ASNT NDT Handbook Volume 10 OVERVIEW

ULTRASONIC PULSE ECHO TECHNIQUES I 383

the incident angle of the peripheral beam exceeds the first FIGURE 5. Ultrasonic beam spread in 44 mm
fl.75 in.) diameter aluminum rod tested
critical angle, shear waves are generated in addition to lon- laterally

gitudinal waves from the straight beam. The screen then BACK SURFACE SECOND MULTIPLE OF
displays nonrelevant signals occurring after the back surface REFLECTION BACK REFLECTION

reflection signal. Such signals generally do not present prob- -I

lems in interpretation because they occur after the back sur- 14 16 18 20 22 24 26 28 30 32
TIME
face reflection (Fig. 5).
Other gating methods can be used to interrogate or (microseconds)

emphasize certain depth ranges in the test object. Some

pulser/receiver units have the ability to use multiple gates.
Gates and variable gain controls may be used concurrently

to focus on certain depth ranges.

Signal processing techniques can be used for waveform
analysis in computerized contact testing systems. The advan-

tage of signal processing over conventional peak detection

and gating techniques is better discontinuity resolution and
discrimination. It is especially useful for achieving a high level
of near surface resolution. In Fig. 6a, a conventional
pulser/receiver contact testing unit shows a minimum near
surface resolution of 0.38 mm (0.015 in.) in graphite-epoxy

FIGURE 4. Ultrasonic test displays for SO-ply FIGURE 6. Longitudinal ultrasonic test results
composite: fa) direct contact transducer; from SO-ply graphite-epoxy composite:
fb) transducer with delay line fa) 0.38 mm f0.015 in.) discontinuit;y
fb) 0.013 mm f0.005 in.) delamination
faJ
faJ

INITIAL PULSE BACK SURFACE
REFLECTION
LU
~-~ ~0
O i--- V1 0 I
:fJ-- .:!:::: -I i'

<( 22 24 26
TIME
-I
(microseconds)
04 20 28

TIME
(microseconds)

fbJ INTERFACE fbJ

() INITIAL PULSE BACK SURFACE LU I
REFLECTION 0

0
~-:fJ-- ~-~ ~f:-J-V1
V1 .:!::::
.:!::::
~~ 0~ (,,--~~~~....-~~

<( <(

-I \ -I
-I
TIME 20 22 24 26 28
(microseconds)
TIME
(microseconds)

384 I NONDESTRUCTIVE TESTING OVERVIEW

composites. Figure 6b shows the radio frequency (rf) wave- applied to the transducer. In contact tests, this event occurs
form of a discontinuity that can be detected with signal pro- at a time just preceding the entry of the sound energy into
cessing techniques at a depth of 0.13 mm (0.005 in.). the front surface of the test object, after leaving the trans-
ducer and passing through the face plate and liquid cou-
Applications of Straight Beam plant. In the A-scan display, this signal is sometimes wrongly
Contact Tests considered to represent the position of the front or entry
surface of the test material.
Manual Scanning
A broadening of this pulse may be noted as the trans-
The pulse echo contact method is often applied manually ducer is placed in contact with the test object. The horizon-
and is rarely used for initial scans of large areas. Its primary tal axis of the A-scan signal display can be defined as (1) the
use is for outlining or mapping discontinuity areas. In a man- time delay followingthe initial pulse or (2) the distance from
ual scan, the technician places the transducer on the test the test object entry surface. The latter interpretation is
object and watches the screen for discontinuity indications. valid only when the sound pulse velocity is constant within
the test material, so that the round trip sound pulse transit
Whenever the display is judged to indicate a discontinu- time is proportional to the depth from which the sound
ity, the corresponding spot on the test object surface is pulse is reflected. The assumption is generally valid for most
marked with a grease pencil or an indelible marker. The uniform, nonporous elastic test materials (metals, ceramics
transducer is moved around the surrounding area and the or glasses). Thermal gradients, nonhomogeneous regions
periphery of the discontinuity is outlined by monitoring and other effects can invalidate this assumption.
where the signal drops below a preset level. Sometimes dis-
continuity areas are traced on a transparent overlay and kept Initial Pulse
as a permanent record of the test.
Widening of the initial pulse indication can result from
Displayed Indications of Discontinuities transducer ringing (inadequate damping), lengthening of the
excitation pulse or (in some cases) use of thick couplant lay-
A typical straight beam ultrasonic contact test oscillo- ers or presence of rough entry surfaces on the test objects.
scope display is shown in Fig. 4. The horizontal displace-
ment indicates the time delay between successive ultrasonic Such prolonged initial pulse indications can obscure or
signals. The first vertical indication, reading from left to hide reflection signals from discontinuities or interfaces that
right, corresponds to the initial pulse of electrical energy lie close to the entry surface. This produces the so-called
dead zone, the distance from the test surface to the nearest
FIGURE 7. Typical ultrasonic test display depth at which discontinuities can be reliably detected. The
showing initial pulse, discontinuity indication extent of a dead zone can be reduced by: (1) selecting a
and back surface reflection higher pulse frequency, (2) selecting a minimal pulse length
and (3) reducing the pulsing time on the instrument. In
-I 24 26 28 30 addition, another transducer may be selected with dual-
22 TIME element construction, a more efficient piezoelectric ele-
ment or more effective damping characteristics. In cases
(microseconds) where pulse length is considered too long for the applica-
tion, it may be possible to scan the material from the reverse
side to cover the obscured areas.

Discontinuity Signals

The second vertical indication in Fig. 7 represents an
echo signal from a discontinuity in the test material. Its
amplitude is relatively low compared to the front surface
signal. The discontinuity reflects only a portion of the inci-
dent sound pulse when the sound beam diameter is greater
than the reflecting cross-sectional area of the discontinuity.

Many discontinuities are not flat or perpendicular to the
incident sound beam and serve to scatter the ultrasonic
energy so that it does not return to the transducer. Such dis-
continuities might be indicated by reduced echo signal levels.
The amplifier type and its gain setting also have strong effects

ULTRASONIC PULSE ECHO TECHNIQUES I 385

on the relative amplitude of discontinuity echo signals. Use of appearing in the gate above a predetermined level triggers a
discontinuity indication device.
time varying signal amplifier gain controls is typically limited
to the far field region of the ultrasonic transducer. Discontinuities between the front and back surfaces
reflect sound energy back to the transducer within the time
Back Surface Reflections period set by the gate. The orientation and geometrical
shape of the discontinuity, with respect to the test object
The third vertical indication (on the right of Fig. 7) cor- surface and the transducer orientation, are important in
responds to the strong echo returned from the far surface of determining the intensity of the ultrasound reflected back
the test material when ( 1) this surface is perpendicular to to the receiving transducer.
the sound beam and (2) the material transmits ultrasound
with little attenuation. This back surface signal could be Figure 9 shows the ultrasonic signals reflected from a
reduced in amplitude or could disappear completely if the series of geometric shapes embedded in plastic. A planar
intervening discontinuity is large enough in area to intercept discontinuity parallel to the front surface is a good reflector,
most of the beam from the transducer. Loss of back surface redirecting energy back to the receiving transducer (Figs. 9a
indications without loss of coupling is an immediate warning and 9b). A spherical discontinuity scatters ultrasound and
that the sound beam from the transducer may have been reflects only a small amount of the impinging energy at the
interrupted by a large discontinuity within the test material apex closest to the test object surface (Fig. 9d). A planar dis-
or scattered by rough or angled discontinuities. continuity perpendicular to the test object surface presents
a small area to the beam, reflecting a small amount of the
The loss of back surface reflected signals can also occur incident energy back to the receiver (Fig. 9c).
in highly attenuative media. In this case, it may be necessary
to reduce the test frequency to decrease the attenuation. To detect clusters of small spherical discontinuities such
Only when clear back surface signals are displayed can it be as porosity in composites, a second gate may be set to moni-
ensured that the ultrasonic beam intensity is adequate for tor the amplitude of the back surface echo. In the case of
producing detectable discontinuity signals. Because lower many small discontinuities (or any discontinuity that diverts
test frequencies decrease the detectability of small disconti- the reflected signal away from the receiving transducer), the
nuities, higher amplifier gains may be required for detection ultrasound reaching the back surface is less than normal and
of small discontinuities. the signal amplitude is diminished. With a continuous auto-
mated recording procedure, this back surface amplitude
Only a small portion of the initial pulse energy is method can be used only in homogeneous material when
returned to the transducer at the end of its first round trip the ultrasonic test system provides a second gate or when
through the test material. As each echo returns to the front digitized waveform analysis is used.
surface, only a small fraction of its energy passes through
the material surface and the couplant to the transducer. FIGURE 8. Multiple ultrasonic reflections in
7 mm (0.27 in.J aluminum block
While the decaying sound beam continues to reflect back
and forth within the test object, the search unit continues to 0
receive weakening signals when each echo reaches the entry
surface under the transducer. Sound pulse energy remaining
within the test material continues to bounce back and forth
between the parallel front and back surfaces. A number of
these multiple back reflections and internal discontinuity sig-
nals may be observed on the screen by extending the time
scale (Fig. 8). Because the signal multiples become more
complex, it is often preferable to adjust the time sweep of the
tube to display only one or two round trip periods.

Discontinuity Discrimination -I 18 20 22 24 26
I6
The most commonly used means of automated disconti- TIME
nuity discrimination is known as amplitude gating. A moni- (microseconds)
toring gate is set by electronically selecting a period of time
in the horizontal time sweep of the pulser/receiver. The gate
is typically set between the front and back surface reflection
signals or in the specific area of interest and any signal

386 I NONDESTRUCTIVE TESTING OVERVIEW

FIGURE 9. Straight beam ultrasonic reflections: DiscontinuitiesDetected by the
(a) from right disk head on; (b) from hexagonal Straight Beam Method
disk head on; (c) from side of right disk;
(d) from sphere Inmetals, common discontinuities are voids, foreign inclu-
sions,cracks and other anomalies. Signalsreflected from these
fa) discontinuities are dependent on shape and orientation of
principal reflecting surfaces. Because few fatigue cracks are
0 .- r oriented with reflective surfaces parallel to the object surface,
angle beam ultrasonic testing is generally used to detect
-I fatigue cracks.

fb) 22 24 26 28 30 Discontinuities commonly seen in composites are delam-
w TIME inations, voids, porosities and ply gaps (Fig. 10). Inclusions in
0 composites are more difficult to detect because their compo-
(microseconds) sition is sometimes close to the acoustic impedance of the
:::) vi' composite and insufficient energy is reflected from the dis-
I- .:!:: .- continuity interfaces. Figure 11 shows signals reflected from
a reference standard with implanted inclusions of synthetic
~-:J ~ fluorine resin and paper. Figure 12 shows signals reflected
from a reference standard with implanted inclusions of
<( polyethylene and polyester films.

-I Sizing Discontinuities

22 24 26 28 30 When discontinuity sizes are larger than the diameter of
TIME the ultrasonic transducer, the outline of the discontinuity
can be obtained by marking positions where the discontinu-
(microseconds) ity signals drop by a preset amount such as 50 percent. With
an appropriate transducer diameter and test frequency, the
fc) contact pulse echo ultrasonic technique is one of the most
accurate means for sizing discontinuities in this way.
0 Ii
In manual scanning, the outline is made by marking the
-I location of the center of the transducer on the test object. In
computerized pulse echo contact scanning, an algorithm
22 24 26 28 30 can be written into the data analysis program to mark and
TIME refine the outline of the discontinuities.

(microseconds) If the size of a discontinuity is smaller than the diameter
of the transducer, a reference standard containing known
fd) discontinuity sizes is necessary for calibration. A common
reference standard configuration has flat bottom holes of
w " ,r graduated sizes machined into assorted metals. Testing
0 specifications of many metallic materials are often classified
:::) vi' 30 in terms of the diameter of these flat bottom holes. Distance
amplitude correction devices are used to standardize the
I- .:!:: reflected signals from different depths in the material.

~-:J ~ Reference standards are also made for composites to aid
in sizing small voids and delaminations. However, inclusions
<( in composites are also a major concern, particularly because
-I detection of inclusions is more difficult than detection of
delaminations. Composite reference standards are made
22 24 26 28 using different materials as implants in a known matrix mate-
rial. Such implant materials generally are synthetic fluorine
TIME

(microseconds)

ULTRASONIC PULSE ECHO TECHNIQUES I 387

FIGURE 1 0. Typical ultrasonic test signals from FIGURE 11 . Ultrasonic test signals from SO-ply
six-ply fabric composite: (a) well bonded material composite: (a) front and back surface reflections
front and back surface reflections; (b) porosities; with no discontinuity; (b) synthetic fluorine resin
(c) delamination at five-ply depth inclusion; (c) paper inclusion (see Fig. 12)

(aJ FRONT SURFACE REFLECTION FRONT SURFACE REFLECTION
BACK SURFACE
BACK SURFACE REFLECTION / /(aJ
REFLECTION

0lJ..J 0 1----~·11w,.-..._,.,.,...,__.l,111l~----

~~,-

<(

-1 .-1 ' . '

20 22 24 26 28 20 22 24 26 28
TIME
TIME
(microseconds) (microseconds)

(bJ (bJ / /FRONT SURFACE REFLECTION
INCLUSION
lJ..J l0J..J
::i Vl BACK SURFACE
0 / REFLECTION
::i Vl ~-f- ~
~rlrA,__~~·0 ~--~11~....,...-,,.. .........~,....
~-f- ~ ::J ~
<(
::J ~
<( -1

-1 22 24 26 28 20 22 24 26 28
20
TIME TIME
(microseconds) (microseconds)

(cJ I I(cJ FRONT SURFACE REFLECTION
PAPER INCLUSION

BACK SURFACE
REFLECTION

lJ..J 0 lJ..J 0 1-----l!IIIJ.-,...,..,_• .,_,,na, ;;

0 0::i Vl .1.
::i Vl f:-:J ~~
n ....,J,IIIIL....-.....,...,,,,,
f- ~ ~- r

~-::J ~ <(

<(

-1 -1

20 22 24 26 28 20 22 24 26 28

TIME TIME
(microseconds) (microseconds)

388 I NONDESTRUCTIVETESTINGOVERVIEW

resins or polyester films. Composite reference standards are the location of the transducer or a sonic system, similar to
more often used for setting of detection levels than for dis- graphic digitizing devices, can be used.
continuity sizing.
Mechanical Transducer Location
Mechanical Scanning
Several forms of mechanical linkage are used to provide
Manual scanning is the most commonly used contact the transducer location. One device uses a pair of perpen-
pulse echo ultrasonic testing method. The disadvantage of dicularly oriented translation bars with drive screws. The
manual scanning is that sometimes the permanent record is transducer slides along one bar, turning a screw. A rotational
an overlay with discontinuity locations traced from mark- encoder attached to the screw provides the angular dis-
ings on the test object. In many cases, permanent C-scan placement and this, in turn, gives the X-direction displace-
records are needed to properly document the ultrasonic ment. After an X-direction scan is completed, the bar is
test procedure. To produce a C-scan recording, the location indexed a given distance, turning the other screw by an
of the contact transducer during scanning must be known angle corresponding to the indexing distance.
at all times. A mechanical linkage can be used to provide
Selection of Ultrasonic Test
FIGURE 1 2. Ultrasonictest signals from SO-ply Frequencies
composite:(aJ polyethyleneinclusion;
(bJ polyesterfilm inclusion(see Fig. 11 J Contact tests are generally limited to frequencies in the
range from 0.1 to 10 MHz. The frequency most frequently
(aJ FRONT SURFACE REFLECTION used in tests of metallic materials is 2.25 MHz in the United
States and 2 MHz elsewhere. Frequencies as low as
POLYETHYLENE INCLUSION 0.4 MHz are used for tests of coarse grained metals and
alloys in order to reduce the scatter of the ultrasonic energy
BACK SURFACE from grain boundaries; the scatter in turn reduces the
energy's penetrating ability.
I IREFLECTION
MULTIPLE ECHO With fine grained materials such as wrought aluminum
alloys, 5, 7.5 or 10 MHz can be used to provide improved
-] resolution of small discontinuities. Transducers labeled as
highly damped 25 MHz are also used to test advanced com-
20 22 24 26 28 posites with improved near surface resolution. Even higher
frequencies (e.g., 100 MHz) are used for noncontact scan-
TIME ning of advanced ceramics.
(microseconds)
Effects of Ultrasonic Transducer
(bJ FRONT SURFACE REFLECTION Diameter
POLYESTER FILM
BACK SURFACE REFLECTION Selection of the transducer face diameter can have sig-
nificant effects on ultrasonic test results. The distribution of
~-L:0:U) .0 sound intensities within the beam close to the face of the
=v==;­ I ll transducer is influenced by the face diameter D as well as by
I:-J- I
~ the selection of the test frequencyf In addition, the velocity

<( -I of sound v in the medium adjacent to the face of the trans-
ducer influences the distance along the ultrasonic beam axis
20 22 24 26 28 in front of the search unit within which nonuniform sound
field intensity variations influence test signal characteristics.
TIME
(microseconds) In the near field or Fresnel field, the amplitudes of echo
signals from discontinuities can vary widely and lead to mis-
interpretation of discontinuity size or location. Considerable

ULTRASONIC PULSEECHO TECHNIQUES I 389

caution should be exercised when interpreting test indica- positions well into the far field, wavefronts can approximate
tions from the near field. those of plane waves and ultrasonic beams diverge in this

Transducer Near Field region. The intensity of the sound beam propagating from
the transducer decreases with increasing distances into the
far field according to an inverse square law.

Equation 2 expresses the half angle of sound beam diver-
gence in the far field:

The length of the near field zone f is given by: 1.22-2:_ (Eq.2)
0
02 sin a

f = 02­ f 4A. (Eq. l) Where:

4v

Where: A = the wavelength (meters) of the ultrasound in the
material; and
D diameter of the transducer (meters);
D = the face diameter (meters) of the transducer.
f test frequency (hertz);
With most practical ultrasonic transducers, the beam
v velocity of sound in the test material ( meters per divergence angle is small, typically under 5 to 10 degrees.
second); and For this reason, Eq. 2 can be approximated by assuming the
sine of the angle to be nearly equal to the angle itself:
A wavelength of sound in the test medium (meters).

The length of the near field zone should be routinely calcu- a = 1.22~ (Eq. 3)
lated to avoid interference with test procedures and signal
interpretation. fD

Also significant in its influence on discontinuity signal Where:
amplitudes is the position of the discontinuity within the
near field or far field of the transducer's ultrasonic beam. As a the ultrasonic beam half angle (radians);
might be expected, a specific loss in ultrasonic beam energy v the velocity (meters per second) of sound in the
results from each unit of distance traveled by the beam. This
rate of attenuation might be constant within a particular test material;
type of material. For example, it might be anticipated that
the farther a discontinuity of specific size is from the front f frequency (megahertz); and
surface, the smaller is its reflection. Likewise, a lesser
amount of reflected energy is received by the transducer D transducer face diameter (millimeters).
and results in a smaller indication height on the oscillo-
FIGURE 13. Typical resolution block calibration
scope. In practice, this is not always true. curve
If the height of the A-scan echo signal from a specific
3.0 ,~I I I I I I
discontinuity is plotted as a function of the discontinuity's I II I I
distance from the transducer, a curve such as that shown in I- -Qg) 2.4 I
Fig. 13 is obtained. Initially, the height of the A-scan signal 2.0 --Pf -·-·- ,_~I
becomes greater. This effect is characteristic when the dis- I II x-.__
continuity lies within the near field of the search unit. Maxi- ---·-x ~xI-xI.....-~1 -~I~ -·
mum A-scan signal height is attained at a discontinuity \w:Jo~.. I III II ---x- t---
depth determined by Eq. 1. The distance at which a given I
reflector produces this maximum A-scan signal height indi- -I I I I ---x
cates the limit of the near field under the specific test condi- I
tions and material. l-==-r~ I
IE 1.4 --I I I I _ ..I.
1.0 :!.-
<z -~__J [\i ..I.
Q.J -I I I I

\:J 1u 32
vi~ 0.4 (l .25)
I II I

0 I II II

11 19 25
1.3 5 7.5 10 13

(0.05) (0.2)10.3)10.4)10.5) (0.75) (1.0)

Divergence of Ultrasonic Beams in DISTANCE FROM SURFACE
the Far Field millimeters {inches)

In the far field, coherent wavefronts tend to be spherical LEGEND
with increasing radii of curvature as distance increases. At
• = 3.25 mm (0. 13 in.) HOLE
x = 2 mm (0.08 in.) HOLE

.i. = l mm (0.04 in.) HOLE

390 I NONDESTRUCTIVE TESTING OVERVIEW

For longitudinal waves in steel with a sound velocity of of a steel block These transducers are particularly useful for
5.84 km-s", the divergence half angle is approximately: tests of thin materials, high resolution C-scan imaging and
determination of bond quality in sandwich structures.
a= 7.11 (Eq.4)
Generally, focused transducers allow the highest possible
fD resolving power with standard equipment because the
energy can be focused at the discontinuity, making its reflec-
For longitudinal waves in steel with a sound velocity of tion very high and providing a ratio up to 10,000:1 between
230,000 in-s+, the divergence angle is approximately: the front surface and the discontinuity echo.

a 0.28 (Eq. 5) Using a cylindrical lens, an improvement in resolving
power can be achieved to a lesser degree without decreasing
Where: fD the horizontal width of the beam. With this construction, a
75 mm (3 in.) wide beam can provide clear resolution of dis-
f frequency (megahertz); and continuities from about 2.0 mm (0.08 in.) below the surface
to a depth of 13 mm (0.5 in.). For standard ultrasonic equip-
D transducer face diameter (inches). ment operating at 10 MHz, such transducers are produced
for tests of thin airframe aluminum extrusions.
For a transducer with a diameter of 25 mprmod(u1ctinJ.)Dan=d a
low contact test frequency of 1 MHz, the 1. Focused beams reduce the effect of surface roughness
and the effect of multiple minute discontinuities such as
For this example, the divergence half angle in the far field in grain boundaries and porosity. In addition, these transduc-
ers produce plane wave behavior at the focal spot and are
steel is about 16 degrees. At a higher test frequency of used in appropriate experimental studies.1

5 MHz, this angle is reduced to about 3 degrees. Focused transducers are more commonly used in the
Under typical ultrasonic contact test conditions, the immersion environment where they are not exposed to wear
or erosive conditions. Such transducers are also preferable
decrease in intensity because of geometrical beam spread- for tests of curved surfaces. An ultrasonic beam encounters
divergence at a curved surface when a flat transducer is used.
ing in the far field is usually relatively small, particularly A focused beam maintains a circular or cylindrical wave front
where the total travel distance is limited. However, when and may minimize distortion by a curved surface (Fig. 14).

the sound beam must travel long distances through a large Ultrasonic Lenses
forging, for example, the geometrical loss in beam intensity
Ultrasonic beams are focused using three basic tech-
could contribute to a significant reduction in echo signal niques: (1) a curved, ground piezoelectric material, (2) a
amplitudes. piano-concave lens cemented to a flat piezoelectric crystal
and (3) a biconcave lens placed in front of the transducer.
Focused Beam Immersion Curved transducers provide a well defined acoustic field
Techniques with limited noise and energy losses. The need to fix a
damper on the back of the crystal (to obtain broad band
Focused Transducers characteristics) makes the use of curved crystals less practi-
cal. Lenses glued to crystals are more commonly used.
Sound can be focused by lenses in a manner analogous to
focusing light. The basic difference between the two is the Spherical lenses can be used in most nondestructive test-
ratio of the lens thickness to the wavelength. In optics the ing applications. When a line focus is needed, cylindrical
lens thickness is 104 to 105 times the wavelength. In ultra- transducers are used. In some special applications, an exter-
sonics, the lens thickness is about 10 times the wavelength. nal lens is attached in front of a flat transducer to focus its
The resulting sound waves with opposite phases, emitted beam and improve test sensitivity. Such lenses produce a
from the surface of an acoustic lens in concentric rings sev- certain amount of disturbance and absorption due to the use
eral millimeters apart, interfere at the focal plane. of materials (such as vulcanized rubber) that contain fillers.

Acoustic lenses can improve testing reliabilityby reducing The main disadvantages of acoustic lenses are aberra-
and controlling certain energy losses. They are usually an tions and the energy loss from reflections and attenuation.
integral part of the transducer assembly.In most applications, Table 1 lists typical lens materials. Most lenses are made
the lens concentrates energy into a long and narrow beam, from plastics for a low reflection coefficient. Unfortunately,
increasing its intensity. Special sharp focused transducers can plastics are highly attenuative. To reduce attenuation, ultra-
be made with a usable test range less than 6.4 mm (0.25 in.). sonic applications use so-called zone lenses, the acoustic
Such a focused transducer can resolve a 0.4 mm (0.015 in.) equivalent to Fresnel lenses in optics.
flat bottom hole located 1 mm (0.04 in.) beneath the surface

ULTRASONIC PULSE ECHO TECHNIQUES I 391

TABLE 1 . Acoustic properties of various lens materials

Material Sound Velocity Density Acoustic Impedance In Water'
fkg·m-3JC Rb
V2 Z2
(km-s-1) f106 kg-mm-2-s-1) {percent)

Acetone 1.16 790 0.92 5.6 1.28
Acrylic 2.75 1, 180 3.25 13.8 0.54
Aluminum 6.25 2,700 16.90 70 0.24
Glass 4 to 6 2,300 to 4,000 13 to 16 63 to 69 0.25 to 0.37
Polyethylene 1.95 1 .75 0.76
Polystyrene 2.35 900 2.5 0.6 0.63
1,060 6.4

a. LISTEDVALUESARE FOR 21 = 1.49 x 107 g-mm-2-s-1 AND V1 = 1.49 x kms+.
b. R :<,; (Z2 - Zd / (Z2 + ZiJ.
c. AT 25 °C (77 °FJ.

FIGURE 14. Improvement obtained by using In a zone lens, rings are scribed on a plate so that every
focused transducer on curved test surface: second ring is in phase, generating constructive interference
(aJ distorted A-scan image obtained by flat at a predetermined point. This point is dependent on the
transducer; (bJ elimination of distortion with frequency and is regarded as the focal spot. The rings that
focused transducer are out of phase are covered to eliminate their contribution.
Zone lenses have found very limited application in nonde-
LJfaJ structive testing.
FLAT TRANSDUCER
Focal Distance
V~ TUBING
The focal distance of a transducer is measured experimen-
• tally with a small ball target from which its reflection is exam-
ined. It is assumed that a spherical wave front is produced
LJfbJ and the surface of the ball behaves as an equal phase reflec-
CONTOURED TRANSDUCER tor. The focal point is inferred to be at the geometric center of
\I/ J the ball. When the ball diameter is larger than the diffraction
QWBING limit, then it is common to add the radius of the ball to the
water path. This path is measured by a pulse echo time-of-
• flight test. The transducer is moved back and forth with the
sphere along its axis, until maximum amplitude is measured.

The focal distance becomes shorter when the ultrasonic
beam propagates from a fluid to a solid material. The reduc-
tion of the focal distance can be determined from a geomet-
rical analysis of the position of the front surface of the
material along the beam path (Fig. 15).

Because of the difference in velocities for metal and
water, small changes of the water path have a relatively
small effect on the focal depth in the metal. The metal sur-
face forms a second lens that is much more powerful than
the acoustical lens itself. This effect pulls the focal spot very
close to the metal surface, compared to the focal length of a
transducer in water.

The second lens has three other important effects (the
composite focusing effect of Snell's law in the metal): it
sharpens the beam, increases the sensitivity to objects in the
focal zone and makes the transducer act as a very directional
and distance sensitive receiver. Large increases in sensitivity

392 I NONDESTRUCTIVETESTING OVERVIEW

are produced by these complex interactions. This makes it the A-scan display (often called grass). Such nonrelevan
possible to locate minute discontinuities and to study areas signals can interfere with signals from significant disconti
that produce very low amplitude reflections, including the nuities and may completely obscure important test results
bond juncture between stainless steel and electroformed In most cases, the only practical solution is to reduce the
copper, for example. magnitude of grain boundary signals by switching to a lowei
frequency. This approach could also reduce the amplitude
Ultrasonic Beam Attenuationby of signals from small discontinuities to the point that the)
Scattering might be missed during ultrasonic testing.

Much more important than beam divergence is the The heights of A-scan indications depend directly on the
effect of ultrasonic beam scattering and incoherence result- transmissibility of ultrasound in specific test materials. Sud
ing from the internal structures of metallic materials. In effects can be significant in tests of large steel castings 01
most cases, the primary cause of scattering is the coarse forgings (copper, nickel based alloys or austenitic stainless
grain in many metals and alloys. Each grain boundary serves steels) whose grain size and ultrasonic scattering effects are
as a small reflector that sends out its own scattered and of much smaller magnitudes. In aerospace aluminum struc-
reflected signals. Large grain boundaries weaken the inter- tures, small discontinuities can be detected and located pre-
rogating sound beam emitted by the transducer, in addition cisely with relative ease.
to attenuating echo signals returning from discontinuities
deep in the test material. The overall effect is that of weak- Effects of Mechanical and Thermal Processes
ening significant signals and lowering the amplitude of the
A-scan signal heights. The microstructures and grain sizes of metals and alloys
can be changed significantly by mechanical and thermal
Coarse grain boundaries can return detectable echoes of processes commonly used in primary mills and in subse-
their own. These appear as numerous vertical indications on quent manufacturing operations. It is essential for the ultra-
sonic test operator to know the actual condition of materials
FIGURE 1 5. Ultrasonicfocus effect in metals, at the time of testing. Hot working and heat treatment oi
demonstratingeffect of secondlens as resultof steels can produce radical changes in grain size and internal
immersionin water microstructure and these can in turn change ultrasonic test
sensitivity and reliability. Forging and other hot working
FOCUSED operations (at temperatures high in the austenitic range)
TRANSDUCER involve plastic deformation and tend to refine grain size
(reduce grains to smaller dimensions).
LENS
On the other hand, prolonged heating at the high tem-
BEAM peratures used for forging can lead to grain size enlarge-
ment. When carbon steels are heated from ambient
WATER temperatures up through the critical ranges 720 to 880 °C
(1,330 to 1,620 °F), grain refinement occurs as the steel is
.----f 'f ·.METAL I converted to its higher temperature microstructure. Such
heating through the critical range occurs in processes such
NEW POINT OF as normalizing, hardening or annealing of steels in heat
FOCUS IN METAL ,,. \ 1 treating operations.

~.·· . --._ DIVERGENCE In some cases, materials are coarse grained and transmit
FOCAL DISTANCE __... BEYOND FOCUS ultrasound poorly as cast or after fusion welding and can be
difficult to test by ultrasonics. After subsequent heat treat-
IF IN WATER ing, mechanical working or reduction operations that refine
their grain size, sound propagation improves, permitting
precise discontinuity detection and location at considerable
depths below the entry surface. It is for these reasons that
some products are tested at particular stages of their manu-
facture when ultrasonic transmissibility is improved .

Effects of Composite Structures

Compared to metals, composite materials generally have
higher sound attenuation because of their heterogeneous

ULTRASONIC PULSE ECHO TECHNIQUES I 393

nature. Thermosets and thermoplastics generally have TABLE2. Typical frequencyranges for
about the same sound attenuation characteristics. Com- conventionalstraightbeam longitudinalpulse
pared to unidirectional tapes, woven fabric composites have echo ultrasonictest applications
higher attenuation than do unidirectional tapes.
Frequency Range Applications
Metal matrix composites have sound attenuation levels
between those of metals and organic composites. Refractory 25 to 100 kHz concrete. wood poles, rock and
composites such as carbon-carbon have the highest sound 0.2 to 2.25 MHz coarse grained nonmetallic materials
attenuation level because of their high porosity. Testing
refractory composites requires special care in frequency 0.4 to 5 MHz castings (gray iron, nodular iron),
selection and signal interpretation. 1 to 2.25 MHz relatively coarse grained metallic
1 to 5 MHz materials (copper. austenitic stainless
Selection of Test Frequencies 1 to 10 MHz steels, nickel alloys), plastics (solid
2.25 to 10 MHz rocket propellants) and grains
Pulsed contact tests use frequencies from 25 kHz to
10 MHz (see Table 2). Because of variations in metallic castings (steel. aluminum, brass) and
structure typically encountered during contact tests in materials with refined grain size
industrial plants, it is often desirable to use the lowest fre-
quency that locates specified minimum sizes and types of welds (ferrous and nonferrous)
discontinuities with consistent results.
wrought metallic products (sheet,
Proper selection of the test frequency requires experi- plate, bars, billets)
ence with similar test materials, careful analysis or experi-
mental tests. Tests with reference standards may be needed forgings (ferrous and nonferrous)
to check the reproducibility of test procedures and the uni-
formity of response. drawn and extruded ferrous and
nonferrous products (bars. tubes,
In fine grained steels, contact tests are usually made at shapes), glass and ceramics
2.25 or 5.0 MHz, when the test is used to detect forging
bursts, flaking, pipe and discontinuities of smaller size. A grains and very good ultrasonic transmissibility. For these
10 MHz frequency is sometimes selected for detection of materials, test frequencies of 5, 7.5 or 10 MHz are preferred
microscopic inclusions and segregations. for high resolution of small discontinuities. Varying grain
sizes may characterize cast and worked aluminum alloys,
Large, medium carbon steel castings are generally tested magnesium, titanium and other alloys, for which frequen-
at l to 5 MHz with ultrasonic beams penetrating 3 m (10 ft) cies of 2.25 to 5 MHz might be used.
or more. Small forgings are tested at 5 to 10 MHz and large
l<>rgings at 2.25 to 5 MHz. The central portion of forgings Austenitic stainless steel and high alloy castings used in
st rch as turbine rotors may have large grain sizes typical of nuclear power systems can often have grain sizes on the
t I 1c original cast shape, whereas the near surface material order of 2.5 mm (O.l in.) or more. These may require com-
rnay have received considerable grain refinement as a con- binations of low test frequencies, high pulse power levels or
s< ·quence of more extensive hot working during forging. focused ultrasonic beams to compensate for the severe scat-
tering and attenuation that occurs. As with most nonde-
High carbon and high alloy steels may require a lower structive tests, when performance or sensitivity is in doubt,
tr-st frequency (500 kHz to 1 MHz) if the ultrasonic beam is results should be compared to reference standards with
lo penetrate over 1 m (3 ft). This also depends on the degree known, similar materials structures or similar ultrasonic
< ,I' working or heat treatment of the material. Lower test fre- transmissibility.
< 111encies are often used for cast iron, in which the flakelike
graphite structure causes scattering of the ultrasonic beam Selection of Test Frequency in Composites
and lowers its penetration, even at low frequencies.
Selection of test frequency for composite materials is
In most metals and alloys, greater grain refinement from generally based on penetrability. For thin composite lami-
I< irging, rolling or heat treating produces a more homoge- nates below 6.4 mm (0.25 in.), special transducers are
I u ·ous metallic structure so that a higher frequency can be required, operating at frequencies from 10 to 25 MHz to
used. Many brass alloy castings have a fine grain structure produce better discontinuity size discrimination and near
I>< -cause of controlled cooling or heat treatment and may be surface resolution. To penetrate thicker laminates, test fre-
l<·sted at 2.25 MHz. Other castings of similar alloy may be dif- quencies from 1 to 5 MHz may be used as a compromise
Iicultto test even at 500 kHz because of their extremely large between penetrability and resolution.
grain size. Most wrought aluminum alloys have relatively fine

394 I NONDESTRUCTIVE TESTING OVERVIEW

For very thick laminates, frequencies below 1 MHz may test system or through-transmission tests. Separate sending
be required to provide sufficient penetrability. Test fre- and receiving transducers may also be considered.
quencies of 2.25 to 5 MHz are generally used for testing
honeycomb structures with composite skins. Effect of Discontinuity Orientation
on Signal Amplitude
Selection of Test Frequency Based on Discontinuity
When a reflecting discontinuity is not oriented normal to
Size the axis of the straight beam longitudinal wave transducer
(and is not parallel to the front surface of the test object),
In addition to grain size, attenuation characteristics of the ultrasonic signal from it is significantly reduced in
the material and the size of the test object, it is also impor- amplitude. This occurs because much of the reflected sound
tant to consider the minimum size and type of discontinuity is directed at angles so that the maximum echo signal does
that must be detected. At higher test frequencies, smaller not return to the transducer where it can be detected. If the
discontinuities can be detected if attenuation and scattering sound beam could be directed at an angle so that it is per-
effects permit penetration and allow echo signals to return pendicular to the face of a tilted, laminar discontinuity, it
from all depths within the material. When test object size or could provide the maximum possible signal from the discon-
grain size dictates lower test frequencies, the response to tinuity.
small discontinuities is reduced and they may not be easily
detected. Theoretically, thin wedge shaped coupling blocks could
be used to angle the longitudinal wave to intercept disconti-
Ingeneral, sensitivity to small discontinuities can best be nuities of known angles and to reflect maximum signals back
determined using ultrasonic reference standards of the to the transducer. In practice, however, when the angle is
same material and grain characteristics as the test object. unknown and can vary widely, such an approach requires
The amplitude of signals from a reference standard can be many tests with wedges graduated in small angles. This pro-
determined experimentally to estimate the size of the small- cedure is costly and time consuming. With angle beam shear
est discontinuity detected reliably at various depths. wave tests, it is common practice to use transducers
equipped with wedges. A choice of beam angles in the test
It must always be recognized that flat bottom holes in material can be made by selecting a transducer with the
reference standards are ideal sound reflectors, including desired wedge. Even with this procedure, it is usually not
those with large grain size and high ultrasonic attenuation. practical to make many tests with different angle beam
The shape, orientations. and depth of discontinuities may transducers to explore each discontinuity.
alter the reflection path and reduce the possibility of detec-
tion. Rough, spherical or tilted discontinuities typically Effect of Geometry of Discontinuity
return smaller signals to the transducer when compared to on Echo Signal Amplitude
flat bottom holes whose faces lie perpendicular to the inci-
dent sound beam. Inindustrial test objects, discontinuity reflecting surfaces
can vary widely. For example, a void may be almost spherical
Echo signal from a rough tilted discontinuity in a large in a casting, fusion weldment or other metallic material that
plate section is typically recognized as low in amplitude but has received little or no mechanical working after solidifica-
broad at the base. It may shift position along the base line as tion. During mechanical working such as forging or hot
the transducer is moved in the direction of tilt. A reference rolling and piercing, a gas hole or porosity in cast metal may
standard's flat bottom holes with diameters smaller than flatten into a laminar discontinuity parallel to the rolling or
minimum size specified for the actual test must be clearly metal flow direction. Because this direction is often parallel
detected. to the surface, as in a plate or sheet, it can offer a good sound
reflecting surface perpendicular to the straight beam longi-
Lowering the test frequency may improve sound trans- tudinal wave in contact or immersion testing.
mission within coarse grained materials by reducing the
magnitude of signals from large grain boundaries. The use On the other hand, if refractory or brittle inclusions are
of different frequencies may alter the sensitivity of the test associated with a discontinuity, the reflecting surface may
to small discontinuities. The need to suppress grain bound- conform to the shape of the inclusion, which could be rough
ary scattering by lowering the test frequency involves a com- or irregular. If a metallic inclusion is bonded into a metal
promise with lower sensitivity to small discontinuities. matrix material, it may reflect only a small fraction of the

In critical cases, it may be necessary to use special tech-
niques such as increasing the pulse power of the incident
sound beam, selecting a transducer of lower damping or
higher efficiency, selecting a smaller area search unit or
devising means for focusing the sound beam to concentrate
its energy at the location of the suspected small discontinuity.
Alternative techniques include the use of double transducer

ULTRASONIC PULSE ECHO TECHNIQUES I 395

11wid1·11t sound beam, depending on the impedance mis- waveform before storing it in memory. The waveforms are
111all'l1 at its surface. If such an inclusion breaks its bonds to then analyzed by setting software gates and discrimination
tlll' surrounding metal and produces a metal-to-air inter- levels to apply accept/reject criteria. Time-of-flight informa-
1.w,-. the unbonded inclusion could be indicated by a strong tion is gathered at the same time. Any of the three modes
, , ·il1 -ction signal and may be easily detected. can be displayed after brief data processing.

lkcause of the factors that affect the height of an A-scan If a computerized testing system is used, different signal
,ig11al from a discontinuity, it is often difficult to determine processing programs can be used to enhance the discontinu-
with precision the actual size of a discontinuity. Ultrasonic ity detecting capability of the system. Waveform averaging
!1·sts are more qualitative than quantitative in this respect. can be used to minimize electronic noise to improve near
To a considerable degree, discontinuities that reflect the surface and far surface resolutions. Back surface reflection
im-ident sound beam have characteristics like new sources amplitude and discontinuity signal amplitude can be moni-
ol" ultrasonic emission. The ratios of their dimensions to the tored simultaneously to improve detection reliability. Image
sound beam wavelengths influence the patterns of the enhancement can improve discontinuity size and shape esti-
sound reflected and scattered from their surfaces, much like mates.
1l1c beam patterns radiated from transducers.
The most significant contribution of computerization to
Data Presentation contact ultrasonic testing is the ability to ensure coupling of
the transducer to the test object surface. Maintaining cou-
Contact tests can be used to produce all three forms of pling is the most difficult part of contact testing, especially
ultrasonic testing data: A-scan, B-scan or C-scan. The in the case of test objects with rough or contoured surfaces.
A-scan is a one-dimensional presentation of ultrasonic sig- Manual testing requires that the technician scrutinize the
nals along the propagation path. If there is a discontinuity system display constantly to ensure coupling. Mechanical
along the path, the occurrence of the ultrasonic signal in the scanning with automatic discontinuity determination must
time sweep of the test unit, in relation to the front and back have a provision to monitor the coupling efficiency to
surface reflections, indicates the depth of the discontinuity. ensure the quality of the testing data.
The amplitude of the signal reflected from the discontinuity
can be a measure of the area of the reflector, if the disconti- Tests of Multilayered Structures and
nuity is smaller than the beam size. Composites

A series of A-scan data obtained along a certain linear One of the major advantages of contact ultrasonic testing
direction across the test object can be pieced together to
form a B-scan. The horizontal axis of a B-scan represents is the portability of the instruments. The technique can be
the position of the transducer with respect to the test object. applied to assembled structures that have adhesives as the
The vertical axis represents the time of flight or swept time
in the A-scan. The time of flight for the front and back sur- joining medium between individual components. Testing
face signals plotted in one direction forms a cross section of the individual components and the adhesive bonds in the
the test object. Any discontinuity cut by the cross section structures is often required. ·
has its depth displayed on the B-scan. The B-scan therefore
provides a picture of discontinuities in a plane parallel to the The components or substrates in an adhesively bonded
direction of wave propagation.
structure can be composite or metallic materials. They can
The C-scan provides a plan view of the position and size be composite bonded to composite, composite bonded to
of discontinuities projected onto a plane that is normal to
the direction of wave propagation. A C-scan may use either metal or composite bonded to honeycomb core materials. It
through-transmission or pulse echo techniques. is essential to know in advance the thicknesses of the sub-

When the test object is scanned in a raster pattern, the strates for better interpretation of signals appearing on the
signals reflected from discontinuities can be used to activate system display. The bondline thickness generally ranges
a plotter showing a top view of the test object with disconti- from 100 to 300 µm (0.004 to 0.012 in.).
nuities indicated in place. Generally, C-scans are used to
provide the location and size of a discontinuity. They are the Dual-Transducer Methods
most popular form of data presentation in ultrasonic testing.
Dual-transducer methods involve transducer pairs that
Modern computerized ultrasonic contact systems can are usually positioned so that signals are directed along well
store all three forms of data simultaneously. The A-scan defined paths. Either contact or immersion coupling may be
results are obtained by digitizing each ultrasonic signal used at normal or oblique incidence. As in the case of the
pulse echo method, there are constraints that need to be

396 I NONDESTRUCTIVE TESTING OVERVIEW

met to produce and acquire meaningful signals. Three the test object collects wave energy scattered out of the
dual-transducer methods for materials characterization are main beam. Inverse analysis is used to infer size and distri-
described here: (1) the through-transmission, (2) the pitch bution data.3
and catch and (3) the acoustoultrasonic method.
Pitch and Catch Method
Through-Transmission Method
The dual-transducer pitch and catch method uses a pair
Through-transmission uses two transducers (sending and of transducers displaced from each other by a fixed distance,
receiving) usually facing each other on opposite sides of a on the same side or opposite sides of a test object, as in the
test object. The object that occupies the space between the leaky Lamb wave method (Fig. 16). The transducers may be
transducers is either in contact with them or is separated by in direct contact at either normal or oblique alignment. In
immersion in a fluid coupling medium. The acoustic beam the latter case, the transducers may be coupled by angle
is directed at normal incidence to test object surfaces. Alter- beam fixtures to excite shear or Rayleigh waves.
natively, with immersion coupling, the test object can be
rotated between the transducers. Oblique incidence can be In a fluid medium, the pitch and catch method can be
used to characterize material properties with shear waves or accomplished with a single focused transducer operating
surface (Rayleigh) waves that are generated when fluid with self-generated and self-intercepted Rayleigh waves
borne longitudinal waves meet a surface at an angle. (Fig. 17).

Through-transmission is often used for making compara- The usual objective with pitch and catch techniques is
tive property measurements with time-of-flight velocity discontinuity location and characterization. The method can
measurements and relative attenuation measurements.2 also characterize material properties. In either case, the
Single-transit, through-transmission testing is used if there positions of the transducers are calculated to recover spe-
is high signal attenuation because of test object thickness. cific signals that have traversed well defined paths along the
The method is often used in a comparator configuration surface or in the bulk. The paths usually involve surface
where the test object's transit time delay is compared with waves or simple back surface reflections that are inter-
the transit time delay in a reference standard, as when mea- cepted by the strategically placed receiving transducer. The
suring relative changes in elastic moduli. pitch and catch method often uses surface waves and guided
waves, such as Rayleigh waves and plate waves, respectively.
To ensure precise attenuation measurements in through- Lamb waves and leaky Lamb waves are used to evaluate
transmission, the transducer pair must be perfectly matched bonds and interfaces by using angle beam immersion tests.4
or folly characterized. Signal modulation properties of the Variations in bonding are observed through variations in the
instrumentation, transducers and interfaces must be elimi- spacing of null zones over a range of frequencies.
nated experimentally or taken into account with signal pro-
cessing. This can be avoided by using the single-transducer FIGURE 1 7. Pitch and catch technique using
pulse echo method where transducer related coupling and focused transducer; wave from rim of lens hits
impedance mismatch factors can be better accounted for in test object surface at critical angle producing
analytical expressions. surface waves that return to lens

Through-transmission also lends itself to forward scatter-
ing measurements. The transducer on the opposite side of

FIGURE 16. Pitch and catch dual transducer WATER
technique using leaky Lamb wave effect TEST OBJECT
r· RAYLEIGH WAVE
~ DIRECT
REFLECTIONS ~

I

O I µs

NULL ZONE LEAKY WAVE TIME

ULTRASONIC PULSE ECHO TECHNIQUES I 397

PART 3

ANGLE BEAM CONTACT TESTING

Although much contact testing is done with longitudinal monitor display. A reflected signal usually indicates a dis-
waves propagating normal to the test object surface, there continuity. Ranging of the discontinuity is not as simple as in
are many cases when an angled beam is preferred. The main straight beam testing. In metals, a standard reference block
reason for angle beam testing is the detection of discontinu- such as the International Institute of Welding (IIW) block
ities with geometries and orientations not parallel to the test can be used to calibrate the distance from the transducer to
surface. Planar cracks normal to the test object surface, the discontinuity. In materials other than steel, a conversion
voidswith small reflective surfaces parallel to the test object must be made, taking into account the ratio of velocities in
surface and discontinuities in welds with uneven top sur- the material to the velocities in steel.
faces are examples of situations that require angle beam
techniques. In range estimating, the beam spread of the angle beam
should also be taken into account. The sound beam radiating

Verification of Shear Wave Angle FIGURE 18. Block standards for angle beam
transducerverification:fa) InternationalInstitute
Angled longitudinal waves, shear waves or surface waves of Welding {IIW) block; {b) miniatureangle
are generated in a test object by mounting the piezoelectric beam block
element at an angle in the contact transducer. The correct
angle can be determined with Snell's law of refraction. Most faJ
of the commercial angle beam contact transducers are
designed to produce shear waves of 30, 45 and 60 degrees in 8 in.
steel. Some produce surface waves. For materials with dif-
ferent sound velocities, the angles of refraction can be cal- 0 40° 50° 60°
culated using Snell's law when standard transducers are in. HOLE
used. The transducer angle can be verified using a calibra- ~0.06
tion block such as the one shown in Fig. 18a. 2 in. DIAMETERHOLE
1 u4 in.
The first step in angle beam calculation is to determine (\.-- 60° 70° 75°
the transducer's beam exit point. If the exit point marking is FOCALPOINT__...
not correct, then the angle cannot be measured accurately. 12 in.
The transducer is placed on the focal point of the calibration
block and is moved back and forth until the monitor reaches I(bJ 3in.~1
maximum amplitude for the reflection from the large out- I : I1~
side radius. The focal point on the block then corresponds
with the beam exit point of the transducer. f--+ -1 (o.75 in.

The transducer is next placed on the other side of the I in. 2 in. f
block to obtain a reflection from the 50 mm (2 in.) diameter
hole. The transducer again is moved back and forth until the -T075 int
reflection from the hole shows maximum amplitude on the 1/5in.
monitor display. At the exit point of the transducer, the 0.06 in.
angle of the sound beam matches the degrees stamped on
the side of the calibration block. ""~FOCALPOINT
.45°. :o
The miniature angle beam block (Fig. 18b) can also be ....... 70°
used to calibrate the transducer angle in far field work.

Ranging in Shear Wave Tests

In angle beam testing with the beam directed away from
the transducer, no back surface reflection is present on the

398 I NONDESTRUCTIVETESTINGOVERVIEW

from the transducer fans out or diverges. The exact location Ultrasonic Tests of Tubes
of the discontinuity represented by the reflected signal on
the monitor display may be difficult to judge because of the Angle beam testing of tubes uses shear waves reflected
beam spread. Another difficulty in discontinuity ranging lies repeatedly along the tube wall. Occasionally longitudinal
in the uncertainty of where the discontinuity may be located waves are used but they suffer much from mode conversion.
relative to the boundary surfaces. As indicated in Fig. 19a, Lamb waves or other guided waves may also be used.
the sound beam undergoes four slanted trips (a double V- Figure 20 shows a sound beam bouncing off inner and outer
path) to complete a round trip to a crack located at or near walls of a tube during a typical ultrasonic test. Longitudinal
the top surface. When the crack is located at or near the bot- cracks in the path of the sound beam reflect the sound and
tom surface, the sound beam only takes two trips to make are indicated on the test system's monitor. To cover the entire
the round trip (Fig. 19b). In Fig. 19c, where the crack is area of the tube, the transducer or the tube must be rotated
located. near the middle of the plate, the sound beam while either one traverses in the longitudinal direction.
bounces off both the upper and lower surfaces and may
return to the transducer, if the plate is not too thick. When- Weld Testing
ever possible, the plate should be tested from both sides to
judge the crack location more accurately. The angle beam technique is extensively used for weld
testing (Fig. 2la). Typically, the weld is tested with a full
FIGURE 19. Possible sound beam paths in skip, making use of the two test directions provided bY_ th.e
angle beam testing angle beam test (as shown in Figs. 2lb and 2lc). A skip 1,s
the surface distance covered by a V-path, from a beams
fa) TRANSDUCER entry point to its exit point. To assist in interpreting the
results of angle beam tests, a direct reading ultrasonic calcu-
BACK SURFACE lator (Fig. 22) may be developed for this purpose. The hori-
zontal scale across the top of the card represents the
fb) TRANSDUCER horizontal distance from the exit point of the transducer.

FIGURE 20. Sound beam bouncing off inner
and outer walls in ultrasonic tube test

FRONT SURFACE ),..

BACK SURFACE DISCONTINUITY
TRANSDUCER
fc)

DISCONTINUITY
BACK SURFACE

ULTRASONIC PULSE ECHO TECHNIQUES I 399

I )istance from the exit point and the center of the weld is Calibrate the horizontal sweep of the monitor to repre-
sent beam travel distance in the test material. The full skip
laid out along this scale. The vertical scale represents dis- distance of the sound beam is obtained by doubling the
86 mm (3.4 in.) intersecting point at the bottom of the plate
tance in the thickness direction. Specimen thickness is indi- and marking the point at 175 mm (6.9 in.) on the upper
plate surface. The 30 degree V weld is next drawn on trans-
cuted on this scale and the arc shows the angle of the sound parent paper positioned over the monitor screen - here at
a value of 140 mm (5.5 in.). The distance between the cen-
beam. ter of the transducer (exit point) and the center of the weld-
ment is then measured, giving 116 mm (4.6 in.). The
As an example of using the calculator, assume a double V transparent paper is moved by the same distance. The posi-
weld with an opening of 30 degrees in a 50 mm (2.0 in.) tion of the discontinuity is indicated and can be evaluated.
steel plate. The weld is to be tested using a 60 degree shear
wave transducer. A line is first drawn from the point of inci-
< lence at the upper left corner of the calculator through the
fiO degree mark on the arc, extending to the 50 mm (2.0 in.)
point representing the plate thickness.

FIGURE 21 . Ultrasonic angle beam weld tests: FIGURE 22. Direct reading ultrasonic
(a) typical sound path and positions for discontinuity location calculator: (a) model;
detecting discontinuities near (b) lower surface (b) example
and (c) top surface
faJ
(aJ SECOND LEG FOURTH LEG
1:rJ-- ....\·!·)& .>•B '.! 111 Ill 1111111111111
FIRSf THIRDLEG ! 40
I
(bJ
: LEG 1 :
DASHED
..,_. FIRST V PATH__..._. SECOND V PATH~I

(bJ

(CJ

A

400 I NONDESTRUCTIVE TESTING OVERVIEW

PART 4

COUPLING MEDIA FOR CONTACT TESTS

Use of Transducer Shoes the interface. Both layers of couplant should be thin, uni-
form in thickness and completely free of voids or bubbles.
Wedge Shaped Shoes
During contact tests with contoured shoes, care should
Wedge shaped shoes are used to refract an ultrasonic be taken to protect the surface of the shoe. This is a sec-
beam to angles off normal from the entry surface, particu- ondary function of the couplant, to lubricate the shoe and
larly in metallic test objects. Introducing straight beams at protect it from damage by a rough test object surface.
an angle to the entry surface also causes mode conversions,
in which surface waves propagate along the surface and Contoured shoes can help maintain contact perpendicular
shear or transverse waves are propagated into the test to a portion of the test object. However, it must be recognized
object, in addition to the refracted straight beam. that sound waves passing through the interface between the
acrylic shoe and the test object are subject to refraction
In most contact tests, the use of a wedge shaped shoe (Fig. 24). When a curved shoe is applied to a convex surface
can reduce signal amplitude or eliminate the usual front (a round bar or tube), the outer portions of the ultrasonic
surface echo signal. In almost all uses of a single search unit, beam tend to diverge within the test material thereby reduc-
the front surface echo is directed at an angle through the ing the beam intensity. The test signal amplitude and the test
wedge so that some portion of it may not return to the trans- sensitivity may also be reduced in the peripheral regions of
ducer. In many dual-transducer techniques, the receiving the spreading sound beam. Focused immersion test tech-
unit is positioned to receive echo signals that do not return niques may be superior for these applications.
to the sending search unit.
Use of Couplantand Membranes
Contoured Shoes for Curved Test Surfaces
Solid or flexible membranes may be attached to a trans-
When testing curved surfaces with contact techniques, ducer for the purpose of directing the sound beam at angles
flat faced transducers may be inappropriate because only a off normal incidence or for cushioning the transducer from a
point (or at best, a line) of the transducer face can actually
touch the curved entry surface. With slightly contoured sur- FIGURE23. Contouredultrasonictesting
faces, viscous couplants may help transmit ultrasonic waves shoes:(a} straightbeam; (b} angle beam
from a larger portion of the transducer face into the test
material, but this is imprecise and variations in signal ampli- faJ
tudes often occur. Also, excess couplant can increase trans-
ducer ringing. For these reasons, the use of contoured shoes (bJ
is preferred for flat faced transducers on curved test sur-
faces. In all cases, the effects of beam refraction are signifi-
cant and must be considered.

As shown in Fig. 23, acrylic coupling shoes can be
shaped to fit the flat surface of the transducer on one side
and the curved surface of a test object on the other side. It is
essential to use a good couplant at both of these interfaces
because ultrasonic waves cannot be transmitted through an
air space. Petroleum jelly or heavier greases can be used to
provide good coupling between the face of the transducer
and the coupling shoe. Oil or another of the commonly used
couplants may serve at the interface between the contoured
surface of the coupling shoe and the test object. In most
applications, the contoured shoe can be locked in place on
the transducer after a layer of couplant has been applied to

ULTRASONIC PULSE ECHO TECHNIQUES I 401

rough or contoured test object surface. Membranes contain- To some degree, this technique may enhance near surface
ing fluid couplant or solid, flexible materials are often used discontinuity resolution.
for ultrasonic tests. Such membranes must be coupled to
the entry surface with a suitable liquid or semiliquid cou- When a sonic delay line is used, it should be long enough
plant layer. so that multiple reflections do not appear in the A-scan sig-
nal trace ahead of the back surface reflection signal from the
When a flexible membrane is used to cover the ceramic test object. Such multiple reflections can mask discontinuity
transducer element, it becomes the front face of the search signals from the test object.
unit. Some transducer cases are designed so that replace-
able membrane covers can be installed when one is dam- High Temperature Applications
aged. A suitable viscous fluid couplant fills the space
between the face of the transducer and the membrane. All A special use of standoffs is testing high temperature
air bubbles must be excluded from this interface. materials, particularly metals. Some delay line configura-
tions can be water cooled to prevent heat from reaching the
Oil, water or another liquid couplant must be applied transducer. Cooled standoffs are also used in research, as
between the membrane and the entry surface to avoid air when measuring the modulus of elasticity in metallic bars
interfaces and to enable transmission of the ultrasound into over a wide temperature range. In such applications, the
the test object. In the 1950s, adaprene rubber membranes test object is placed in a furnace so that temperature varia-
were developed in the United States to couple and, in some tions can be controlled. The stand-off extends beyond the
applications, to match with the impedance of porous heat wall of the furnace and is cooled so that the transducer
shield structures. These membranes were liquid coupled to operates near normal temperatures.
the transducer but dry coupled to the structure because of
its porous nature. A notched standoff was introduced in 1947 by J.R. Fred-

Use of Delay Lines erick to obtain modulus data at high temperatures. Wire
standoffs were used in 1957 by F.W Bell for similar applica-
Plastic coupling blocks are often used between the face tions.
of the contact transducer and the front surface of the test
object. Such blocks are also known as delay lines, buffer Selection and Use of Coupling
rods or standoffs. Their function is analogous to that of an Media
equivalent water path distance in ultrasonic immersion
tests. Much of the near field zone can be confined to the A contact ultrasonic test ordinarily cannot function with-
delay line. Inaddition, the ringing of the transducer falls off out a suitable couplant to transmit ultrasound between the
before the front surface echo returns from the test object. transducer and the test material. A couplant may be liquid,
semiliquid or paste that does the following things:
FIGURE 24. Beam spreading in convex test
object caused by off-normal beam incidence on 1. provides positive acoustic coupling for reliable testing
curved surface; signal amplitudes are reduced (consistent back surface echo amplitude);
and less clear than discrete reflections obtained
with flat surfaced test objects 2. wets both the surface of the test object and the face of
the transducer, excluding air between them;

3. can be easily applied;
4. does not run off the surface too quickly;
5. provides adequate lubrication for easy movement of

transducer over the test object surface;
6. is homogeneous and free from solid particles or

bubbles;
7. is free of contaminants (such as from lead or sulfur)

and is not corrosive, toxic, flammable or otherwise
hazardous or polluting;
8. does not freeze or evaporate under test conditions; and
9. is easily removed or evaporates after testing is com-
plete.

Another critical characteristic for a couplant is that its
acoustic impedance is between that of the transducer face

402 I NONDESTRUCTIVE TESTING OVERVIEW

and that of the test material or is identical to that of the test experimental situations, propylene glycol serves as a conve-
material. nient liquid couplant. In some cases a small amount of wet-
ting agent helps the glycerine adhere to the surface.
For rough or porous test surfaces, soft rubber sheets are
sometimes effective as coupling materials. Resins and Special Purpose Commercial Products

Selection of Couplants Hair grooms, cellulose gums (gravy thickeners), chewing
gum and petroleum jelly have found applications as sonic
It is critical to select the proper couplant for specific couplants in special circumstances, because of their adher-
ultrasonic testing applications. ence to vertical, overhead or rough surfaces and because of
their availability and economy.
Water l\s a Couplant
High viscosity resins and honey are useful in shear wave
Water is widely used as a couplant for ultrasonic tests. contact tests at normal incidence.
Wetting agents or detergents are sometimes added to
ensure good surface wetting and the elimination of air films. Pressure Coupling

However, water's viscosity is so low that it will not stay on Pressure coupling has been used to dry couple longitudi-
some test surfaces long enough to complete the ultrasonic nal and shear waves at normal incidence to test objects at
test procedures. For example, water cannot be used as a temperatures well above 1,000 °C (1,800 °F). Pressure cou-
couplant on vertical or angled surfaces unless it is continu- pling has also been used with contact tests at oblique inci-
ously replenished with a hose and pump setup or with a dence at cryogenic temperatures.
water coupled dolly transducer system. Water is not suitable
for tests of absorbent materials or those that react to it Nuclear Component and Shear Wave Couplants
adversely.
Couplants on nuclear components are required to meet
Water Based Gelatin Couplants stringent specifications to ensure that no harmful or corro-
sive effects on test materials result from their use.
Water based gelatin is most widely used on test objects
made of advanced composites. Such materials absorb water Couplants for vertically polarized shear waves at oblique
and as a result experience critical property degradation. incidence can be the same as for longitudinal waves.
Because of its higher viscosity,a gel can also serve as a filler
for rough composite surfaces. Such gels are water soluble OperatorTechniques to Ensure
and can be easily cleaned after testing is completed. Good Coupling

Oil and Grease Couplants Operator technique can be a significant factor in success-
ful coupling for contact tests. Before applying couplants, test
More frequently than water, various grades of oil are materials should be wiped clean and free from grit, metal
used for contact ultrasonic tests, mainly because they stay chips or liquids. Touching the surface with the fingers and
on the test surface longer. Oils containing wetting agents, as sliding them about can ensure that no wedges or burred
in many commercial motor oils, are most desirable for test edges of metal are present to prevent good, flat contact.
applications.
For general testing, a surface finish of 250 root mean
Heavier oils and greases are used as couplants on hot square or better is required. For higher sensitivity and relia-
surfaces, on vertical surfaces and to fill in irregularities on bility, smoother surfaces are needed. Waviness exceeding
very rough test surfaces. These heavier grades are retained 1.5 mm (0.06 in.) over a 50 mm (2 in.) span is unacceptable.
on the test surface much longer than lighter grades.
Ringing Technique
Glycerine Based Couplants
A technique of ringing the transducer to the couplant
Glycerine is often used as a contact test couplant coated surface (similar to the placing of precision gage
because it adheres to surfaces more effectively than water or blocks in intimate contact) is often used. The operator
light oil and because it is a better acoustic impedance match places the transducer gently on the test surface then rotates
for transducers and test objects. it or moves it back and forth while watching the test signal
indication responses. The procedure is complete when con-
On forged component surfaces, higher amplitude echo sistent, maximum amplitude signals are obtained.
signals are obtained with glycerine than with water. In many

ULTRASONIC PULSE ECHO TECHNIQUES I 403

Constant pressure on the transducer during this move- The type of couplant used in contact testing of forged
ment tends to expel air bubbles and to attain a more uni- surfaces can have an effect on detection of discontinuities.
form thickness of the couplant film. Forged and shot blasted surfaces can reduce the amplitude
of discontinuity signals when surface roughness exceeds
Test Object Surface Preparation 6.5 µm (2.5 x 10-4 in.). However, no significant reduction in

Liquid couplants are used successfully on test surfaces echo signal amplitude is observed with surface finishes less
created by swing grinding, rolling, forging or sand blasting than 5 urn (2.0 x 10-4 in.).
and other elaborate surface preparations. Smaller A-scan
signals are received from internal discontinuities when test- Rough machined test objects sometimes produce spurious
ing through the grooved, milled surface of a steel block. echo signals between the front surface and back surface indi-
This effect is frequency dependent and can be severe at cer- cations in A-scan presentations. Such signals typically disap-
tain test frequencies. Proper choice of the couplant tends to pear if the rough surface can be smoothed with a hand
minimize this effect. grinder and fine grit abrasive. Rough surface effects can be
severe when test objects are thin and the resolving power of
the ultrasonic system is nearly half the test object thickness.

404 I NONDESTRUCTIVETESTINGOVERVIEW

PART 5

IMAGING OF PULSE ECHO CONTACT TESTS

To ensure the quality of welds, film radiography and Position is recorded with optical encoders or by use of
ultrasonic testing are the two nondestructive methods gen- stepping motor pulses. Because stepping motors are prone
erally used. Film radiography is a repeatable method that to skip counts under certain loading conditions, optical
provides a permanent record of discontinuity location and encoders provide a more accurate measurement of trans-
size, with the following disadvantages: (1) a health hazard ducer position. A scan grid is generated for recording the
may be posed to personnel in the vicinity of the test, (2) the transducer position (Fig. 26). Each cell in the grid (called a
sensitivity of the method diminishes for thick sections, grid spacing) determines the spatial resolution of the scan.
(3) discontinuity depth information is not provided and For each grid spacing, the transducer location, time-of-
(4) access to both sides of the test object is required. flight and amplitude information are recorded and images
may then be generated in real time.
Conventional ultrasonic testing does not present a per-
sonnel hazard. It provides depth information and may FIGURE25 .. Measurementof amplitudeand
detect discontinuities that radiography could overlook. time-of-flightdata
However, the conventional ultrasonic methods have these
potential disadvantages: (1) variance in test results because \\---~---1I THRESHOLD
of variables controlled by the operator, (2) requirements for
a high degree of operation subjectivity, (3) lack of a suitable TIME· OF- FLIGHT MEASUREMENT
method for producing a repeatable, permanent record and
(4) low probability of repeated test results on small but
rejectable discontinuities.

Ultrasonic imaging helps overcome these limitations and
provides the advantages of both nondestructive techniques.
For example, the automation of the data acquisition process
and the recording of transducer location and test parameters
provide a repeatable, permanent record. In the text below,
the application of ultrasonic imaging techniques for weld
tests is discussed. Several examples of discontinuities com-
monly found in welded joints are presented, including lack of
fusion, porosity and intergranular stress corrosion cracking.

Ultrasonic Imaging Procedures FIGURE26. Scangrid for positionmeasurement

To determine the acceptability of weld indications, two
measurements are taken from ultrasonic signals: amplitude
and time of flight. Figure 25 shows a video display of a typi-
cal ultrasonic signal (an A-scan). The display represents a
plot of amplitude as a function of time. A gate is set to
record signals in the region of interest while a threshold is
set so that noise is not recorded.

Signals that occur in the gate and exceed the threshold
trigger the measurement circuits to digitally record the time
of flight and signal amplitude. These measurements can be
done in hardware on analog signals or in software on digi-
tized signals. Several imaging systems record the first signal
in the gate while others record the peak signal. The resolu-
tion of the electronics determines the accuracy of the mea-
surements.

ULTRASONIC PULSE ECHO TECHNIQUES I 405

In addition to saving the scan data, most ultrasonic imag- collected when the transducer is further from the weld. In
ing systems record other information with each scan, includ- applications where the scan is limited by weld configuration
ing a list of test parameters that detail the test setup. These or transducer access, the data are collected from a third
parameters include administrative information, scanner position. At this point, the sound beam has diverged signifi-
setup, calibration data and the ultrasonic equipment set- cantly and small discontinuities are more difficult to resolve.
tings. This information serves as the basis for generating
repeatable scans. Figure 27 also shows the use of two transducers operat-
ing in tandem, one transmitting the ultrasonic signal and the
Contact Weld Tests second receiving the reflected signals. The configuration
lends itself to detection of discontinuities that are poorly ori-
Test Frequency ented for single-transducer applications.

The majority of weld tests are done with angle beam Ultrasonic Images of Welds
transducers. The test angle is selected to obtain the opti-
mum signal from the expected indications. Most weld test- The images in Figs. 28 to 30 are from tests conducted
ing is done with frequencies between 2 and 5 MHz. There with an angle beam transducer operating in the pulse echo
are special cases for testing outside the typical frequency mode. Tests were made on three welded objects containing
range: thick, coarse grained materials require lower test fre- lack of fusion, porosity and intergranular stress corrosion
quencies (0.5 to 2 MHz) while thin, fine grained materials cracking.
can tolerate higher frequencies (5 to 15 MHz).
The first test object is a high yield strength carbon steel
When choosing test frequencies, several considerations plate having a double V butt weld containing lack of fusion.
should be noted: high frequencies provide better resolution The scan shown in Fig. 28 was made with a 5 MHz trans-
and sensitivity while low frequencies offer better probability ducer aimed toward the weld. Data were collected from the
of detection. first and second positions by setting the ultrasonic gate to reg-
ister signalsover the appropriate skip distances. Normally, the
Transducer Positioning gate is set to record data from a single position. To compen-
sate for signal loss from beam divergence and attenuation,
Weld discontinuities with specific orientations can be time controlled gain was applied to the received signal.
detected at specific transducer positions relative to the weld
(Fig. 27). The sound path of the transducer is divided into The ultrasonic image from lack of fusion is a C-scan dis-
zones that correspond to multiples of the test object thick- play of the ultrasonic signal amplitude. Data from the sec-
ness. The sound path between the transducer and the back ond position are shown at the top of the image. As the
surface is the first and is commonly referred to as a halfskip transducer is moved toward the weld centerline, data from
or half V test. The sound path from the back surface to the the first position come into view. High signal amplitudes are
top surface is second and is called afull skip orfull V test. shaded light and low signal amplitudes are darker. The

Data collected from the first position occurs when the FIGURE 28. Scan area for weld tests
transducer is closest to the weld. Second position data are
DATA FROM
FIRST POSITION

FIGURE 27. Weld testing with pulse echo
contact transducer

SECOND FIRST

~IPOSITION POSITION WELD

TRANSDUCER

(I

WELD

DISCONTINUITY

406 I NONDESTRUCTIVETESTINGOVERVIEW

white area surrounding the indications corresponds to data Figure 30 is an ultrasonic time-of-flight map of intergran-
that did not exceed the trigger threshold. In angle beam ular stress corrosion cracking. Thick sections are shaded light
tests, signals are returned only from echoes that reflect back and thin sections are darker. The horizontal indication at the ·
to the transducer - if no signals are present, no data are
recorded. top is caused by signals from the counterbore and has a
thickness near the nominal thickness of the pipe. The darker
The slice at the bottom of the display represents a hori- indication is from intergranular stress corrosion cracking and
zontal cross section of amplitude through the cursor on the is easily distinguished from the counterbore.
C-scan display. During analysis, viewing thresholds can be
set to various amplitude levels and the indications appropri- Ultrasonic imaging of weldments is a practical method
ately sized for acceptance or rejection according to the test for obtaining accurate volumetric information about discon-
criteria being used. tinuities in welds. Because the area is typically well-defined,
the test can be performed by automated scanning processes.
The second test object has the same weld geometry and Images can be archived and retrieved for comparison as
material composition as the first but contains porosity in the additional data are collected. The technique lends itself to
weld. Porosity is typified by low amplitude signals that are periodic monitoring for assessing weld integrity. With this
clustered together, as the image shows. The horizontal slice knowledge of discontinuity shape, size and location, the crit-
at the bottom of the display shows similar indications. ical nature of the discontinuity can be determined.

The third test object is an austenitic stainless steel pipe FIGURE30. Testof intergranularstress
weld containing intergranular stress corrosion cracking. The
weld has a single V full penetration butt configuration as ~~·I ~Icorrosioncracking:(a) ultrasonicimage;
shown in Fig. 29. Intergranular stress corrosion cracks prop-
agate in a branching manner along the grain boundaries of (b) testingsetup
the material. The cracking occurs in the heat affected zone fa) 2
of a weld when the microstructure is sufficiently sensitized
and exposed to critical stress-strain conditions in a corrosive 1,l~\'1 ~,·-
environment. Typically, such cracking is next to the counter-
bore and the toe of the weld. Signals can be difficult to dif-
ferentiate because of this proximity.

COUNTERBOR~E -. . _
.
INTERGRANULARSTRESS _/
FIGURE29. lntergranularstresscorrosioncrack CORROSIONCRACKING 4
testingof butt weld 3

(b)

"r-J..L:.TRANSDUCER WELD

I

------CO_U_N_T-ER_B_O-RE-~ CRACK FROM GENERAL ELECTRICNUCLEAR SERVICES. REPRINTED WITH
PERMISSION.

ULTRASONICPULSEECHO TECHNIQUES I 407

PART 6

ULTRASONIC PULSE ECHO WATER
COUPLED TECHNIQUES.

An ultrasonic wave, propagating from a transducer to a immersion technique provides good acoustic impedance
test object and back, crosses several interfaces with differ- matching to composite materials. In addition, the trans-
ent acoustic impedances (the product of density and acous- ducer is not exposed to wear during use and immersion
tic velocity). The mismatch of the acoustic impedances methods are ideal for automated testing.
causes a loss of energy due to reflections, which can be very
high if air is present in the wave path. Water is the fluid most commonly used for immersion
coupling because of its availability and low cost. To inhibit
To improve the energy transfer from the transducer to corrosion or other chemical or biological aqueous reactions,
the test object, two coupling techniques are used: contact various types of inhibitors can be added to water.
and immersion. The difference between these two tech-
niques is related primarily to the length of the couplant To prevent air bubbles from forming and accumulating
medium between the transducer and the test object. on the surface of test objects, detergent additives are com-
monly mixed with the water. Such additives reduce the sur-
Immersion Coupling face tension between the water and test object, making the .
formation of air bubbles less likely.
Immersion coupling uses a long fluid delay line. The dis-
tance between the transducer and the test object is long When filling a water tank or conducting operations that
enough to separate in the time domain the reflections from move the water (scanning or filtering, for example), air tends
the test object front surface and the transducer's excitation to dissolve in the water and increase the attenuation. This
signal. In addition, a separation is maintained between the attenuation increase can be as high as 6 dB and can be
test object's internal reflections and the repetitive reflec- avoided by keeping water movement slow or by allowing the
tions in the water path. This adjustment is necessary to air to be outgassed before testing. Outgassing can last from
avoid interference between the various reflections and to 0.25 to 1.0 h after filling the water tank Generally, water is fil-
simplify the evaluation of the response. tered before and during use to ensure that it does not contain
particles that create false test indications.
As a rule in immersion coupling, the velocities ratio (cou-
plant to test media) is about 1:6 for some ceramics, 1:4 for The primary advantages of immersion scanning are listed
metals and 1:2 for composites and plastics. A constraint over below.
the path length of the coupling medium limits the practical
usage of the immersion method to relatively thin sections. 1. No special transducer adapters or shoes are required
Because most aerospace structures meet this requirement, when changing the size or shape of the test object.
the immersion technique is very popular for testing aircraft
components. 2. Simple continuous adjustment of the incidence angle
of the sound beam is permitted. This capability is
Other limitations of the immersion technique are as fol- essential for contour following of complex shaped
lows: (1) because of weight, immersion systems lack porta- structures or when developing a test procedure.
bility and are impractical in some field applications;
(2) some hardware is complex and relatively expensive; 3. The coupling liquid is continuously available.
(3) the technique is not recommended for test objects sus- 4. Because intimate contact is not required, testing is
ceptible to corrosion by the immersion fluid; and (4) tests
are practical on relatively thin objects. significantly faster.
5. The immersion technique is not influenced so greatly
Immersion testing has significant flexibility - it is possi-
ble, for instance, to use immersion techniques in most con- as the contact method by loss of coupling due to ovality
tact coupling applications. The immersion technique of tubing, surface conditions or dimensional variations.
provides coupling uniformity and simplicity of changing the 6. Total immersion in a water bath aids in suppressing
insonification angle without changing the transducer. The surface waves that inordinately increase signals from
minor outside surface discontinuities.
7. The water path provides a delay line that allows the
very strong initial signal to pass through the amplifier
before the weaker signals return to the instrument.
This is particularly advantageous when testing small
tube sizes and thin plates.

408 I NONDESTRUCTIVE TESTING OVERVIEW

Immersion CouplingDevices An immersion tank can be used to test many shapes,
including plates, wires and contours. Computer controlled
The key to immersion coupling is the presence of a con- systems can follow complex shapes by changing the insonifi-
tinuous fluid medium in the path between the transducer cation angles to maintain a constant angle of incidence.
and the test object. This condition can be maintained by the
various devices detailed below. While immersion of the The immersion testing of stiff materials with a constant
transducer and the test object in a water tank is the most cross section (such as pipes and rods) can be simplified to
widely used form of immersion coupling, other forms are avoid the high cost of large tanks.5 These materials may be
also finding widespread usage, particularly the water jet. tested by passing them through a short tank with two win-
dows that match the test object cross section (Fig. 32). To
The cost of an ultrasonic test increases with the complex- prevent water leakage, the windows contain rubber seals in
ity of the coupling device. Therefore, choosing the device is the gap between the test object and the window. After
a budgetary decision that must be weighed against the inserting the test object through the windows, the water
intention of the test procedure. Any of the coupling devices level in the tank is brought higher than the transducer and
below can be used for automated testing. the ultrasonic test is performed.

Immersion Tanks Bubbler Devices

The method of coupling a transducer to a test object by The device known as a bubbler contains a transducer and
submerging both in a water tank has been in use for ultra- a captured water column. The bubbler maintains a constant
sonic testing since the early 1940s. In the 1980s, the use of flow of water discharged through the gap between the bub-
immersion tanks increased substantially with the develop- bler adapter andthe test object. The transducer mounting
ment of automated scanning systems. unit is designed to provide the desired angle of incidence
for the beam.
In a typical configuration, scanning systems are assem-
bled on the immersion tanks and the transducer is moved The test object is positioned above or below the water
sequentially in at least two normal directions, either manu- column opposite the transducer. For continuous tests, it is
ally or automatically, following a programmed scanning preferable to couple the bubbler to the test object's bottom.
plan. A manipulator permits adjustment of the beam angles With a weak water flow, this arrangement makes it easier to
and remote control of the distance between the transducer ensure that the area between the transducer and the test
and test object. A scanning system with an immersion tank is
shown in Fig. 31. FIGURE 32. Short immersion tank stuffing box
for scanning test objects with constant cross
FIGURE 31 . Typical ultrasonic scanning system section
with immersion tank
MANIPULATOR

FROM TESTECH CORPORATION. REPRINTED WITH PERMISSION. RUBBER SEf\L

ULTRASONIC PULSE ECHO TECHNIQUES I 409

object is always filled with water. When the test object is the water jet is smooth. These reflections can sometimes
placed below the bubbler, a strong water flow is required to make the squirter coupling unsatisfactory for the pulse echo
expel air from the system. Water is fed to the bubbler cavity technique. Squirters are more commonly used with the
through a pipe nipple. If this is done at sufficient pressure, a through-transmission technique, where water jets are
water cushion is formed and the bubbler can slide over the applied on both sides of the test object.
test object without intimately contacting its surface.
When using squirters with the pulse echo technique, the
The bubbler is used in a variety of field applications. As response is very sensitive to the angle of incidence. To
an example, a bubbler device has been installed on a manual ensure a sufficient signal-to-noise ratio, the water jet should
scanner to conduct a normal beam test of glass-epoxy tub- be within 2 degrees of normal to the surface. The surface
ing.6 The bottom section of the bubbler was machined to fit needs to be smooth and free of scratches or wrinkles, a diffi-
the outside diameter of the tube and a fixture was designed culty when testing composite laminates.
to hold the bubbler in a steady position and constant angle
while scanning. Wheel Transducers

Water Jet Devices Wheel transducers (Fig. 34) consist of a plastic tire filled
with coupling fluid under pressure. During an ultrasonic
If sufficient pressure is available, a water jet can provide test, the tire rolls over the test object and maintains a con-
noncontact coupling of the transducer to a test object over a tinuous coupling between it and the transducer. The trans-
distance of 120 mm (4. 75 in.) or more. Special hydrodynamic ducer is attached solidly to the wheel's shaft and is
considerations are used in the design of the squirter to positioned a few millimeters from the surface of the tire.
ensure a minimum of bubbles or turbulence. This capability The transducer can be manipulated to transmit at an angle
is important for rapid automated testing where the manipu- that excites shear waves in the test object. The angle
lator might collide with an uneven surface or with a projec- between the plane of incidence and the rolling direction of
tion from the test object. Water jet coupling has an the transducer can be adjusted to any rotation angle
advantage over immersion when testing large structures (no between O and 90 degrees.
large volume of water or tank is required). Coupling is
maintained by pressurizing a water column and draining the Testing with the transducer wheel is performed by
water after it strikes the test object. Examples of a water jet rolling the transducer with light pressure while scanning the
are shown in Fig. 33. test object manually or automatically. The tire creates
reflections that need to be discriminated from the signifi-
Water jets are widely used in the aerospace industry, cant reflections of the test object.
where many assemblies have a large volume of air in their
internal structures. This air causes flotation capability and FIGURE 34. Wheel transducers: (a) for straight
immersion is not practical. Furthermore, water can pene- ultrasonic beams; (b) for angle beams
trate into such structures and may later induce corrosion.
(aJ LIQUID FILLED TIRE
A squirter produces many disturbing reflections at the
contact point behind the front surface reflection, even when STATIONARYAXLE

FIGURE 33. Schematic diagram of water jet for PIEZOELECTRIC CRYSTAL
ultrasonic tests

TO ULTRASONIC INSTRUMENT

fbJ

41 0 I NONDESTRUCTIVE TESTING OVERVIEW

Boot Attachment changed. Generally, such systems are similar in perfor-
mance speed to the analog systems.
The boot attachment uses a rubher or plastic enclosure
to maintain a water path between the transducer and the Back Surface Reflection Amplitude
test object (as in the wheel transducer).
Back surface reflection amplitude serves as a measure of
A rubbe_r cup is attached to the transducer assembly and the material attenuation and as a detector for anomalies that
the cup is filled with fluid. Either a flat or a focused trans- affect t~e energy of a traveling acoustic wave. This parame-
?u~er can be used with the boot attachment. The angle of ter is a fast indicator of discontinuities and is widely used to
incidence can he controlled by the manipulator on which detect delaminations, porosity and microcracks.
the transducer is mounted.
Back surface reflection amplitude can be sensitive to
Pulse Echo Immersion Test irrelevant so_ur~es. ~o~ example, geometry, surface rough-
Parameters ness and variations in front and back surface conditions can
c_ause ~hanges in back surface reflection amplitude. Rejec-
Parameter Analysis tion of the test object on_ the basis of a change in this param-
eter can be clone only after careful consideration because of
Pulse echo immersion test systems can use four parame- the uncertainty in determining the source of the change.
ters to detect and characterize discontinuities: (1) hack sur-
face reflection amplitude, (2) amplitude of extraneous The loss or absence of back surface reflection is evidence
reflections, (3) time-of-flight measurements and (4) spectral that the transmitted sound is being absorbed, refracted or
response. reflected so that the energy does not return to the trans-
ducer. Loss of back surface reflection can result from many
A schematic view of a typical pulse echo response is causes and it is difficult to provide a quantitative measure of
shown in Fig. 3.S. Time gates, superimposed on an A-scan material properties on the basis of this parameter. With the
display, are used to examine the first three parameters. Win- gain levels and frequencies used in ultrasonic tests, at least
dows are used in combination with a fast Fourier transform several back surface reflections are ususally obtained in
to analyze the frequency domain response. acceptable materials, particularly in metals.

The analysis of a parameter can he done with analog Loss of back surface reflection may be determined by
hardware or a computer based digital system. Analog sys- measuring the· ratio of the number of back surface reflec-
te1~s have high speed performance but also have predefined tions in a reference material of equivalent thickness to the
options that are relatively limited. If signals are digitized and number of back surface reflections in the test material. The
analyzed by a microprocessor, a large variety of options l~ss n~ay also be evaluated by reducing the gain setting to
become available. For real time performance, hard coded give slightly less than a saturated (maximum undistorted) sig-
firmware is used with a program that can acquire specific nal from the first back surface reflection in a reference mate-
r_a~ameters. This approach provides high speed data acqui- rial. ~he amplitude of thi~ signal is then compared with the
sition but the test parameters are predefined and cannot be amplitude of the back surface reflection in the test object.

FIGURE 35. Schematic view of typical pulse Amplitude of Extraneous Reflections
echo test parameters
. The detection of discontinuities can be performed
FRONT REFLECTION directly by examining their reflections from the bulk of the
AMPLITUDE material. These reflections appear between the reflections
SURFACE from the front_ and the back surfaces (Fig. 35). Time gates
I GATE I --1 are set in this region and signals above a preset threshold
I indicate the presence of discontinuities. The reflection pat-
BACK REFLECTION tern can indicate the type of discontinuity. The reflection
-- r amplitude provides a measure of the discontinuity size .
_Ii------_-___j · I 'AMPLITUDE
. Individual discontinuities that are small compared to the
TIME OF FLIGHT -I effective beam diameter are evaluated by comparing their
GATE 3 amplitude of reflection with the amplitude of reflection
from a standard hole in a reference block Extensive experi-
ence in the aerospace industry has shown this technique to
have acceptable reliability.

When using reference blocks to evaluate small disconti-
nuities, the estimated discontinuity size is generally smaller

ULTRASONIC PULSE ECHO TECHNIQUES I 411

than the actual discontinuity size. The surface of a test determine grain flow in various parts of the test object.
object and the surface of a discontinuity in a test object are Results of such destructive procedures help determine the
usually not so flat and smooth as the surface of the reference most likely orientation of discontinuities for subsequent
block and the flat bottom hole in the reference block. In ultrasonic test setups.
addition, the attenuation in the test object is commonly dif-
ferent than the one in the reference standard. Test Indications Requiring Special
Consideration
Individual discontinuities that are large compared to the
effective beam diameter are also detected by the presence of Contoured Surfaces
reflections between the front and back surfaces. However,
the discontinuity size cannot be determined by comparing it Reflections from fillets and concave surfaces may produce
with a reference block. The extent of the discontinuity can test indications between the front and back reflections; these
be obtained by moving the transducer along the surface of indications can be confused with indications from discontinu-
the test object and finding the points at which the disconti- ities. Such spurious indications result from sound reflected
nuity indication is still maintained. back to the transducer at a time equivalent to the time of

The boundaries of the discontinuity are determined by FIGURE36. Discontinuitiesorientedin
marking the points where the reflection drops 6 dB below transversedirectionin 7075-T6 aluminumalloy
the amplitude of the reflection at the center of the disconti- rolledplate: (aJ A-scanindication;(bJ cross
nuity. As an alternative approach, a C-scan system can be sectionof discontinuity
used with its receiver gain or attenuation calibrated so that,
when scanning a reference standard, the exact sizes of dis- (aJ DISCONTINUITY BACK REFLECTION
continuities are displayed.
-,(bJ
Time-of-Flight Measurements

Time-of-flight measurements typically are made between
the test object's front surface and the next significant reflec-
tor. When this reflector is the object's back surface, there
are no discontinuities. Pulse echo measurements of time of
flight will provide very good resolution. A near surface reso-
lution of 0.5 mm (0.02 in.) and a far surface resolution of
0.25 mm (0.01 in.) can be reliably obtained with a 5 MHz,
highly damped transducer.

Plate and Extrusions

Grain direction (the direction the metal flows during
working) is relatively simple to determine in a plate. Discon-
tinuities are generally parallel to the plate surface and elon-
gated in the direction that received the most rolling,
although there are variations to this generalization (Fig. 36).

Discontinuities in extrusions are nearly always elongated
in the direction of extrusion (along the long axis of the extru-
sion). In the case of plate and extrusions, it is very important
to note recurring discontinuity indications when scanning
parallel to the direction of grain flow.

Die Forging

Grain flow is a complex process in die forgings. Disconti-
nuities are not necessarily oriented parallel to the surface or
elongated in the long dimension of the test object. With the
typically complicated geometry of forgings, these factors
make detection and evaluation of discontinuities difficult.

In some instances where large, complex die forgings are
being tested, it is possible to section a sample forging to

412 I NONDESTRUCTIVETESTINGOVERVIEW

flight from a discontinuity at a given distance below the sur- false indication is caused by surface waves reflected from a
face. It is not always possible to distinguish between disconti- nearby edge on the extremely smooth surface.
nuities and false indications from curved surfaces, but the
four methods listed below can help discern the difference. Such false indications can be eliminated by slightly dis-
turbing the entry surface - coating with wax crayon or a
Frequently, if a false indication results from a contoured thin film of petroleum jelly. One of the distinguishing char-
surface, the amplitude of the indication is related to the acteristics of this false indication is its consistency. It is good
amplitude of the reflection from the front surface. In this practice to be suspicious of any indication that is unusually
case, the amplitude of the front surface echo diminishes as consistent in amplitude and appearance when the trans-
the false indication increases. ducer is passing over the test object.

A false indication tends to be consistent as the trans- Location of Discontinuities
ducer is moved along the contoured surface. A reflection
from a discontinuity tends to be strongly localized. False Because of the near field zone effect and equipment
indications from contoured surfaces are more likely to result recovery time, discontinuities that are very close to the test
in a broad based indication. Discontinuity indications are surface cannot always be detected while scanning at angles
typically sharp spikes. normal to the surface. However, indications of discontinu-
ities are sometimes evident at a distance slightly less than
If false indications result from reflections around a con- that at which a definite individual peak is observed. The
toured surface, it is sometimes possible to distinguish them sound wave reflected from the discontinuity near the sur-
by interrupting the ultrasonic beam between the transducer face interferes with sound waves reflected from the front
and the surface of the test object with a foreign object (a surface but the ultrasonic equipment is unable to resolve or
piece of sheet metal, for example). If the indication is a separate the energy into two distinct signals.
reflection from a curved surface, shielding a portion of the
curved area may eliminate the false indication and allow A slight variation in the appearance of the front reflec-
most of the beam to enter the test object. tion is not necessarily an indication of a near surface discon-
tinuity: a variation in the flatness or roughness of the test
Edge Effect object surface can also produce a variation in the indication
from the front surface. Roughness or flatness variations suf-
Irrelevant indications are sometimes produced near the ficient for causing fluctuations in a front surface indication
edges of rectangular shapes. This type of indication is can usually be detected by touch. When fluctuations of the
observed when the transducer is placed close to the test front reflection cannot be attributed to surface condition,
object's edge. This effect is the result of reflections from the the possibility of a discontinuity near the surface should be
edges, even though the ultrasound enters the top of the investigated by testing from the opposite surface. Near sur-
object and is not refracted by the comer. face discontinuities may also cause a loss of back surface
reflection. To improve the detection of near surface discon-
One distinguishing characteristic of the edge effect is its tinuities, double transducer techniques can be used.
consistency. There can be some variation in the distance
below the surface (typically one quarter to one half the test Discontinuities Oriented at an Angle to the Surface
object thickness) but the location and characteristics of edge
effect indications are consistent. As the transducer travels Discontinuities oriented at an angle to the front surface
parallel to the edge of the test object, the indication remains may be difficult to detect and evaluate if care is not exercised.
relatively uniform in appearance and amplitude. In contrast Generally, it is desirable to scan first at a comparatively high
to this, an indication from a discontinuity generally shows gain level to detect discontinuities oriented at an angle to the
variation in amplitude because of roughness in the disconti- test surface. It has been shown that a 2.0 mm (0.08 in.) diam-
nuity's surface. In addition, discontinuities that give a con- eter flat bottom hole oriented at 25 degrees to the front sur-
tinuous indication over several millimeters of transducer face is not discernible on the display if the transducer is
travel are generally big enough to give a significant reduc- parallel to the surface.7 This test was conducted using a gain
tion in back reflections. level giving a peak height 50 percent of the screen from a
2.0 mm (0.08 in.) flat bottom hole at normal incidence.
Surface Conditions
It is necessary to manipulate the transducer when evalu-
Occasionally, test objects with smooth, shiny surfaces ating discontinuities oriented at an angle to the surface so
produce irrelevant or false indications. When testing plates that the sound beam strikes the plane of the discontinuity at
with smooth finish surfaces, for example, consistent indica- right angles. Even though manipulation is accomplished,
tions may exist beyond the front surface reflection. The
indications remain relatively uniform in shape and magni-
tude when the transducer is moved around the edge. The

ULTRASONIC PULSE ECHO TECHNIQUES I 41 3

discontinuities oriented at angles to the surface result in FIGURE 38. Effect of grain size on ultrasonic
indications with magnitudes slightly lower than those for indications from 4340 steel (both A-scans
discontinuities parallel to the surface. This difference is not obtained with same gain): (a) photomicrograph
large. Figure 37 shows indication amplitude as a function of of large grain material; (b) ultrasonic A-scan
angle between the discontinuity and the front surf~ce. In indication for (a); (c) photomicrograph of fine
this instance, the discontinuities were 2.0 mm (0.08 m.) flat grain material; (d) ultrasonic A-scan indication
bottom holes. The instrument gain level was constant for all for (c)
tests and the transducer was manipulated to obtain maxi-
mum indication height. (a)

Grain Size Discontinuities (b)

In an ultrasonic test of 4340 chrome-molybdenum steel
at 5 MHz, an unusually high noise level was detected.8 A
study of this material showed very large grain size compared
with ASTM grain size standards of 1 to 4 (Fig. 38a). The
large grains found in the as-received condition resulted
from: (1) high temperature during hot working and (2) sub-
sequent improper annealing.

Ultrasonic testing was performed on a similar sample
that was again heat treated to attain grain refinement. The
sample showed an absence of noise on the reference line,
indicating that with proper heat treatment a finer grain size
was obtained (Fig. 38c). Microscopic examination revealed
a refined grain size of ASTM standards 6 to 8.

In another test performed on forgings of a nickel based
alloy, a frequency of 5 MHz was used. One forging pro-
duced seven back reflections, while another showed no back

FIGURE 37. Variation of indication amplitude (c)
with discontinuity angle for 2.0 mm (0.08 in.) (d)
flat bottom hole using 5 and 2 5 MHz
transducers; transducer is oriented to give
maximum indication amplitude in each
instance; angle of discontinuity is with respect
to front surface of test object

JOO

L:Q::U>-.;c:;- 75
50
- uf- (lJ 25

_J I.....
Q.. (lJ

~<..9:

O IO I 5 20 25 30 35 40 45

DISCONTINUITY ANGLE
(degrees)

LEGEND
D = 5 MHz 9.5 mm (0.38 in.) QUARTZ TRANSDUCER

o = 25 MHz 9.5 mm (0.38 in) QUARTZ TRANSDUCER

414 I NONDESTRUCTIVE TESTING OVERVIEW

reflections with the same gain level, transducer and test fre- degrees to normal. A schematic description of the A-scan at
quency. Microscopic examination was made of these and this point is shown in Fig. 39b. The initial pulse (emitted by
other forgings to determine if any internal discontinuities the transducer) is shown at the left and the reflection from
were responsible for the ultrasonic pattern and to compare the front surface is shown next. As the transducer moves in
the grain size. The forgings with the unusually large grain an axialdirection, the refracted beam (position 1) is just cut-
size showed a loss of back reflections even though no inter- ting across the corner of the rabbit groove and producing an
nal discontinuity was present. Further investigations insignificant indication. The reflection from a rear surface is
revealed that prolonged or improper forging temperature shown at the right side.
could cause the abnormally large grain size.
Crack Indications
Interpretationof Indications from
Rotor Wheels In Fig. 40a, the transducer has been moved farther to the
right in an axial direction and the sound beam (position 2) is
The tests discussed below use as an example the straight reflected from a crack. This produces an indication on the
beam immersion test of an aircraft component. Figure 39a A-scan (Fig. 40b). The rabbit groove indication has disap-
illustrates the area tested ultrasonically in an aluminum peared because the ultrasonic beam has been repositioned.
compressor rotor wheel. The ultrasonic beam from the Because the ultrasonic pulse requires about 8 µs to make a
transducer is directed at an incidence angle of 5 degrees at round trip through 25 mm (1.0 in.) of aluminum, the time
the surface periphery. interval between the indication from the front surface and
that from the crack is 2 µs. At the position shown in Fig. 40,
According to Snell's law, the angle of the refracted beam the crack starts atabout 6.5 mm (0.25 in.) below the surface.
01 is about four times greater than 0, or about twenty
As the transducer is moved axially, the refracted ultra-
sonic beam is reflected from the face of the crack at a depth

FIGURE 39. Ultrasonic testing of aluminum FIGURE 40. Ultrasonic testing of aluminum
compressor rotor wheel at beam position 1 : compressor rotor wheel at beam position 2:
(a) beam hits corner of rabbit groove; (b) A-scan (a) beam misses rabbit groove and strikes
from position 1 shows indication from corner of crack; (b) A-scan indication from position 2,
rabbit groove showing crack reflection

fa) TRANSDUCER fa) TRANSDUCER

fbJ INITIAL FRONT REAR fbJ INITIAL FRONT REAR
PULSE SURFACE SURFACE PULSE SURFACE CRACK SURFACE

CORNER OF
RABBIT GROOVE

ULTRASONIC PULSE ECHO TECHNIQUES I 41 5

of about 3 mm (0.125 in.) below the surface. The crack indi- in the rim's heat affected zone. These occur often enough to
cation moves toward the front surface indication and the require 100 percent testing.
time interval between the two indications becomes about
The wheel cracks lie in a plane parallel to the face of the
1 us. As the transducer moves, the crack indication seems to wheel and extend in a radial direction. Ultrasonic testing is
the only means to detect cracks in this orientation. Fig-
move toward the front surface reflection. This indicates that ure 4la shows the direction of the ultrasonic waves and the
the crack is not parallel to the surface but instead is position of the discontinuity in the cross section of the wheel.
approaching the surface (the sound beam's angle of inci-
dence has not changed with the change in the transducer's During testing, the wheel is rotated on a turntable
axial position). immersed in water. The A-scan display is shown in Fig. 4lb
for the condition in which the transducer is oriented over a
Indications of Weld Cracks solid metal path. Both the A-scan and B-scan presentations
A cross section of a welded turbine rotor is shown in are shown (in Fig. 4lc and 4ld) with the transducer over an
area containing a crack. The discontinuity is large enough in
Fig. 41. In this wheel, a rim of forged stainless steel is this case to cause a reflection of high amplitude and a com-
welded to a hub of forged ferritic material. Despite plete loss of back surface reflection (Fig. 4ld).
advanced welding techniques, cracks occasionally develop
The distance between front surface reflection and the
FIGURE 41 . Ultrasonic testing of welded reflection from the crack indicates the cracks depth below
turbine rotor: (a) beam position for crack in the surface. Care must be exercised when angulating the
heat affected zone; (bJ A-scan indication over sound beam close to the interior interface of weld and rim
sound material; (cJ A-scan indication over crack, material - the beam can be reflected off these faces and is
also showing loss of back reflection; (dJ 8-scan rapidly attenuated. Consequently, this area is not always
indication over crack effectively tested by the ultrasonic method. Radiographic
(a) tests are also used for the weld area.

'·: u(b) FRONT SURFACE Indications of Metallic Inclusions and Segregations
REAR SURFACE
Metallic inclusions in the same plane as cracks give similar
indications in the heat affected zone. They are found most
frequently in the rim and farther away from the weld area.

Ultrasonic testing is also used for solid forged turbine
rotor wheels of certain stainless steel alloys fabricated with-
out welding. The technique is used chiefly to detect sharp
cracks resulting from the forging process (Fig. 42).

FIGURE 42. Photomicrograph of turbine wheel
showing sharp discontinuity caused during
forging

[__(c) FRONT SURFACE

(d) -------- FRONT SURFACE

-- CRACK

--- REAR SURFACE FROM GENERAL ELECTRIC COMPANY. REPRINTED WITH PERMISSION.

416 I NONDESTRUCTIVE TESTING OVERVIEW

Forgings may also contain segregates that reflect ultra- Figure 45 shows a B-scan presentation of a complete contour
sonic beams (Fig. 43). These indications cannot be distin- forged rim with reflections from the outside diameter and the
guished from those caused by cracks. Metallurgists have inside diameter surfaces and from the familiar clustered
determined that mechanical properties of materials contain- group of indications from forged bursts.
ing segregates meet normal service requirements for room
temperature tensile and elevated temperature stress rup- FIGURE 44. Ultrasonic testing of stainless steel
ture properties. These segregates can cause indications with rotor wheel with forging bursts: (a) beam
amplitudes of 15 to 90 percent of the front surface reflec- position on fillet area; (bJ A-scan indications
tion (Fig. 4lc). More highly concentrated ones cause higher from forging bursts between front surface and
amplitude indications. These peaks can appear between the rear surface
first and second multiples of the rear surface reflection.
faJ FORGING BURSTS
Discontinuities in this stainless steel occur in either a loose
or tightly bound condition. Fortunately, these discontinuities fbJ INITIAL FRONT
do not occur frequently. Although they do not adversely PULSE SURFACE
affect mechanical properties, they cannot be accepted. If they
should occur in a critical area like serrations of the rotor FORGING BURSTS
wheel, they might propagate and cause catastrophic failure.
REAR SURFACE
Indications of Forging Bursts
FIGURE 45. Ultrasonic 8-scan presentation of
Figure 44 illustrates the cross section of a stainless steel contour forged turbine wheel rim containing
rim of a turbine wheel containing forging bursts. These are forging bursts
irregularly shaped cavities caused by rupture of the material
during forging. Forging bursts are rejectable, are likely to be
clustered in groups and produce many A-scan indications of
varying degrees of amplitude. The reflections from inclu-
sions may also be of varying amplitude but are more likely to
be widely scattered. By comparison, the indication from a
crack is sometimes continuous for as much as one fourth the
circumference of the rim, with complete loss of reflection
from the rear surface.

FIGURE 43. Photomicrograph of segregate in
forged turbine wheel

FROM GENERAL ELECTRICCOMPANY. REPRINTED WITH PERMISSION.

ULTRASONIC PULSE ECHO TECHNIQUES I 41 7

Indications of Small Inclusions surface. The depression has been exaggerated to illustrate
the condition more clearly. In actual cases these depressions
The rotor wheel rims are also tested by ultrasonics for may be so slight that they cannot be seen under water and
nonmetallic inclusions too small to be detected by radiogra- their presence can be verified only by touch.
phy. These inclusions do not seriously affect the mechanical
properties of the materials as do cracks and forging bursts. Under certain conditions, such depressions can produce
false indications as shown in the A-scan in Fig. 46b. If the
The inclusions are often randomly located. If they are sound waves enter the surface at the edge of the blended
exposed on a machined surface such as a serration used for area, they can be scattered in such a way that part of the
the insertion of a bucket, a costly rejection is the result. The beam travels across the surface of the blend, reflects from
rim acceptance standard does not permit any ultrasonic the opposite edge and appears as a discontinuity indication.
indications between front and rear surface reflections. Because sound travels much slower in water than in metals,
the amplitudes and position of false indications with respect
Indication from Surface Contour Blending to front and back surface reflections depend largely on the
size, depth and contour of the blended area and the path
Surface conditions may cause false indications during length in the metallic test object.
ultrasonic tests. The safest way to avoid this occurrence is to
have the surface prepared to completely avoid ultrasonic A good practice is to machine the as-forged rotor wheel
wave scattering. A common practice is to require that the to a finish of 2 urn (8 x 10~5 in.) root mean square with oppo-
surface of test objects have a finish of at least 2 µm (8 x site faces parallel wherever possible so that the effects of
10--5 in.) root mean square. varying shape and contour can be minimized. This is more
necessary when testing materials that may contain inclu-
Surface treatment such as blending or grinding can pro- sions such as carbide bands whose reflections can be con-
duce misleading ultrasonic test results. Figure 46 shows the fused with those from serious discontinuities.
cross section of a turbine rotor where a slight depression has
been made by a blending operation in the otherwise smooth In these cases, the forging vendor (who also tests the
wheels ultrasonically) is hired to prepare the test objects by
machining. Enough stock is provided in the forging and

FIGURE 46. False indication from surface FIGURE 4 7. False indications from heat
blending: (a) path of ultrasonic beam treating scale: (a) path of ultrasonic beam
producing false indication (dashed); (b) A-scan entering scaled surface; (b) A-scan indications
showing false indication between front surface showing false indications between front
and rear surface indications surface and rear surface indications

fa) TRANSDUCER fa) TRANSDUCER

fbJ FRONT SURFACE REAR SURFACE fbJ FRONT SURFACE REAR SURFACE

FALSE INDICATION

FALSE INDICATION

418 I NONDESTRUCTIVETESTINGOVERVIEW

after machining, before ultrasonic testing, so that the the cross section of a machined turbine rotor wheel as
machine finish contours and tolerances can be maintained. obtained from a forging vendor.
This costly procedure pays for itself in greater reliability of
ultrasonic testing for critical components. After preliminary The transducer is directed onto the surface of the rotor
machining and ultrasonic testing, the rotor wheels are heat in an area containing a thin scale. The size of this scale has
treated before shipment from the forging vendor's plant. been exaggerated to illustrate the condition more clearly.
The indications from this surface condition are depicted on
Indications from Heat Treating Scales the A-scan in Fig. 47b and are generally characterized by an
increased number of scattered signals next to the particular
Heat treating can produce a thin imperceptible scale or surface reflection.
film on the surface of rotor wheels and this can in turn pro-
duce confusing ultrasonic test indications. Figure 47a shows Slight surface etching can remove the cause of these
reflections. They are a concern because they can mask indi-
cations from an actual discontinuity.

ULTRASONIC PULSE ECHO TECHNIQUES I 419

PART 7

IMMERSION TESTING OF COMPOSITE
MATERIALS

Discontinuitiesin Composite peel plies are introduced into the laminate bulk. Generally,
Laminates the presence of such inclusions prohibit sufficient bond
between the plies.
Composites are used in applications requiring materials
with high stiffness-to-weight ratios or high strength-to- During layup, ply gap can occur in a laminate if the vari-
weight ratios. Composites possess a complex failure mecha- ous impregnated composite tapes are not properly positioned
nism that causes difficulties in establishing design criteria and a gap is left between them. This gap is filled with a pocket
and nondestructive testing methods. of resin and causes a thickness reduction at its center.

Most composites are made of layers containing many Discontinuities such as delaminations, inclusions, poros-
fibers bonded with a matrix of different or equal composi- ity and ply gap can lead to property degradation not
tion. The layers are stacked at various fiber orientations, accounted for in design and ultimately can shorten the com-
depending on design requirements. The resulting nonho- posite's service life.
mogeneous, anisotropic, layered characteristics hamper the
use of some nondestructive testing methods that are well Results of Composite Discontinuities
established for homogeneous, isotropic materials.
At less than 2 percent of volume, porosity provides
In aerospace, three composite systems are commonly improved fracture toughness. However, porosity also reduces
used: graphite-epoxy (for critical structures), glass-epoxy compressive and interlaminar shear strength and compro-
and aramid-epoxy. The diameters of the fibers vary from 1 mises the fatigue life of the material. Porosity can produce an
to 10 µm for graphite and 5 to 15 µm for glass. Several types increase in the moisture equilibrium level and aggravates the
of discontinuities are commonly induced during the manu- thermal spike phenomenon. Both of these conditions lead to
facture and service life of a composite structure. deterioration in the material's elastic properties.

Causes of Composite Discontinuities Delaminations are a more severe discontinuity because
they do not transfer interlaminar shear stresses. Under com-
Composite systems have a tendency to nucleate porosity pressive loading, delamination can cause rapid and catas-
if the volatile components in the resin are not properly trophic buckling failure. The presence of a peel ply inside
removed during cure. At curing, trapped air is pushed out, the laminate is harmful because of the low interlaminar
typically along the fibers and between composite layers shear it provides.
because of the high resin content in these regions. Once the
curing composite passes the gel stage, it begins to harden The effect of ply gaps depends on the stacking order and
and air is trapped in porosity or voids. In addition to poros- the discontinuity location. As an example, for [O, +45, 90,
ity, improper cure can lead to a partial delamination of the -45]2s laminate, a 2.5 mm (O.l in.) gap in the O degree layers
composite plies. reduces the tensile strength by 8 percent. The same size gap
in the 90 degree layer reduces the strength by 17 percent.9
Generally, delaminations result from hole drilling and
impact damage. Delaminations can also be the result of The most severe anomaly is impact damage; even at lev-
stresses at the free edges of the composite, when the trans- els barely visible from the outside surface, impact damage
verse tensile or shear strength is exceeded. During the layup can degrade the material strength by more than 50 percent.
process, foreign materials tend to be introduced, particu-
larly the plastic carrier film and the release paper on which Ultrasonic Testing of Composite
impregnated (uncured) composite plies are delivered. Laminates

Another foreign material that can be left in the compos- Composites are usually tested for delamination using the
ite is peel ply. Peel plies are used to prevent bonding of the straight beam immersion technique. Time-of-flight measure-
laminate to the mold during cure and sometimes pieces of ments are used to map the depth distribution of discontinu-
ities. Loss of back surface reflection provides a profile of

420 I NONDESTRUCTIVETESTINGOVERVIEW

severity. Time-of-flight measurements can be used to detect Reflector Plate
less than 1 mm (0.04 in.) diameter delaminations in graphite-
epoxy laminates with ±0.2 mm (0.008 in.) accuracy.-" In some thin structures, the ultrasonic beam attenuation
may not be high enough for testing. When this occurs, it is
Figure 48 shows A-scans of a 16-layer graphite-epoxy common to use a reflector plate to detect small discontinu-
composite bonded to an aluminum honeycomb through a ities. The reflector plate may be highly polished metal
protective layer of glass-epoxy and adhesive bond. The designed for the purpose. The bottom of the submersion
ultrasonic test uses short duration pulses in the range of tank may also be used.
100 ns.
The plate is used with the straight beam immersion
Figure 49 shows time-of-flight C-scan imaging of delam- method and the test object is placed between the transducer
inations that resulted from impact damage. Using the com- and the plate. Tests are conducted by monitoring the
puter to conduct a three-dimensional rotation, the depth changes in reflection amplitude from the front surface of
distribution of the delaminations is clearly identified. the plate after passing twice through the test object.

The examination of back surface reflection amplitude Tests of Composite Tubing
provides a measure of the attenuation and can reveal the
presence of material changes and discontinuities. Typical Composite tubing is commonly manufactured by a fila-
discontinuities that affect attenuation are porosity, voids, ment winding process rather than by stacking layers to build
deviations in resin-to-fiber volume ratio and presence of a laminate. During the manufacture of composite tubing,
impact damage. several types of discontinuities can be induced and some of
these may cause deterioration in performance. For such
Many factors that are not related to the material quality tubes, straight beam ultrasonic tests can be performed with
may also affect attenuation, so that attenuation changes water jet or bubbler instruments.6
serve as a discontinuity indicator only when a severe change
is observed. Porosity, for example, is typically detected by its Tubing Discontinuity Types and Their Characteristic
effect on attenuation. However, at low volume (below 3 per- Response
cent), the effect of porosity can have about the same effect
as surface roughness or geometrical variations. The discontinuities below can be induced in a filament
winding process. The characteristic responses when tested
FIGURE48. Delaminationdetectionin with straight beam ultrasonics are also detailed below. Each
graphite-epoxy(depth of discontinuitiesis
indicatedby time-of-flightvalue): (a) front FIGURE49. ComputerizedC-scanof impact
surface reflection;(b) reflectionfrom fiberglass damage in graphite-epoxylaminate
layer; (c) back surfacereflectionfrom adhesive
layer; (d) reflectionfrom 1 mm (0.04 in.)
delamination

(aJ - --

II

J (bJ

vI IVf\v r-»; - -/\/\ {\ -
(cJ
~ (aJ

- \ ~I/\ I/\ - ~ I/\ vr - ----
"' '
(dJ

\

I

ULTRASONIC PULSE ECHO TECHNIQUES I 421

TABLE 3. Characteristic parameters of various discontinuities in filament wound tubes

Discontinuity Additional Increaseof Shift Change in Velocity Changes
Reflection Attenuation of First Time of Change near
Reflection Fligh~
Discontinuity b

Concealed cut .I
Knot
Lack of rovings .I .I .I .I .I
Impact damage
Resin starved layer .I .I .I .I .I
Flexible resin .I
Low modulus fibers .I .I

.I

.I .I .I

.I .I

a. CHANGE IN TIME OF FLIGHT CAN BE INDICATION OF THICKNESS VARIATION WITHOUT CORRESPONDING CHANGE IN ULTRASONICVELOCITY.
b. SEVERAL DISCONTINUITIES IN ONE LOCATION CAN BE DETECTED ULTRASONICALLYBEFORE POSITIONING TRANSDUCER AT DISCONTINUITY LOCATION.

of these discontinuities has distinct characteristics that can off from the front reflection. The additional reflection
be used with computer software to identify the discontinuity advances gradually with respect to the front reflection and
automatically (see Table 3). its amplitude grows steadily. In parallel, there is a decrease
in the amplitudes of the other echoes. When the center of
Concealed cut or ply gap is a discontinuity that produces the discontinuity is reached, the reflection pattern acquires
a resin pocket. The mismatch of acoustic properties in the the appearance of a standard echo train, advanced in time
discontinuity area, as well as the presence of increased and strongly attenuated (Fig. 51).
resin, cause a loss of back surface reflection (Fig. 50).
FIGURE 51 . Ultrasonic A-scan indications of
Knot appears as a local increase in tube thickness. As the knot (dashed line indicates location of front
discontinuity is approached, an additional reflection splits surface reflection): (aJ 10 mm (0.4 in.J from
discontinuity center; (bJ 4 mm (0.15 in.J from
FIGURE 50. Ultrasonic A-scan indication of discontinuity center; (cJ 2 mm (0.08 in.J from
discontinuity center; (dJ at discontinuity center
concealed cut: (aJ 3 mm (0.1 in.J away from
z
discontinuity center; (bJ at discontinuity center
(a) 60

(b) (a)
0....J

sz~

(b) iz==

0u

V)

0
0
(c)
(!,'.'.

~

0

1--

\z..'.J
(d) z0
~

0

<(

422 I NONDESTRUCTIVE TESTING OVERVIEW

Lack of rovings appears in a helical configuration around between the discontinuity reflection and the front reflection
the tube as a local decrease in tube thickness. When the depends on the depth of the resin starved layer. This discon-
transducer is near the discontinuity, a decrease of the echo tinuity has some of the characteristics of delaminations in
train amplitude is observed and simultaneously the whole laminates (Fig. 54).
train is displaced from the main radio frequency pulse
because of the increase in the time delay. Lack of rovings Flexible resin is a nonlocalized discontinuity caused by
also involves shorter time of flight but with no significant the use of the wrong resin, an incorrect amount of hardener
effect on the ultrasonic velocity. Whenever the severity of or by unsatisfactory curing processes. Ultrasonic pulses are
this discontinuity increases, the ultrasonic changes are more attenuated in flexible resin much more than in a properly
distinct. Reflection patterns from lack of rovings are shown hardened resin (17 dB in glass-epoxy, for example). In addi-
tion, flexible resin is associated with some decrease in the
in Fig. 52. ultrasonic velocity. The identification of this discontinuity
Impact damage involves the appearance of local cracks requires that reference velocity and attenuation values be
compared with those of the tubing being tested.
and delaminations. When the transducer is near the discon-
tinuity, additional reflections appear and the reflection from Low modulusfibers constitute another nonlocal disconti-
the inner diameter surface is decreased. This decrease nuity. It is caused by use of the wrong fibers in the compos-
reaches a maximum at the center of the damage and small ite. This discontinuity exhibits a decrease in ultrasonic
reflections may also appear in the interval between the first velocity by as much as 10 percent of the original velocity. No
and second reflections. The amplitude of the reflection significant attenuation change is observed.
from the outer tube surface is not changed at the disconti-
nuity location and increased attenuation is measured. The FIGURE 53. ,Ultrasonic A-scan indications from
reflection pattern of impact damage is shown in Fig. 53. impact damage: (a) 10 mm (0.4 in.) from
discontinuity center; (bJ 4 mm (0.15 in.) from
A resin starved layer has a nonlocalized nature, extend- discontinuity center; (cJ 2 mm (0.08 in.) from
ing through a tube over a large area. This discontinuity discontinuity center; (d) at discontinuity center
causes the appearance of additional reflections originating
at the surface of the starved layer. The time of flight

FIGURE 52. Ultrasonic A-scan indications of

lack of rovings (dashed line indicates location

of front surface reflection): (a) 10 mm (0.4 in.) fa)
z
from discontinuity center; (b) 4 mm (0.15 in.)
06
from discontinuity center; (c) 2 mm (0.08 in.)

from discontinuity center; (d) at discontinuity

center 0

z _J

60 0 fbJ s~z

fa) _J iz=

~zs 0u
Vi51
fcJ 0
fbJ zi= Cl::'.
~
0u zIo0-
Vi51 0z
0 §

Cl::'.

fcJ ~

I0- fdJ 0
oz
<(

fdJ 0z
§
0
<(

ULTRASONICPULSEECHO TECHNIQUES I 423

Laminate Test Indications Focused Beam Bond Testing

Focused Beam Tests of Laminates When a reflected ultrasonic wave impinges on a bonded
Focused beam techniques have proven valuable for test- structure, its amplitude is reduced. This occurs because of a
partial transmission of energy to the adhesive media rather
ing multilayered structures composed of aluminum sheets than complete reflection by the solid-to-air interface of the
bonded with layers of aramid. Such laminate structures tend unbonded plate.
to suffer from disbanding in various layers. Focused beams
can be used to detect delaminations and to determine their A focused transducer and a normal beam can detect rel-
depth. atively smal.l disbands in both metals and composite assem-
blies.t- The test object is scanned and its response is
With the aid of a pulse echo technique and a broad band analyzed by comparison to a reference standard.
high frequency focused transducer, images of delaminations
at the various layers have been made.'! Conclusion

FIGURE 54. UltrasonicA-scan indicationsof Ultrasonic techniques are implemented primarily in the
resin starvedlayer: fa) reference pattern; pulse echo and through-transmission modes, with contact or
fbJ discontinuitycharacteristics immersion coupling. This section covers pulse echo immer-
sion, a coupling method not commonly used in field applica-
faJ tions because of the complexity associated with maintaining
the couplant. Several means are commercially available to
fbJ overcome the limitations of immersion coupling, including
wheels and boots.

Pulse echo techniques provide very detailed information
about test objects. This capability is attributed to the various
parameters that can be analyzed, including time of flight,
amplitude of back surface reflection and amplitude of extra-
neous reflections. Significant improvements in the capabil-
ity of the method have been made with the introduction of
microprocessor controlled pulser/receivers, signal analyzers
and computerized C-scan controllers. Tests have been made
more reliable, data are easier to interpret and systems are
capable of testing complex object shapes.

Pulse echo immersion may be used with materials made
of metal, plastic, composites and ceramics - the raw mate-
rials for most engineering structures. A wide variety of dis-
continuities can be detected and characterized with pulse
echo methods, where the location, depth, size and disconti-
nuity type can be determined.

424 I NONDESTRUCTIVE TESTING OVERVIEW

REFERENCES

1. Born, M. and E. Wolf. Principles of Optics. Oxford, 7. Kleint, R.E. "Relationship between Defect Orien-
United Kingdom: Pergamon Press (1970). tation and Ultrasonic Indications." Nondestructive
Testing. Vol. 15, No. 1. Columbus, OH: American
2. Papadakis, E. "Ultrasonic Velocity and Attenuation: Society for Nondestructive Testing (January-
Measurement Methods with Scientific and Industrial February 1957): p 30-34.
Applications." Physical Acoustics ­ Principles and
Methods. Vol. 12. W Mason and R. Thurston, eds. 8. Metallurgical Investigation of Sample of 4340 Steel.
New York, NY: Academic Press (1976): p 277-374. Report No. 330. Seattle, WA: Boeing Airplane
Company (1955).
3. Tittmann, B., L. Ahlberg and K. Fertig. "Ultrasonic
Characterization of Microstructures in Powder 9. Bar-Cohen, Y. "NDE of Fiber Reinforced Compos-
Metal Alloy." Materials Analysis by Ultrasonics. A. ite Materials - A Review." Materials Evaluation.
Vary, ed. Park Ridge, NJ: Noyes Data Corporation Vol. 44, No. 4. Columbus, OH: American Society
(1987): p 30-46. for Nondestructive Testing (April 1986): p 446-454.

4. Bar-Cohen, Y. "Ultrasonic NDE of Composites - 10. Bar-Cohen, Y., U. Amon and M. Meron. "Defect
A Review." Solid Mechanics Research for Quantita­ Detection and Characterization in Composite
tive Non­Destructive Evaluation. J. Achenbach and Sandwich Structures by Ultrasonics." SAMPE Jour­
Y. Rajapakse, eds. Dordrecht, Netherlands: Marti- nal. Vol. 14, No. 1. Covina, CA: Society for the
nus Nijhoff Publishers (1987): p 188-201. Advancement of Material and Process Engineering
(1978): p 4-9.
5. Krautkramer, J. and H. Krautkramer, Ultrasonic
Testing of Materials, first edition. New York, NY: 11. Bar-Cohen, Y. "Nondestructive Characterization of
Springer Verlag (1969): p 238. Defects Using Ultrasonic Backscattering Measure-
ments." Proceedings of the Ultrasonic International
6. Bar-Cohen, Y. et al. "Ultrasonic NDE Methods of Conference. Surrey, United Kingdom: Butterworth
Detection and Identification of Defects in Filament and Company (July 1987).
Wound Glass Fiber Reinforced Plastic Tubes."
Materials Evaluation. Vol. 37, No. 8. Columbus, 12. Hagemaier, D. and R. Fassbender. "Nondestructive
OH: American Society for Nondestructive Testing Testing of Adhesive Bonded Structures." SAMPE
(August 1979): p 51-55. Quarterly. Vol. 9. Covina, CA: Society for the
Advancement of Material and Process Engineering
(July 1978): p 36-58.

12SECTION

VISUAL TESTING

Stanley Ness, Mission Viejo, California

426 I NONDESTRUCTIVE TESTING OVERVIEW

PART 1

DESCRIPTION OF VISUAL AND OPTICAL
TESTS

Nondestructive tests typically are done by applying a Geometrical Optics
probing medium (such as acoustic or electromagnetic
energy) to a material. After contact with the test material, Image Formation
certain properties of the probing medium are changed and
can be used to determine variations in the characteristics of Most optical instruments are designed primarily to form
the test material. Density differences in a radiograph or images. In many cases, the manner of image formation and
location and peak of an oscilloscope trace are examples of the proportion of the image can be determined by geometry
means used to indicate probing media changes. and trigonometry without detailed consideration of the
physics of light rays.
Ina practical sense, most nondestructive tests ultimately
involve visual tests - a properly exposed radiograph is use- This practical technique is called geometrical optics and
ful only when the radiographic interpreter has the vision it includes the formation of images by lenses and mirrors.
acuity required to interpret the image. Likewise, the accu- The operation of microscopes, telescopes and borescopes
mulation of magnetic particles over a crack indicates to the also can be partially explained with geometrical optics. In
inspector an otherwise invisible discontinuity. Visual and addition, the most common limitations of optical instru-
optical tests use probing energy from the visible portion of ments can be similarly evaluated with this technique.
the electromagnetic spectrum. Changes in the light's prop-
erties after contact with the test object may be detected by Light Sources
human or machine vision. Detection may be enhanced or
made possible by mirrors, magnifiers, borescopes or other The light source for visual tests typically emits radiation
vision enhancing accessories. of a continuous or noncontinuous (line) spectrum. Mono-
chromatic light is produced by use of a device known as a
Luminous Energy Tests monochromator, which separates or disperses the wave-
lengths of the spectrum by means of prisms or gratings and
Visual testing was probably the first method of nonde- selects a narrow portion of that spectrum by an aperture.
structive testing. It has developed from its ancient origins
into many complex and elaborate optical investigation tech- Less costly and almost equally effective for routine tests
niques. Some visual tests are based on the simple laws of are light sources emitting distinct spectral lines. These
geometrical optics. Others depend on properties of light, include mercury, sodium and other vapor discharge lamps.
such as its wave nature. A unique advantage of many visual Such light sources may be used in combination with glass, liq-
tests is that they can yield quantitative data more readily uid or gaseous filters or with highly selective interference fil-
than other nondestructive tests. ters, for transmitting only radiation of a specific wavelength.

Luminous energy tests are used primarily for two pur- Stroboscopic Sources
poses: (1) testing of exposed or accessible surfaces of
opaque test objects (including a majority of partially assem- The stroboscope is a device that uses synchronized
bled or finished products) and (2) testing of the interior of pulses of high intensity light to permit viewing of objects
transparent test objects (such as glass, quartz, some plastics, moving with a rapid, periodic motion. A stroboscope can be
liquids and gases). For many types of objects, visual testing used for direct viewing of the apparently stilled test object
can be used to determine quantity, size, shape, surface fin- or for exposure of photographs.
ish, reflectivity, color characteristics, fit, functional charac-
teristics and the presence of surface discontinuities. The timing of the stroboscope also can be adjusted so
that the moving test object is seen to move but at a much
slower apparent motion. The stroboscopic effect requires an

VISUAL TESTING I 427

accurately controlled, intermittent source of light or may be time. Photographic film emulsions can be selected to meet
achieved with periodically interrupted vision. specific test conditions, sensitivities and speeds.

Light Detection and Recording Fluorescence Detection

Once light has interacted with a test object (been ab- A material is said to fluoresce or phosphoresce when
sorbed, reflected or refracted), the resulting light waves are exposure to radiation causes the material to produce a sec-
considered test signals that may be recorded visually or pho- ondary emission of longer wavelength than the primary,
toelectrically. Such signals may be detected by means of exciting light. Visual tests based on fluorescence or phos-
photoelectric cells, bolometers or thermopiles, photomulti- phorescence play a part in qualitative and quantitative inor-
pliers or closed circuit television systems. ganic and organic chemistry, as a means of quality control of
chemical compounds, for identifying counterfeit currency,
Electronic image conversion devices often are used for tracing hidden water flow and other methods of leak detec-
the invisible ranges of the electromagnetic spectrum tion and for detecting discontinuities in metals and pave-
(infrared, ultraviolet or X-rays) but they also may be used to ment. Collectively, fluorescence and phosphorescence are
transmit visual data from hazardous locations or around referred to as luminescence and can usually be phenomeno-
obstructions. logically distinguished by their lifetimes. If the lumines-
cence apparently stops immediately after removal of the
Occasionally, intermediary photographic recordings are exciting radiation, then the luminescence is probably fluo-
made. The processed photographic plate can subsequently rescence, but if it continues, then the luminescence is
be evaluated either visually or photoelectrically. Some almost certainly phosphorescence.
applications take advantage of the ability of photographic
film to integrate low energy signals over long periods of

428 I NONDESTRUCTIVE TESTING OVERVIEW

PART 2

VISION AND LIGHT

The Physiology of Sight similarity that the brain momentarily retains as the eye
moves from one group to the other. On the other hand, the
Visual Data Collection tilted Ts share no edge orientations with the upright Ts,
making them stand out in the figure.
Human visual processing occurs in two steps. First the
entire field of vision is processed. This is typically an auto- Differentiation of colors is more difficult when the dif-
matic function of the brain, sometimes called preattentive ferent colors are in similarly shaped objects in a pattern.
processing. Secondly, focus is localized to a specific object in The recognition of geometric similarities tends to over-
the processed field. Studies at the University of Pennsylva- power the difference in colors, even when colors are the
nia indicate that segregating specific items from the general object of interest. Additionally, in a grouping of different
field is the foundation of the identification process. shapes of unlike colors, where no one form is dominant, a
particular form may hide within the varied field of view.
Based on this concept, it is now theorized that various However, if the particular form contains a major color vari-
light patterns reaching the eyes are simplified and encoded, ance, it is very apparent. Experiments have shown that such
as lines, spots, edges, shadows, colors, orientations and ref- an object may be detected with as much ease from a field of
erenced locations within the entire field of view. The first thirty as it is from a field of three.3
step in the subsequent identification process is the compar-
ison of visual data with the long-term memory of previously Searching the Field of View
collected data. Some researchers have suggested that this
comparison procedure is a physiological cause of deja vu, The obstacles to differentiation discussed above indicate
the uncanny feeling of having seen something before.' that similar objects are difficult to identify individually. Dur-
ing preattentive processing, particular objects that share
The accumulated data are then processed through a common properties such as length, width, thickness or ori-
series of specific systems. Certain of our light sensors entation are not different enough to stand out. If the differ-
receive and respond only to certain stimuli and transmit ences between a target object and the general field is
their data to particular areas of the brain for translation. dramatic, then a visual inspector requires little knowledge
One kind of sensor accepts data on lines and edges; other of what is to be identified. When the target object is similar
sensors process only directions of movement or color. Pro- to the general field, the inspector needs more specific detail
cessing of these data discriminates between different com- about the target. In addition, the time required to detect a
plex views by analyzing their various components.2 target increases linearly with the number of similar objects
in its general field.
By experiment, it has been shown that these areas of sen-
sitivity have a kind of persistence. This can be illustrated by When an unspecified target is being sought, the entire
staring at a lit candle, then diverting the eyes toward a blank field must be scrutinized. If the target is known, it has been
wall. For a short time, the image of the candle is retained. shown statistically that only about half the field must be
The same persistence occurs with motion detection and can searched.
be illustrated by staring at a moving object, such as a water-
fall, then at a stationary object like the river bank. The bank FIGURE 1 . Pattern changes illustrating
will seem to flow because the visual memory of motion is boundary and edge detection
still present.

Differentiation in the Field of View

Boundary and edge detection can be illustrated by the
pattern changes in Fig. 1. When scanning the figure from
left to right, the block of reversed Ls is difficult to separate
from the upright Ts in the center but the boundary between
the normal Ts and the tilted Ts is easily apparent. The diffi-
culty in differentiation occurs because horizontal and verti-
cal lines comprise the L and upright T groups, creating a

VISUAL TESTING I 429

The differences between a search for simple features changing its focal length. Eye muscles aid in the altering of
and a search for conjunctions or combinations of features the lens shape as well as controlling the point of aim.
can also have implications in nondestructive testing environ-
ments. For example, visual inspectors may be required to This configuration achieves the best and sharpest image
take more time to check a manufactured component when for the entire system. The retina consists of rod and cone
the possible errors in manufacturing are characterized by nerve endings that lie beneath the surface. They are in
combinations of undesired properties. Less time could be groups that represent specific color sensitivities and pattern
taken for a visual test if the manufacturing errors always recognition sections. These areas may be further subdivided
produced a change in a single property.4 into areas that collect data from lines, edges, spots, positions
or orientations.
Another aspect of searching the field of view addresses the
absence of features. The presence of a feature is easier to The light energy is received and converted to electrical
locate than its absence. For example, if a single letter O is signals that are moved by way of the optic nerve system to
introduced to a field of many Qs, it is more difficult to detect the brain where the data are processed. Because the light is
than a single Q in a field of Os. The same difficulty is appar- being reflected from an object in a particular color or com-
ent when searching for an open O in a field of closed Os. In bination of colors, the individual wavelengths representing
this case statistics show that the apparent similarityin the tar- each hue also vary. Each wavelength is focused at different
get objects is greater and even more search time is necessary. depths within the retina, stimulating specific groups of rods
and cones (see Figs. 2 and 3). The color sensors are grouped
Experimentation in the area of visual search tasks in specific recognition patterns as discussed above.
encompasses several tests of many individuals. Such experi-
ments start with studies of those features that should stand To ensure reliable observation, the eye must have all the
out readily, displaying the basic elements of early vision rays of light in focus on the retina. When the point of focus
recognition. The experiments cover several categories, is short or primarily near the inner surface of the retina clos-
including quantitative properties such as length or number. est to the lens, a condition known as nearsightedness
Also included are search tasks concentrating on single lines, (myopia) exists. If the focal spot is deeper into the retina,
orientation, curves, simple forms and ratios of sizes. All farsightedness (hyperopia) occurs. These conditions exist
these tests verify that visual systems respond more favorably because the axis of the eyeball is too long and/or the refrac-
to targets that have something added ( Q versus O) rather tive power of the lens too high ( myopia) or the axis is too
than something missing. short and/or the refractive power of the lens is not high
enough (hyperopia). Astigmatism is a form of abnormal
In addition, it has been determined that the ability to vision in which a variable degree of axis distortion exists in
distinguish differences in intensity becomes more acute different meridians of the eyeball and/or a variable degree
with a decreasing field intensity. This is the basis of Weber's of refractive power exists in these two meridians.
law. The features it addresses are those involved in the early
visual processes: color, size, contrast, orientation, curvature, FIGURE 2. Components of the human eye in
lines, borders, movement and stereoscopic depth. cross section

Vision Acuity SCLERA (WHITE OUTER COVERING/

LAYER OF RODS AND CONES

Vision acuity encompasses the ability to see and identify
what is seen. Two forms of vision acuity are recognized and
must be considered when attempting to qualify visual abil-
ity. These are known as near vision andfar vision.

Components of the Human Eye PUPIL
IRIS
The components of the human eye (Fig. 2) are often
compared to those of a camera. The lens is used to focus OPTIC NERVE
light rays reflected by an object in the field of view. This
results in the convergence of the rays on the retina (film),
located at the rear of the eyeball. The cornea covers the eye
and protects the lens. The quantity of light admitted to the
lens is controlled by the contraction of the iris (aperture).
The lens has the ability to become thicker or thinner, thus

430 I NONDESTRUCTIVE TESTING OVERVIEW

Determining Vision Acuity size. Each group is a few lines long and the lettering is black.
In a vision examination using this chart, visual testing per-
The method normally used to determine what the eye sonnel may be required to read, for example, the smallest
can see is based on the average of many measurements. The letters at a distance of 300 mm (12 in.). Near vision acuity
average eye views a sharp image when the object subtends examinations that are more clinically precise are described
an arc of five minutes, regardless of the distance the object below.
is from the eye. The variables in this feature are the diame-
ter of the eye lens at the time of observation and the dis- Far Vision Examinations
tance from the lens to the retina. Conditions are the same as those for near vision exami-

When vision cannot be normally varied to create sharp nations, except that the chart is placed 6 m (20 ft) from the
clear images, then corrective lenses are required to make eye plane. Again, each eye is tested independently.
the adjustment. While the eye lens is about 17 mm (0.7 in.)
from the retina, the ideal eyeglass plane is about 21 mm Grading Vision Acuity
(0.8 in.) from the retina. Differences in facial features must The criterion for grading vision acuity is the ability to see
therefore be considered when fitting for eyeglasses. Under
various working conditions, glass lenses may not stay at their and correctly identify 7 of 10 optotypes of a specific size at a
ideal location. This can cause slight variations when evaluat- specific distance. The average individual should be able to
ing minute details and such situations must be individually read six words in four to five seconds, regardless of the letter
corrected. size being viewed.

For the majority of visual testing applications, near vision The administration of a vision acuity examination does
acuity is required. Most visual inspections are performed not necessarily require medical personnel, provided the
within arm's length and the inspector's vision should be administrator has been trained and qualified to standard
examined at 400 mm (16 in.) distance. Examinations for far and approved methods. In some instances specifications
vision are done at distances of 6 m (20 ft). may require the use of medically approved personnel. In
these cases, the administrator of the examination may be
Vision Acuity Examinations trained by medically approved personnel for this applica-
tion. In no instance should any of these administrators try to
Visual testing may occur once or more during the fabri- evaluate the examinations.
cation or manufacturing cycle to ensure product reliability.
For critical products, visual testing may require qualified FIGURE 3. Magnified cross section showing
and certified personnel. blind spot of human eye

Certification of the visual test itself may also be required BIPOLAR CELLS
to document the condition of the material at the time of
testing. In such cases, testing personnel are required to suc- OPTIC
cessfully complete vision acuity examinations covering spe- NERVE
cific areas necessary to ensure product acceptability. For
certain critical inspections, it may be required for the eyes BLIND SPOT
of the inspector to be examined as often as twice per year. (WHERE FIBERS

Near Vision Examinations LEAVE RETINA
TO FORM
The examination distance should be 400 mm (16 in.)
from the eyeglasses or from the eye plane, for tests without OPTIC NERVE)
glasses. When reading charts are used, they should be in the
vertical plane at a height where the eye is on the horizontal LIGHT
plane of the center of the chart. Each eye should be tested RECEPTORS
independently while the unexamined eye is shielded from
reading the chart but not shut off from ambient light.

For near vision acuity examinations, the Jaeger eye chart
is widely used in the United States. The chart is a 125 x
200 mm (5 x 8 in.) off-white or grayish card with an English
language text arranged into groups of gradually increasing

VISUAL TESTING I 431

If an applicant does not pass the examination (fails to TABLE 1. Eye examination system conversion chart
give the minimum number of correct answers required by
specification), the administrator should advise the applicant Eye Slide Slide Display with
to seek a professional examination. If the professional Chart Display OcularSystem
responds with corrective lenses or a written evaluation stat-
ing the applicant can and does meet the minimum stan- 360 mm (14 in.J
dards, the applicant may be considered acceptable for
performance of the job. I JO 20/20
28 20/28
Vision Acuity Examination Requirements 36 20/33
45 20/38
There are some basic requirements to be followed when 5 4.6 20/42
setting up a vision acuity examination system. The distances 6 4.3 20/50
mentioned above are examples but there are also detailed 74 20/55
requirements for the vision chart. 83 20/60
9 2.5 20/63
The chart should consist of a white matte finish with 10 2 20/65
black characters or letters. The background should extend
at least the width of one character beyond any line of char- The room lighting for examinations using charts should
acters. Sloan letters as shown in Fig. 4 were designed to be be 800 1x (75 ftc). Incandescent lighting of the chart is rec-
used where letters must be easily recognizable. Each char- ommended to bring the background luminance up to 85 ±
acter occupies a five stroke by five stroke space. 5 cd-m". Fluorescent lighting should not be used for vision
acuity examinations. Incandescent lamps emit more light in
The background luminance of the chart should be 85 ± the yellow portion of the visible spectrum. This makes read-
5 cd-m". The luminance is a reading of the light reflected ing more comfortable for the examinee. Fluorescent lamps,
from the white matte finish toward the reader. especially those listed as full spectrum, are good for color
vision examinations.
When projected images are used, the parameters for the
size of the characters, the background luminance and the Many of the lighting conditions for vision acuity exami-
contrast ratio are the same as those specified for charts. In nations can be met by using professional examination units.
no case should the contrast or illumination of the projected With one such piece of equipment, the examinee views
image be changed. A projection lamp of appropriate slides under controlled, ideal light conditions.
wattage should be used. When projecting the image, room
lighting is subdued. This should not cause any change in the Another common design is used both in industrial and
luminance of the projected background contrast ratio to medical examinations. With this unit, the individual looks
that of the characters.
into an ocular system and attempts to identify numbers, let-
FIGURE 4. Letters used for acuity examination ters or geometric differences noted in illuminated slides.
charts (measurements in stroke units) The examinee is isolated from ambient light.

-y/ 37 DEGREES The slides and their respective data were developed by
the Occupational Research Center at Purdue University,
I based on many individuals tested in many different occupa-
tions. Categories were developed for different vocations and
~1- are provided as guides for examinations required by various
jf--5-I industries. Such equipment is expensive and accordingly
eye charts are still very popular. Table 1 compares the
D} K-:~H N1.5 1_-STl5t results of these three vision acuity examination systems.
125 DEGREES -I -
There are slight differences between the reading charts
v "R2z1,0 C41 -11- I and the slides. The reading chart distance for one popular
letter card is 400 mm (16 in.). The simple slide viewer is set
l for near vision testing at 330 mm (13 in.).

There also are some differences between individual
examination charts. Most of the differences are the result of
variances in typeface, ink and the paper's ink absorption
rate. Regardless of the examination system, requirements
for lighting and contrast remain the same.

432 I NONDESTRUCTIVETESTINGOVERVIEW

Visual Angle Visual Testing Viewing Angle

Posture The angle of view is very important during visual testing.
The viewer should in all cases attempt to observe the target
Posture affects the manner in which an object is on the center axis of the eye. The angle of view should not
observed - appropriate posture and viewing angle are vary more than 45 degrees from normal.
needed to minimize fatigue, eyestrain and distraction. The
viewer should maintain a posture that makes it easy to main- The same principle applies to objects being viewed
tain the optimum view on the axis of the lens. through accessories such as mirrors or borescopes. The fiel?
of view should be maintained much in the same way that it
Peripheral Vision is when viewed directly.

Eye muscles may manipulate the eye to align the image On reflective backgrounds, the viewing angle should be
on the lens axis. The image is not the same unless it off normal but not beyond 45 degrees. This is done so that the
impinges on the same set of sensors in the retina. (see light reflected off the surface is not directed toward the eyes,
Fig. 5). As noted above, different banks of sensors basically reducing the contrast image of the surface itself. It also allows
require different stimuli to perform their functions with the evaluation of discontinuities without distorting their size,
optimum results. Also, light rays entering the lens at angles color or location. This is very important when using optical
not parallel to the lens axis are refracted to a greater degree. devices to view areas not available to direct line of sight.
This changes the quality and quantity of the light energy
reaching the retina. Even the color and contrast ratios vary Color Vision
and depth perception is altered.5
There are specific industries where accuracy of color
The commonly quoted optimum, included angle of five vision is important: paint, fabrics and photographic film are
minutes of arc is the average in which an individual encloses examples. Surface inspections such as those made during
a sharp image. There are other angles to be considered metal finishing and in rolling mills are to determine manu-
when discussing visual testing. facturing discontinuities. Color changes are not indicative of
such discontinuities and therefore, for practical purposes,
The angle of peripheral vision is not a primary consider- color is not as significant in these applications.6 However,
ation when performing detailed visual tests. It is of value heat tints are sometimes important and colors may be cru-
under certain inspection conditions: (1) when surveying cial in metallography and failure analysis.
large areas for a discontinuity indication that (2) has a high
contrast ratio with the background and (3) is observed to When white light testing is performed, it must be remem-
one side of the normal lens axis. The inspector's attention is bered that white light is composed of all the colors (wave-
drawn to this area and it can then be scrutinized by focusing lengths) in the spectrum. If the inspector has color vision
the eyes on the normal plane of the lens axis. deficiencies, then the test object is being viewed differently
than when viewed by an inspector with normal color vision.
FIGURE 5. Vision acuity of peripheral vision
Color deficiency may be as critical as the test itself. Dur-
ing visual testing of a white or near white object, slight defi-
ciencies in color vision may be unimportant. During visual
testing of black or near black objects, color vision deficien-
cies make the test object appear darker.7

Color Vision Examinations

Ten percent of the male population have some form of
color vision deficiency. The so-called color blind condition
affects even fewer people - truly color blind individuals are
unable to distinguish red and green. But, there are many
variations and levels of sensitivity between individuals with
normal vision and those with color deficiencies.

There are two causes of color deficiency: inherited and
acquired. And each of these may be subdivided into specific
medical problems. Most such subdivisions are typically dis-
covered during the first vision examination.