LASER GENERATION OF FOCUSED ULTRASONIC WAVE
Kyungyoung Jhang1, Hongjoon Kim2, Hyunmook Kim2, and Yob Ha2
School of Mechanical Engineering, Hanyang University, Hangdang-dong 17,
Seongdong-gu, Seoul 133-791, Korea
Department of Precision Mechanical Engineering, Graduated School of Hanyang
University, Hangdang-dong 17, Seongdong-gu, Seoul 133-791, Korea
ABSTRACT. An arc-shaped line array slit is used for the laser generation of focused laser-ultrasonic
wave. The spatially expanded Nd: YAG pulse laser is illuminated through the arc-shaped line array slit
on the surface of a sample to generate the ultrasonic wave of the same pattern as the slit. Then the
generated ultrasonic wave is focused at the focal point of which distance from the slit position is
dependent on the curvature of slit arc. The relationship between the characteristics of the generated
ultrasonic wave including the focusing performance and several design parameters such like as slit
width and slit interval are investigated. By using the focused wave we can upgrade the inspection
ability for the small size defect with the improvement of spatial resolution. Also this method can be
combined with the scanning mechanism to get image just like we can get by the scanning acoustic
microscope(SAM)
INTRODUCTION
Lamb waves are guided elastic waves propagating in plate structures parallel to the
boundary surfaces[l]. The propagation of these wave modes allows for the inspection of
plates and strip materials along the length of the sample thus providing for the possibility
of a considerable saving in time over a C-scan of the sample[2]. Recent years have seen
considerable application of Lamb wave testing applied, for example, to defect detection in
metal[3], composite plates[4,5], sheet thickness measurement[6], elastic constant
evaluation[7] and bonded metal joint inspection[8]. The usual method of launching Lamb
waves is to utilize a compression transducer coupled to sample plate with a matching
wedge such that effective spatial matching into the desired Lamb wave mode is
obtained[l]. An additional technique is to employ laser generation[9]. Attractive features
of a laser-ultrasound source include its non-contact nature, reproducibility, broad
bandwidth and capability to launch a variety of acoustic modes. Preliminary work on laseracoustic sources was conducted by White[10] using a pulsed ruby laser. But, in these
methods, laser generated Lamb waves had wide bandwidth frequency, low directivity, and
low sensitivity. Several techniques have been investigated to increase the amplitude and
narrow the bandwidth of laser generated signal. One approach is temporal modulation of
the laser source[ll]. Another method is controlling the shape and phase of the laser source
to generate Lamb waves with narrow band and high directivity. Line array slit was used to
generate narrow-band, high-directivity wave[12]. In his paper, arc-shaped line array slit
was used to generate focused Lamb wave. This technique can upgrade the inspection
ability of conventional line array slit for the small size defect with the improvement of
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/S20.00
305
specimen(aluminum,
2mm)
specimen(aluminum, 2mm)
760
760
Nd-YAG LASER
Slit
Beam
expander
Broadband
pinducer
Osilloscope
Osilloscope
Computer
Computer
FIGURE
FIGURE1.1.Experimental
Experimental setup.
setup.
spatialresolution
resolution and
and be
be combined
combined with the scanning mechanism to
spatial
to get
get image
image just
just like
like we
we
canget
getby
byscanning
scanning acoustic
acoustic microscope(SAM).
microscope(SAM).
can
EXPERIMENTAL ARRANGEMENT
ARRANGEMENT
EXPERIMENTAL
FIGURE 11shows
shows the
the experimental
experimental arrangement for the
FIGURE
the laser
laser generation
generation of
of focused
focused
Lambwaves
waves in
inthe
the sample
sample plate.
plate. The
The source
source laser was a Q-switched Nd:YAG
Lamb
Nd:YAG laser
laser system.
system.
The diameter
diameter of
of the
the laser
laser beam
beam was extended to about 20mm
The
20mm by
by aa beam
beam expander.
expander.
Extended laser
laser beam
beam transmitted
transmitted through an arc-shaped line array
Extended
array slit
slit and
and generated
generated
ultrasonic source
source as
as same
same shape
shape of
of the slit
slit on the sample plate. The
ultrasonic
The sample
sample was
was an
an
aluminum plate
plate with
with 2mm
2mm thickness.
thickness. The
The ultrasonic
ultrasonic waves
aluminum
waves generated
generated by
by the
the laser
laser source
source
was detected
detected by
by aa pinducer,
pinducer, and
and the
the wave
wave signal
signal was
was
was monitored
monitored by
by oscilloscope.
oscilloscope. AA
standard GPIB
GPIB data
data interface
interface allowed
allowed signals
signals to
standard
to be
be transferred
transferred from
from the
the oscilloscope
oscilloscope toto aa
personalcomputer
computer for
for subsequent
subsequent data
data processing
personal
processing and
and storage.
storage.
DESIGNOF
OFARC-SHAPED
ARC-SHAPED SLITS
SLITS
DESIGN
Three parameters
parameters of
of arc-shaped
arc-shaped slits
slits were
were determined.
determined. The
Three
The prarmeters
prarmeters are
are slit
slit space,
space,
slit
number,
and
focusing
length
of
arc.
In
first
step,
we
determined
slit
space.
In
this
slit number, and focusing length of arc. In first step, we determined slit space. In this step,
step,
taking the
the sensitivity
sensitivity of
of detector
detector into
into consideration,
consideration, we
taking
we determined
determined the
the frequency
frequency of
of
ultrasonic wave
wave generated
generated by
by slit.
slit. From
From the
the frequency,
frequency, we
ultrasonic
we determined
determined the
the wave
wave length
length
with dispersion
dispersion curve
curve (FIGURE
(FIGURE 2).
2). Generally,
Generally, AO
with
A0 mode
mode is
is generated
generated dominantly
dominantly so
so that
that
we
have
tried
to
generate
AO
mode,
and
its
wavelength
was
determined
from
we have tried to generate A0 mode, and its wavelength was determined from the
the
relationshipbetween
between the
the frequency
frequency and
and wavelength
relationship
wavelength which
which is
is given
givenby
by
V = (λ / d ) × ( f × d )
306
(1)
(1)
10
A4
9
S2
S4
P has e veloc ity (k m /s )
8
«
to
£
S1
7
A3
A2
S3
6
S0
5
A1
44
33
22
A0
11
00
0
22
44
6
8
6
8
FFre
r eqquueennccyy --TThhi icckknneessss pprroodduucctt ((MMHHzz**mmm
m ))
10
FIGURE 2. Lamb wave phase velocities
velocities in
in an
an aluminum
aluminumplate.
plate.
Where,
Where, V
V is
is phase
phase velocity
velocity of
of the
the wave,
wave, λA is
is wavelength,
wavelength, dd isis thickness
thickness of
of the
the plate,
plate, ff isis
frequency
frequency of
of the
the wave.
wave. As
As frequency
frequency is
is selected
selected and
and the
the thickness
thickness of
of the
the plate
plate isis
determined when we select the plate, from the dispersion curve
curve at FIGURE
FIGURE 1,
1, we
we can
can find
find
the
thickness of
the plate
plate is
the velocity
velocity of
of the
the wave.
wave. In
In here,
here, the
the thickness
of the
is 2mm,
2mm, frequency
frequency is
is 1.75MHz.
1.75MHz.
At
At the
the dispersion
dispersion curve,
curve, phase
phase velocity
velocity of
of the
the A0
AOmode
mode Lamb
Lamb wave
wave is
is about
about 2.8km/s.
2.8km/s. By
By
substitute
substitute velocity,
velocity, and
and frequency
frequency in
in equation
equation (1),
(1), the
the wavelength
wavelength was
was determined
determined as
as
1.5mm
1.5mm and
and this
this become
become space
space between
between slits.
slits.
In
In the
the next
next step,
step, number
number of
of slits
slits was
was determined.
determined. Many
Many slits
slits make
make frequencyfrequencybandwidth
bandwidth narrow,
narrow, but
but laser
laser spot
spot diameter
diameter was
was limited
limited as
as 10~20mm.
10~20mm. So,
So, too
too many
many slits
slits
were
were not
not needed.
needed. Generally,
Generally, the
the number
number of
of slits
slits N
N is
is
N = Ds / λ + 2 ,
(2)
(2)
where,
where, D
Dss is
is laser
laser spot
spot size
size on
on the
the slit
slit array.
array, because
because the
the wavelength
wavelength was
was determined
determined in
in the
the
former
former step,
step, we
we can
can determine
determine the
the number
number of
of slits
slits by
by the
the equation
equation (2).
(2).
Lastly,
Lastly, The
The focal
focal length
length of
of the
the slits
slits was
was determined
determined from
from radius
radius of
of arc.
arc.
LAMB
LAMB WAVE
WAVE SIGNAL
SIGNAL GENERATED
GENERATED BY
BY ARC-SHAPED
ARC-SHAPEDSLIT
SLIT
Typical
Typical signals
signals recorded
recorded at
at the
the focusing
focusing point
point are
are shown
shown in
in FIGURE
FIGURE 33 with
with STFT
STFT
contour
contour plot.
plot. Peak
Peak frequency
frequency of
of signal
signal generated
generated by
by slit
slit space
space 1.5mm
1.5mm was
was about
about 1.75MHz,
1.75MHz,
and
and itit was
was about
about 600kHz
600kHz when
when the
the slit
slit space
space was
was 4mm.
4mm. ItIt shows
shows that
that frequency
frequency of
of the
the
Lamb
Lamb wave
wave generated
generated by
by the
the arc-shaped
arc-shaped slit
slit array
array was
was determined
determined by
by slit
slit space
space as
as we
we
designed.
designed. From
From the
the contour
contour plot
plot of
of STFT,
STFT, we
we can
can find
find that
that A0
AOmode
mode Lamb
Lamb waves
waves were
were
dominantly
dominantly generated.
generated. More
More intensive
intensive ultrasound
ultrasound was
was generated
generated from
from the
the slit
slit array
array which
which
have
than 1.5mm
have 4mm
4mm slit-space
slit-space than
1.5mm slit-space.
slit-space.
To
To verify
verify that
that the
the arc-shaped
arc-shaped line
line array
array slits
slits were
were generating
generating focused
focused wave
wave in
in aa
satisfactory
satisfactory manner,
manner, experiments
experiments were
were performed
performed on
on an
an aluminum
aluminum test
test plates
plates of
of 2mm
2mm
thickness.
thickness. A
A 1.75MHz
1.75MHz pinducer
pinducer was
was moved
moved from
from point
point to
to point
point as
as seen
seen in
in FIGURE
FIGURE 44 to
to
detect
detect the
the Lamb
Lamb wave
wave signals
signals and
and peak
peak to
to peak
peak value
value of
of the
the signal
signal was
was measured
measured at
at each
each
points.
points. Simulation
Simulation using
using simple
simple scalar
scalar diffraction
diffraction theory
theory was
was performed
performed to
to compare
compare with
with
this
this experiment.
experiment. FIGURE
FIGURE 55 shows
shows the
the result.
result. The
The experimental
experimental result
result and
and simultation
simulation
307
Timef micro-seel
2
Frequency-Thickness Product (MHz*mm)
3
4
5
6
7
8
Frequency-Thickness Product (MHz*mm)
Slit
space 1.5mm
Slit
Slit space
space 1.5mm
1.5mm
Slit
space
1.5mm
Slit space4mm
4mm
Slit
Slit
space
Slitspace
space4mm
4mm
FIGURE
3.3.Lamb
wave signal
and its
plot of
of Short
Short Time
TimeFourier
Fourier transform.The
The focusinglength
length of
FIGURE
FIGURE 3.
3. Lamb
Lamb wave
wave signal
transform.
focusing
ofof
FIGURE
Lamb
wave
signal and its contour
contour plot
plot of
of Short
Short Time
Time Fourier
Fouriertransform.
transform. The
Thefocusing
focusing length
lengthof
the
arc
was
35mm,
the
thickness
of
aluminum
sample
plate was
was2mm.
2mm.
the
arc
was
35mm,
the
thickness
sample
plate
thearc
arcwas
was35mm,
35mm,the
thethickness
thicknessof
ofaluminum
aluminumsample
sampleplate
platewas
was2mm.
2mm.
the
5mm
5mm
5mm
5mm
„«*
<>
••j >
o
<>
o
(>
<)
(>
*• \ < >
- © center
© back
©
Front
Front
Front
Front center
center back
back
FIGURE
4.
Lamb
wave
detection
points
generating
focused
wave.
FIGURE
to
verify
generating
focused
wave.The
Thepinducer
pinducerlocated
locatedon
onthese
these
FIGURE4.4.
4.Lamb
Lambwave
wavedetection
detection points
points to
to verify
focused
wave.
The
pinducer
located
on
these
FIGURE
Lamb
wave
detection
points
to
verify
generating
focused
wave.
The
pinducer
located
on
these
points
and
amplitudes
of
the
Lamb
wave
signals
were
measured.
The
spaces
between
the
points
were
5mm.
points
and
amplitudes
of
the
Lamb
wave
signals
were
measured.
The
spaces
between
the
points
were
5mm.
points
and
amplitudes
of
the
Lamb
wave
signals
measured.
The
spaces
between
the
points
were
5mm.
points and amplitudes of the Lamb wave signals were measured. The spaces between the points were 5mm.
999
0.05
0.05
0.05 1
U.U&
0. 045
888
0.
0.045
045
7
0.045iB<-x
0.04x^^^v \
77
0.04
0.04
U
M4
"<v
6
0. 035
0.
035
66
0.
035
^,ua&^
0.03
55
t-—
-2S
0.03
0.03
5
0.
025
*X.Jk %
4
0.
025
0.
025
434
Z^* U.Uk!t»
0.02
Bv '
0.02
0.02
————
Jl * ," 0. 015
U.U2
323
T*~>x!^.^
^^
0.
015
0.
015
g;*
U.U1&
0.01
js^>^
^
*->NX
212
0.01
1
0.01
U.U1
0.
005
r ------*
101
0.
005
i
0.
005
U.UU5
00
1
0 0 -10 -8 -6 -4 -2 0 2 4 6 8 10
u0 0
-10
-8
-6
-4
-2
0
2
4
6
8
10
10
5
5
10
-10
--55
00
55
1010
-10 -8- 8-6- 6-4 - -24 -0 2 20 24 4 66 8 10
10 -- 10
10
1 0
- 5
0
5
1
Simulation
Experiment
Simulation
Experiment
Simulation
Experiment
Simulation
Experiment
FIGURE
intensity
lines.
( dashed :: front,
solid :: center,
FIGURE5.5.
5. Wave
Waveintensity
intensity distribution
distribution measured
measured at
at 55
5 points
points along
along three
three
FIGURE
three lines.
lines.(((dashed
dashed ::front,
front,solid
solid ::center,
center,
FIGURE
5.Wave
Wave
distribution measured
measured at
at 5 points
points along
along three
lines.
dashed
front,
solid
center,
dashdot
:: back
)) intensity distribution
dashdot
back
dashdot : back )
y/^U^)3
dashdot: back)
308
1
2
3
1
4
Time(|j,sec)
2
3
Time (jisec)
..J............J...........J............J............J...
i.............i.............i..............
i..
|....«
1
1
Freq. (kHz)
Freq. (kHz)
Single slit(Slit width 2mm)
Slit array(Slit space 4mm, 5 slits)
FIGURE 6. Variation of frequency bandwidth as slit numbers.
result are similar in crossing points between the intensity distribution lines and side lobes.
It shows that focused waves could be generated by the arc-shaped line array slits effectivly.
FIGURE 6 illustrates Lamb wave signal which was generated by arc-shaped single
slit to compare with waves generated by arc-shaped slit array. The wave generated by slit
array have relatively narrow bandwidth and low frequency area was eliminated which
represented in wave signals generated by single slit.
CONCLUSION
Arc-shaped slits were designed for generating the focused ultrasound. Intensity of
ultrasound increases around the designed focusing point and it shows that the focused
Lamb wave can be generated by arc-shaped array slits. It is expected that the focused
Lamb wave generated by arc-shaped slit will be more effective to detect small size defect
than conventional linear line array slit.
ACKNOWLEDGEMENTS
This work was supported by Domestic Co-work Program of the New Technology
Development and Research Center of Laser Application (RRC).
REFERENCES
1. Vikotorov, LA. "Rayleigh and Lamb waves physical theory and applications"(Plenum
Press, New York, 1967), Chapter 2, pp. 67-121
2. Karautkramer, J. and Karautkramer, H., "Ultrasonic testing of materials"(Springer
Verlag, 1977,2nd edn), Chapter 22, pp.399-426
3. Alleyne, D.N. and Cawley, P., "The interaction of lamb waves with defects", IEEE
Trans. Ultrason. Ferroelectr. Freq. Control., 1992, 39, (3), pp. 381-397
4. Guo, N. and Cawley., P., "Lamb wave reflection for the quick non-destructive
evaluation of large composite laminates", Mater. Eval. 1994, 52, (3), pp. 404-411
5. Pierce, S.G Staszevski, W. J. Gachagan, A. James, I.R. Philp, W.R.Worden, K.
Culshaw, B. Mcnab, A. Tomlinson, G.R. and Hayard, G., "Ultrasonic condition
309
monitoring of composite structures using a low profile acoustic source and an
embedded optical fibre sensor", Proc. SPIE Int. Opt. Eng., 1997, 3041, pp. 437-448
6. Dewhurst, RJ. Noui,I. and Shan, Q., "Polymer film thickness measurement using
laser-ultrasound techniques", Rev. Sci. Instrum., 1990,61,(6), pp. 1736-1742
7. Rogers, W.P., "Elastic property measurement using Rayleigh Lamb waves", Res.
Nondestru. Eval., 1995, 6, pp. 185-208
8. Bork, U. and Challis, R.E., "Nondestructive evaluation 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, (1), pp. 72-84
9. Scruby, C.B. and Drain, L. E., "Laser ultrasonics: technique and applications" (Adam
Hilger, 1990) Chapter 5, pp. 223-324
10. Hite, R. M., "Generation of transient elastic waves by transient surface heating", J.
Appl. Phys., 1963, 34, (12), pp.3559-3567
11. Pierce, R. Ume, C. and Jarzynski, J., "Temporal Generation of a laser source for the
generation of ultrasonic waves", Ultrasonics, 1995, 33, (2) pp. 133-137
12. Cerniglia, D. Jhang, K. Y. and Djordjevic, B. B." Non-Contact Ultrasonic testing of
Aircraft Lap Joint", Proceedings of 15th World Conference on Non-Destructive
Testing, edited by G. Nardoni et al., Italian Society for NDE, 2000, 161st in CD title
310