PSI - Issue 79

H.G.E. da Silva et al. / Procedia Structural Integrity 79 (2026) 97–104

98

and stiffness, good impact mitigation, energy dissipation, corrosion resistance, and low thermal and acoustic conductivity (da Silva et al. 2025). The advantages of using two faces co-operating and separated by a core was first introduced by Duleau (1820). Nowadays, sandwich structures are widely used in the lightweight design, especially in aircraft (stabilizers, flaps, wing boxes, radomes, and floor panels), ships (hull, bulkheads, and deck), and wind turbines (blades, spinners nacelle covers, and generator) (Hsu 2009, Tarlochan 2021). For a widespread usage of sandwich structures, it is necessary to make reliable prediction tools available. A comprehensive review covering early developments was published by Noor et al. (1996). The common numerical approach for structural modelling is the finite element method (FEM), currently integrated into simulation software (He 2011, Szabó and Babuška 2021). To model damage initiation and propagation along pre-established paths of fracture, CZM evolved as a powerful technique (Ramalho et al. 2020, Tserpes et al. 2022). Sandwich skins and core can be modelled as solid or shell elements (Thiagarajan and Munusamy 2020): 4- or 8-node shell elements (Gao et al. 2020) or solid hexahedral elements (Sayahlatifi et al. 2020) can be applied to model the mesh of skins and core. Zhang et al. (2020) developed a refined 3D FEM for honeycomb sandwich panels, based on continuum damage mechanics and combined with physically- based Puck’s composite failure criteria . The simulation results showed a good agreement with experimental data and the model can be used to predict the low-velocity impact response and impact damage effectively. Thiagarajan and Munusamy (2020) used the CZM approach to capture the behavior of Anabond ® EFA 960 adhesive film, used to bond aluminum honeycomb and carbon-fiber reinforced plastic (CFRP) skins. The failure criteria for the laminates include the Hashin criterion to model fiber failure, the Puck criterion for matrix failure, and the Tsai-Wu criterion for the skin (Gupta et al. 2017). The numerical outcomes presented good correlation with the tensile, flexural, edgewise compression, and flatwise compression experimental tests. Farrokhabadi et al. (2020) experimentally and numerically investigated a multilayer sandwich panel with glass fiber fabric laminates and a corrugated core under three-point bending loads. The Hashin- Puck’s and Puck and Schurmann criteria were applied to predict damage initiation in the skins. The debonding between skins and core was modelled using a bi-linear triangular CZM. The applied numerical approach successfully predicted the bending behavior with deviations under 10% compared with experimental data. Zhu et al. (2023) evaluated the response of CFRP trapezoidal corrugated sandwich panels under 4PB load. The Hashin failure criterion was utilized to evaluate the damage of the composites, and the adhesive failure was modelled by CZM. Comparing experimental and numerical data, the sample stiffness was overpredicted by 11.9% at the most, and the superimposed buckling modes (-14.9%) and critical loads (14.6%) were accurately predicted. Djama et al. (2020) studied a sandwich structure with a glass fiber reinforced truss core under shear, and three-point bending tests. The Hashin criterion was used to numerically model the skins. Comparing experimental and numerical data in the compression, shear and bending tests, stiffness errors of 13.0%, 6.1% and 12.4%, respectively, were found. Sayahlatifi et al. (2021) experimentally and numerically investigated hybrid corrugated composite/balsa core sandwich panels under 3-point bending load. Composite damage was modelled in accordance with a combination of Hashin and Puck failure criteria. To simulate the adhesive layer between the face sheets and core, CZM was utilized. Numerical results presented good agreement with the experimental findings The corrugated core combined with common solid balsa contributes to a substantial increase in strength (34.7%) and stiffness (28.2%) compared to the conventional balsa core sandwich structure. This research studies the behaviour of sandwich structures under 4PB tests. The experimental testing campaign allows to extract the load-displacement ( P -  )curves, to compare with numerical models. The numerical work considered CZM for the adhesive, the fractured foam model for the core, and the Tsai-Wu criterion for the skins. 2. Experimental and numerical procedure 2.1. Geometry and test setup In this study, two distinct sandwich structure configurations, differing in skin layups but sharing the same material composition, were analyzed. Square sandwich panels with an initial dimension of 300 mm per side were fabricated and subsequently sectioned into test specimens measuring 250 mm in length and 36 mm in width. These dimensions were selected to comply with standard requirements, ensuring the width-to-height ratio satisfied the condition >2ℎ, where b represents the specimen width and h its overall thickness. The detailed geometric parameters of the sandwich specimens are illustrated in Fig. 1 (a).