PSI - Issue 77

L.A.S. Maia et al. / Procedia Structural Integrity 77 (2026) 87–94 Maia et al. / Structural Integrity Procedia 00 (2026) 000–000

88

2

ratio, more uniform stress distribution, and ability to bond dissimilar materials (metal/wood/polymer) without compromising its integrity (Saeedifar et al. 2023). Adhesives are often used in structural applications of industrial sectors such as aerospace, aeronautical, civil construction, automotive, aerospace, and wind energy (Fan et al. 2024). Powerful computational tools are nowadays available to simulate the behaviour of adhesive joints (He 2011). Abaqus ® and Ansys ® are the leading Finite Element Method (FEM) software allowing the analysis of stress distribution, deformation, failure modes, and strength, helping in the joint design and structural integrity optimization (Khan et al. 2021). CZM is a numerical technique used to simulate the fracture behaviour of adhesive joints, which represents the adhesive layer as a cohesive zone with specific traction-separation laws that describe the relationship between stresses and displacements in the adhesive. This model is particularly useful to evaluate crack initiation and propagation in adhesive joints, as it can capture complex interactions between the adhesive and the adherends (Perrella et al. 2024). The performance of adhesive joints depends on the adhesive properties, surface preparation of the adherends, and design parameter such as the thickness of the adherends, the overlap length, inner/outer chamfers and adhesive fillets (Liao et al. 2013). Other way to increase the joint strength and improve the stress distribution can be achieved through the combination of a brittle and a ductile adhesive. This method, known as bi-material adhesive joint or DAJ, was presented by Raphael in 1966 (Raphael 1966). The brittle and high modulus adhesive is placed at the overlap centre and the ductile adhesive at the edges where higher stress concentrations occur. Several authors have demonstrated that DAJ exhibit superior performance when compared to single-adhesive joints (SAJ). Öz and Özer (Öz and Özer 2017) performed a study of DAJ with two different adhesive combinations (AV138+2015, AV138+DP-8005) obtaining improved P m when comparing to SAJ. An experimental and numerical CZM study of aluminium stepped joints considering a double adhesive approach was carried out by Carvalho et al. (Carvalho et al. 2024) The DAJ showed a remarkable improvement in U and damage tolerance, particularly in joints with high overlap lengths ( L O ). Carvalho et al. (2022) presented a CZM work using a DAJ methodology to evaluate the P m in T-joints. The numerical analysis showed an increase in P m and U when applying a combination of adhesives in T-joints, compared to a SAJ layout. Kurennov et al. (2023) addressed the topological optimization of a symmetric DAJ using a generalized Goland Reissner model. The outer adherend shape was optimized via a Fourier series expansion, determining optimal adhesive section lengths. A genetic algorithm minimized the adherend cross-sectional area while constraining maximum stresses. The stress state was analysed using the finite difference method, and results were validated by FEM. This work aims to improve the DAJ impact behaviour with steel adherends by applying geometrical modifications (outer and inner chamfers, and adhesive fillets) and considering different adhesive combinations. A numerical study was conducted using CZM, encompassing an analysis of σ y and τ xy stresses, P m , and U . 2. Methods 2.1. Joints and geometries The overlap SAJ experimental data from Valente et al. (2019) was used to validate the impact CZM approach. The dimensions presented in Fig. 1 include the total length between grips ( L T ) of 200 mm, adherend thickness ( t P ) of 2 mm, width ( b ) of 25 mm, adhesive thickness ( t A ) of 0.2 mm and L O =25 mm.

Fig. 1. SAJ geometry and boundary conditions.

The DAJ geometry is presented in Fig. 2. t P , t A , and L O were kept constants in all trials. The parameters subject to variations, and dedicated analysis, are the angle of the outer chamfers ( α ), inner chamfers ( β ), and adhesive fillets ( θ ), varying from 7.5 to 90º. The joint dimensions are (in mm): L T =215, t A =0.2, t P =2, b =25, and L O =25. The brittle adhesive was placed in the middle of the overlap (4/6 L O ), while the low-modulus adhesive was applied at the edges (2/6 L O ).