The ability to detect defects in materials is of great importance in industry. The presence of defects alters material properties, often compromising the structural performance of critical components. Ultrasonic detection techniques mainly rely on the fact that defects alter incoming sound waves, either through scattering and attenuation (linear) or through the activation of dynamical phenomena such as clapping and friction, giving rise to non-linear effects in the wave propagation. In the current work, we exploit the fact that defects act as localized non-linear sources within a material and couple this behavior to the transfer of acoustic energy into thermal energy in order to locate their position by means of full-field thermography. The basic idea is to apply the principle of time-reversal to focus vibrational energy at a specific region in the material. In combination with filtering techniques (most notably the scaling subtraction method), it is possible to direct this energy on a defect without prior knowledge about its location. As a result, the defect will get strongly activated and converts part of the acoustic energy into heat (due to clapping and friction), acting as a localized thermal source. The heat signature produced by thermal diffusion from this internal heat source can then be recorded at the surface with an IR-camera, revealing the location of the defect. Each step in this procedure will be illustrated by 2D and 3D simulations for thin plates and guided waves. In addition, the coupled thermosonic methodology will also be numerically tested for multiple defects and selective focusing.

The growing emphasis on hydrogen within the European Green Deal has intensified the need for lightweight, high performance composite structures, particularly for aviation grade liquid hydrogen storage. A persistent challenge for such storage tanks is the formation of matrix cracks in manufacturing and service loading, which can affect the structural integrity and leak tightness required for safe hydrogen containment.

Nonlinear ultrasonic behavior caused by the breathing cracks, commonly referred to as contact acoustic nonlinearity (CAN) offers a pathway to assess this form of micro damage. While CAN has been examined extensively in metallic systems and simplified composite arrangements, the nonlinear ultrasonic behaviour associated with transverse cracking in cross ply laminates has not been examined in sufficient depth, highlighting the clear gap that this study addresses.

In this study, guided wave propagation in cross ply ([0/90]_S) CFRP laminates containing matrix cracks was numerically investigated using finite element simulations. The fundamental symmetric (S₀) mode was excited by radially outward excitation around a 10 mm diameter circle, representing an ultrasound transducer, and the waves behavior was examined along propagation paths both parallel and perpendicular to the crack direction. Cracks with varying numbers, positions, and orientations were investigate to evaluate their influence on wave damage interaction and the resulting nonlinear responses.

The results demonstrate that crack breathing produces clear nonlinear features in the guided wave response. Furthermore, variations in nonlinearity for different excitation directions reveal that propagation path selection can enhance sensitivity to specific crack characteristics, thereby enabling an improved assessment of damage extent.

These findings advance the understanding of CAN mechanisms in cross ply CFRP laminates and contribute essential knowledge for the development of nonlinear guided wave approaches aimed at detecting and characterizing matrix cracking in hydrogen storage composite structures, supporting the future integration of structural health monitoring.

The use of composite materials, and particularly laminated composites, in aircraft structures is now widespread. For safety reasons, their application to structural components requires appropriate monitoring to ensure their integrity. However, damage can occur during operation, particularly due to impacts, leading to barely visible impact damage (BVID). Ultrasonic guided wave-based monitoring methods developed in recent years are capable of locating such defects, provided that reference measurements are available before their occurrence. In this paper, we focus on the development of a Structural Health Monitoring (SHM) method that does not rely on such a reference state. This method combines a vibrational excitation, referred to as a pump, with interrogation by ultrasonic guided waves (referred to as probes), following an approach proposed by [1]. When a contact defect is present in the thin structure, its state is modulated by the pump, as is the amplitude of the probe signals that have interacted with the defect. A baseline is then reconstructed from the ultrasonic signals recorded for numerous different pump states and stored in a matrix called the differential matrix. The method is applied to an artificial contact defect on an anisotropic Carbon-Fiber-Reinforced laminated composite plate. The pump is generated by a shaker, while the probe signals are emitted by thin PZT patches, with the plate fixed to a frame at two points to reproduce realistic boundary conditions.

The defect is probed using the qA0 mode and localized by two imaging algorithms: the incoherent Delay-and-Sum algorithm and an anisotropic Beamforming algorithm [2] that compensates for both dispersion and anisotropy. The method is compatible with a multi-transmitter and multi-receiver setup used in Round-Robin, even without synchronization between the pump and probe, provided the vibration is stationary and appropriate adjustments are applied.

Furthermore, with the aim of eventually replacing the external pump with ambient vibrations experienced by the structure during operation (e.g., engine noise), an alternative method for extracting the amplitude modulation of the defect response from the differential matrix signals, leveraging a Singular Value Decomposition, is proposed. It is then demonstrated that the artificial defect subjected to a pink noise vibration with a bandwidth of [100 Hz, 1000 Hz] is successfully imaged, with higher contrast using the Beamforming algorithm compared to Delay-and-Sum.

[1] M. Terzi et al. « Pump-probe localization technique of varying solid contacts ». The Journal of the Acoustical Society of America, 2021

[2] P. Goislot et al. « Sparse guided wave imaging in highly anisotropic plates with phase skewing and amplitude focusing compensation”, The Journal of the Acoustical Society of America, 2025.

This article examines the application of nonlinear acoustic methods for detecting and localizing damage in structural materials. Thanks to their high sensitivity and ability to identify early-stage damage, these techniques are gaining traction in material monitoring and non-destructive testing. These methods are based on entirely different phenomena from classical ultrasonic techniques. For example, in the case of guided waves, reflection, attenuation, and diffraction play crucial roles. In contrast, nonlinear acoustics focuses on phenomena associated with damage dynamics, such as the relative motion of fatigue crack surfaces, fluctuating temperature fields, or the opening and closing of a fatigue crack under external excitation. A key factor in this process is the selection of an appropriate excitation method and amplitude to activate specific mechanisms that generate nonlinear effects observable in the response signal spectrum. The paper presents the theoretical background of the method, along with sample experiments and analyses that enable material characterization based on nonlinear effects resulting from contact damage. Techniques based on vibro-acoustic modulation and the Luxembourg–Gorky effect will be presented. The results of analyses that enable both global damage detection and methods allowing for damage localization will be presented. Potential sources of nonlinear effects will be presented, based on which damage identification is possible. The research was conducted on both metallic structures and complex composite structures. Signal acquisition was carried out in a non-contact manner using laser-based techniques. It also addresses challenges in interpreting the resulting data and in identifying the sources and types of damage-related nonlinearity.