Você está na página 1de 27

PII: 80963~8695(96)00050-3

NDT&E International, Vol. 30, No. 2, pp. 93~98, 1997


0 1997 Elsevier Science Ltd.
Printed in Great Britain
0963-8695/97 $17.00 + 0.00

Nondestructive
testing of
adhesively-bonded
joints
R. D. Adams

and B. W. Drinkwater

Department of Mechanical

Engineering, University of Bristol, Bristol, UK

One approach
to testing the suitability
of an adhesive
joint for a particular
application
is to build and test to destruction a representative
sample of the joint. In
this way the best adhesive and surface treatment for a given application can be
found.
To reduce the costs of this approach, the designer will wish to call on previous
experience with adhesives, surface treatments, and joint designs so as to reach a
high
probability of success before he builds and tests a structural prototype. If
structures are
expensive, it will be difficult to justify more than a very limited series of
prototype
tests before production
begins. During the production
phase, and also in service
with critical structures, it is essential to use nondestructive
tests to assess the quality
and fitness for purpose of the product. The nondestructive
test will not measure
strength directly but will measure a parameter which can be correlated to strength.
It
is therefore, essential that a suitable nondestructive
test is chosen and that its results
are correctly interpreted.
In this paper, typical defects found in adhesive joints are
described together with their significance.
The limits and likely success of current
physical nondestructive
tests are described, and future trends outlined.
0 1997
Elsevier Science Ltd.
Keywords:

adhesively-bonded

structures,
nondestructive

testing,

disbonds

Poor adhesive strength is often the result of poor surface


preparation
or the presence of a contaminant
on an
adherend. Adhesive strength is very difficult to measure
nondestructively
since it is an interfacial
phenomenon
involving a very thin (usually less than 1O~l.m) layer of
material.
This aspect of quality control is, therefore,
usually reduced to assessing the nature of the adherend
surfaces prior to bonding.

The objective of any form of nondestructive


test is to
correlate the joint strength with some physical, chemical
or other parameter
which can be measured
without
causing damage. Anything which may affect the short- or
long-term
strength
is required
to be detected.
The
successful
detection
of defects is not the same as
knowing whether they are significant,
as this depends
on their extent, position, and the nature of the applied
loads. The presence of defects, however, is more likely to
be indicative
of poor joint
manufacture
than
of
impending
failure, especially
for short-term
loading.
single lap
For example, Wang et al.[‘] used aluminium
joints and an epoxy adhesive with a disbonded
area
achieved by inserting a polypropylene
disc. The presence
of this large ‘defect’ caused virtually no change in the
joint strength.

Poor cohesive strength results from either incomplete


mixing, incorrect formulation,
or from insufficient cure
of the adhesive. Unlike the adhesive strength the cohesive
strength
of the adhesive
can be estimated
with a
reasonable
degree of confidence
by a number
of
nondestructive
methods.
Complete
voids, disbonds
and porosity
are the most
simple forms of defect to detect nondestructively
and the
majority of nondestructive
testing performed on bonded
structures is aimed at detecting such defects. Porosity is
caused by entrapped air and volatiles in the adhesive, and
is present in most bond-lines
to some extent. Cracks in
the adhesive are formed by incorrect
curing (and/or
thermal
shrinkage)
or by the applied stresses, either
monotonic
or cyclic. Voids in the adhesive are caused by

There are three basic types of defect which may occur in


a simple adhesive joint, shown schematically
in Figure 1,
and which need to be monitored:

(9 poor

adhesion,
i.e. a weak bond between
the
adhesive and adherend;
(ii) poor cohesive strength, i.e. a weak adhesive layer;
(iii) complete voids, disbonds or porosity.

93
R. D. Adams

and B. W. Drinkwater

which has to take both shear and direct (tensile or


compressive)
loading,
holds the structure
together.
Figure 2 illustrates
some of the possible forms of the
skin/honeycomb
bond. Figure 2a shows a well-filled joint
in which there is a generous adhesive fillet. When the
adhesive forms such a fillet, this is due to a property
called reticulation.
However, Figure 2b shows that the
adhesive layer on the facing skin has not wetted the cell
wall and has not formed a fillet, so giving a weak bond.
Defects such as those shown in Figure 2b may be found
by monitoring
the whole bond area. Other forms of
defect in honeycomb
sandwich construction
are due to
lack of attachment
between the core and the skin. This
may be due to several causes, such as locally crushed
honeycomb,
skin defects, or lack of adhesive, causing
skin-core
disbonds.
Alone, none of these defects may
seriously reduce the short-term joint strength. However,
they show poor preparation
and manufacture
and will
provide
sites for fatigue
crack propagation,
hence
reducing the life of the structure.

zero thickness
void (no adhesion)

poor’cure

por&ity

void at surface
(finite thickness)

Figure

Defects
in adhesively-bonded

structures

(a)

Adhesive with
a good fillet
(good reticulation)

hon>yc&nb
cell walls

Tests prior to bonding

(b)

Before applying the adhesive to the joint, the adherend


surfaces
should
have been prepared
by washing/
degreasing,
abrasion,
chemical
etching
etc.[2’31. The
condition
of the surfaces of the adherend is crucial in
making a good bond. The presence of excessive amounts
of water vapour, hydrocarbons,
and other contaminants
will all result in poor adhesive strength.

No reticulation

Adhesive

of adhesive to
form fillet

cellwalls
Figure2
no fillet

Skin-honeycomb

A simple test involves the wettability of the surface; this


is, in effect, a subjective measurement
of the contact
angle. Clean surfaces are readily wetted and a drop of
water will spread over a large area. A simple but
quantifiable
test involves measuring
the spread of a
liquid drop of constant
volume through a transparent
gauge placed over the drop. The Fokker contamination
tested41 uses an oscillating probe to measure the electron
emission energy of the surface. This varies greatly with
the degree of contamination,
and can even be used to
detect residues from alkaline cleaning operations.

adhesive joints; (a) with fillet, (b) with

air or gases becoming trapped during the application


of
the adhesive, or because insufficient adhesive has been
applied.
Large voids cannot
be caused by volatiles,
unless something is very wrong with the adhesive system.
Surface unbonds
are a form of void which can occur
when an uneven coating of adhesive is applied to one
adherend
such that when the second
adherend
is
introduced
some areas have insufficient adhesive present
to form a bond between the surfaces. This type of defect
can also occur when some degree of cure has already
taken place in the adhesive before the adherends
are
brought together. Disbonds or zero-volume unbonds can
occur during manufacture
due to the presence of a
contaminant,
such as grease, on an adherend.
The
surfaces of such a disbond are in close proximity or even
touching, but will not transfer load from the adherend
to the adhesive. Disbonds can also occur as a result of
impact or environmental
degradation
of the adhesive
and adhesiveeadherend
interlayer after manufacture.

None of these methods are totally satisfactory


and the
best means of ensuring that a ‘good’ surface exists prior
to bonding is carefully to control the processes leading to
its preparation,
having previously
established
‘good
practice’. Maintaining
good control over the manufacturing process will increase the probability
of achieving a
defect-free joint and is particularly
important
in the case
of the adhesive
strength
for which no satisfactory
nondestructive
testing method currently exists.

Tests after bonding

As well as simple adhesively bonded lap-joints, the other


major form of joint used with structural adhesives is that
used in bonding a honeycomb
core to skins to form a
sandwich construction.
The resulting mesh of fine joints,

Conventional

Ultrasonics

94

ultrasonic

techniques

is one of the most widely

used methods

of
NDT of adhesively-bonded

joints

generally
done using contact
transducers
in which
coupling is achieved by applying a layer of grease or
gel to the surface of the structure.

nondestructive
examination.
In standard testing, pulses
of shear or compressional
waves at frequencies of l-20
MHz are generated by a piezoelectric
transducer.
After
the pulse has passed through the object, it is picked up by
either the transmitting
transducer (pulse-echo mode) or a
separate receiving transducer
(pitch-catch
mode). The
pulse is modified by the path taken and energy is reflected
by discontinuities
such as material
boundaries.
Large
differences in the acoustic impedance
of the materials
cause a large proportion
of the energy to be reflected.
Since a defect containing
air or any other low density
substance
will have a very low acoustic
impedance
relative to the adhesive or adherend, the ultrasonic pulse
will be almost totally
reflected.
Standard
testing is
performed with the transducer
positioned
such that the
ultrasonic
beam emitted by the transducer
is incident
perpendicular
to the surface of the structure. With the
transducer
in this configuration,
information
can be
obtained about the area of the structure illuminated
by
the ultrasonic
beam (usually
around
lo-20mm
in
diameter).
Focused transducers
can be used if greater
spatial resolution is required.

When using a liquid couplant,


care must be taken to
ensure that the couplant
(or any other liquid such as
water or fuel which may be present
in the test
environment)
is not allowed to penetrate
the defect,
which usually contains air. The presence of the liquid
reduces the acoustic impedance
difference between the
adhesive and the defect which effectively reduces the
contrast between good and defective areas, thus making
the defect much more difficult to detect, The liquid may
also remain in the joint region or even be absorbed by the
structure
if, for example,
the structure
is a fibre
reinforced resin. This moisture can then cause environmental degradation,
and so is to be avoided wherever
possible. In some cases, especially when using woven
material and absorbent fibres such as Kevlar, it is best to
test dry.
A map of the defects can be produced by scanning the
surface of the structure. The amplitude
of a particular
echo, such as that from the bottom adherend-adhesive
is measured
as the probe
traverses
the
interface,
structure.
Changes in the echo amplitude
indicate the
presence of a defect, and a record of defect location can
be obtained
by plotting
amplitude
against
position
(Figure 4). Such a plot is called a C-scan. The resolution
of small defects, such as porosity, can be improved by
decreasing the distance between the scan lines, but this
also increases the inspection time.

A display of the variation


of the magnitude
of the
reflected echoes with time can be used to indicate the
presence of defects; such displays are commonly
called
A-scans. Figure 3 shows an A-scan from a good singlelap adhesive joint. If there
were a disbond below the
adherend nearest to the transducer,
the reflections from
the lower interface would no longer be present. It is these
relatively
small signals which need to be detected to
reveal disbonds.

Because of the inconvenience


of using liquid couplants, a
great deal of research effort has been put into air coupled
and solid coupled alternatives.
Much work has been
done in recent years on laser generation
and detection
systems[5-71 and, although progress has been made, such
systems have yet to become easily applicable outside the
laboratory.
Also, the cost of laser based ultrasonic
systems is very high and continues
to limit their
application.
Work has also been done to develop
ultrasonic
transducers
which will operate
with air
coupling and, although
recent improvements
in ultrasonic technology
have demonstrated
the feasibility
of
airborne ultrasonic propagation
in the low MHz rangeL8],
few attempts
have been made to monitor
other than

A standard piezoelectric transducer


cannot operate with
air coupling due to the large difference in the acoustic
impedance
between air and solid materials.
A liquid
couplant
is required
between the transducer
and the
structure to allow transmission
to, and reception from,
the test structure. The most reliable coupling method is
immersion,
the testpiece being fully immersed in a water
bath. In such a system, the ultrasound
propagates across
the water-filled gap (typically 25-100 mm depending on
the transducer)
into the testpiece. This configuration
is
frequently not practical, particularly with large structures
or for testing in the field. Large aerospace structures are
often tested using water jet probes in which the ultrasound propagates
along a moving column
of water.
In-service testing and the inspection of irregular shapes is

Reflections from
3
.s

adherend interface

zva
6
z

Reflections from bottom


adhesive/adherend interface

S
2

Top surface reflection


Time

Figure

A-scan

from good single-lap

adhesive

Figure

joint

95

Schematic

diagram of a C-scan examination

of a flat sheet
R. D. Adams

and B. W. Drinkwater

surface reflections.
Significantly
higher frequencies
are
required if accurate structural
detail is to be imaged.
Work
is continuing
in this area19P’31 and future
developments
may offer potential
solutions
to a
significant
range of inspection
problems.
Other noncontacting
ultrasonic
transducers
include
EMATs
(electro-magnetic
acoustic
transducers)1141.
However,
EMATs
can only operate on electrically
conducting
surfaces and have much lower sensitivity
than piezoelectric transducers.

shapes
change
with propagation
distance
due to
dispersion.
If many modes are excited, the signal at the
receiver quickly becomes very complex, thus making the
detection of defects difficult.
Leaky Lamb waves are usually excited and received by
transducers
operating in pitch-catch (send-receive) mode
at oblique incidence, in adjustable angle probes designed
accurately to adjust the angle between the transducer and
the surface of the structure so as to enable the selection of
the required mode.

Solid coupling, in which a soft rubber is used to couple


the transducer
to the surface of the structure was, until
recently, limited to frequencies
below 1 MHz, but it is
now possible at higher frequenciesl’51 because of the
development
of a low loss rubber. Solid coupled transducers come in the form of a wheel probe
(or roller
probe) which rolls across the surface of the structure,
coupling occurring between the tyre and the structure.

Oblique

incidence

Due to the complexity


of the system, considerable
understanding
is required in order to set up a leaky
Lamb wave inspection
strategy on a given structure. A
mode which is sensitive to the defect of interest must be
found and excited in a region where it is both nondispersive and separate from
other modes. An alternative
approach
was used by Bork and Challisl241 who
successfully
used leaky Lamb waves to interrogate
a
steel-steel
T-peel joint, employing
a ‘neural network’
approach
to extract information
from the complex
signals generated.

ultrasonics

In oblique incidence testing, two transducers


are inclined
at equal and opposite angles with respect to the normal
of the surface. This has two main advantages.
First, the
reflections become separated in space as well as time,
which is useful if the echoes are closely spaced and
difficult to separate in time at normal incidence. Second,
several authors have suggested that obliquely incident
waves are more sensitive to interfacial
weakness than
are normally
incident
waves [‘6p’81 This technique
is
currently
not sensitive
enough
to detect interfacial
weakness (other than gross disbonds)
outside laboratory conditions.
Nagylisl has suggested
the oblique
incidence
technique
as a solution
to the detection
of
‘kissing’ bonds, which is a form of zero-volume
disbond
under compressive stress. The basis of the method is that
whilst the normal stiffness of the bond is high enough to
make it ‘invisible’, its transverse stiffness is significantly
lower than a well bonded interface.

Lamb

Sonic

vibrations

Sonic vibration techniques,


using frequencies of 1 to 30
kHz, measure
the local stiffness
of the structure.
Disbonds
reduce
the local
stiffness
as measured
perpendicular
to the surface. Usually, only disbonds or
voids can be found,
the minimum
detectable
size
depending on the size and depth of the defect. Although
the detectable
defect size is larger than that obtained
using ultrasonic
techniques,
the tests are often more
convenient
since they do not require a couplant between
the transducer
and the test structure.
Commercially
available mechanical
impedance measuring instruments
generally take measurements
at a single
pre-set frequency. The impedance
is highest over good
areas, and lowest over disbonded areas. The test becomes
unreliable when the structure itself becomes more flexible
as the impedance
of a defective zone can be higher or
lower than that of a good zone, depending
on the
frequency 125~261
One advantage of this type of test is that,
instead of using a couplant, a dry point contact is used
between the transducer
and structure. This dry contact
has a finite stiffness[271 which must be kept as high as
possible, to maximize the sensitivity of the technique.

waves

Lamb waves, or more precisely leaky Lamb waves (as a


true Lamb wave does not leak energy to the surrounding
medium and so cannot be excited or detected), are waves
which propagate
along plate-like
structures.
Such a
wave in an adhesive joint will propagate along a layered
system consisting,
for example, of air/adherend/adhesive/adherend/air.
It has been shown that the velocity
and wavelength
of leaky Lamb waves are sensitive to
the mechanical
properties of the adhesivel*‘] and to the
boundary
conditions
between the adhesive and adherend[*‘.**]. Many leaky Lamb modes exist in a given
system and these modes are usually described in plots of
phase velocity versus frequency called dispersion curves.
If care is not taken in exciting a single mode[*‘] in a region
where its velocity is not dependent
on frequency (nondispersive), it is easy to excit many modes, all of which
are propagating
at different velocities and whose pulse

Time
Figure 5
Low velocity impact test (coin tapping):
region; (b) over a defect

96

(a) over a good


NDT of adhesively-bonded

sensitive,
narrowband
piezoelectric
transducers
are
positioned
on the surface of the structure which record
the stress waves emitted
by crack propagation
and
micro-cracking.
The time of the first arrival enables the
defect to be located. The amplitude of the signals gives an
indication
of the severity of the defect and an indication
of the future life of the joint. The technique is essentially
destructive as high loads must be applied, but there are
currently few alternative measures of adhesion strength.

Tapping the structure with a hammer or coin is one of


the oldest methods of nondestructive
inspection.
This is
essentially
a subjective
method
and there has been
considerable
uncertainty
about the physical principles
involved. When a structure is struck with a hammer or
coin, the impact depends on the local impedance
of the
structure
and the hammer used. The local change in
structural
stiffness produced
by a defect changes the
nature of the impact. This is indicated by the force-time
history of the impact on the structure
which can be
measured
by incorporating
a force transducer
in the
hammer. Impacts on good and disbonded
areas of an
adhesively-bonded
structure
are shown in Figure 5.
Characteristically,
the impact on the sound structure has
a higher peak force and a shorter duration than that on
the damaged area. Thus, a measure of either peak force
or duration
can be used to locate defects. Since the
method uses only measurements
of impact force, no
transducers
need to be attached to the structure,
thus
avoiding
the coupling
and alignment
problems
which
arise with ultrasonic
techniques12*l. The sensitivity and
reliability
can be increased
by further processing
the
force-time
histories such as by taking Fourier transforms
of the signal to present the data in the frequency domain
(force vs. frequency).
Spectroscopic

Thermal

methods

By heating
one surface of a bonded
structure
and
observing
the temperature
rise of the opposite
face,
areas of disbond, which resist the transfer of heat, show
as cool areas. Alternatively,
if the heated face is scanned,
disbonds will show as hot areas. Temperature
sensing is
normally
done with a scanned infrared
camera (e.g.
AGA Thermovision).
More recently,
heat pulses or
moving heat sources have been used in conjunction
with
video recording
of the transient
thermal response[33].
Temperature
sensitive
paints or liquid crystals,
and
thermoluminescent
coatings are also used.
An alternative is to cause the structure to vibrate at one
of its resonant frequencies such that defective locations
produce frictional heating, thus leading to a local rise in
temperature [341
. The structure
is then viewed with an
infrared camera.

methods

Ultrasonic
spectroscopy
or the measurement
of the
through-thickness
vibration
characteristics
of an adhesive joint can be used for the nondestructive
testing of the
cohesive properties of the adhesive layer[291. The modulus
of the adhesive can be determined
from measurement
of
the through-thickness
natural frequencies if the thickness
of the adhesive layer is known. The Fokker Bond Test
MK 1113’]which uses a spectroscopic
measurement
is the
only commercially
available instrument
which claims to
indicate the cohesive properties of the adhesive in a joint.
Frequency and amplitude changes in the first two modes
of through-thickness
vibration
of a system comprising
the transducer
and the joint
are measured.
These
parameters
depend on both the adherend and the bond
line thicknesses
as well as the material properties.
The
range of frequencies over which the instrument
operates
is typically between 0.3 and 1.0 MHz. Small voids and
disbonds at different depths in a multilayer joint can be
detected reliably. However, prediction
of the cohesive
strength
of the adhesive is questionable[3’l
since the
frequency
shifts, resulting
from a change in cohesive
properties
or bond line thicknesses,
are small and of a
similar magnitude
to each other. Thus, to obtain a true
measure of the cohesive properties with this instrument,
the bond line thickness must be kept constant
(or be
measured separately).
Acoustic

joints

Radiography

Conventional
X-ray techniques are of little use on metalto-metal bonded joints since the
polymeric adhesive is
much less dense than the adherends.
Metallic fillers[351
can be used to enhance the contrast and show tapering or
voids. However, the density of fibre reinforced plastics
adherends is of a similar order to that of the adhesive and
so X-rays can be used, by choosing a suitable energy and
flux. For honeycomb-cored
panels, X-rays are used for
checking the position of the core and whether it has been
locally crushed or otherwise damaged.
Optical

holography

Holographic
interferometry
can be used to measure
surface displacements
to less than 0.5 pm. A laser beam is
split into two signals, one of which is reflected off the
surface of the structure and the other is taken directly
from the source. The phase difference between these
signals is then measured. Load is applied to the structure
by vibration,
vacuum cup, pressure or heat. Defects
show as local perturbations
in the holographic
interferogram[361. The technique has found application
with
sandwich structures, but not with lap or similar joints.

emission

Conclusions

Acoustic
emission[321 can be used to detect adhesion
failure prior to fracture, but the joint has to be loaded to
approximately
50% of its failure load. A number
of

Nondestructive
adhesively-bonded

97
testing
poses special
problems
for
joints, owing to the multi-layered
R. D. Adams

and B. W

nature of the construction,


involving
very different
materials and, consequently,
large differences in mechanical impedance. Also the presence of some defects which
may appear to be large has been shown to have only a
small effect on the final strength, although their presence
may indicate poor quality of manufacture.

10

11

12

No robust nondestructive
test for measuring the adhesive
or cohesive properties
of an adhesive joint currently
exists despite a large amount of research effort. Cohesive
properties can be measured to reasonable accuracy under
laboratory
conditions
using high-frequency
normalincidence ultrasonics.
Measurement
of the velocity of
leaky Lamb waves and guided waves is another future
possibility.
The low frequency through-thickness
vibration characteristcs
can also be used to measure
the
cohesive
properties.
The measurement
of adhesive
strength
is still very much the subject of research
interest,
the best possibilities
probably
being highfrequency
oblique incidence
ultrasonics
and interface
waves.

13
I4

15

16
17

18

Most nondestructive
testing techniques
are aimed at
detecting voids, cracks, porosity or lack of adhesive. The
most appropriate
nondestructive
testing method for a
given application
is dependent on many parameters,
the
most important
being the size of the defect that it is
required to detect, the geometry of the structure, and the
cost of implementing
the method.
Normal
incidence
ultrasonics
is the most generally applicable method and
is able to detect small defects in most materials. If it is
only required to detect large defects in thin structures
then the low-frequency
techniques
are a good, cheaper
alternative.
Thermography
becomes particularly
attractive if it is required to inspect large areas quickly, but it is
expensive to implement.

I9
20
21

22

23
24

25

References

26
27

Wang, T. T., Ryan, F. W. and Schonhorn, H. J. Appl. Polymer


Science, 1972, 16, 1901
Adams, R. D. and Wake, W. C. Structurul Adhesive Joints in
Erzgineering, Applied Science, London,
1986
Kinloch, A. J. Durability of StructuralAdhesives, Applied Science,
London, 1983
Bijlmmer, P. F. A. Adhesion 2, ed. K. W. Allen, Applied Science,
London, 1978, p 45
Hutchins, D. A. Ultrasonic generation by pulsed lasers. In Physicul
Acoustics, Vol XVIII, eds. Mason, W. P. and Thurston,
R. N.,
Academic Press, New York, 1988, pp 21-123
Monchalin, J.-P. Progress towards the application
of laser ultrasonics in industry.
Review of Progress in QNDE, Vol 12, eds.
Thompson,
D. 0. and Chimenti,
D. E., Plenum Press, New
York,*l993, pp 495-506
McKie. A. D. W. and Addison, R. C. Rapid inspection of composites using laser based
ultrasound.
In-Review
of Progress in
QNDE, Vol. 12, eds. Thompson,
D. 0. and Chimenti,
D. E.,
Plenum Press, New York, 1993, pp 507-516
Yano, T., Tone, M. and Fukumoto, A. Range finding and surface
characterization
using high frequency
air transducers.
IEEE
Transactions on Ultrasonics. Ferroelectrics and Frequency Control,
1987, 34, 232-236
Farlow, R. and Hayward,
G. Real-time
ultrasonic
techniques
suitable for implementing
non-contact
NDT systems employing
piezoceramic
composite
transducers.
British fournul of NDT,
1994,36, 926-935

28
29

30

31

32

33
34

35
36

98

Drinkwater

Farlow, R., Kelly, S. P. and Hayward, G. Advances in air coupled


NDE for rapid scanning applications.
In Proceedings of the 1994
IEEE Ultrasonics Symposium, Vol. 2, eds. Levy, M., Schneider,
S. C. and McAvoy, B. R., 1994, pp 1099-l 102
Schindel, D. W. and Hutchins,
D. A. Applications
of micromachined capacitance
transducers
in air-coupled
ultrasonics and
nondestructive
evaluation.
IEEE Transactions on Ultrasonics,
Ferroelectrics and Frequency Control, 1995, 42(l), 51-58
Hailer, M. I. and Khuri-Jakuh,
B. T. A surface micromachined
electrostatic
ultrasonic air transducer.
In Proceedings of the 1994
IEEE Ultrasonics Symposium, Vol. 2, eds. Levy, M., Schneider,
S. C. and McAvoy, B. R., 1994, pp 1241-1244
Carr, H. and Wykes, C. Diagnostic measurements
in capacitive
transducers. Ultrasonics, 1993, 13(l), 13-20
Thompson, R. B. Physical principles of measurements
with EMAT
transducers.
In Physical Acoustics, eds. Mason, W. P., Thurston,
R. N., Vol. XIX, Academic Press, New York, 1990, pp 157-200
Drinkwater, B. W. and Cawley, P. An ultrasonic wheel probe alternative to liquid
coupling. British Journal of NDT, 1994,36(6), 430433
Pilarski, A. and Rose, J. L. Ultrasonic
oblique incidence for
improved sensitivity in interface weakness determination.
NDT
International, 1988, 21(4), 241-246
Plarski, A., Rose, J. L. and Balasuhramian, K. The angular and
frequency
characteristics
of reflectivity
from a solid layer
embedded
between two solids with imperfect boundary
conditions. Journal of the Acoustical Society of America, 1990, 87(2),
532-542
Rokhlin, S. 1. and Marom, D. Study of adhesive bonds using lowfrequency
obliquely
incident ultrasonic
waves. Journal of the
Acoustical Society of America, 1986, 80(2), 585-590
Nagy, P. B. Ultrasonic
classification
of imperfect interfaces%
Journal of NDE, 1992, 11, 127-139
Bar-Cohen, Y., Mal, A. K. and Yin, C.-C. Ultrasonic evaluation of
adhesive bonding. Journal of Adhesion, 1989, 29, 257-274
Rokhlin, S. I., Hefets, M. and Rosen, M. An ultrasonic interfacewave method for
predicting
the strength
of adhesive bonds.
Journal of Applied Physics, 1981, 52(2), 2847-2851
Nagy, P. and Adler, L. Interface characterisation
by true guided
modes. In Review of Progress in QNDE, Vol. 10, eds. Thompson,
D. 0. and Chimenti, D. E., Plenum Press, New York, 1991, pp
1295-1302
Alleyne, D. N. and Cawley, P. Optimisation
of Lamb wave inspection techniques. NDT and E International, 1992, 25, 1 l-22
Bark. U. and Challis, R. E. NDE of the adhesive fillet size in a Tpeel joint using
ultrasonic Lamb waves and a linear network for
data discrimination.
Meas. Sci. Technol.. 1995. 6. 72 84
Cawley, P. The impedance method of nondestructive
inspection.
NDT International, 1984, 17, 59-65
Cawley, P. The sensitivity of the mechanical impedance method of
NDT. NDT International, 1988, 20(4), 209-215
Lange, Y. V. and Teumin, I. 1. Dynamic flexibility of dry point
contact. Soviet J. NDT, 1971, 7, 1577165
Cawley, P. and Adams, R. D. Sensitivity of the coin-tap method of
nondestructive
testing. Muter&
Evaluation, 1989, 47, 558-563
Cawley, P. and Hodson, M. J. The NDT of adhesive joints using
ultrasonic
spectroscopy.
In Review of Progress in QNDE, Vol.
8B, eds. Thompson,
D. 0. and Chimenti, D. E., Plenum Press,
New York, 1988, pp 1377-1384
Schliekelmann, R. J. Nondestructive
testing of bonded joints ~
recent developments
in testing systems. Nondestrucfive
Testing,
1975, 100 103
Guyott, C. C. H., Cawley, P. and Adams, R. D. Use of the Fokker
bond tester on joints with varying adhesive thickness.
Proc.
IMechE, 1987, 201(Bl), 41-49
Curtis, G. J. Nondestructive
testing of adhesively-bonded
structures
with acoustic methods. In Ultrasonic Testing ~ Non-conventional
Testing Techniques, Wiley, Chichester, 1982
Reynolds, W. N. Inspection of laminates and adhesive bonds by
pulse-video thermography.
NDT International, 1988, 21, 229-232
Pye, C. J. and Adams, R. D. Heat emission from damaged composite materials and its
use in nondestructive
testing. J. Phys D: Appl.
Phys. I98 I, 14, 927-941
Segal, E. and Kenig, S. Acceptance criteria for NDE of adhesively
bonded structures. Material Evaluation, 1989, 47(8), 921-927
Rao, M. V., Samuel, R. and Ramesh, K. Dual vacuum stressing
technique for holographic
NDT of honeycomb
sandwich panels.
NDT International, 1990, 23(5), 267-270

Você também pode gostar