NONLINEAR ACOUSTIC TECHNIQUES FOR IMPACT DAMAGE DETECTION IN COMPOSITE MATERIALS

LOI, GABRIELA

2022-04-08

---

Abstract

Due to their high stiffness–to–mass ratio, fiber-reinforced materials have been intensively used in several engineering branches. Nevertheless, their low resistance to low-velocity impacts may question their use in critical load-bearing structures. Impacts may result in the emergence of barely visible internal damage. Although not severe enough to cause the catastrophic failure of the inspected component, the impact-induced damage may compromise its load-bearing capability. The need to detect the onset of internal damage has led to an ever-rising interest in Structural Health Monitoring (SHM). Nowadays, although the underlying physics is still not clear, the SHM approaches based on nonlinear acoustic phenomena are widely used to provide information on the structural integrity of composite materials. This thesis focuses on two nonlinear acoustics techniques, namely the Scaling Subtraction Method (SSM) and the Nonlinear Vibro-Acoustic Modulation (VAM), to further assess their effectiveness in detecting the occurrence of impact damage in composite beams. The Scaling Subtraction Method accounts for the global nonlinear content of the system response to infer information on its integrity and ability to withstand critical loads. Mainly applied to granular media, the SSM was proven to be effective in identifying the onset of internal damage also in metals and composites. However, the quality of the provided indications was highly affected by the selection of the interrogating frequency, usually chosen among the resonance frequencies of the inspected system. The need for preliminary modal analysis and the lack of a robust algorithm to pick the resonances with the highest sensitivity to damage prevent the SSM from being a reliable NDT tool. In an attempt to overcome these limitations, this thesis proposes a novel SSM-based approach relying on the use of a broadband impulsive excitation. To demonstrate the feasibility of a pulse-based SSM approach, a composite beam has been alternatively excited through impulsive or pure-tone harmonic excitations tuned at different natural frequencies in both pristine and damaged conditions. The results showed the proposed approach to be a rather promising option for detecting the onset of low-velocity impact damage in composite materials. The Nonlinear Vibro-Acoustic Modulation relies upon simultaneously driving the inspected structure with two waves of distinct frequencies and amplitudes. The presence of defects perturbates the propagation of the two impinging waves leading to the emergence of modulation sidebands that can be exploited to infer information on the material properties degradation. In recent years, the analysis of modulation sidebands has been applied to detect the onset of internal damage in both metal and non-metallic structures. Nonetheless, since some critical issues still have to be investigated, this thesis aims to further assess whether the selection of some testing parameters may affect the effectiveness and the sensitivity of the VAM. For this purpose, three identical composite beams have been subjected to multiple low-velocity transverse impact loads to induce the emergence of a pattern of modulation sidebands around the probe frequency peak and, subsequently, tested under different boundary conditions and through different actuation-sensing scenarios. The obtained results showed that the amplitude of the modulation sidebands tends to increase with the damage severity. However, the trial of a set of pump frequencies revealed the VAM to be highly sensitive to the selected natural frequency. Similarly, the pump excitation amplitude was shown to affect the quality of the provided indications. In addition, the method capability to identify the changes in the sample integrity slightly varied with the considered sensing-actuation scenario, even though it was the sensor positioning to be found the factor susceptible to affect the method performance to a greater extent.