*2.2. Experimental Procedure*

The experimental measurements of guided waves were carried out by the scanning laser Doppler vibrometry (SLDV) method using the set-up presented in Figure 3a. The input wave signal was generated by the arbitrary function generator AFG 3022C (Tektronix, Inc., Beaverton, OR, USA) and magnified with the use of the high-voltage amplifier PPA 2000 (EC Electronics, Krakow, Poland). The excitation of the guided wave field was provided by the plate piezoelectric actuators NAC2024 and NAC2025 (Noliac, Kvistgaard, Denmark) bonded to the surface of each examined specimen by

the petro wax 080A109 (PCB Piezotronics, Inc., Depew, NY, USA). The signals of propagating waves (out-of-plane velocity components) were collected by the non-contact method using the scanning head of the laser vibrometer PSV-3D-400-M (Polytec GmbH, Berlin, Germany) equipped with the VD-07 velocity decoder. The sampling frequency for each measurement was set to 2.56 MHz. The area of scanning was covered with a retro-reflective sheeting to improve the light backscatter.

**Figure 3.** Experimental investigations: (**a**) setup for scanning laser Doppler vibrometry (SLDV) measurements; (**b**) scheme of measurements for dispersion curves determination; and (**c**) scheme for damage identification and visualization measurements.

The samples #1 and #5 were firstly tested to determine dispersion curves in a single-layer and three-layer medium, according to the scheme presented in Figure 3b. For this purpose, the propagation of antisymmetric and symmetric modes was induced by actuators P1 and P2, respectively, both attached to the steel plate. The input signal had the form of a wave packet obtained by the Hanning window modulation of the single-cycle sine function. To obtain a wide frequency spectrum for both excitation types, measurements were conducted five times with the different carrier frequency, equal to 50, 100, 150, 200, and 250 kHz, consecutively. The time-domain wave signals were acquired in 91 points spaced evenly over a straight line with a length of 90 mm, resulting with a resolution equal to 1 mm. The influence of potential reflections of propagating waves at the boundaries, induced by the short path length in relation with beam thickness, was further eliminated by analyzing only the initial part of the collected signals. The main investigations aimed at damage imaging in all beams #1–#5 were conducted with respect to the scheme shown in Figure 3c. Each specimen was tested twice, changing the location of the excitation point (P3 and P4). Excitation signal was a five-peak wave packet obtained from a sinusoidal burst with a central frequency equal to 100 kHz. The signals of propagating waves were collected in 2323 points distributed on the top surface of the steel plate, covering the area of 440 × 88 mm2. The scanning was conducted point by point in the regular mesh of 23 rows and 101 columns with a resolution of 4 mm in both directions.

Additional measurements with a contact method were performed to identify the damaged specimens. The input signal was identical to the one used for damage imaging with SLDV (five-cycle sine wave packet with a central frequency of 100 kHz). The piezoelectric actuators were attached at the ends of each beam to excite and acquire wave signals in points P3 and P5, respectively (Figure 3c). The excitation signal was created by the arbitrary waveform generator AFG 3022C (Tektronix, Inc., Beaverton, OR, USA) and then amplified by the high-voltage amplifier A400DI (FLC Electronics AB, Partille, Sweden). The collection of signals was provided by the digital oscilloscope PicoScope 4824 (Pico Technology, St Neots, Great Britain) with a sampling frequency of 20 MHz.
