PSI - Issue 51

R.B.P. Barros et al. / Procedia Structural Integrity 51 (2023) 17–23 R.B.P. Barros et al. / Structural Integrity Procedia 00 (2022) 000–000

18 2

al. 2011). Park and Dillard (2007) proposed the asymmetric tapered double-cantilever beam (ATDCB) test, which has one DCB adherend bonded to a TDCB adherend. The single-leg bending (SLB) test was addressed in the work of Yoon and Hong (1990). This test has similarities to the ENF test, but one adherend has smaller length, giving rise to both modes I and II at the crack tip. Reeder and Crews (1990) proposed the mixed-mode bending (MMB) test for mixed-mode analysis of composite materials. This test manages to vary  by changing the load ( P ) application site in the testing jig, always using the same specimen. Brussat et al. (1977) is the responsible for creating the CLS test. The associated test specimen is composed of two adherends bonded together, but one being shorter. Tensile loading is applied only to the lengthier adherend. This test promotes constant G C owing to the adherends’ high length/thickness ratio. On the other hand, loading parallel to the specimen’s length gives rise to mode II component, which adds to adherend mode I separation due to P eccentricity, creating a mixed-mode stress state at the crack tip. Due to these test characteristics, flexure of the adherends should be considered in analytical formulations and numerical modelling. The CLS test was analysed by different researchers (Datla et al. 2010). Tong and Luo (2008) compared several analytical techniques applicable to the CLS specimen, regarding stresses and failure prediction. The fully-coupled nonlinear formulation gave the best match to the finite element method (FEM). This work experimentally evaluates the CLS test for mixed-mode G C assessment of two epoxy adhesives. Different theoretical reduction methods were evaluated. The work consisted of testing CLS specimens with the different adhesives, and carrying out the respective data processing. 2. Experimental work 2.1. Joint and material description The CLS adhesive joint geometry was experimentally addressed for the mixed-mode analysis of two structural adhesives. Fig. 1 shows the joint geometry, in which a tensile displacement (  ) is applied at the edge of the lengthier adherend. The specimen includes a pre-crack ( a 0 ) equal to 50 mm.

Fig. 1. CLS adhesive joint (all dimensions are in mm).

The following dimensions were considered, together with the width ( b =15 mm): L BA – length of the lower adherend, L OA – length of the upper adherend, e 1 – thickness of the lower adherend, e 2 – thickness of the upper adherend, and t – adhesive thickness. The carbon-epoxy adherends are composed of stacked unidirectional pre-preg tape (with 0.15 mm ply unit thickness) from Seal ® (Texipreg HS 160 RM; Legnano, Italy). Elastic orthotropic characterization is given in a former work (Ribeiro et al. 2016). Two adhesives from Araldite ® were tested: brittle Araldite ® AV138 and ductile Araldite ® 2015. Table 1 identifies the measured properties in a previous work (Campilho et al. 2011). 2.2. Experimental and testing details Specimen fabricated initiated by manufacturing a prepreg plate from 16 layers with dimensions of 300×300 mm 2 , to be piled by hand with the same orientation. The plate at 130ºC and 4 bar in a hot-plates equipment during 60 minutes, resulting in 2.4 mm thick plate. The plate was cut into specimens, the surfaces receiving the adhesive were abraded with sandpaper, followed by duly acetone cleaning before bonding. Calibrated shims were used to assure the desired t . Adhesive application was accomplished in the lower adherend, followed by application of the upper adherend and joint assembly. The set was held by grips during one week at room temperature (RT) up to complete curing. A Shimadzu machine with electromechanical drive and 100 kN cell was employed for the tests, which were conducted at 20ºC. The testing speed depended on the adhesive: either 0.25 mm/min (AV138) or 0.5 mm/min (2015).