducers is for testing metal products at located about 6 m (20 ft) upstream from
the fabrication mill. The transducers must the ultrasonic testing station.
not only be made of materials that
withstand high temperatures but also be Tests with Moving
powerful enough to overcome the high Transducers
attenuation in metals near their melting
points. Pipeline Tests

Laboratory studies of sound To achieve electromagnetic coupling, an
propagation in steel show that electromagnetic acoustic transducer coil
longitudinal waves at a frequency near and magnet need only be brought close to
1 MHz can penetrate hot steel132 and this a metal surface. There is no need for an
information was used to design a compact operator to adjust the transducer
transmitter and receiver pair of alignment or to optimize the thickness of
electromagnetic acoustic transducers133 on the coupling fluid layer. In-line tests of
each side of a 100 mm (4 in.) square block buried gas pipelines are possible without
of stainless steel.134 interrupting the flow of the gas (Figs. 77
and 78). In such an application, no
For both of these transducers, the coupling liquid is available and the test
magnetic field is supplied by flat, circular device must move unattended through
coils of heavy copper wire oriented with many kilometers of moderate diameter
their planes parallel to the surface of the pipe at speeds averaging 6.7 m·s–1 (22 ft·s–1
steel. The coils are positioned in heat or 15 mi·h–1). An inline testing robot
exchangers less than 2.5 mm (0.1 in.) (called a pig) is suited for performing this
from the steel.129 A pulse of high current test. The device is designed to test
(about 1.5 kA) through these coils applies 750 mm (30 in.) diameter pipelines using
a magnetic field of 0.5 T (5 kG) to the eight electromagnetic acoustic transducers
steel. Separate, spiral coils sandwiched distributed around its circumference.135
between the heat exchanger and a Because these transducers can excite and
ceramic cover plate form the coils that detect lamb waves in plates, the system
induce or detect eddy currents in the steel can locate corrosion in the walls of a
surface. All of the coils and heat pipeline. This location is done by
exchangers are surrounded by ceramic detecting echoes reflected from areas of
tubes about 300 mm (12 in.) in length to irregular pitting in the path of a lamb
protect the wires and coaxial cables from
the radiant heat generated by the nearby FIGURE 77. Pipe and tube testing with electromagnetic
hot ingots. acoustic transduction.

FIGURE 76. Diagram of permanent magnet, longitudinal
wave, high temperature electromagnetic acoustic transducer
that is also sensitive to transverse waves. Not shown is
cooled shield on the four open sides.

Copper Iron
Magnets

3 mm (0.125 in.) diameter cooling tube

122 Ultrasonic Testing

wave propagating around the pipe‘s When the inside of the pipe is
circumference.136 unavailable, a hand or motor propelled
carriage can be used to drive a wheeled
Eight sensors ensure adequate electromagnetic acoustic transducer probe
sensitivity in those areas where coal tar along the length of long, above-ground
coatings on the outside of the pipe cause pipes such as those found in refineries or
high attenuation of the waves. Because chemical processing plants. Here, again,
the device must withstand the pressure of the inspecting ultrasonic waves are guided
gas in the line, its central structure wave modes that propagate around the
comprises a heavy walled tube containing circumference. This technique is
batteries, signal processing electronics and particularly useful for detecting corrosion
data recording devices. The magnetic field under pipe supports from a transducer
required by the electromagnetic acoustic placed on the easily accessible top of the
transducers is generated by special pipe.
permanent magnets mounted on wheels.
These magnetize only the region around If the operator can steer the wheels of
each of the eight transducer coils. the probe, guided wave inspection
Acquired test data are recorded as the pig techniques can be applied to the
is transported along the pipe. After a test, inspection of large areas such as the
offline computers analyze the test data bottom and sides of liquid storage tanks.
recorded by an onboard computer to Figure 79 shows a mobile electromagnetic
produce a map, showing the location of acoustic transducer probe held against the
suspect areas. side of a large storage tank by magnetic
wheels. In this configuration, the operator
Pipe and Tube Testing can control the path of the probe up,
down and around the tank surface.
Electromagnetic acoustic transducers can Because the area tested is the rectangular
be mounted on wheels so that they can area under the probe itself, anomalous
move through the inside of pipes or tubes areas can be established and mapped by
in a wide variety of industrial settings. guiding the probe along a well defined
Figure 78 shows a crawler motor designed scanning pattern over the outer surface of
to pull an electromagnetic acoustic the tank.
transducer package through a small
diameter pipe in a natural gas distribution Moving Tests in Steel Mills
system. Ultrasonic guided waves are sent
from the probe around the pipe Another important application of moving
circumference to detect corrosion pits or electromagnetic acoustic transducers is
cracks by both pulse echo and found in steel mills. One example is the
through-transmission techniques. In thickness measurement discussed above.
either case, the signals are generated and Another application is based on rayleigh
detected by tiny circuits inside the probe waves, surface acoustic waves easily
and sent through cables to external signal
processing systems far outside the pipe. FIGURE 79. Mobile electromagnetic acoustic transducer test
probe held in place by magnetic wheels is being driven
FIGURE 78. Small tractor motor pulls electromagnetic around surface of large liquid storage tank. Corrosion
acoustic transducer test probe into small diameter pipe. anywhere in rectangular area under probe is detected.
Guided waves propagate around circumference to inspect
pipe wall.

Generation and Detection of Ultrasound 123

excited by electromagnetic acoustic discrete elements rather than the series
transducers using a meander coil.137 connected meander, multiple-period
radiofrequency coils that excite
By directing these waves around the narrowband, obliquely propagating bulk
circumference of tubular products, waves and guided modes when driven by
common discontinuities such as laps, gated bursts of radiofrequency current.
seams and pits can be detected. More When the discrete elements are
important, simple signal processing appropriately driven, the electromagnetic
techniques can be used to obtain a acoustic phased arrays exhibit a
quantitative measure of the depth of laps unidirectional directivity pattern and
and seams138 at production line speeds so produce broad band signals of controlled
that the manufacturer can immediately direction and focusing.139 These provide
segregate materials according to the
amount of rework needed for specified FIGURE 80. Coil configurations of
quality. electromagnetic acoustic transducer arrays:
(a) nonsegmented meander coil;
Sheet metal is supplied to rolling mills (b) segmented coil; (c) staggered,
in the form of individual coils that have segmented coil.
to be welded together to form a
continuous strip to support the (a)
continuous processing of useful products.
The weld must be of high quality to λS
maintain the integrity of the strip along
the entire production line because a IHF P
failure will shut down the line and can
harm the machines and personnel in the (b)
immediate vicinity of the break.
Therefore, it is imperative that the weld W
be tested for discontinuities as soon as it
is formed and before it begins its journey IHF
down the production line.
(c)
Electromagnetic acoustic transducers
are well suited to performing this test IHF
because they can interrogate the weld line
with guided waves produced and detected Legend
from a point downstream from the weld IHF = high frequency current
using a pulse echo technique. Figure 80 P = segmented transducer period
shows a drawing of the butt weld that W = transducer width
joins the tail of one coil to the head of λS = nonsegmented transducer period
the next coil. Two clamps hold the head
and tail sections in contact while the weld
is made and while a transmitter/receiver
pair of electromagnetic acoustic
transducers scans across the width
dimension looking for echo signals
emanating from discontinuities in the
weld. The transmitter and receiver ride on
wheels so that they can be driven across
the width of the sheet by a screw
mechanism attached to a stationary
bridge that spans the production line.

Phased Arrays for Testing of
Seamless Tube

Given the ability to operate when the
metal test object is moving past the
electromagnetic acoustic transducer at
production speeds, it would be helpful to
be able to rapidly change the
characteristics of the interrogating
radiation, to obtain as much information
as possible. One approach uses phased
arrays whose ability to excite and detect
ultrasonic beams with electronically
controlled angles of propagation and
focusing is familiar from piezoelectric
applications. Electromagnetic acoustic
transducers have been used in a similar
way. These arrays consist of several small,

124 Ultrasonic Testing

great flexibility in both the selection of (charge coupled devices). For guided
wave mode and angles of propagation. modes, this delay corresponds to the time
One important application of the latter is of flight that the ultrasonic wave requires
the need to interrogate inclined to propagate from one element to the
discontinuities from both sides to ensure other. For obliquely propagating bulk
that the largest possible signal is obtained. waves, it corresponds to the time of
propagation to an equiphase plane
The meanderlike multiple-period coil inclined at the desired angle. The signals
(Fig. 80a) is segmented into several received by the receiver elements are time
elements (Fig. 80b), each with a dipole shifted analogously. If a second set of
radiation pattern. The radiofrequency staggered coils is added (Fig. 80c) and
pulses energizing the different elements driven with the appropriate phase,
are time shifted by electronic delay lines radiation in the desired direction can be
further enhanced. The advantage of the
FIGURE 81. Echoes from 5 percent phased array technique is that the
calibration discontinuity as observed with incident angle and the direction of
phased array: (a) angle incidence for inside radiation (backward or forward) can be
diameter discontinuity; (b) angle incidence switched rapidly following computer
for outside diameter discontinuity; controlled settings of the time delays.
(c) rayleigh wave.
(a) A system that can generate and detect
signals under a selectable angle promises
Discontinuity echo high testing speed (because no couplant is
needed) and easy automation (because no
(b) mechanical adjustments are needed to
obtain the required beam pattern and
Discontinuity echo direction). Such systems have been
designed for industrial applications for
(c) ultrasonic testing of seamless ferritic tubes
with outer diameters ranging from 185 to
Discontinuity echo 650 mm (7 to 25 in.) and wall thicknesses
between 6 and 50 mm (0.25 and
2.0 in.).140-142 The sensitivity is that
needed to satisfy European standards. This
means that longitudinal notches must be
detected that reach a depth of 5 percent
wall thickness or 1.5 mm (0.03 in.).
Transverse notches must be detected that
reach a depth of 10 percent of the wall
thickness or 3 mm (0.1 in.), where again,
the minimum of both has to be chosen.

Figure 81 shows echo signals of these
calibration discontinuities obtained by the
system in a static experiment. These
exhibit the signal-to-noise ratio of 20 dB
achieved with the phased array transducer
at a circumferential speed of 2 m·s–1
(6.5 ft·s–1). To distinguish echoes from
discontinuities on the outer tube surface
from those on the inner tube surface, the
electromagnetic acoustic transducer can
be switched to another mode of operation
where only surface waves are generated.
The system was tested online in a tube
mill up to circumferential tube speeds of
2.2 m·s–1 (7.2 ft·s–1).

Moving Tests of Railroad Rail

Further applications of moving tests are
found in the railroad industry. Modern
railroads realize significant fuel savings by
applying a lubricant onto their rails to
reduce friction. Unfortunately, this
complicates conventional ultrasonic tests
because the liquid layer needed for
coupling the piezoelectric transducer with
the rail no longer wets the surface
sufficiently for transmitting the acoustic

Generation and Detection of Ultrasound 125

energy. Electromagnetic acoustic current from the motor and generator
transducers can excite and detect mounted in the center of the trailer. The
ultrasonic vibrations through the power unit also provides alternating
lubricant layer without difficulty and high current for the instrumentation housed in
speed scanning of rails can be done the box on the front of the trailer. Motive
without removal of the lubricant or power comes from a towing vehicle that
interruption of normal train schedules. houses the operator and the recording
Figure 82 shows a prototype testing equipment.
trailer143 on a section of track at the
Transportation Test Center, Pueblo, Moving Tests of Railroad Wheels
Colorado, where the Association of
American Railroads is conducting tests on In wheel rail systems, the undercarriage
the wear rate of lubricated rail under and the wheels are also exposed to high
severe loads. loads. Through dynamic working, cracks
(thermal and fatigue) occur and
This system has two electromagnetic sometimes result in total wheel failure.
acoustic transducers supported on There is no simple way to predict how
individual carriages in front of and these discontinuities evolve and their
behind the automobile tire in the center timely detection is instrumental for safety.
of the trailer. The front unit tests the head In such applications, visual tests are time
of the rail by sending transverse consuming and do not allow detection of
horizontal waves fore and aft, so that a critical discontinuities in time,
discontinuity is detected twice as it passes particularly subsurface discontinuities.
under the carriage. The transducers in the
rear send a normal beam transverse wave An ultrasonic system has been
directly to the base of the rail where it is developed for the German Railway Society
reflected back to the transducer. This wave to provide in-service tests of the wheel
is used with a pulse echo procedure to treads on its high speed trains.147,148 This
detect discontinuities in the rail web. system can detect and classify critical
discontinuities in motion using an
Both of these units use small, high ultrasonic rayleigh wave, a surface wave
efficiency electromagnets mounted on excited by electromagnetic acoustic
their carriages to supply the magnetic transducers integrated with the rail when
fields required by the electromagnetic the wheel is in contact with the
transduction process. For transverse transducer. The wave pulse travels along
horizontal waves, a magnetostrictive the surface of the wheel with little
coupling mechanism144,145 requires high attenuation. The wheel‘s response to the
tangential magnetic fields. These fields are pulse is detected by a receiving transducer
obtained by pulsing a large current either as a discontinuity echo in the pulse
through the electromagnet whenever a echo mode or as a through-transmission
test of the rail head is required. signal after several round trips.

Because the transverse horizontal The transducer shown in Fig. 83
waves can be directed at an angle near consists of an electromagnet in the lower
grazing along the surface146 a region in part of a housing mounted in a
front of and behind the transducer can be mechanical support fixed at the rail. The
tested with each pulse. Power for these magnet produces a magnetic field normal
electromagnets is supplied as direct to the wheel surface. A meander
radiofrequency coil is located on top of
FIGURE 82. Trailer supporting the central pole piece. The transducer
electromagnetic acoustic transducers for period is 7 mm (0.3 in.), selected to equal
testing head and web of installed railroad
rail. The rubber tire in the center of the FIGURE 83. Schematic diagram of electromagnetic acoustic
vehicle can be lowered for highway towing. transducer installed in railroad rail for wheel tread testing.

Wheel

λS Rail
B0 Radiofrequency coil

Magnetizing coil Magnetic yoke

126 Ultrasonic Testing

the wavelength of a rayleigh wave at chamber immediately after the weld is
about 430 kHz. formed. This is possible because the
electromagnetic coupling mechanism can
The complete system, which operated operate across a vacuum and because the
for two years in the field, was fully transducer can be made of materials that
automatic, including a microcomputer for withstand the temperature immediately
system control and for storing and after a weld — as high as 350 °C (660 °F).
evaluating test data. Two probes are
attached to each rail (see Fig. 84), one The mechanical system scans a pair of
covering the dead zone of the other. The high frequency electromagnetic acoustic
probe support is slightly flexible and transducers around the weld line on a test
allows the probe to touch the wheel as it object as it sits inside a vacuum
passes. The test is done in the pulse echo chamber.149 The object is a cylinder with a
mode. circumferential weld line near its top. Two
electromagnetic acoustic transducers
The ultrasonic frequency has been straddle the weld and test it with angle
optimized for sensitivity of discontinuity beam transverse waves, from both above
detection and attenuation, so that the and below the weld, using a pulse echo
wave examines a distance of two technique.150 Because the wall thickness
circumferences of the wheel tread during of the cylinder is only about 5 mm
testing. The lab trials have shown that a (0.2 in.) and the sensitivity specification
1 mm (0.04 in.) deep transverse saw cut in requires detection of a 0.75 mm (0.03 in.)
the center of the tread can be detected diameter flat bottom hole, the
with a signal-to-noise ratio of 15 dB, electromagnetic acoustic transducer wires
independent of the distance of the saw were curved and positioned to focus
cut from the probe. 7 MHz ultrasonic waves at the center of
the wall thickness dimension.151
An optimized sizing algorithm
distinguishes discontinuities up to a depth This focal spot is scanned over the total
of 3 mm (0.1 in.) and cross sections larger weld volume by translating the
than 10 mm2 (0.02 in.2) into three classes transducers perpendicularly to the weld
based on cross section. The classes are line while the cylinder is rotated. A
used to specify subsequent action for the personal computer controls the operation
tested wheels. and takes about 300 s to collect the echo
amplitude and position data and to make
Ultrasonic Testing in an accept/reject decision. Hardcopy
Vacuum C-scan images of the joint can be
prepared offline from data in the
Many special alloys used for rocket computer‘s memory.
engines can only be welded in a vacuum
chamber with electron beams or high High Speed Self-Aligning
power lasers. Like all welds, the full Probes
thickness of these joints must be tested
with radiographic or ultrasonic To test small objects produced at rates
techniques. This requires removal of the approaching one per second, the
test object from the chamber, testing at a ultrasonic transducer and the object must
special facility and returning to the be accurately positioned relative to one
chamber if a reweld is necessary. This slow another and then a 100 percent scan must
process has been eliminated by using be performed in a fraction of a second. By
electromagnetic acoustic transducers to using electromagnetic acoustic transducers
perform the test inside the vacuum in air instead of piezoelectric transducers
in water, the test object can be kept dry
FIGURE 84. Detail of electromagnetic and the positioning mechanism does not
acoustic transducer test probes for wheel have to be as precise (there is no
tests. refraction of the ultrasonic beam at the
surface of the test object). An example of
these simplifications is a system for
testing machine gun projectiles as they
are produced.152

The system consists of a rotating drum
that picks up each projectile from a feed
line and delivers it rotating to the gap
between the pole pieces of a direct current
electromagnet. Three electromagnetic
acoustic transducer coils mounted on
compliant substrates slide into light
contact with the projectile surface as it
enters the gap. Because the compliance of

Generation and Detection of Ultrasound 127

the substrate allows the coils to conform independent velocity difference directly
to the curved surface, the ultrasonic waves proportional to the stress is obtained. The
enter the test object over a well defined theory behind this approach and its
profile inside the object. Each sound experimental confirmation demonstrates
beam interrogates a different, critical area that the procedure produces reliable
of the projectile with a pulse echo numbers for applied stresses.153
technique while the rotation ensures that
every part of the circumference is Most practical ultrasonic stress and
inspected several times during a test cycle. texture measurements have been made
using electromagnetic acoustic transducers
If any one of the transducers detects an because the errors associated with
echo larger than a preset limit for its transducer coupling are much smaller in
channel, the object is rejected and the such a system. A small transducer that has
control computer records the event so been used successfully for stress and
that at the end of a production run, a texture measurements is shown in Fig. 85.
printout of the number of rejections in The travel time for a transverse horizontal
each channel is available. plate wave propagating between the two
receivers can be obtained within 1 ns by
Stress Measurements calculating the cross correlation function
of the two receiver signals.
Several ultrasonic velocity or transit time
techniques have been used to determine Measurement repeatability is within
both applied and residual stresses in ±3 ns with most of the error coming from
metals. The basis for these techniques is a lack of reproducibility in the transducer
stress induced shift in the velocity. The coupling to the surface. For a 30 mm
major source of error in most methods lies (1.2 in.) propagation distance, this
in determining the contribution of texture corresponds to an error of about 1 m·s–1
(material anisotropy) to the elastic (200 ft·min–1) in the velocity or an error
velocities being measured. This in stress measurements of about ±10 MPa.
contribution is a result of the In other words, at 3 mm·µs–1, 30 mm is
polycrystalline character of the material. covered in 10 µs. If the error is 30 ns or
The individual grains are essentially small, 3000 µs, the error is 300× the reading.
single crystals and the ultrasonic velocity
in each depends on its orientation. If the Conclusion
grains have a preferred orientation
(texture), the velocity in the metal Among the desirable capabilities
depends on propagation direction. The demonstrated for electromagnetic acoustic
problem is to differentiate velocity shifts transducers are operation in vacuums or
due to this effect from those due to stress. high temperature; high speed or moving
In one solution, it has been found that if tests; self-alignment; phased array
two transverse velocities can be measured compatibility; and the excitation of
where the polarization and propagation horizontally polarized transverse waves
directions are interchanged, then a texture for the measurement of stress or tests of
anisotropic weldments. The major
FIGURE 85. Horizontally polarized transverse wave, periodic drawback is lower efficiency than
permanent magnet transmitter and two horizontally piezoelectric transducers. Careful
polarized transverse wave line receivers used for velocity modeling is sometimes required to design
measurements to determine stress and texture. an optimum ultrasonic electromagnetic
acoustic transducer test. Considerable
Preamplifier knowledge about the principles and
engineering characteristics of these
devices is available to draw on in
designing custom applications.118,131,154-156

Power supply

Radiofrequency coil
Magnet

25Scale,Pmivomt point Radiofrequency coils
(1) (25)0 Magnet
0
(0) (in.)
75
(3)

128 Ultrasonic Testing

PART 10. Air Coupled Transducers157

Ultrasonic transducers operating in liquids (41) P = Kγq
and solids have many applications in
signal processing devices such as delay where K is a constant representing the
lines, resonators, convolvers and ratio of ambient pressure to density, γ is
correlators and in systems for medical the ratio of the specific heat coefficients
imaging, nondestructive testing and at constant pressure and volume (1.4 for
underwater sensing. In air, ultrasonic air) and ρ is the total density (kilograms
transducers are used for robotic and per cubic meter).
metrology applications. They have a more
limited potential at high frequencies (over An expression is derived for the
500 kHz) because of the low impedance of velocity v by substituting Eq. 41:
air and the high attenuation of sound
waves. (42) v2 = dP
dρ
The physics of sound propagation in
air is reviewed below and the parameters Using the ideal gas equation:
that control the design and
implementation of air coupled transducers (43) P = ρRT
are highlighted. Examples are given for
the most popular transducer designs and It is found that:
there is discussion of the requirements
that could increase the usefulness of air (44) v2 = γRT
coupled ultrasonics.
where R is a constant (value dependent on
Physical Principles of Air the gas) and T is temperature of the gas
Coupling (kelvin).

The theoretical framework for the Using the above equation for dry air at
propagation of sound waves in air has 0 °C (32 °F), v = 331 m·s–1. From Eq. 44, it
been known for centuries and the was found by simple taylor expansion
fundamental thermodynamic principles around 0 °C (32 °F) that:
explaining the physics of sound
propagation in air are widely detailed in (45) v = 331 + 0.1 Tc
the literature.158-161 The discussion below
sketches the basic principles for sound where Tc is the temperature in celsius.
propagation in air and gives examples of Measurements of the speed of sound in air
transducers, their performance and and its temperature dependence agree
limitations. In addition, novel materials well with these calculations. The low
such as silica aerogels are considered and velocity of sound in air results in a
their impact on the performance of air smaller wavelength at a given frequency
transducers is discussed. compared to sound waves in water or
solids. The smaller wavelength provides
Sound Velocity in Air improved test resolution for metrology
and imaging systems.
Assume that air is a classical newtonian
fluid. Euler‘s equation of motion and Mechanical Impedance of Air
mass conservation is used to set up the
wave equation: The mechanical impedance Z of a
material is defined as the ratio of the
(40) ∇2P = 1 δ2P sound wave stress or pressure to its
v2 δt 2 velocity. This leads to the more simple
and useful expression:162,163
where P is the air pressure (pascal) and v is
the wave velocity in air (meter per (46) Z0 = pv
second). The P term is related to the
density by Laplace‘s adiabatic assumption For air at 0 °C (32 °F),
for an ideal gas: Z0 = 400 kg·m–2·s–1. For comparison, the
impedance of water is 1.5 × 106 kg·m–2·s–1

Generation and Detection of Ultrasound 129

and for most solids it is in the range of the lower velocity and resulting shorter
3 to 100 × 106 kg·m–2·s–1. This value for wavelength, a 5× better resolution occurs
the impedance of air is the source of the while operating in air at the same
difficulty in coupling and exciting high frequency.
frequency waves in air. This difficulty can
be explained by noting that the power Low Frequency
excited by a transducer is proportional to Transducers
the product of the impedance of the
medium, the square of the frequency and The earliest air transducers operated under
the square of the displacement on the 100 kHz and typically belonged to one of
surface of the transducer. the following classes: modulated airflow
units, mechanical vibrating sources
One way to overcome this difficulty is (whistles), electroacoustic transducers (a
to use what is known as matching layers piezoelectric tube operating in its length
between the transducer and the air. For resonant mode), flexurally vibrating
instance, if a piezoelectric ceramic with an transducers (cantilevers clamped on one
impedance of 35 × 106 kg·m–2·s–1 is to or two ends), electrostatic transducers
excite waves in air, a matching layer with (including the automatic focus transducer
0.1 × 106 kg·m–2·s–1 is nearly ideal for described below) and microphones.160 All
coupling into the air at a single frequency, these devices operate at low frequency
because it is halfway between 35 × 106 with large displacements to generate large
and 400. power density in the air.

Sound Attenuation in Air Figure 86 is a schematic diagram of a
low frequency (under 500 kHz) bimorphic
There are three sources of sound
attenuation in air: viscous, thermal and FIGURE 86. Details of ultrasonic transducer: (a) piezoelectric
vibrational losses. Viscous losses are layering; (b) poling; (c) mode of displacement; (d) final
caused by frictional damping and are device assembly.
proportional to the coefficient of viscosity
and the square of the frequency. Thermal (a) Metal disk
losses result from the conversion to heat Barium titanate
of some energy in the sound wave and its Singular
conduction from elevated temperature circular
(high pressure) gases to low temperature electrode
(low pressure) gases. This loss depends on
the thermal conductivity and the square (b)
of the frequency. Vibrational losses are
caused by coupling the sound wave into +–
resonances of the constituent molecules
of air. These losses occur at specific (c)
frequencies and are negligible elsewhere
in the frequency domain.158.159 Nodal
diameter
Attenuation can be plotted as a
function of frequency for air at a (d) Nodal
temperature of 20 °C (68 °F), at diameter
atmospheric pressure and relative Housing
humidity of 20 percent, corresponding to Resilient washer
a water volume fraction of 4.7 × 10–3. Foam rubber
Relaxation frequencies correspond to
resonances in the molecules of oxygen Back closure Terminal plug
and nitrogen, at 12.5 kHz and 173 Hz
respectively.

At frequencies over 20 kHz, this loss
mechanism is not important and viscous
and thermal losses dominate with their
frequency square dependence. At a
frequency of 1 MHz, the attenuation is
calculated to be 101 dB·m–1 but actual
measurements yield 165 dB·m–1.161 This
loss limits a pulse echo measurement in
air at a frequency of 1 MHz to about
250 mm (10 in.) for a system with a
reasonable signal-to-noise ratio. It is
useful to compare the attenuation in air
to that of water (0.22 dB·m–1 at 1 MHz) to
understand the difficulty in using high
frequency ultrasonics in air. It is also
important to remember that because of

130 Ultrasonic Testing

ultrasonic transducer. With proper Ceramic TransducerAmplitude (dB)
electrical tuning, good band width
characteristics are obtained in both A device has been developed that operates
transmitting and receiving modes. at 200 kHz but with good band width.167
The transducer has a compact impulse
Electrostatic Transducer response that reflects the large band width
at the center frequency of 200 kHz. This
An electrostatic transducer has been good performance is the result of two
developed and is widely used in the range novel ideas in the design of the
finder of an autofocusing commercial film transducer. First, it uses a composite
camera.164 The transducer is shown in material with a ceramic volume fraction
Fig. 87 and operates as a conventional of about 1 percent. Second, it uses
electrostatic transducer. A special foil is ceramics resonant in the cross
stretched over a grooved plate, forming a polarization direction.
moving element that transforms electrical
energy into sound waves. The returning Composite ceramics are popular for
echo is transformed into electrical energy. water immersion applications because it is
possible to tailor their impedance,
A grooved metallic back plate is in dielectric constant and coupling
contact with the foil and forms a coefficient for the application. In present
capacitor. When charged, the capacitor applications, a reduction in the
exerts electrostatic force on the foil.164 impedance improves the match into air
The frequency response of the transducer with no decrease in the electromechanical
is shown in Fig. 88, highlighting the coupling coefficient. The ceramic is
broad band width essential for operation damped along its length in the resonant
of the range finder. direction and this also improves the band
width of the transducer. A focused, higher
Air Coupled System frequency, version of this device is of use
in imaging and nondestructive testing.
A 250 kHz air coupled ultrasonic system
has been used for robotic range sensing High Frequency
and wind velocity measurements.165,166 Transducers
The transducer for this work is similar to
conventional immersion transducers and High frequency transducers operate above
consists of a resonant disk of piezoelectric 500 kHz and are limited to short distances
ceramic. Because of the small
diameter-to-thickness ratio (2:1) used for FIGURE 88. Transmit and receive frequency
the piezoelectric ceramic, particular care responses of electrostatic acoustic transducer
must be taken to dampen the lateral shown in Fig. 87: (a) typical transmit
modes that interact with the casing and response; (b) typical free field receive
reduce the band width of the transducer. response.

A matching layer of epoxy resin is used (a)
to improve coupling into the air. Though
the impedance of the matching layer is 40
not ideal, the device is functional and a 35
successful metrology system has been 30
demonstrated. Such devices are typically 25
narrow band and their two-way insertion 20
loss is in the range of 40 to 50 dB. 15
10
FIGURE 87. Construction details of an
electrostatic acoustic transducer. 5

Inner ring Housing 10 20 30 40 50 60 70 80 90 100Amplitude (dB)
Grooved plate
Frequency (kHz)
Special foil
(b)
Retainer
40
35
30
25
20
15
10

5

10 20 30 40 50 60 70 80 90 100

Frequency (kHz)

Generation and Detection of Ultrasound 131

because of the high attenuation of sound necessary to have materials with low
waves in air. The traditional plane piston, impedance and low attenuation.
water immersion transducer for air
applications has been well One such material is silica
documented.168,169 In this design, the aerogel.171,172 Samples of silica aerogels
main problem is that of matching the have a measured impedance of
impedances of the ceramic and air, which 0.01 × 106 kg·m–2·s–1 and a mechanical Q of
differ by six orders of magnitude. 50.
Conventional design rules require
matching layers with impedances on the (47) Q = k
order of 1 × 104 kg·m–2·s–1. 2α

Impedance Matching Layers where k is the wave number and α is the
attenuation (neper per unit length).
In one high frequency application, a
silicone rubber matching layer with an With this material, it is possible to
impedance of 1 × 106 kg·m–2·s–1 is used to make devices that have low two-way
match a ceramic transducer to air.168 The insertion loss (less than 20 dB) and broad
two-way insertion loss of the device is 3 dB band width (greater than
about 35 dB with a fractional band width 30 percent).170 Unfortunately, silica
of 3 percent. In another application, a aerogels are difficult to work with — they
mixture of silicone rubber and glass disintegrate when exposed to water and
microbubbles is used to obtain a matching are difficult to machine to small
layer with an impedance of thicknesses. It is possible that such layers
0.3 × 106 kg·m–2·s–1. This material is used in may be cast to final thickness over a
transducers with single and double piezoelectric ceramic.
matching layers for the purpose of
imaging threads in cloth.169 Ligneous materials such as cork and
balsa wood have been studied as possible
The insertion loss of a device with two matching layer materials.172 The
matching layers is 50 dB, with a large 6 dB impedance of cork is 1.5 × 105 kg·m–2·s–1
fractional band width of 38 percent at a and the impedance of balsa wood is 8 ×
center frequency of 1 MHz. The insertion 104 kg·m–2·s–1. Cork and balsa wood are
loss is 15 dB higher than expected because ideal for matching lead zirconate titanate
of errors in controlling the thickness of ceramics to air (these ceramics have
the matching layer and, more impedances around 3.5 × 106 kg·m–2·s–1).
importantly, because of attenuation in the However, the advantage is compromised
second matching layer. when it is noted that the attenuation
expressed by the mechanical Q is 1.98 for
Mismatch and Attenuation cork and 1.52 for balsa wood. For high
frequency devices, the search continues
A study of mismatch and attenuation in for better matching layer materials.
matching layers has demonstrated that
attenuation controls the insertion loss
and the band width of the device.170 The
limit on the band width is set because the
mechanical band width (due to
attenuation) is far narrower than the
electrical band width caused by the
impedance mismatch between air and the
ceramic.

Attenuation also controls the insertion
loss — because of the impedance
mismatch, large stresses are set up in the
matching layer leading to large
attenuation. A design with multiple
matching layers is preferred because the
stress fields set up in the matching layers
are weaker and the influence of
attenuation in each matching layer is
reduced.

Matching Layer Materials

A design criterion optimizes two matching
layers when the lowest impedance layer is
not optimum.170 Not having the necessary
low impedance is a common problem
with such transducers —to obtain low
insertion losses and large band width, it is

132 Ultrasonic Testing

References

1. Allen, T.L., F.A. Bruton, D. Jolly and 10. Hayward, G. “The Influence of
R. Roch. Section 4, “Ultrasonic Pulser Parameters on the
Transducers and Piezoelectric Transmission Response of
Characteristics.” Nondestructive Piezoelectric Transducers.”
Testing Handbook, second edition: Ultrasonics. Vol. 23, No. 3. Saint
Vol. 7, Ultrasonic Testing. Columbus, Louis, MO: Elsevier (1985):
OH: American Society for p 103-112.
Nondestructive Testing (1991):
p 65-100. 11. DeSilets, C.S., J.D. Frasier and
G.S. Kino. “The Design of Efficient
2. Piezoelectric Technology Data for Broadband Piezoelectric
Designers. Bedford, OH: Morgan Transducers.” Transactions on Sonics
Matroc, Vernitron Division (1965). and Ultrasonics. SU-25, No. 3. New
York, NY: Institute of Electrical and
3. Brown, L.F. and J.L. Mason. Electronics Engineers (1978):
“Disposable PVDF Ultrasonic p 115-125.
Transducers for Nondestructive
Testing Applications.” IEEE 12. Hazony, D. and T. Kocher.
Transactions on Ultrasonics, “Finite-Response Ultrasonic
Ferroelectrics, and Frequency Control. Transducers.” Journal of the Acoustical
Vol. 43, No. 4. New York, NY: Society of America. Vol. 71, No. 1.
Institute of Electrical and Electronics New York, NY: Acoustical Society of
Engineers (1996): p 560-568. America (1982): p 203-206.

4. Mason, W.P. “An Electromechanical 13. Buchler, J., M. Platte and
Representation of a Piezoelectrical H. Schmidt. “Electronic Circuit for
Crystal Used As a Transducer.” High Frequency and Broadband
Proceedings of the IRE. New York, NY: Ultrasonic Pulse-Echo Operation.”
Institute of Radio Engineers Ultrasonics. Vol. 25, No. 2. Saint
(October 1935): p 1252-1263. Louis, MO: Elsevier (1987):
p 112-114.
5. Mason, W.P. Electromechanical
Transducers and Wave Filters, second 14. Mattila, P. and M. Luukkala. “FET
edition. New York, NY: Van Pulse Generator for Ultrasonic
Nostrand (1948). Pulse-Echo Applications.”
Ultrasonics. Vol. 19, No. 5. Saint
6. Struetzer, O.M. “Impulse Response Louis, MO: Elsevier (1981):
Measurement Technique for p 235-236.
Piezoelectric Transducer
Arrangements.” IEEE Transactions on 15. Rose, J.L. Ultrasonic Waves in Solid
Sonics and Ultrasonics. Vol. SU-15, Media. London, United Kingdom:
No. 1. New York, NY: Institute of Cambridge University Press (1999).
Electrical and Electronics Engineers
(January 1968): p 13-17. 16. MacDonald, D. Private
communication. Palo Alto, CA:
7. IEEE 176, IEEE Standard on Electric Power Research Institute
Piezoelectricity. New York, NY: (1989).
Institute of Electrical and Electronics
Engineers (1987). 17. Wickramasinghe, H.K. “Acoustic
Microscopy: Present and Future.”
8. Fortunko, C.M. Section 5, IEE Proceedings. Vol. 131, Part A,
“Ultrasonic Testing Equipment.” No. 4. London, United Kingdom:
Nondestructive Testing Handbook, Institution of Electrical Engineers
second edition: Vol. 7, Ultrasonic (June 1984): p 282.
Testing. Columbus, OH: American
Society for Nondestructive Testing 18. Ilic, D.B., G.S. Kino and
(1991): p 101-129. A.R. Selfridge. “Computer Controlled
System for Measuring
9. Posakony, G.J. “Influence of the Two Dimensional Acoustic Velocity
Pulser Parameters on the Ultrasonic Fields.” Review of Scientific
Spectrum.” Materials Evaluation. Instruments. Vol. 50, No. 12.
Vol. 43, No. 4. Columbus, OH: Woodbury, NY: American Institute
American Society for Nondestructive of Physics (1979): p 1527-1531.
Testing (March 1985): p 413-419.

Generation and Detection of Ultrasound 133

19. Hughes, M.S., D.K. Hsu and 30. Green, E.R. “Worst-Case Defects
D.O. Thompson. “Characteristics of Affecting Ultrasonic Inspection
a Prototype Unipolar Pulse-Echo Reliability.” Materials Evaluation.
Instrument for NDE Applications.” Vol. 47, No. 12. Columbus, OH:
Review of Progress in Quantitative American Society for Nondestructive
Nondestructive Evaluation [Brunswick, Testing (December 1989):
ME, July 1989]. Vol. 9A. New York, p 1401-1407.
NY: Plenum Press (1990): p 917-925.
31. Introduction to Phased Array Ultrasonic
20. Motchenbacher, C.D. and Technology Applications. Waltham,
F.C. Fitchen. Low-Noise Electronic MA: R/D Tech [Olympus NDT
Design. New York, NY: Wiley Canada] (2004).
Interscience (1973): p 47-59.
32. Wüstenberg, H., A. Erhard and
21. Krautkrämer, J. and H. Krautkrämer. G. Schenk, “Some Characteristic
Ultrasonic Testing of Materials, fourth Parameters of Ultrasonic Phased
edition. Berlin, Federal Republic of Array Probes and Equipments.”
Germany: Springer-Verlag (1990): NDT.net. Vol. 4, No. 4. Kirchwald,
p 204-205. Germany: NDT.net (April 1999).

22. Sevick, J. Transmission Line 33. Clay A.C., S.-C. Wooh, L. Azar and
Transformers. Newington, CT: J.-Y. Wang. “Experimental Study of
American Radio League (1987). Phased Array Beam Characteristics.”
Journal of NDE. Vol. 18, No. 2. New
23. Clarke, K.K. and D.T. Hess. York, NY: Plenum (June 1999): p 59.
Communication Circuits: Analysis and
Design. Reading, MA: 34. Lafontaine, G. and F. Cancre.
Addison-Wesley Publishing “Potential of Ultrasonic Phased
Company (1971): p 16-64. Arrays for Faster, Better and Cheaper
Inspections.” NDT.net. Vol. 5, No. 10.
24. Selfridge, A.R., R. Baer, Kirchwald, Germany: NDT.net
B.T. Khuri-Yakub and G.S. Kino. (October 2000).
“Computer-Optimized Design of
Quarter-Wave Acoustic Matching 35. Lareau, J.P and R.M. Plis. “Phased
and Electrical Matching Networks Array Imaging First Use
for Acoustic Transducers.” Ultrasonics Qualification Effort: BWR Feedwater
Symposium Proceedings. New York, Nozzle Inner Radius Inspection from
NY: Institute of Electrical and Vessel OD for a US Nuclear Power
Electronics Engineers (1982): Plant.” NDT.net. Vol. 7, No. 5.
p 644-648. Kirchwald, Germany: NDT.net
(May 2002).
25. Cuthbert, T.R. Circuit Design Using
Personal Computers. New York, NY: 36. Whittle, A.C. “Phased Arrays —
Wiley-Interscience (1983): Panacea or Gimmick?” Insight.
p 189-194. Vol. 46, No. 11. Northhampton,
United Kingdom: British Institute of
26. Sloan, W.W. “Detector-Associated Nondestructive Inspection
Electronics.” Infrared Handbook. Ann (November 2004): p 674-676.
Arbor, MI: Environmental Research
Institute of Michigan (1989): 37. Ginzel, E.A. and D. Stewart.
p 16.2-16.33. “Photo-Elastic Visualisation of
Phased Array Ultrasonic Pulses In
27. Lewin, P.A. and A.S. DeReggi. “Short Solids.” 16th World Conference on
Range Applications.” The Nondestructive Testing [Montreal,
Applications of Ferroelectric Polymers. Canada, August-September 2004].
New York, NY: Chapman Hall Hamilton, Ontario, Canada:
Publishing (1988): p 165-167. Canadian Institute for
Nondestructive Evaluation (2004).
28. Green, E.R. “The Effect of
Equipment Bandwidth and Center 38. Ciorau, P., D. MacGillivray,
Frequency Changes on Ultrasonic T. Hazelton, L. Gilham, D. Craig and
Inspection Reliability: Modeling and J. Poguet. “In-Situ Examination of
Experimental Results.” Review of ABB 1-0 Blade Roots and Rotor
Progress in Quantitative Nondestructive Steeple of Low-Pressure Steam
Evaluation [Brunswick, ME, Turbine, Using Phased Array
July 1989]. Vol. 9A. New York, NY: Technology.” 15th World Conference
Plenum Press (1990): p 901-908. on NDT [Rome, Italy, October 2000].
Brescia, Italy: Associazione Italiana
29. Posakony, G.J. “Experimental Prove non Distruttive [Italian
Analysis of Ultrasonic Responses Society for Nondestructive Testing
from Artificial Defects.” Materials and Monitoring Diagnostics] (2000).
Evaluation. Vol. 44, No. 13.
Columbus, OH: American Society
for Nondestructive Testing
(December 1986): p 1567-1572.

134 Ultrasonic Testing

39. Dubé, N. “Electric Resistance 50. Achenbach, J.D. “Theory of
Welding Inspection.” 15th World Ultrasound Propagation in Solids.”
Conference on NDT [Rome, Italy, Topics on Nondestructive Evaluation:
October 2000]. Brescia, Italy: Vol. 1, Sensing for Materials
Associazione Italiana Prove non Characterization, Processing, and
Distruttive [Italian Society for Manufacturing. Columbus, OH:
Nondestructive Testing and American Society for Nondestructive
Monitoring Diagnostics] (2000). Testing (1998): p 3-21.

40. Bar-Cohen, Y. Section 8, “Ultrasonic 51. Kwun, H. and K.A. Bartels.
Pulse Echo Immersion Techniques.” “Magnetostrictive Sensor
Nondestructive Testing Handbook, Technology and Its Applications.”
second edition: Vol. 7, Ultrasonic Ultrasonics 36. Guildford, Surrey,
Testing. Columbus, OH: American United Kingdom: Elsevier (1998):
Society for Nondestructive Testing p 171-178.
(1991): p 219-266.
52. Rose, J.L. and X. Zhao. “Anomaly
41. Born, M. and E. Wolf. Principles of Throughwall Depth Measurement
Optics. Oxford, United Kingdom: Potential with Shear Horizontal
Pergamon Press (1970). Guided Waves.” Materials Evaluation.
Vol. 59, No. 10. Columbus, OH:
42. Lees, S. “Useful Criteria in American Society for Nondestructive
Describing the Field Pattern of Testing (October 2001): p 1234-1238.
Focusing Transducers.” Ultrasonics.
Vol. 16, No. 5. Saint Louis, MO: 53. Badi, M.H., G.G. Yaralioglu,
Elsevier (1978): p 219-436. A.S. Ergun, S.T. Hansen and
B.T. Khuri-Yakub. ”Capacitive
43. Tarnoczy, T. “Sound Focusing Lenses Micromachined Ultrasonic Lamb
and Waveguides.” Ultrasonics. Vol. 3. Wave Transducers Using Rectangular
Saint Louis, MO: Elsevier (July 1965): Membranes.” IEEE Transactions on
p 115-127. Ultrasonics, Ferroelectrics, and
Frequency Control. Vol. 50, No. 9.
44. Madsen, E., M. Goodsitt and New York, NY: Institute of Electrical
J. Zagzebski. “Continuous Waves and Electronics Engineers (2003):
Generated by Focusing Radiators.” p 1191-1203.
Journal of the Acoustical Society of
America. Vol. 70, No. 5. Melville, NY: 54. Holland, S.D. and D.E. Chimenti.
American Institute of Physics, for “Air-Coupled Acoustic Imaging with
the Acoustical Society of America Zero-Group-Velocity Lamb Modes.”
(1981): p 1508-1517. Applied Physics Letters. Vol. 83,
No. 13. Melville, NY: American
45. O‘Neil, H. “Theory of Focusing Institute of Physics (2003):
Radiators.” Journal of the Acoustical p 2704-2706.
Society of America. Vol. 21, No. 5.
Melville, NY: American Institute of 55. Telschow, K.L., V.A. Deason,
Physics, for the Acoustical Society of R.S. Schley and S.M. Watson. Journal
America (1949): p 516-526. of the Acoustical Society of America.
Vol. 106. Melville, NY: American
46. Turner, J. Development of Novel Institute of Physics (1999)
Focused Ultrasonic Transducers for p 2578-2587.
NDT. Report No. 313/1986.
Cambridge, United Kingdom: The 56. Viktorov, I.A. Rayleigh and Lamb
Welding Institute (1986). Waves — Physical Theory and
Applications. New York, NY: Plenum
47. Lemons, P. and C. Quate. “Acoustic (1967).
Microscope.” Physical Acoustics:
Principles and Methods. Vol. 14. New 57. Achenbach, J.D. Wave Propagation in
York, NY: Academic Press (1979): Elastic Solids. Amsterdam,
p 1-92. Netherlands: North-Holland/Elsevier
(1984).
48. Lamb, H. “On Waves in an Elastic
Plate.” Proceedings of the Royal Society 58. Auld, B.A. Acoustic Fields and Waves
of London. Series A, Vol. 93. London, in Solids, second edition. Malabar,
United Kingdom: Royal Society FL: R.E. Krieger (1990).
(1917): p 114-128.
59. Graff, K.F. Wave Motion in Elastic
49 Achenbach, J.D. Wave Propagation in Solids. New York, NY: Dover
Elastic Solids: With Applications to (1963, 1991).
Scattering by Cracks. Amsterdam,
Netherlands: North-Holland/Elsevier 60. Nayfeh, A.H. Wave Propagation in
(1973). Layered Anisotropic Media with
Applications to Composites.
Netherlands: North-Holland/Elsevier
(1995).

61. Shull, P.J. Nondestructive Evaluation:
Theory, Techniques and Applications.
New York, NY: Marcel Dekker
(1999).

Generation and Detection of Ultrasound 135

62. Rose, J.L. “A Baseline and Vision of 71. Monchalin, J.-P. and J.W. Wagner.
Ultrasonic Guided Wave Inspection “Optical Generation and Detection
Potential.” Transactions of the ASME: of Ultrasound.” Nondestructive
Journal of Pressure Vessel Technology. Testing Handbook, second edition:
Vol. 124. New York, NY: ASME Vol. 7, Ultrasonic Testing. Columbus,
International (2002): p 273-282. OH: American Society for
Nondestructive Testing (1991):
63. Sun, Z., L. Zhang and J.L. Rose. p 313-325.
“Flexural Torsional Guided Wave
Mechanics and Focusing in Pipe.” 72. Monchalin, J.-P. “Laser-Ultrasonics:
ASME Transactions Journal of Pressure From the Laboratory to Industry.”
Vessel Technology. Vol. 127, No. 4. Review of Progress in Quantitative
New York, NY: ASME International Nondestructive Evaluation. Vol. 23A.
(2005): p 471-478. New York, NY: American Institute of
Physics (2004): p 3-31.
64. Ditri, J.J., J.L. Rose and A. Pilarski.
“Generation of Guided Waves in 73. Scruby, C.B., R.J. Dewhurst, D.A.
Hollow Cylinders by Wedge and Hutchins and S.B. Palmer. Research
Comb Type Transducers.” Review of Techniques in Nondestructive Testing.
Progress in Quantitative Nondestructive Vol. 5. New York, NY: Academic
Evaluation [La Jolla, CA: July 1992]. Press (1982): p 281-327.
Vol. 12A. New York, NY: Plenum
(1993): p 211-218. 74. Hutchins, D.A. “Ultrasonic
Generation by Pulsed Lasers.”
65. Rose, J.L., J. Ditri and A. Pilarski. Physical Acoustics. Vol. 18. New York,
“Wave Mechanics in NY: Academic Press (1988): p 21-23.
Acousto-Ultrasonic Nondestructive
Evaluation.” Journal of Acoustic 75. Doyle, P.A. “On Epicentral
Emission. Vol. 12. Los Angeles, CA: Waveforms for Laser-Generated
Acoustic Emission Group (1994): Ultrasound.” Journal of Physics D:
p 23-26. Applied Physics. Vol. 19. Melville, NY:
American Institute of Physics (1986):
66. Rose, J.L., K.M. Rajana and F.T. Carr. p 1613-1623.
“Ultrasonic Guided Wave Inspection
Concepts for Steam Generator 76. Schleichert, U., K.J. Langenberg,
Tubing,” Materials Evaluation. W. Arnold and S. Fassbender.
Vol. 52, No. 2. Columbus, OH: “A Quantitative Theory of
American Society for Nondestructive Laser-Generated Ultrasound.” Review
Testing (February 1994): p 307-311. of Progress in Quantitative
Nondestructive Evaluation [La Jolla,
67. Alleyne, D.N. and P. Cawley. “Long CA, July-August 1988]. Vol. 8A. New
Range Propagation of Lamb Waves York, NY: Plenum Press (1989):
in Chemical Plant Pipework.” p 489-496.
Materials Evaluation. Vol. 55, No. 4.
Columbus, OH: American Society 77. McDonald, F.A. “On the Precursor in
for Nondestructive Testing Laser-Generated Ultrasound
(April 1997): p 504-508. Waveforms in Metals.” Applied
Physics Letters. Vol. 56, No. 3.
68. Rose, J.L., Z. Sun, P.J. Mudge and Melville, NY: American Institute of
M.J. Avioli. “Guided Wave Flexural Physics (1990): p 230-232.
Mode Tuning and Focusing for Pipe
Inspection.” Materials Evaluation. 78. Aindow, A.M., R.J. Dewhurst and
Vol. 61, No. 2. Columbus, OH: S.B. Palmer. “Laser Generation of
American Society for Nondestructive Directional Surface Acoustic Wave
Testing (February 2003): p 162-167. Pulses in Metals.” Optics
Communications. Vol. 42.
69. Li, J. and J.L. Rose. “Implementing Amsterdam, Netherlands: Elsevier
Guided Wave Mode Control by Use (1982): p 116-120.
of a Phased Transducer Array.” IEEE
Transactions on Ultrasonics, 79. Cielo, P., F. Nadeau and
Ferroelectrics, and Frequency Control. M. Lamontagne. “Laser Generation
Vol. 48. New York, NY: Institute of of Convergent Acoustic Waves for
Electrical and Electronics Engineers Material Inspection.” Ultrasonics.
(2001): p 761-768. Vol. 23. Saint Louis, MO: Elsevier
(1985): p 55-62.
70. Rose, J.L. and L.E. Soley. “Ultrasonic
Guided Waves for the Detection of 80. Yamanaka, K., O.V. Kolosov,
Anomalies in Aircraft Components.” Y. Nagata, T. Koda, H. Nishino and
Materials Evaluation. Vol. 59, No. 10. Y. Tsukahara. “Analysis of Excitation
Columbus, OH: American Society and Coherent Amplitude
for Nondestructive Testing Enhancement of Surface Acoustic
(September 2000): p 1080-1086. Waves by Phase Velocity Scanning
Method.” Journal of Applied Physics.
Vol. 74. Amsterdam, Netherlands:
Elsevier (1993): p 6511-6522.

136 Ultrasonic Testing

81. Conant, R.J. and K.L. Telschow. 89. Delaye, P., A. Blouin, D. Drolet,
“Longitudinal Wave Precursor Signal L.-A. de Montmorillon, G. Roosen
from an Optically Penetrating and J.-P. Monchalin. “Detection of
Thermoelastic Laser Source.” Review an Ultrasonic Motion of a Scattering
of Progress in Quantitative Surface by Photorefractive InP:Fe
Nondestructive Evaluation [La Jolla, under an Applied DC Field.” Journal
CA, July-August 1988]. Vol. 8A. New of the Optical Society of America.
York, NY: Plenum Press (1989): Vol. 14. Washington, DC: Optical
p 497-504. Society of America (1997):
p 1723-1734.
82. Hoffman, A. and W. Arnold.
“Modeling of the Ablation Source in 90. Monchalin, J.P., R. Héon and
Laser-Ultrasonics.” Review of Progress N. Muzak. “Evaluation of Ultrasonic
in Quantitative Nondestructive Inspection Procedures by Field
Evaluation [Montréal, Canada, Mapping with an Optical Probe.”
July 1999]. Vol 19A. Melville, NY: Canadian Metallurgical Quarterly.
American Institute of Physics (2000): Vol. 25. Willowdale, Ontario,
p 279-286. Canada: Pergamon of Canada
(1986): p 247-252.
83. Hébert, H., F. Vidal, F. Martin,
J.-C. Kieffer, A. Nadeau, 91. Monchalin, J.-P., M. Choquet,
T.W. Johnston, A. Blouin, A. Moreau C. Padioleau, C. Néron, D. Lévesque,
and J.-P. Monchalin. “Ultrasound A. Blouin, C. Corbeil, R. Talbot,
Generated by a Femtosecond and a A. Bendada, M. Lamontagne,
Picosecond Laser Pulse near the R.V. Kolarik II, G.V. Jeskey,
Ablation Threshold.” Journal of E.D. Dominik, L.J. Duly,
Applied Physics. Vol. 98, Paper K.J. Samblanet, S.E. Agger, K.J. Roush
033104. Amsterdam, Netherlands: and M.L. Mester. “Laser Ultrasonic
Elsevier (2005). System for On-Line Steel Tube
Gauging.” Review of Progress in
84. Wagner, J.W. “High Resolution Quantitative Nondestructive Evaluation
Holographic Techniques for [Bellingham, WA, July 2002].
Visualization of Surface Acoustic Vol. 22A. Melville, NY: American
Waves.” Materials Evaluation. Vol. 44, Institute of Physics (2003):
No. 10. Columbus, OH: American p 264-272.
Society for Nondestructive Testing
(September 1986): p 1238-1243. 92. Monchalin, J.-P., J.-D. Aussel,
P. Bouchard and R. Héon.
85. Monchalin, J.-P. “Optical Detection “Laser-Ultrasonics for Industrial
of Ultrasound.” Transactions on Applications.” Review of Progress in
Ultrasonics, Ferroelectrics and Quantitative Nondestructive Evaluation
Frequency Control. UFFC-33. New [Williamsburg, VA, June 1987].
York, NY: Institute of Electrical and Vol. 7B. New York, NY: Plenum
Electronics Engineers (1986): (1988): p 1607-1614.
p 485-499.
93. Turner, W., T. Drake, M. Osterkamp,
86. Monchalin, J.-P., J.D. Aussel, D. Kaiser, J. Miller, P. Tu and
R. Héon, C.K. Jen, A. Boudreault and C. Wilson. “Using Computer Vision
R. Bernier. “Measurement of to Map Laser Ultrasound onto CAD
In-Plane and Out-of-Plane Geometries.” Review of Progress in
Ultrasonic Displacements by Optical Quantitative Nondestructive Evaluation
Heterodyne Interferometry.” Journal [Bellingham, WA, July 2002].
of Nondestructive Evaluation. Vol. 8, Vol. 22A. Melville, NY: American
No. 2. New York, NY: Plenum Institute of Physics (2003):
(1989): p 121-133. p 340-347.

87. Monchalin, J.-P., R. Héon, 94. Fiedler, C.J., T. Ducharme and
P. Bouchard and C. Padioleau. J. Kwan. “The Laser-Ultrasonic
“Broadband Optical Detection of Inspection System (LUIS) at the
Ultrasound by Optical Sideband Sacramento Air Logistics Center.”
Stripping with a Confocal Review of Progress in Quantitative
Fabry-Perot.” Applied Physics Letters. Nondestructive Evaluation [Brunswick,
Vol. 55. Melville, NY: American ME, July-August 1996]. Vol. 16A.
Institute of Physics (1989): Vol. 16A, New-York, NY: Plenum
p 1612-1614. Press (1997): p 515-522.

88. Blouin, A. and J.-P. Monchalin.
“Detection of Ultrasonic Motion of a
Scattering Surface by Two-Wave
Mixing in a Photorefractive GaAs
Crystal.” Applied Physics Letters.
Vol. 65. Melville, NY: American
Institute of Physics (1994):
p 932-934.

Generation and Detection of Ultrasound 137

95. Pétillon, O., J.-P. Dupuis, 102. Dewhurst, R.J., C. Edwards,
H. Voillaume and H. Trétout. A.D.W. McKie and S.B. Palmer. “A
“Applications of Laser Based Remote Laser System for Ultrasonic
Ultrasonics to Aerospace Industry.” Velocity Measurement at High
Proceedings: 7th European Conference Temperatures.” Journal of Applied
on Non-Destructive Testing Physics. Melville, NY: American
[Copenhagen, Denmark, May 1998]. Institute of Physics (1988):
Vol. 1. Broendby, Denmark: p 1225-1227.
7th ECNDT: p 27-33.
103. Aussel, J.D. and J.-P. Monchalin.
96. Choquet, M., R. Héon, C. Padioleau, “Precision Laser-Ultrasonic Velocity
P. Bouchard, C. Néron and Measurement and Elastic Constant
J.-P. Monchalin. “ Laser-Ultrasonic Determination.” Ultrasonics. Vol. 27,
Inspection of the Composite No. 3. Saint Louis, MO: Elsevier
Structure of an Aircraft in a (1989): p 165-177.
Maintenance Hangar.” Review of
Progress in Quantitative Nondestructive 104. Lindh-Ulmgren, E., M. Ericsson,
Evaluation [Snowmass Village, CO, D. Artymowicz and B. Hutchinson.
July-August 1994]. Vol. 14A. New “Laser-Ultrasonics As a Technique to
York, NY: Plenum (1995): p 545-552. Study Recrystallisation and Grain
Growth.” Materials Science Forum.
97. Fomitchov, P.A., A.K. Kromine, Vol. 467-470. Zurich, Switzerland:
S. Krishnaswamy and Ütikon, Trans Tech Publications
J.D. Achenbach. “Ultrasonic Imaging (2004): p 1353-1362.
of Small Surface-Breaking Defects
Using Scanning Laser Source 105. Moreau, A., D. Lévesque, M. Lord,
Technique.” Review of Progress in M. Dubois, J.-P. Monchalin,
Quantitative Nondestructive Evaluation C. Padioleau and J.F. Bussière.
[Brunswick, ME, July-August 2001]. “On-Line Measurement of Texture,
Vol. 21A. Melville, NY: American Thickness and Plastic Strain Ratio
Institute of Physics (2002): Using Laser-Ultrasound Resonance
p 356-362. Spectroscopy.” Ultrasonics. Vol. 40.
Saint Louis, MO: Elsevier (2002):
98. Ochiai, M., D. Lévesque, R. Talbot, p 1047-1056.
A. Blouin, A. Fukumoto and
J.-P. Monchalin. “Visualization of 106. Kruger, S.E., G. Lamouche,
Surface-Breaking Tight Cracks by A. Moreau and M. Militzer. “Laser
Laser-Ultrasonic F-SAFT.” Review of Ultrasonic Monitoring of
Progress in Quantitative Nondestructive Recrystallization of Steels.” Materials
Evaluation [Bellingham, WA, Science and Technology 2004
July 2002]. Vol. 22A. Melville, NY: Conference Proceedings [New Orleans,
American Institute of Physics (2003): LA, September 2004]. Warrendale,
p 1497-1503. MI: Association for Iron and Steel
Technology (2004): p 809-812.
99. Lévesque, D., A. Blouin, C. Néron
and J.-P. Monchalin. “Performance 107. Kruger, S.E., A. Moreau, M. Militzer
of Laser-Ultrasonic F-SAFT Imaging.” and T. Biggs. “In-Situ,
Ultrasonics. Vol. 40. Saint Louis, MO: Laser-Ultrasonic Monitoring of the
Elsevier (2002): p 1057-1063. Recrystallization of Aluminum
Alloys.” Thermec 2003 International
100. Monchalin, J.-P., R. Héon, Conference on Processing &
J.F. Bussière and B. Farahbakhsh. Manufacturing of Advanced Materials.
“Laser-Ultrasonic Determination of Part 1. Zurich, Switzerland: Ütikon,
Elastic Constants at Ambient and Trans Tech Publications (2003):
Elevated Temperatures.” p 483-488.
Nondestructive Characterization of
Materials II. J.F. Bussière, 108. Dubois, M., A. Moreau and
J.P. Monchalin, C.O. Ruud and J.F. Bussière. “Ultrasonic Velocity
R.E. Green, Jr., eds. New York, NY: Measurements during Phase
Plenum Press (1987): p 717-723. Transformations in Steels Using
Laser-Ultrasonics.” Journal of Applied
101. Bresse, L.F., D.A. Hutchins and Physics. Vol. 89, 11. Melville, NY:
K. Lundgren. “Elastic Constant American Institute of Physics (2001):
Determination Using Ultrasonic p 6487-6495.
Generation by Pulsed Lasers.”
Journal of the Acoustical Society of 109. Truell, R., C. Elbaum and B.B. Chick.
America. Vol. 84, No. 5. Melville, NY: Ultrasonic Methods in Solid State
American Institute of Physics, for Physics. New York, NY: Academic
the Acoustical Society of America Press (1969).
(1988): p 1751-1757.

138 Ultrasonic Testing

110. Aussel, J.D. and J.P. Monchalin. 119. Wilbrand, A. “Quantitative
“Measurement of Ultrasound Modeling and Experimental Analysis
Attenuation by Laser Ultrasonics.” of the Physical Properties of
Journal of Applied Physics. Vol. 65, Electromagnetic-Ultrasonic
No. 8. Melville, NY: American Transducers.” Review of Progress in
Institute of Physics (1989): Quantitative Nondestructive Evaluation
p 2918-2922. [Williamsburg, VA, June 1987].
Vol. 7A. New York, NY: Plenum
111. Dubois, M., M. Militzer, A. Moreau (1988): p 671-680.
and J.F. Bussière. “A New Technique
for the Quantitative Real-Time 120. Alers, G.A. and H.G.N. Wadley.
Monitoring of Austenite Grain “Monitoring Pipe and Tube Wall
Growth in Steel.” Scripta Materialia. Properties during Fabrication in a
Vol. 42, No. 9. Oxford, United Steel Mill.” Intelligent Processing of
Kingdom: Elsevier (2000): p 867-874. Materials and Advanced Sensors.
Warrendale, PA: American Institute
112. Jeskey, G., R. Kolarik II, E. Damm, of Metallurgical Engineers (1987):
J.-P. Monchalin, G. Lamouche, p 17-27.
S.E. Kruger and M. Choquet. “Laser
Ultrasonic Sensor for On-Line 121. Alers, G.A. ”Electromagnetic
Seamless Steel Tubing Process Induction of Ultrasonic Waves:
Control.” Proceedings of the 16th EMAT, EMUS, EMAR.” 16th World
World Conference on Nondestructive Conference on Nondestructive Testing
Testing [Montreal, Canada, [Montreal, Canada,
August-September 2004] Hamilton, August-September 2004]. Hamilton,
Ontario, Canada: Canadian Institute Ontario, Canada: Canadian Institute
for NDE (2004). for Nondestructive Evaluation
(2004).
113. Kruger, S.E., G. Lamouche,
J.-P. Monchalin, R.V. Kolarik II, 122. Kupperman, D.S. and K.J. Reimann.
G.V. Jeskey and M. Choquet. “Ultrasonic Wave Propagation in
“On-Line Monitoring of Wall Austenitic Stainless Steel Weld
Thickness and Austenite Grain Size Metal.” Transactions on Sonics and
on a Seamless Tubing Production Ultrasonics. Vol. SU-27, No. 1. New
Line at the Timken Co.” Iron and York, NY: Institute of Electrical and
Steel Technology. Vol. 2, No. 10. Electronics Engineers (January 1980):
Warrendale, PA: Association for Iron p 7-15.
and Steel Technology (2005):
p 25-31. 123. Hirsekorn, S. “Directional
Dependence of Ultrasonic
114. Maris, H.J. “Picosecond Ultrasonics.” Propagation in Textured
Scientific American. New York, NY: Polycrystals.” Journal of the Acoustical
Scientific American (1998): p 86-89. Society of America. Vol. 79, No. 5.
Melville, NY: American Institute of
115. Rogers, J.A., A.A. Maznev, M.J. Banet Physics, for the Acoustical Society of
and K.A. Nelson. “Optical America (May 1986): p 1269-1279.
Generation and Characterization of
Acoustic Waves in Thin Films.” 124. Hubschen, G. “Results for Testing
Annual Review of Material Science. Austenitic Welds and Cladding
Vol. 30. Palo Alto, CA: Annual Using Electromagnetically Excited
Reviews (2000): p 117-157. SH-Waves.” Proceedings of the Sixth
International Conference on
116. Carrion, L., A. Blouin, C. Padioleau, Nondestructive Evaluation in the
P. Bouchard and J.-P. Monchalin. Nuclear Industry. Materials Park, OH:
“Single-Frequency Pulsed Laser ASM International (1984):
Oscillator and System for p 238-289.
Laser-Ultrasonics.” Measurement
Science and Technology. Vol. 15. 125. Silk, M.G.A. “A Computer Model for
Bristol, United Kingdom: Institute of Ultrasonic Propagation in Complex
Physics (2004): p 1939-1946. Orthotropic Structures.” Ultrasonics.
Vol. 19, No. 9. Saint Louis, MO:
117. Alers, G.A. and B.W. Maxfield. Elsevier (September 1981):
“Electromagnetic Acoustic p 208-212.
Transducers.” Nondestructive Testing
Handbook, second edition: Vol. 7, 126. Ogilvy, J.A. “Ultrasonic Beam
Ultrasonic Testing. Columbus, OH: Profiles and Beam Propagation in an
American Society for Nondestructive Austenitic Weld Using a Theoretical
Testing (1991): p 326-340. Ray Tracing Model.” Ultrasonics.
Vol. 24, No. 11. Saint Louis, MO:
118. Thompson, R.B. “Physical Principles Elsevier (November 1986):
of Measurements with EMAT p 337-347.
Transducers.” Physical Acoustics.
Vol. 19. New York, NY: Academic
Press (1990): p 157-200.

Generation and Detection of Ultrasound 139