PSI - Issue 68

P.M.D. Carvalho et al. / Procedia Structural Integrity 68 (2025) 398–404

399

2

P.M.D. Carvalho et al. / Structural Integrity Procedia 00 (2025) 000–000

control, and the need for careful joint preparation (Adams et al. 1997). The properties of the adhesive and adherends, the geometry of the joint, and other factors influence the strength of an adhesive bond. Single-lap joints (SLJ) are the most common due to the ease of manufacture and predominant shear loading. However, the non-collinear transmitted forces cause significant peel stresses. Double-lap joints (DLJ) prevent substrates from bending. However, they are more time-consuming to manufacture than SLJ. Step and scarf joints are alternatives that minimize peel stresses. In T-joints, the elements to join are usually at a 90° angle (Ebnesajjad and Landrock 2014). These joints are applied in aircraft for different structural parts, e.g., fuselage skin/bulkhead and fuselage skin/rib joints (Masoudi Nejad et al. 2022). In the maritime industry, T-joints find application in hull/deck joints, bulkhead joints, and frame-to-plate joints (Choffat et al. 2023). Door panel assembly, hood and trunk lid bonding, glass bonding, and lightweight composite bonding are applications for the automotive industry (Lambiase et al. 2021). The widespread use of adhesive bonding as a joining technique supposes the existence of reliable design tools. Analytical methods provide simplified solutions based on theoretical principles. The Volkersen’s model (Volkersen 1938), for example, is a widely-used analytical approach to calculate stresses and displacements within the joint, although it oversimplifies the SLJ behavior. Continuum mechanics methods treat the adhesive layer as a continuous medium, and use the elasticity theory to describe stress and strain distributions within the joint. Finite element analysis (FEA) allows modelling complex joint geometries and material nonlinearities, making it a versatile tool (Ye et al. 2018). Fracture mechanics methods predict adhesive joint failure by considering the propagation of cracks within the adhesive layer. The critical stress intensity factor ( K C ) and energy release rate ( G C ) are conventional approaches. Linear elastic fracture mechanics (LEFM) is commonly applied to predict failure in brittle adhesive joints. CZM simulates crack initiation and propagation in adhesive joints. CZM accounts for the nonlinear behavior and can capture cohesive and adhesive failures (Huang et al. 2021). The extended finite element method (XFEM) is another advanced numerical technique that extends traditional FEA. XFEM introduces enrichment functions to represent the crack tip displacement discontinuities, allowing crack growth in complex geometries (Sadeghi et al. 2020). XFEM is well-suited for analyzing adhesive joints with irregular crack paths and interfaces. Different techniques are studied that take advantage of geometrical or material modifications to improve the load transfer efficiency of bonded structures. Ferreira et al. (2019) numerically assessed the static strength of tubular joints using an external chamfer in the region where the tubes overlap. For this study, the authors used FEA to analyze peel and shear stresses in the adhesive layer, and CZM was used to predict the strength. The use of external chamfers promoted a gradual reduction in peak stresses, and higher maximum load ( P m ). Material modifications may involve stiffness grading of the adhesive or adherends, or using DAJ. Raphael (1966) is the precursor in the application of DAJ. This technique consists of placing a flexible and ductile adhesive at the overlap edges, due to stress concentrations, and a stiff and stronger adhesive in-between, since stresses are smaller. Akhavan-Safar et al. (2022) conducted a comprehensive review addressing the advantages and manufacturing complexities associated with DAJ. By evaluating various joint architectures, it was concluded that DAJ techniques can notably increase P m . Furthermore, the study offered valuable manufacturing insights and design recommendations aimed at optimizing the performance of DAJ. The work of Ferreira et al. (2020) experimentally and numerically addressed aluminum stepped-lap DAJ, considering different overlap lengths ( L O ), and assessed the possible improvements over SAJ. The triangular CZM approach was highly accurate for different adhesive combinations. While P m improvements by the DAJ technique were modest, it was possible to accomplish substantial absorbed energy at failure ( U ) increases, which is relevant for the damage monitoring and tolerance of these joints. Gajewski et al. (2021) presented a novel approach to analyze DAJ under uniaxial tensile loading, employing both numerical modelling and neural networks. The DAJ configuration comprised adhesives with distinct stiffness and strength. A point bonded joint with higher stiffness and strength was positioned along the overlap axis, while a bonded joint with lower stiffness and strength extended to the overlap edges. The joint undergoes a two-stage degradation process, wherein damage to the stiff adhesive is followed by the flexible adhesive. Numerical modelling used FEA and CZM to simulate the joint behavior. The simulations provided input data for an Artificial Neural Network (ANN), facilitating the investigation of individual parameters’ impact on P m and energy. This integrated approach offered insights into optimizing dual adhesive joints. Aiming to improve the strength of adhesive T-joints, in this work the DAJ technique is studied. To test this method in T-joints, various combinations of adhesives are evaluated. Various relative lengths between adhesives along the bondline are tested, always keeping joint symmetry. To evaluate these combinations, a CZM study was carried out. Initial validation of the CZM approach was undertaken with experiments.