PSI - Issue 77

H. Lopes et al. / Procedia Structural Integrity 77 (2026) 673–680 H. Lopes/ Structural Integrity Procedia 00 (2026) 000 – 000

675

3

field and its spatial derivatives (Hung, 1999; Lopes et al., 2014). A general sign of internal damage is a symmetrical set of fringes resembling a “butterfly” shape. Specific defect types are correlated with distinct patterns: a localized “butterfly” or co ncentric fringes signifies delamination or disbond in CFRP components or bonded joints (Hung, 1999; Burkov et al., 2017); a symmetrical or closed fringe pattern corresponds to voids, inclusions, or foreign material beneath the surface (commonly found in composite structures and tires) (Yang and Xie, 2003; Tao et al., 2025); non uniform or distorted fringes are produced by crushed cores in sandwich panels (Burkov et al., 2017); and a discontinuity or break in fringes along a line is associated with lack of adhesion or cracks in composite laminates (Hung, 1999; Hung et al., 2000). The evaluation of the strain field from phase maps measured with DS typically involves phase filtering, unwrapping, and scaling (Araújo dos Santos and Lopes, 2024; Yang and Xie, 2003). These steps are designed to enhance the visibility and quantifiability of damage signatures such as delaminations, crushed cores, cracks, and voids. To further enhance damage detectability, the strain field can be post-processed by computing its spatial derivatives. This is achieved through numerical and smoothing procedures that emphasize perturbations associated with internal damage while simultaneously minimizing the amplification and propagation of experimental noise (Reis Lopes et al., 2011). The main limitation of this approach is its high computational cost, as well as the need to carefully balance the derivative order and filtering level to optimize the damage visibility signatures without introducing artifacts. This work presents a novel methodology that circumvents the limitations of traditional digital shearography (DS) post-processing. The current methodology begins by applying a band-pass filter directly to the set of phase maps recorded during the structure's thermal-load cooling stage. This crucial step effectively isolates damage-related perturbations, which are often confined to a specific spatial frequency band. Subsequently, summing the filtered phase maps leads to a significantly improved signal-to-noise ratio (SNR). This summation effectively enhances both the visibility and detectability of the damage signatures. Compared to existing published methodologies, the proposed technique not only achieves significantly improved damage identification but also offers practical advantages: it is less computationally intensive, and the selection of appropriate cutoff frequencies is less critical. 2. Materials and Methods The subsequent sections detail the inspected CFRP plate with two impact damages, the experimental setup for the measurement of the phase maps using the DS technique and the novel methodology for enhancing the damage detectability. 2.1. Inspected Structure A rectangular plate of carbon fiber-reinforced polymer (CFRP) was used as a test sample, with dimensions of 0.2765 × 0.198 m and a thickness of 1.825 mm (Reis Lopes et al., 2011). It was attached to a supporting structure by clamping its two shorter edges. To induce the internal damage (delamination), the plate was subjected to two low energy impacts by dropping a steel sphere. The plate's surface was pre-coated with a thin white powder to enable precise identification of the impact locations. The two impacts are characterized as follows: a lower energy of 13.5 J at coordinates (74 mm, 147 mm), and a higher energy of 26.2 J at coordinates (215 mm, 38 mm). Following the procedure, a visual inspection was performed. The absence of white powder enabled the precise identification of both impact locations, but no signs of damage were visible on the surface. Figure 1 details the location of both impacts and the clamped region of the plate.

2.2. Experimental Setup

The measurements of the plate's strain field during the cooling stage after applying a thermal pulse excitation were performed using an in-house DS system. This system uses a Michelson interferometer, which employs a beam splitter to divide and then recombine the laser light reflected from the plate's surface. A lateral shift (shearing amount) is introduced by slightly tilting one mirror of the interferometer. The resulting superposition in the image plane creates

Information classification: Internal