PSI - Issue 75

J. Filho et al. / Procedia Structural Integrity 75 (2025) 353–362 J.Filho, L. Wittevrongel, F. Pieron, P. Lava / Structural Integrity Procedia (2025)

354 2

DIC has been widely used in material identifications (Pierron and Grédiac (2021)) and few works focused on numerical model validations (Lava et al. (2020)). In the context of fatigue tests, high-frequency signals are often employed to excite specimens and high-speed machine vision cameras are commonly used. However, to achieve high frequencies, the image has to be cropped which is detrimental for the signal-to-noise ratio. To overcome this limitation, the signal-locking principle has been proposed in the literature. Lorenzo et al. (2020) and Fruehmann et al. (2015) have demonstrated the effectiveness of the signal locking principle in structural dynamics and structural fatigue applications, respectively. The signal-locking principle allows the use of low-cost, high-resolution and limited-frame-rate devices to artificially reconstruct the high frequency periodic signals using aliasing. With respect to fracture mechanics, full-field-fitting approaches have been used to investigate fatigue crack growth (Becker (2024)). Williams (1957) has proposed a high-order series expansion that can be employed to estimate stress intensity factors and determine the position of the crack-tip considering linear elastic materials. The full-field displacements obtained by DIC can be used in Williams’ formulation and the parameters can be identified using an optimization routine to minimize the difference between the measured and reconstructed full-field displacements (Becker (2024)). Moreover, the energy release rate due to crack propagation can also be extracted in fracture experiments. Rice (1968) has first proposed a path-independent integral, called J -Integral, as an integral criterion to estimate the energy release rate over a closed counterclockwise contour surrounding the crack-tip. Later, Li et al. (1985) proposed an equivalent domain-independent J -Integral to measure the energy release rate using a closed domain. This work addresses the capabilities of a fully integrated commercial system (MatchID (https://www.matchid.eu/)), encompassing signal-locking, the estimation of some fracture properties using DIC and verification based on numerical models. 2. Signal-locking principle Considering high-frequency periodic fatigue signals, high-speed devices are often used to acquire data at the same frequency used to excite specimens. In the context of DIC, high-speed or ultra high-speed cameras can be used to grab images that can be later processed to obtain full-field displacements and deformations. These cameras are known to be costly and, in order to achieve high frequencies, the record length and camera resolution must be reduced due to hardware limitations. Therefore, temporal resolution is increased with the cost of reducing the DIC spatial resolution. On the other hand, high-speed or ultra high-speed cameras can be replaced by low-cost, high-resolution and limited frame-rate cameras. This can be done by means of the signal-locking procedure, which is commonly used in several fatigue applications. Reference points can be defined in the periodic signal flow, generating lock pulses that can be used to send sync signals to trigger the cameras, as schematically illustrated in Fig. 1. As a result, images can be acquired at the peaks, or the source signal can be reconstructed using aliasing. Signal-locking has been implemented into the MatchID quasi-static grabber, where low-speed machine vision cameras can be used to grab images of specimens subjected to high-frequencies. 3. Synthetically deformed images In order to verify the crack analysis implementations and study the experimental uncertainties, FEA-based synthetic images with predefined values can be used (Rossi et al. (2015) and Balcaen (2017)). The synthetically deformed images can be considered a digital twin, i.e. , a digital representation of how a specimen is deformed considering a given load condition or prescribed displacements. A benchmark test can be performed to validate the test setup in both 2D and Stereo experiments. The inherent systematic errors in DIC can be estimated when correlating the synthetic images and comparing the full-field DIC results with the imposed FEA model. For illustrative purposes, Fig. 2 shows the synthetic image generated by the MatchID FEDEF module using a stereo configuration with an imposed analytical solution proposed by Williams (1957) under mixed-mode loading. Here, the crack growth and the stress intensity factors can be simulated and are exactly encoded in the virtual images. Accordingly, it can be verified how well these can be reproduced by the analysis chain.