OPTICAL ABSORPTION OF EPOXY RESIN AND ITS ROLE IN
THE LASER ULTRASONIC GENERATION MECHANISM IN
COMPOSITE MATERIALS
T. Stratoudaki, C. Edwards, S. Dixon and S. B. Palmer
Department of Physics, University of Warwick, Coventry, CV4 7AL, UK
ABSTRACT. Epoxy resins are used in various applications and are essential to the fabrication
of carbon fibre reinforced composite materials (CFRCs). This paper investigates laser generated
ultrasound in epoxy resins using three different lasers, a TEA CO2, a Nd:YAG and a XeCl
excimer. In these partially transparent materials the ultrasonic generation mechanism is directly
related to the optical absorption depth which can therefore be measured directly from the
ultrasonic waveforms using for example a Michelson interferometer as detector. The present
work aims firstly to relate the observed amplitude of the longitudinal wave to the optical
absorption depth of the epoxy and secondly to evaluate the role of the epoxy resin to the
generation of the ultrasound in CFRCs. For the latter, comparative results of generation
efficiency between the three wavelengths are presented and an attempt is made to understand the
way that the resin matrix influences the generation mechanism of ultrasound in composite
materials.
INTRODUCTION
Composite materials are being increasingly used in the aircraft industry, bringing
Laser Based Ultrasound (LBU) [1] to the focus of attention, as it is able to test nondestructively Carbon Fibre Reinforced Composites (CFRCs). It offers important
advantages being non-contact (and practically remote), couplant free and able to deal with
complex surfaces.
In LBU, the ultrasound is generated using a high power pulsed laser (usually a TEA
CO2) and is often detected by a Fabry-Perot interferometer [2, 3]. According to the incident
laser energy used for the ultrasonic generation, two regimes are distinguished [4],
In the low power thermoelastic regime, there is no damage to the material. The
laser beam is focused on to the surface of the sample, causing it to expand rapidly, in times
that are comparable to the rise time of the laser pulse [5]. In metals, the incident laser
radiation is absorbed in the thin electromagnetic skin depth, which is of the order of 5lOnm. Thermal diffusion in the material extends the ultrasonic source to a total of ~ljiim
[6]. In the case of non-metallic materials (such as CFRCs), the absorption of the laser
radiation is determined by the absorption coefficient of the material and it takes place
within the optical absorption depth [4, 7]. The latter is much greater and is a function of the
optical properties of the material itself and the laser wavelength. As a consequence, a
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
965
bigger volume of the material is affected, the temperature rise is less and the amplitude of
the longitudinal wave is greater [8]. This condition is referred to as "a buried thermoelastic
source". Since CFRCs are made of carbon fibers and epoxy resin, the optical absorption
depth of both these materials influences the ultrasonic generation mechanism. Thus,
different sorts of buried sources are produced on the same material due to the different
laser wavelengths used, as will be shown in this study.
In the ablation regime (higher energy densities), the ultrasonic generation
mechanism differs [9], there is plasma formation and the sample is damaged. The
ultrasound is due to the recoil force exerted by the ejected matter and the pressure from the
expanding plasma [10]. As a result, stronger longitudinal amplitudes are produced
compared to the thermoelastic regime.
For the present study, three different types of pulsed lasers were used. A mid-IR
TEA CO2 (10.6|im), a near-IRNd:YAG (1064nm) and a UV XeCl excimer (308nm).
EXPERIMENTAL SETUP AND PREPARATION OF EPOXY RESIN SAMPLES
The experimental setup (described in [11]) consists mainly of a laser generator and
a laser detector. For the generation three lasers were used: a) a TEA COa (Coherent, Hull,
LaserbrandlSO), b) a Q-switched fundamental Nd:YAG (Spectron Laser Systems) and c) a
XeCl excimer (Lamda Physik). The generation beam spot size was kept between 0.020.03cm2, at all times. Filters and variable beamsplitters were used to control the energy.
The ultrasound was detected using a modified Michelson Interferometer [13] to record the
absolute epicentral displacements on the opposite side of the sample. In order for the
Michelson to work the samples had to be polished at the back side. All results were taken
using a single laser shot on a fresh spot. After the irradiation, the samples were inspected
under an optical microscope.
The CFRC sample (provided by Rolls Royce) had a thin superficial layer of resin
with a mean thickness of ~12^m. For all the experiments presented in this paper the same
unidirectional sample was used (orientation of the fibers: 0°/0°, average thickness ~8mm).
Pure epoxy resin samples were prepared using resins commonly used for the
manufacture of composites (supplied by Vantico). The first was a cold-curing epoxy resin
and the second a warm-curing epoxy resin. They were used to make two samples which
were slightly yellow (the cold curing sample was slightly darker than the warm). The
samples were cured according to manufacturer's instructions. Information about the
samples can be found in table 1.
EXPERIMENTAL RESULTS
A good quality beam is needed for accurate determination of the damage threshold
[12]. The Nd:YAG operated in TEMoo mode, its pulse duration at FWHM was 10ns and its
spot size was ~0.02cm2. The FWHM of the TEA CO2 was 50ns and the spot size was
~0.03cm2. The XeCl laser had a "top hat" beam profile according to the manufacturer, its
FWHM was 40ns and its spot size was ~0.02cm2.
TABLE Liable of samples used.
Name
Epoxy Resin
Hardener
Thickness
Comments
Cold Curing Epoxy
Araldite LY 5052
Aradur 5052
16.2mm
Slightly yellow
Warm Curing epoxy
Araldite LY 3505
XB 3403 (clear)
9.8mm
Slightly yellow
966
Nd;YAG (1064nm)
Epicentral waveforms recorded above and below the damage threshold are shown
in figure 1. Figure 2(a) shows the variation in the amplitude of the longitudinal
displacement with respect to the energy. To identify the damage threshold for the IR lasers
(TEA CO2 and Nd:YAG), two parameters were used: a) during the experiment, the
appearance of plasma ("blue flash") and b) after the experiment, microscopic observation
of the first exposed fibers and comparison with images taken before laser irradiation.
TEA COi^ a0.6um) and XeCl (308nm)
Waveforms recorded with these lasers on the CFRC sample above and below the
damage threshold have been presented in [11]. The damage threshold for the TEA CCh has
been identified and shown in figure 2(b) and the ablation threshold for the XeCl is shown in
figure 2(c). The term "damage threshold" was not used in this case because there was no
exposure of fibers up to the laser energies used. The criteria for identifying the ablation
threshold in the case of the UV laser were: a) during the experiment the appearance of a
"blue flash" denoting plasma formation b) the change of gradient in figure 2(c) and c) the
change of the shapes of the waveforms recorded (as can be seen in [11]). There was no sign
of yellow discoloration of the irradiated spots (as was the case with the IR lasers) even at
the highest energies used and the difference was hardly noticeable under the microscope.
Pure Epoxv Samples
CFRCs are made of carbon fibers embedded in an epoxy resin matrix. The laser
generation mechanism in non-metals relies on the optical absorption depth of the material
irradiated, which depends on the laser wavelength used. In order to determine the role of
the epoxy resin in the generation mechanism, a series of experiments were carried out on
pure epoxy resin samples.
Waveforms recorded with the TEA COa and the XeCl can be seen in figure 3. There
are no waveforms recorded with the Nd:YAG because the epoxy is practically transparent
at 1064nm (the optical absorption depth was measured by means of a joulemeter on the
cold epoxy resin and was found to be approximately 4mm).
From figure 3 it can be seen that the shape of the waveforms produced by the two
lasers is similar, especially in the case of the cold curing epoxy.
No Damage
a-l.lmJ
b-1.7mJ
c-2.6mJ
0
5
10
Time (ys)
FIGURE 1. Nd:YAGi epicentral waveforms. L = longitudinal, S = shear, 3L= longitudinal echo (Waveform
(d) has a time offset of +0.1 us to facilitate the reader).
967
12
302520-
50-
•
3.0-
•
2.5-
Damage
threshold
i2-0:
•
:
0
IP'
"
<0.5-
J..
10
•
10
* 8
Damage
threshold
4
2-
0.015
20
25
0
30
Ablation
6- threshold
I.--'
4
6
8
Energy (mJ)
Energy (mJ)
Energy (ml)
10
(b)
(a)
(C)
FIGURE 2. L-amplitude vs. mean laser energy (a) Nd: YAG: The damage threshold occurs at (6.2mJ, 1.6nm),
(b) TEA CO2: The damage threshold occurs at (37.8mJ, 1.4nm), (c) XeCl: The ablation threshold occurs at
(3.28mJ50.17nm).
10
a-17.3mJ
b-38.8mJ
c - 54.4mJ
a - 12.4mJ
b-31.1mJ
c - 66.9mJ
10
Time (urn)
(3b)
e-3.4mJ
f - 5.4mJ
g - 7.7mJ
E
<
10
15
20
Time (us)
(3c)
FIGURE 3.. L = longitudinal, 3L= longitudinal echo (3a) and (3b) TEA CO2 laser on cold curing and warm
curing epoxy resin respectively, (3c) and (3d) XeCl excimer on cold and warm curing epoxy resin respectively.
CALCULATION OF OPTICAL ABSORPTION DEPTH FROM EPICENTRAL
WAVEFORMS
The ultrasonic generation mechanism in non-metals is so closely dependent on the
optical absorption depth that the latter can actually be calculated from the features of the
longitudinal wave of the recorded waveforms. The absorbed laser radiation in the material
can be deduced from Beer's Law. According to this law the extinction of the radiation in
the material is exponential and the absorbed energy is the difference between light entering
and exiting a layer, hence the negative differential is needed.
(1)
where Iabs is the intensity of the absorbed radiation, I is the incident radiation, R is the
reflectance, a is the absorption coefficient and z is the direction normal to the surface. The
968
temperature distribution (T) in the affected volume decays exponentially with depth in the
material also:
(2)
pC
where p is the density and C is the specific heat capacity.
In the thermoelastic regime the temperature rise is associated with the thermal
expansion of the material which generates elastic waves that propagate into its bulk. In this
way the rise time of the longitudinal pulse reflects the exponential decay with depth of the
temperature distribution and consequently of the optical absorption [14].
In order to calculate the optical absorption depth (8) from the recorded waveforms,
we need to know the longitudinal velocity (DI) and the rise time (tr) of the longitudinal
pulse. The former is calculated by measuring the time difference between the first
longitudinal pulse and its echo:
2d
,_
u,=-——
(3)
l
l-ll
where d is the sample thickness and tL-3L is the time required for the longitudinal wave to
travel a length equal to twice the thickness of the sample (figure 4a). For all the epoxy
samples the longitudinal velocity was found approximately 2600m/s.
The rise time of the longitudinal pulse is measured directly from the recorded
waveforms as the time interval between the maximum of the longitudinal amplitude and its
1/e value (figure 4b). The optical absorption depth is calculated as:
S = vr*r
(4)
Table 2 summarizes the results from the recorded waveforms for the epoxy samples.
TABLE 2. Measured rise time of L-pulse and calculated optical absorption depth for the two epoxy samples.
L - Pulse Rise Time (tr) in ns
Name
5inum
TEACO2
XeCl excimer
TEAC02
XeCl excimer
Cold Curing Epoxy
100
81
262
209
Warm Curing Epoxy
60
100
151
261
0
5
10
15
20
Time (urn)
3.0
3.5
4.0
4.5
5.0
Time (urn)
(a)
(b)
FIGURE 4. Calculation of 8 in non-metals. Warm curing epoxy resin waveform using the XeCl excimer
(energy=6.9mJ) (a) L=first longitudinal pulse, 3L=longitudinal echo, (b) longitudinal rise time measurement.
969
HH
HH
ID-
1*1
KH
S'
KH
HH
W
W
W
W
6
°CFRC
f ;
• Epoxy
i. 4KH
2ij-j_ii_i
<***»" *
H
0
20
10
0-
W
30
40
50
||g)t«^l*l
*** (** ***"^
0 1
60
Mty
2
3
4
Energy (mJ)
5
6
7
8
9
1 0
Energy (mJ)
(a)
(b)
FIGURE 5. (a) TEA CO2: results taken on the cold curing epoxy resin and on a CFRC, (b) XeCl: results taken
on the cold curing epoxy resin and on a CFRC.
0.6
fU
2.5-
M
o°o
30
25
^
" '
^2.05
10
15
20
i|
N
< 0.5-
0.0-
.C02
A XeCl
!
« ^
.« 1.0-
W
,
'
W
r-—— ------^--i
°XeCl
0-
^..........................J
20
°Nd:YAG
|20
0.1
$& •
30
40
0
50
10
Energy (mJ)
(a)
20
30
40
50
60
Energy (mJ)
(b)
FIGURE 6. (a) Results taken on the cold curing epoxy resin sample with the TEA CO2 and the XeCl. The
insert is an enlargement of the area marked in dashed lines, (b) Comparative results: longitudinal amplitude vs.
mean laser energy for all three wavelengths.
TABLE 3. Comparative results: Damage/Ablation threshold for all three wavelengths.
Nd:YAG
TEA CO2
XeCl
Damage Threshold
Ablation Threshold
Microscopical observation of the first exposed fibres
(Ablative PhotoDecomposition)
No exposed fibres
Ablation rate at the corresponding energies:
~12 um/pulse
Ablation rate at 6mJ:
~0.15 um/pulse
L-Amplitude (nm)
1.62
1.15
0.17
Energy (mJ)
6.2
37.8
2
DISCUSSION
The term "ablation threshold" was used instead of "damage threshold" for the XeCl
excimer (figure 2(c)). This is because of the different way that UV radiation interacts with
the superficial epoxy resin. The 308nm photons of the XeCl have enough energy to directly
970
excite and break the intermolecular bonds of the polymers, which is a photochemical
phenomenon [15].The IR radiation has excite the vibrational modes and rely on thermal
energy to ablate the material. The mechanism that the UV radiation uses is called "Ablative
Photodecomposition". The ablation threshold occurs at lower energies (as can be seen in
figure 2(c)) but the ablation rate (i.e. the rate of material removal from the surface of the
sample) is very low [16]. This rate was measured by profilometry on the cold curing epoxy
sample and was found to be approximately 0.12fim/pulse for 6mJ incident UV energy.
From the results shown in table 2 for the cold curing epoxy, it can be seen that the
optical penetration depths for the IR and the UV wavelength are very close. This is a
coincidence, demonstrated also by the similarity of the shapes of the waveforms in figure 3.
Comparisons were made between results of the cold curing epoxy and the CFRC.
CFRC vs. Cold Curing Epoxv Resin Sample
The first comparison was between results recorded using either the TEA CO2 or
XeCl excimer on the cold curing epoxy and CFRC sample. The energy vs. amplitude graphs
can be seen in figure 5. The results have been corrected for the different thickness of the two
samples. For this reason, a series of experiments was carried out on pure epoxy resin
samples of various thicknesses. The recorded longitudinal amplitude was found to follow a
1/thickness law which agrees with [17].
From figure 5(a) it can be seen that with the IR laser the behavior both on the CFRC
and on the epoxy sample follow similar curves, bearing in mind that the data have not been
corrected for the different attenuation of the two samples. On the other hand, for the excimer
laser, figure 5(b) shows that the two curves are not similar. It is obvious that there is a
change of gradient on the results from the CFRC sample (due to the change from the
thermoelastic regime to the photo-ablative regime). From this figure it is noted that the
presence of the carbon fibers plays a significant role on the laser generation mechanism of
ultrasound.
IR vs UV laser wavelength
The second comparison was between results on either the cold curing epoxy resin or
CFRC sample and different lasers (TEA CO2 and XeCl).
In figure 6 results recorded with the two lasers on the pure epoxy resin sample are
compared. Because of experimental limitations (damage of the joulemeter at higher UV
laser energies), the energy range of the experimental data was limited between 0.5 and
lOmJ. Nevertheless, as the inset of figure 6 shows, the curves of the two wavelengths are
quite close (very close to the experimental error). This is mainly because for these two
wavelengths the optical absorption depths on this epoxy resin are similar (table 2).
Comparative results of the 3 wavelengths on the same unidirectional CFRC
Results taken on the same CFRC sample are compared in figure 6(b). The data from
the TEA CC>2 laser have been corrected for the small difference in spot size. In the case of
the Nd:YAG (1064nm), most of the energy is absorbed in the first layer of carbon fibers [7]
as the resin is almost transparent. At 10.6fim (TEA COa ), both the superficial resin and the
carbon fibers absorb strongly [7, 18]. The same happens at 308nm (XeCl) but in this case,
because of the different way that the UV radiation interacts with the superficial epoxy resin
(Ablative Photo-Decomposition), the generation mechanism is different. The fact that the
three lasers produce three different sorts of buried source is most easily observed in the
generation efficiency graph shown in figure 6(b) and the shape of the recorded waveforms
971
[11]. Table 3 summarizes the results shown in figure 6(b) and emphasizes the criteria used
for the determination of the damage threshold. Given the fact that our CFRC sample had a
12|im layer of superficial epoxy resin and that our measurements were single shot, the
ablation rate at the damage threshold was calculated. For the UV laser an insignificant
amount of material was removed.
CONCLUSIONS
From the results shown both in figure 6(b) and table 3, the NdrYAG at 6mJ produced
similar amplitude longitudinal waves to the TEA CCb at 40mJ. This shows that the NdrYAG
is at least 6 times more energy efficient than the TEA C(>2. In the case of the excimer laser,
although the ablation threshold was found to be very low, the damage was localized at the
superficial resin layer of the CFRC. Exposure of fibers was never observed, up to the
maximum energy densities used and the ablation rate was found to be very low. At the same
laser wavelength it has been observed that the presence of the carbon fiber layer plays a
significant role in the generation mechanism. It is possible that absorption of energy by the
carbon fibers seeds the photo-ablative decomposition in the overlying epoxy layer.
ACKNOWLEDGEMENTS
This work is supported by the E.U. (Marie-Curie fellowship HPMF-CT-2000-00999).
REFERENCES
1. Davies S. J., Edwards C., Taylor G. S. and Palmer S. B, J. Phys. D, 26, pp.329-348
(1993).
2. Taylor G. S., Hutchins D. A., Edwards C. and Palmer S. B., Ultrasonics, 28, pp.343349(1990).
3. Monchalin J.-P., Neron C., Bouchard P., Heon R. and Padioleau C., Canadian
Aeronautics and Space Journal, 43, No. 1., (1997).
4. Scruby C. B. and Drain L. E., Laser-Ultrasonics: Techniques and Applications, Adam
Hilger, Bristol, UK, 1990.
5. Rose L. F. R., J Acoustic Soc. Am. 75 pp.723-733 (1984).
6. Doyle P.A., J. Phys. D: Appl.Phys, 19, pp.1613-1623 (1986).
7. McKie A D W and Addison R C Jr., Ultrasonics, 32, pp.333-345 (1994).
8. Taylor G., Edwards C. and Palmer S. B., Nondestr. Test. Eval, 5, pp.135-143 (1990).
9. Krehl P., Schwirzke F. and Cooper A. W., /. Appl. Phys., 46, pp.4400-4406 (1975).
10. Dewhurst R. J., Hutchins D. A., Palmer S. B., Scruby C. B., J. Appl. Phys., 53, pp.40644071 (1982).
11. Stratoudaki, T Edwards C., Dixon S. and S. B. Palmer, in Review of Progress in QNDE,
Vol.21 eds. D. O. Thompson and D. E. Chimenti, (AIP, 2002), pp. 316-323.
12. Edwards C., Stratoudaki T., Dixon S. and S. B. Palmer, in Review of Progress in
QNDE, Vol.20 eds. D. O. Thompson and D. E. Chimenti, (AIP, 2001), pp. 220-227.
13. Bushell A. C., Edwards C. and Palmer S. B., in Review of Progress in QNDE, Vol.l 1
eds. D. O. Thompson and D. E. Chimenti, (Plenum Press, 1992), pp. 569-573.
14. Edwards C., Stratoudaki T. and S. B. Palmer, in Review of Progress in QNDE, Vol.22
eds. D. O. Thompson and D. E. Chimenti, (AIP, 2003), to be published.
15. Garrison B. J., Srinivasan R., J. Appl. Phys., 57, pp. 2909-2914 (1985).
16. Sutcliffe E. and Srinivasan R., J. Appl. Phys., 60, pp. 3315-3322 (1986).
17. Knopoff L. J. Appl. Phys., 29, pp. 661-670 (1958).
972