PSI - Issue 28

Sergey Smirnov et al. / Procedia Structural Integrity 28 (2020) 234–238 Author name / Structural Integrity Procedia 00 (2019) 000–000

235

2

testing of a cylindrical specimen by a load uniformly distributed along the generatrix. Such tests are in the scientific literature referred to as Brazil disk, Brazil-nut sandwich or Brazilian test. In the normative documentation (GOSTs, office operating instructions, etc.), the Brazilian test is often referred to as the splitting test; the main requirements to the experiment are regulated by ASTM D3967–95a (Molotnikov et al., 2014 and Zajcev et al., 2016). One of the modifications of this method is the use of an interlayer along the diametral plane and estimation of the processes occurring therein (figure 1, a) in works Wang et al., (1990), Volinsky et al., (2002), Borroso et al., (2012) and Vicentini et al., (2012). This paper proposes to consider this method as an experimental method for estimating the effect of the stress state on the adhesive strength of a glued sandwich. A cylindrical specimen with an interlayer of thickness h positioned along the meridional plane of symmetry is placed between flat parallel strikers, the specimen length being several times its diameter 2a . Through these strikers, the specimen is loaded to failure. The stress state in the joint is controlled by the tilt angle θ . The main advantage of this test is the fact that, by varying the tilt angle θ within one experiment, we can obtain both pure shear and pure cleavage, or a combination of both. The aim of this paper is to discuss the possibility of applying shear-compression testing to studying ultimate adhesive properties of a glued sandwich under a complex stress state. a b

Fig. 1. Brazil-nut specimens mounted on the test machine strikers: (a) load application diagram; (b) angle θ = 9.4°

2. Materials and methods The object of the study is the ETP-2 glue developed by the I. Ya. Postovsky Institute of Organic Synthesis, UB RAS (Suvorov et al., 2003). The ETP-2 glue solution was prepared in tetrahydrofuran. The paint-and-varnish compositions were solidified according to optimized conditions. The test specimens were glued-together half cylinders, 20 mm in diameter, made of the AMg6 aluminum magnesium alloy, with the interlayer lying on the symmetry plane. The view of the lateral surface of the specimens before testing is shown in figure 1, b. The area of the glue-metal boundary was 700 mm 2 , the glue layer thickness for each specimen is found in Table 1. The testing was performed at 25 °С with the use of an Instron 8801 universal test machine. The specimens were mounted on the machine strikers at different orientation angles of the plane of the glued sandwich relative to the direction of compression (figure 1, a). The loading was performed through the upper striker. Loading diagrams were continuously recorded during the testing. After the testing, the fracture surface of the glue layer was examined in a Wyko NT 1100 noncontact optical profilometer-profilograph. As is known, when θ = 0°, the destruction of a glued sandwich occurs by cleavage; therefore, in order to reveal the mechanisms of shear-induced adhesive failure under the effect of normal compressive stresses, the testing was performed at θ > 9°.

Table 1. Experimental data for the “AMg6 – ETP-2 glue – AMg6” glued sandwich obtained in shear-compression tests at +25 °С