PSI - Issue 41

J.E.S.M. Silva et al. / Procedia Structural Integrity 41 (2022) 36–47 Silva et al. / Structural Integrity Procedia 00 (2019) 000 – 000

38

3

analytical method; the proposed mathematical representation consists of linear and exponential functions employed to model the elastic-plastic behaviour, usually observed in structural adhesives. Regarding the use of FEM models, Labbé and Drouet (2012) studied the influence of geometrical parameters aiming to optimise tubular lap joints subjected to axial loads concluding that the ideal adhesive TLJ configurations are characterised by larger adhesive bond lengths, a thin adhesive layer, and a large adherend inner diameter. The torsional behaviour of tubular steel adhesive joints was studied using the Ramberg-Osgood plasticity model, obtaining a good agreement for various overlap lengths (Hosseinzadeh et al., 2007). Ferreira et al. (2019) studied geometrical and material optimization of tensile loaded TLJ using CZM, concluding that CZM is a valid tool for predicting joint strength in TLJs; also, the joint behaviour is influenced by the geometry and material properties of the adhesive. Later, Kaiser and Tan (2020) studied TLJs with Carbon Fiber Reinforced Polymer (CFRP) and Titanium substrates, subject to a traction load, through CZM and FEM. The analysis was parametric and evaluated the effect of adhesive layer thickness, a bi adhesive bond-line, and spew angles on joint strength. The findings agree with previous research on SLJs. Furthermore, leaving a 45° spew in the adhesive layer together and a bi-adhesive bond-line yielded the best performance. From the descriptions above, it can be observed that extensive research is available for TLJs but not much about TSJs, which are expected to have higher strength than TLJs. This study aims to compare the tensile performance of TSJs with aluminium (AW6082-T651) adherends, considering the variation of the scarf angle (  ) from 45° to 3.43°. First, the cohesive zone model (CZM) technique and respective cohesive parameters were validated through the comparison of experimental tests and numerical analysis of TLJs. A CZM analysis was then performed on the TSJ to analyse peel ( σ y ) and shear stresses ( τ xy ) in the adhesive layer. Damage, joint strength ( P m ), and dissipated energy ( U ) analyses were carried out for all the studied joint configurations.

2. Experimental work 2.1. Geometry and dimensions

In this section, all the geometrical parameters and dimensions used in the experimental and numerical works are described. Fig. 1 illustrates the tubular overlap joint (TLJ) geometry used in the experimental work and to validate the numerical methodology followed in this work. The adhesive thickness ( t A ) is the same for all samples (0.20 mm), as well as the thickness of the inner ( t SI ) and outer ( t SE ) tubes (2 mm). The same applies to the external tube diameter: inner tube ( d SI ) of 20 mm and outer tube ( d SE ) of 22.40 mm. Two overlap lengths ( L O ) were considered: 20 and 40 mm. The adherends’ length ( L S ) is variable depending on L O to result in a total length up to the gripping fixtures of the testing machine ( L T ) of 80 mm. A total of ten samples were manufactured, being five samples for each tested L O .

Fig. 1 – Schematic representation of the TLJ.

The geometry of the TSJs studied in the numerical work is presented in Fig. 2. The value of t A is the same for all samples (0.20 mm) and the tube thickness ( t S ) is equal to 2 mm. The external tubes’ diameter ( d S ) is 20 mm, and L S is 80 mm. The geometric parameter subject to variation was the scarf angle (  ), which assumed values of 45°, 30°,