JEC COMPOSITES MAGAZINE - Issue #116 - October 2017 - 60

TECHNOLOGY NDT
(Figure 2). A water-confined regime was used to maximize the applied
pressure. Back-face velocity monitoring was achieved using a PDV system [8] to follow the main back and forth travels of shock waves within
the thickness of the sample.
Procedure
The disbonding thresholds were investigated for both types of mechanical strength. Using two samples for each mechanical grade also
made it possible to investigate both TA6V and composite front-face
configurations, i.e. to see how the impacted face could influence the
disbonding threshold. The side dimensions of the samples enabled
about 10 laser shots per sample (Figure 3). With this distribution,
there were about 20 mm between each laser spot, ensuring that the
shots did not influence one another. A 6 mm focal spot diameter was
used to maximize the size of the induced disbondings, and hence their
detectability. With this geometry, intensities just beyond the water
breakdown threshold could be reached (see Equation 1), thus maximizing the ablation pressures used. In addition, the diameter-to-thickness ratio of the focal spot stayed relatively high, ensuring 1D shock
propagation. In each bonding quality/front face configuration, gradual
intensities were applied from 0.5 to 6 GW/cm².
After the LASAT experiments, an ultrasonic inspection in water immersion with transmission configuration was used to detect laser-induced disbonding. An emitting transducer worked at 15 MHz and
reception was achieved with the help of a 10 MHz sensor.

Results
Disbonding thresholds
The laser shots for each configuration are referenced in Figure 4 with
the corresponding intensity and ultrasonic (US) evaluation of the
bond state. The disbonding thresholds are provided in Table 2. The
threshold uncertainty is considered to be only impacted by the laser
energy uncertainty, which is about 10%.
First, it was noticed that whatever the illumination side, the disbonding thresholds are clearly different for healthy bonds and weak
bonds. This shows the potential of the LASAT technique to identify
weak bonds by applying a calibrated laser pulse, above the weak bond
threshold and beyond the nominal bond threshold. For example, illu-

Fig. 3: Front face of a TA6V/composite sample after laser ablation
Tab. 2: Disbonding thresholds identified by US inspection after LASAT

TA6V illumination

Composite illumination

Healthy bond

5.2 GW/cm²
(+/- 0.5 GW/cm²)

4.25 GW/cm²
(+/- 0.4 GW/cm²)

Weak bond

3.25 GW/cm²
(+/- 0.3 GW/cm²)

2.25 GW/cm²
(+/- 0.2 GW/cm²)

jec composites magazine / N°116 October 2017

Weak Bonding
TA6V Illumination

Healthy Bonding
TA6V Illumination

Weak Bonding
Good Bonding
Composites Illumination Composites Illumination

Fig. 4: Summary of laser shots for each configuration with US disbonding diagnostic (green circles - no disbonding detected, red circles - disbonding detected)
and disbonding thresholds as black lines (threshold uncertainty is shown with
dotted lines)

minating such a sample with a 4 GW/cm² pulse on the TA6V side will
leave a healthy bond intact and induce a disbonding in a weak bond.
Post-shock US scanning would reveal a possible disbonding and hence
determine if the sample is a healthy bond or not.
A slight difference was seen whether the shock was generated on the
TA6V side or on the composite side. This difference was most probably due to different wave propagation patterns within the samples.
Time-position diagrams
At the same time, numerical simulations were realized in order to better understand phenomena during LASAT tests aiming to be able to
predict bonding mechanical behaviour and to optimize experimental
parameters (intensity, pulse time, pulse form, spot diameter,...).
Simulations were conducted using a 1D-hydrodynamic code with
simple material models. These simulations produce the time-position
(X-t) diagram corresponding to 3GW/cm² experiments. The same
laser loading (ablation pressure) is applied at t=X=0 for both configurations: TA6V illumination and composite illumination. Tensile
stresses in the joint appear at different times, depending on the type
of configuration used, in relation with an impedance mismatch in the
three layers (ZEpoxy = 3.106 Pa.s-1, ZComposite = 5.106 Pa.s-1 and
ZTA6V = 15.106 Pa.s-1).
In the case of composite irradiation (Figure 5), the incident shock
waves travelling in the composite interact with the composite/epoxy
interface (X=-3.8mm). Due to relatively close acoustic impedances on
each side of this interface, the amplitude of the reflected release wave
(ZComposite>ZEpoxy) is rather small and most of the incident energy is transmitted as a shock wave in the epoxy layer. The resulting
wave transmitted in the epoxy layer is a shock wave that then interacts
with the epoxy/TA6V interface (X=-3.95mm). The reflected wave is
a shock wave (ZEpoxy < ZTA6V) going backward in the epoxy layer and then in the composite. The wave transmitted at this interface
is a shock wave propagating in the TA6V, which will then reflect itself
as a release wave at the free surface on the back face of the material
(X=- 4.35mm - t=1100 ns). This release wave is partially reflected as a
shock wave at the TA6V/epoxy interface as it goes backward and a corresponding transmitted release wave goes back in the epoxy layer and
then in the composite. Quick back and forth travels of the shock waves
trapped within the TA6V layer can be seen due to the high impedance

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #116 - October 2017