PSI - Issue 65

Sergey Smirnov et al. / Procedia Structural Integrity 65 (2024) 263–268 Sergey Smirnov, Irina Veretennikova, Dmitry Vichuzhanin / Structural Integrity Procedia 00 (2024) 000–000

264

2

et.al. (2022), and Povolny et.al. (2022). However, there are modifications of the conventional technique and attempts to widen its applicability. Particularly, Smirnov et.al. (2014) used the Brazilian test to study the strength of the interface in a metal composite produced by explosive welding and Smirnov et.al. (2024) studied the effect of the failure mode of a metal-polymer composite as dependent on the proportion of normal and tangential stresses. The advantage of applying this technique and enlarging the applications to study interfaces between dissimilar materials (besides the relatively simple experimental procedure) is that various combinations of the values of shearing and compressive normal stresses can be achieved by selecting the interface tilt angle α relative to the direction of loading (Fig. 1). In the Brazilian test, as applied to metal-polymer materials, the failure of the adhesive joint occurs due to shear under conditions of acting compressive stresses when α > 0. A cylindrical specimen with an adhesive layer of thickness h , placed in the meridional plane of symmetry, is between parallel flat dies. The dies compress the specimen, thus loading it to failure. The stress state of the interlayer is determined by the value of the tilt angle α. The aim of this study is to perform the Brazilian test to make a comparative assessment of the adhesive strength of adhesive bonding between aluminum alloy surfaces at different test temperatures.

Fig. 1. An outline of the transverse upsetting test

2. Materials and experimental procedure

The specimens were made from a hot-pressed AlMg5 aluminum-magnesium alloy bar, 20 mm in diameter. They were joined in the shared research facilities center at the I. Ya. Postovsky Institute of Organic Synthesis UB RAS (Ekaterinburg) The ED-20 epoxy 4,4’-isopropylidenediphenol resin (Sverdlov Plant, Russia), with an epoxy number of 21.1%, and various curing agents, or hardeners (Chimex Limited, Russia) were used, the composition and curing conditions being shown in Table 1. The polyamidepolyamine curing agent (PAPA) is an acylated derivative of the product of the oligomerization of ethylenimine with an average molecular weight of 200 g/mole by higher fatty acids. The diethylenetriamine curing agent (DETA) is an aliphatic polyamine hardener. Coring was performed at 25 °C for 24 h. The thickness of the epoxy layer in each composite is 0.2 ± 0.02 mm. The application of the epoxy layer is schematically shown in Fig. 2. The test was performed on an Instron 8801 universal testing machine (Instron, UK) with a maximum load of 100 kN. The experimental procedure is schematically shown in Fig. 2 b . As the specimens were placed between the flat dies of the test machine, the angle α between the direction of compression and the plane of the adhesive joint was varied between 0 and 20°. A series of tests was performed at 50, +25 and +50 °C, this being the range of climatic temperatures on the Eurasian continent. A climate chamber from the Instron 8801 equipment set was used for this purpose. Cooling in the chamber was performed by liquid nitrogen vapors. The temperature was controlled automatically with an accuracy of ±2° by means of the standard software of the machine on the basis of the measurement data from chromel-alumel thermocouples. In Zhang et. al. (2017), Wu et. al. (2018) it was discussed how the rate of loading may affect the propagation of an interface delamination crack, and some recommendations on selecting the rate were given. Accordingly, the rate of loading was selected to be 0.000025 m/s. The loading was performed up to the moment of bond failure, which was recorded by means of the StrainMaster system of video