A COMPARISON OF THE DETECTABILITY OF DRY CONTACT
KISSING BONDS IN ADHESIVE JOINTS USING LONGITUDINAL,
SHEAR AND HIGH POWER ULTRASONIC TECHNIQUES
C. J. Brotherhood, B. W. Drinkwater, and F. J. Guild
Department of Mechanical Engineering, Queens Building, University Walk, Bristol, BS8
1TR, UK
ABSTRACT. This paper details a study on the detectability of dry contact kissing bonds in adhesive
joints using three ultrasonic inspection techniques. Conventional normal incidence longitudinal and
shear wave inspection were conducted on dry contact kissing bonds using a standard immersion
transducer and an EMAT respectively. The detectability of the dry contact kissing bonds was
assessed by calculating the reflection coefficient of the interface at varying loads for a number of
surface roughnesses. A high power ultrasonic method was also employed to determine the non-linear
behavior of the adhesive interface. The non-linearity of the interface was determined by the ratio of
the amplitudes of the first harmonic and fundamental frequencies of the transmitted waveform. It was
found that the high power technique showed the greatest sensitivity to kissing bonds at low contact
pressures, however at high loads conventional longitudinal wave testing was more sensitive. It was
also noted that a combination of two or more techniques could provide enhanced information about
the kissing bond compared to a single technique alone.
INTRODUCTION
A large amount of literature has been published on the subject of kissing bonds[l4], however not all definitions are consistent which has lead to confusion regarding exactly
what a kissing bond is and how they are formed. This paper concerns dry contact kissing
bonds which are adhesive disbonds in which the disbonded surfaces are compressively
loaded thereby providing intimate kissing contact.
Conventional longitudinal wave [5, 6] and shear wave [7] techniques have both
been used to inspect imperfect interfaces in metal-metal contacts, whilst recent years have
seen an increase in the interest surrounding the use of high power ultrasonic techniques to
determine the non-linear behavior of structures [8-10]. This paper looks at the ultrasonic
response of dry contact kissing bonds with respect to determining their detectability. The
detectability of dry contact kissing bonds will be assessed for both longitudinal and shear
wave inspection by determining the degree of reflection from the disbonded interface. For
the high power inspection of the kissing bonds, the detectability is determined by
comparison of the non-linearity of the disbonded interface to that of a perfectly bonded
interface. The detectability of the kissing bonds using all three techniques is then
compared.
CP657, Review of Quantitative Nondestructive Evaluation Vol. 22, ed. by D. O. Thompson and D. E. Chimenti
© 2003 American Institute of Physics 0-7354-0117-9/03/$20.00
1033
EXPERIMENTAL SET-UP
Longitudinal Wave and Shear Wave Inspection
The experimental set-up for both the longitudinal (Figure la) and shear (Figure
Ib) wave tests was based on the experimental technique used by Drinkwater et al [6] for
the study of aluminum-aluminum contact. Both experiments were conducted in pulseecho looking at the adhesive bondline from the disbonded side.
Longitudinal wave testing was conducted using a 10MHz wideband immersion
probe with a spherical focal length of 76.2mm in water. The longitudinal wave ultrasound
was coupled to the bond using a water bath with the standoff set to focus on the first
aluminum-adhesive interface which was where the disbond was created. Shear wave
testing was conducted using a 4MHz Electro-Magnetic Acoustic Transducer (EMAT).
The EMATs used produced normal incidence radially symmetric shear waves. The
adhesive bond samples were loaded in compression using a Zwick 1478 mechanical
testing machine. The time traces were converted to the frequency domain using a Fast
Fourier Transform (FFT) and then divided through by a reference trace. The reference
trace is obtained by capturing the reflection from an aluminum-air interface for which the
degree of reflection is known to be 100%. The Reflection Coefficient (RC) of the
interface is therefore determined by equation 1.
(1)
where £7exp and Uref are the amplitudes of the reflection from the kissing bond at a given
load and the reference traces respectively.
The same experimental technique used for the longitudinal wave testing was used
for the shear wave inspection. The amplitude of the shear waves produced was found to
be highly sensitive to the standoff of the EMAT from the surface of the aluminum disc.
Load
i l l
•
Aluminium
/ Cylinder
Larger diameter
upper aluminium
/ cylinder
Water-
Focused longitudinal
wave immersion
/
transducer
//
EMAT shear
wave transducer
(a)
FIGURE 1. Loading test experimental set-up. Longitudinal (a) and shear (b) wave loading apparatus
schematic diagrams.
1034
Load
Ball jointed
jointed loading
loading
Ball
platten
^^ platten
To Scope for
capture
Standard immersion
immersion
Standard
transducer ~~~~
transducer
Adhesive bond
bond under
under
Adhesive
inspection ~~~~~
inspection
High power
power
____— High
ultrasonic transducer
transducer
ultrasonic
_
Fixed lower
lower loading
loading
Fixed
platten \.
platten
From High
High
From
Power
Power amplifier
amplifier
FIGURE
FIGURE 2.
2. Schematic diagram of high power ultrasonic inspection
inspection setup.
setup.
The
The effect
effect of
of the signal amplitude variation was reduced by dividing each of the signals,
both
both reference
reference and
and for
for each
each load
load increment,
increment, by
by the
the signals
signals obtained
obtained from
from the
the second
second
reverberation
reverberation within
within the lower aluminum disc. The reflection
reflection coefficient
coefficient therefore
becomes,
becomes,
U exp 2 U expl
exp 1
RC =
=RC
U ref 2 U ref
•ef\1
(2)
(2)
where
where the
the numerical subscripts refer
refer to the first
first and second reverberations within the lower
aluminum disc.
High Power Ultrasonic Inspection
For the high power ultrasonic inspection, through transmission was used requiring
the
the addition
addition of an upper loading cylinder to accommodate a second transducer for
reception
reception of the through transmitted signal (Figure 2).
To obtain the large amplitudes of oscillation necessary to create non-linear
behavior at the adhesive-aluminum interface a narrow band ultrasonic transducer was
manufactured
manufactured using a 20mm diameter PZT8 focal bowl element with a spherical focal
length of 80mm in water. Input to the high power transducer was a 6 cycle gated burst at
RAM 10000 high power amplifier. The peak-to-peak amplitude
1.85MHz from
from a RITEC RAM10000
of the signal input to the transducer was 300V. The signal transmitted across the bond was
received by an off-the-shelf
off-the-shelf 5MHz wideband ultrasonic immersion transducer and fed
through an amplifier and captured by a digital oscilloscope. An FFT
FFT was then performed
on the signal and the non-linearity of the system calculated by
by taking the
the ratio of the
amplitudes of the first harmonic and fundamental frequencies ((A
A1l/AAQ0).).
EXPERIMENTAL RESULTS
Longitudinal Wave Inspection
Figure 3 shows loading curves for three dry contact kissing bond tests inspected
graph shows
shows the reflection coefficient plotted
using longitudinal wave ultrasound. The graph
against contact pressure for the three bonds. Also plotted is the reflection coefficient
loading curve for a perfectly bonded joint.
1035
0
20
40
60
80
Contact Pressure (MPa)
FIGURE 3. Longitudinal wave loading plots. Surface roughness given in Ra.
It can be seen from Figure 3 that, for all roughnesses of interface, as the contact
pressure at the interface is increased the reflection coefficient reduces. This reduction in
reflection coefficient is due to an increase in the physical contact area. The increase in the
actual contact area increases the interfacial stiffness and hence increases the percentage of
ultrasound transmitted. It can also be seen that for the smooth interface the reflection
coefficient tends towards that of the perfect contact bond at lower contact pressures than
for the two rougher interfaces.
Shear Wave Inspection
Figure 4 shows the multiple loading cycle plots for both a longitudinal and shear
wave test. Both samples were produced to the same nominal roughness and the two
loading curves calculated for an interrogating frequency of 5MHz. Although not the
center frequency of either the longitudinal or shear wave transducers, both have sufficient
bandwidth for the 5MHz point to lie within their -6dB points. It can be seen from Figure
5 that the essential form of the two loading curves is the same. It is however noticeable
that because of the much lower signal amplitudes produced by the EMAT shear wave
transducers, the loading curve shows a much "noisier" response than the longitudinal wave
immersion transducer.
Figure 5 shows the plot of contact pressure against Reflection Coefficient Ratio
(RCRatio) for three surface roughnesses of sample. The RCRatio is defined as the ratio
of the shear wave reflection coefficient to the longitudinal wave reflection coefficient.
RCRatio = RC Long
(3)
It can be seen from Figure 5 that the RCRatio at low contact pressures is very similar.
As the contact pressure is increased, the RCRatios for the separate interfaces become
different. The RCRatio of the smoother interfaces tends to be larger for a given contact
pressure indicating a larger degree of shear wave reflection for a given longitudinal
reflection. This suggests that dry contact kissing bonds with smoother surfaces tends
towards the case of a slip bond in which the shear interfacial stiffness is much reduced
1036
20
40
60
80
100
Contact Pressure (MPa)
FIGURE 4. Comparison of shear and longitudinal loading tests.
0.4pm
£ 0.9
6.2pm
U
.2 0.8
Contact Pressure (MPa)
FIGURE 5. RCRatio plotted against contact pressure for 3 roughnesses. Surface roughness given in Ra.
due to the ability of the surfaces to slip across one another. In contrast the larger
asperities of the rougher surfaces will tend to give a higher shear stiffness due to
interlocking of the surface asperities.
High Power Ultrasonic Inspection
An example of the high power ultrasonic response of a dry contact kissing bond
is shown in Figure 6. The plot shows the non-linear ratio plotted against contact
pressure for a two cycle loading test. It can be seen from the plot that the non-linear
ratio of the system is at a maximum at very low loads. This can be attributed to
clapping behavior of the interface. At low loads the interface will have large areas
which are only just or very nearly in contact. When the large amplitude of oscillation is
incident on these contact areas, their localized stress-strain curve may look something
akin to that in Figure 7. The large difference between the tensile and compressive
stiffnesses results in a highly non-linear system creating large amplitude harmonics in
the transmitted signal. As the contact pressure is increased, the percentage of contact
area that is only lightly loaded will decrease and hence the non-linearity of the interface
will also decrease.
1037
6
8
Pressure (MPa)
10
FIGURE 6. High power ultrasonic response from dry contact kissing bond.
Compressive
load (arb.)
FIGURE 7. Non-linear stress-strain response of clapping contact.
Comparison of Shear, Longitudinal and High Power Inspection Techniques
Figure 8 shows the sensitivity of each of the three techniques to the presence of
three roughnesses of interface. The sensitivity is determined by the percentage of the
perfect contact value for each technique. In the case of longitudinal and shear wave
inspection, this is determined by percentage change in the reflection coefficient from the
disbonded interface relative to the reflection coefficient for the perfectly bonded interface.
In the case of the high power inspection, this has been measured by the percentage change
in the non-linear ratio of the disbonded interface relative to the non-linearity of the entire
inspection system when inspecting a perfectly bonded joint.
It can be seen from Figure 8a and Figure 8b that the sensitivity of the high power
technique drops off very quickly with load. Once a small degree of pressure is applied to
the system the contact non-linearity reduces. Because the non-linearity of the system is
also influenced by any non-linearity in the equipment and measuring technique, there may
still be some contact non-linearity at higher pressures but it may be overwhelmed by the
inherent system non-linearity.
Although the longitudinal wave sensitivity to this type of kissing bond is less at
low pressures than for the high power inspection technique, it can also be seen that the
sensitivity is spread over a much larger range of pressures. This suggests that longitudinal
wave inspection is more sensitive to dry contact kissing bonds at high pressures.
1038
I HLongitudi
Longitudinal
Shear
High Power
Percentage of perfect
contact value
250
200
150
100
50
0
0.4]um
0.4µm
0.9]um
0.9µm
2.0]um
2.0µm
FIGURE 8a.
8a. Sensitivity
Sensitivity at
at 0.04MPa
0.04MPa contact
contact pressure.
pressure.
FIGURE
.ongitudinal
Longitudinal
0 Shear
Shear
EHighPow
High Power
_i
Percentage of perfect
contact value
250
200
150
100
50
0
0.4]um
0.4µm
0.9]um
0.9µm
2.0]um
2.0µm
FIGURE 8b.
8b. Sensitivity
Sensitivity at
at 10MPa
lOMPa contact
contact pressure.
pressure.
FIGURE
| H Longitudinal
Longitudinal
B Shear
Shear
B High Power |
High Power
Percentage of perfect
contact value
250T
250
200
150
100
50
0
0.4]um
0.9]um
0.4µm
0.9µm
2.0]um
2.0µm
FIGURE 8c.
8c. Sensitivity
Sensitivity at
at 50MPa
50MPa contact
contact pressure.
pressure.
FIGURE
CONCLUSIONS
CONCLUSIONS
Longitudinal wave,
wave, shear
shear wave
wave and
and high
high power
power ultrasonic
ultrasonic inspection
inspection has
has been
been
Longitudinal
carried out
out on
on dry
dry contact
contact kissing
kissing bonds.
bonds. Comparison
Comparison was
was made
dry contact
carried
made between
between dry
contact
kissing bonds
bonds with
with different
different interface
interface roughnesses
roughnesses using
using all
all three
three techniques.
techniques. ItIt was
was found
found
kissing
that
at
very
low
contact
pressures
the
greatest
sensitivity
is
gained
by
use
of
the
high
that at very low contact pressures the greatest sensitivity is gained by use of the high
power
inspection
technique.
As
the
contact
pressure
increases,
the
sensitivity
of
this
power inspection technique. As the contact pressure increases, the sensitivity of this
technique
reduces
rapidly
making
it
less
effective
at
higher
contact
pressures.
For
higher
technique reduces rapidly making it less effective at higher contact pressures. For higher
contact pressures,
pressures, conventional
conventional longitudinal
longitudinal wave
wave inspection
inspection offers
offers the
the greatest
greatest contrast
contrast
contact
between
kissing
bonds
of
different
surface
roughness
and
contact
pressures,
thereby
between kissing bonds of different surface roughness and contact pressures, thereby
showing
the
greater
potential
for
their
detection.
showing the greater potential for their detection.
1039
Although each single technique shows promise for detection in certain
circumstances, a combination of two or more ultrasonic techniques could potentially
provide enhanced information about the quality of an adhesive bond.
REFERENCES
1.
Rose, J. H., "Ultrasonic Reflectivity of Diffusion Bonds," in Review of Progress in
QNDE, edited by D. O. Thompson and D. E. Chimenti, Plenum N.Y., (1989), Vol. 8,
pp. 1925-1931.
2.
Nagy, P. B., Journal of Adhesion Science and Technology, 5, pp. 619-630 (1991).
3. Jiao, D. and Rose, J. L., Journal of Adhesion Science and Technology, 5, pp. 631-646
(1991).
4.
Adams, R. D. and Drinkwater, B. W., NDT&E International, 30, pp. 93-98 (1997).
5.
Arakawa, T., Materials Evaluation, 41, pp. 714-719 (1982).
6. Drinkwater, B. W., Dwyer-Joyce, R. S. and Cawley, P., Proc. Royal Soc. London,
452(1996).
7. Wooldridge, A. B., CEGB Research Division, Report Number NW/SSD/RR/42/79,
(1979)
8.
assbender, S. U., Kroening, M. and Arnold, W., Materials Science Forum, 210-213,
pp. 783-790(1996).
9. iu, G., Qu, J., Jacobs, L. J. and Li, J., "Characterizing the Curing of Adhesive Joints
by a Nonlinear Ultrasonic Technique," in Review of Progress in QNDE, edited by D.
O. Thompson and D. E. Chimenti, Plenum N.Y., (1999), Vol. 18, pp. 2191-2199.
10. Rothenfusser, M., Mayr, M. and Baumann, J., Ultrasonics, 38, pp. 322-326 (2000).
1040