Fracture mechanisms of hybrid adhesive bonded joints: Effects of the stiffness of constituents

https://doi.org/10.1016/j.ijadhadh.2020.102649Get rights and content

Abstract

In this study, different single-lap hybrid joints are used to analyse the effects of the stiffness of the adherends and the adhesive on the failure mechanism. The hybrid joints include a combination of (a) different adherends: aluminium (6082 T6) and PolyPhtalamide (PPA) reinforced with 50% of glass fibre (grade HTV-5H1 from Grivory) and (b) different adhesives: epoxy-based adhesive (Loctite EA 9497) and silane-modified polymer-based adhesive (Teroson MS 9399). Six different single-lap joints are fabricated and analysed. The cohesive parameters of different adhesives against different adherends are determined respectively using single-mode coupons and validated with finite element modelling. Single-lap shear tests are conducted to understand different fracture mechanisms of the joints. Finite element (FE) models using the Cohesive Zone Method (CZM) are developed to simulate the failure of the joints and validated by the testing results. Different failure processes obtained from different hybrid joints combinations are discussed further by analysing the stress distributions along the interfaces of the joints. Finally, the relationship between the stiffness of the constituents of a hybrid adhesive joint and its failure mechanism is summarised. The load vs displacement behaviour of the single-lap joints demonstrate that the stiffness of adherends affects the maximum failure load of the joints with rigid adhesive (epoxy). However, the joint with flexible adhesive (polyurethane) is not sensitive to the stiffness of the adherends. In addition, higher shear stress distribution occurs in the interface adjacent to the adherend with lower stiffness, leading to the failure initiation at the PPA side regardless of adhesive types.

Introduction

With the rapid development of new engineering materials, multi-material structures are now widely used to achieve the desired performance. Consequently, the use of adhesive joining techniques is increasing due to their advantages over traditional joints, such as easy manufacturing, more uniform stress distribution, and the possibility of joining dissimilar adherends. However, there are still some barriers in using adhesive joining techniques in practice due to a lack of an accepted theory, which describes the fracture mechanism of the hybrid joints and summarises the factors affecting the performance of the joints. As the hybrid joint involves the combination of two different adherends with different mechanical properties, leading to a more complicated fracture mechanism in practice, for instance, mixed-mode failure (crack may be initiated from the interface and grow into the adhesive layer, or vice versa).

In recent years, several experimental works have been conducted that explore the factors affecting the strength of adhesive joints, such as the type of adhesives, the materials of adherends and the joint configurations (overlap length, adherend and adhesive thickness). Wu et al. [1] corrected Goland and Reissner’s solution by modifying their classical equation for analysing the adhesive layer in dissimilar adherends with different thicknesses and lengths. Sawa et al. [2] analysed the single-lap joint of dissimilar adherends (aluminium bonded to mild steel) under a tensile shear loading. Their results show that the stress singularity increases at the free edges of the interface with lower stiffness, and the thinner adherend. Pinto et al. [3] evaluated the tensile strength of single-lap joints with different adherends (polypropylene (PP), polyethene (PE), carbon-epoxy, and glass-polyester composites). They found that increasing the adherends’ stiffness reduces the joint bending and diminishes stress at the overlap edges and, consequently, increases the joint strength. Reis et al. [4] studied the influence of the adherend’s stiffness on the shear strength of the single-lap adhesive joint by using three different adherends (laminated composite, high elastic limit steel, and the 6082-T6 aluminium alloy). Their studies concluded that the effects of the overlap length on the shear strength depend on the stiffness of the adherends. Pereira et al. [5] showed that the increase in the thickness of the adherend decreases the rotation angle of the joint and the peak plastic strain. Da Silva et al. [6] and Nunes et al. [7] studied the influences of the adhesive type (epoxy and ductile adhesives) and thickness of the bond-line on the single-lap joint strength. It can be concluded that the shear strength of SLJ increases by decreasing the adhesive thickness or increasing the adhesive toughness.

Cohesive Zone Modelling (CZM) has been widely used in the simulation as it allows multiple failure paths in the middle of the adhesive or along the interface to predict failure. There are various techniques (direct and indirect methods) to obtain CZM parameters (tn, GIC, ts, GIIC) by using double cantilever beam (DCB), end notch flexure (ENF) and single-lap joint (SLJ) tests. Zhu et al. [8] used the direct method (J-integral) to obtain the traction-separation laws of both mode 1 and mode 2 with sandwich specimens for polyurea/steel interfaces. Their results show that the traction values in both cases depend on the loading rates. An increase in the loading rate increases the cohesive peak stress, while the critical opening displacement decreases. Ruadwska [9] and Alves et al. [10] analysed the tensile strength of the bonded joint between similar and dissimilar material by using an indirect method considering both experimental and CZM approaches for fracture predictions. Katsivalis et al. [11] noted that the validated cohesive parameters depend on several factors, including the bond-line thickness (tA), the adherends’ stiffness and surface chemistry. Wang and Qiao [12] compared shear-mode (model II) fracture toughness of the wood-wood and wood-FRP by using tapered end-notched flexure (TENF) specimens. Their results show that the fracture toughness of the wood-FRP interface is lower than the value of wood-wood bonded interfaces. Tvergaard et al. [13] noted that the interface roughness and crack growth along the bond-line of the dissimilar joints under a mixed-mode loading condition strongly depend on the elastic modulus ratio E1/E2 of adherends.

Most of the previous numerical works used a single layer of the cohesive element in the bond-line to simulate the adhesive layer, which is accurate enough for identical adherend joints. Nonetheless, the method cannot describe the failure process for the hybrid joint and estimate the strength of the joint accurately. Since the change of the adherend changes the interaction between adhesive and adherend due to roughness and chemical links [14].

The objectives of this work are to predict joints strength and analyse stress distributions along bond-lines, and to understand the failure mechanisms of the single-lap joints geometry with dissimilar adherends by comparing to the performances of identical single-lap joints. Finite element models were developed to predict the strength of the hybrid joints by considering the effects of their adherend stiffness. Experimental works on the six different kinds of single-lap joints were conducted, which consist of three categories of adherend combinations (AL bonded to AL, polyphthalamide (PPA) bonded to PPA, and AL bonded to PPA) using two kinds of adhesives (Loctite EA 9497 epoxy adhesive and Terson MS 9399 polyurethane adhesive), to understand their failure performances as well as to validate the FE models. The innovation of the FE models is to use two layers of cohesive elements along the different interfaces between the adhesive bulk and the adherends with different cohesive properties measured from single-mode coupons using the relevant adherends, respectively. This method is approved to provide a more concise strength prediction regarding the hybrid joint combinations. Stress distribution analysis, stiffness degradation analysis, as well as failure surface observations, were also carried out to obtain a better understanding of the failure mechanism of the hybrid joints.

Section snippets

Material properties of adherends and adhesives

The adherends used in this study were aluminium alloy 6082 T6 (AL) and polyphthalamide (PPA). The PPA material, commercially named Grivory HTV-5H1 black 9205, is a glass fibre (50%) reinforced engineering thermoplastic material based on a semi-crystalline, partially aromatic polyamide. Tensile tests were carried out for both AL and PPA materials based on the ISO EN 485–2:2016 to characterise their mechanical properties, as shown in Table 1. The Young’s modulus and elongation at fracture of the

Cohesive parameters

Cohesive zone model (CZM) laws are based on a relationship between cohesive forces, and displacement jumps along the material surface, and it is one of the most commonly used methods that allows simulating the degradation and eventual failure of the adhesive bond-line. The adhesive bond-line behaves elastically until contact stress reached the nominal traction stress (t consists of two components (tn and ts) in two-dimensional in normal and shear directions respectively). The elastic behaviour

Load vs displacement of single-lap joint

Five SLJ specimens of each design category were tested under tensile load, and three representative results were presented in the below figures. Fig. 6, Fig. 7 present comparisons between experimental and numerical results for the joints with epoxy and polyurethane adhesives, respectively. In general, there are good agreements between experimental and numerical results.

The results show that the maximum failure load in samples with epoxy adhesive is more sensitive to the stiffness of adherends

Conclusion

In this work, the effects of the stiffness of the constituents of an adhesive joint on its fracture mechanism were studied. Joints with different combinations of adherends and adhesives were analysed using both numerical and experimental methods. According to the analysis of the results, the following conclusions could be summarised:

  • A novel FE model is developed to describe the mechanical performance of the adhesive joint by introducing two layers of the cohesive element at the individual

Cited by (0)

View full text