Issue34

M. Ševčík et al, Frattura ed Integrità Strutturale, 34 (2015) 216-225; DOI: 10.3221/IGF-ESIS.34.23 217 1 mm in the civil engineering application the adhesive layer thickness can reach up to a few centimeters. This nonnegligible adhesive thickness behaves like additional layer in the adhesive joint and should be taken into account when designing the joint against failure. The main failure mode of the composite materials is delamination of the layers under bending or tension loading and buckling under compression loading [3]. If the crack propagates in the material interface of different materials the crack tip stress field is inherently under mixed-mode behavior [4]. This in fact brings some difficulties for the fracture mechanics analysis of the fiber reinforced polymer materials. The delamination behavior of adhesively bonded GFRP joints is usually studied by variety of specimens - Double- Cantilever-Beam (DCB) specimen for pure mode I, End-Loaded-Split (ELS) or End Notched-Flexure (ENF) specimen for pure mode II and Mixed-Mode Bending (MMB) for mixed modes tests. Nowadays, the MMB test is one of the standardized test method for evaluation of the interlaminar fracture toughness of unidirectional fiber reinforced polymer matrix composites [5]. The concept of the strain energy release rate (SERR) is commonly used for the fracture mechanics description of the failure behavior of adhesively bonded GFRP laminates [6,7]. There exist various expressions for the calculation of the total strain energy release rate for the symmetric MMB specimens but mostly valid only for symmetrical specimens [8]. The asymmetrical specimens are, however, often accompanied by relatively thick adhesive layer or asymmetry of the stacking sequence either in the laminate or in the adhesive joint and therefore there is a need for the development of the new procedures for the calculation of the strain energy release rate components. Another way is to describe the delamination behavior using the complex stress intensity factor introduced in [4,9]. In this work the concept of the strain energy release rate as a fracture parameter is utilized. For the accurate description of the crack behavior in asymmetric joints it is necessary to decompose the mode components and explicitly estimate their contribution to the total fracture energy [10–12]. Various procedures exist in the literature for the mode separation. Most of them are based on the differences between moments [6], displacements [13] or bending stiffnesses [14] of the two arms of the joint. However, the differences between moments or displacements are applicable only for symmetric specimens. The mode separation based on the difference in bending stiffness of the two arms is applicable for the asymmetric specimens, nevertheless, the bending stiffnesses are usually measured on the cracked specimen and it is therefore difficult to predict the behavior of the joint before the test. This paper presents new easily applicable relationship for the calculation of the strain energy release rate of the asymmetric MMB specimens that is proposed based on the beam theory and its ability is validated through comparison with experimental results. M ATERIALS DESRIPTION OF THE MMB SPECIMEN he pultruded GFRP laminates, supplied by Fiberline A/S, Denmark, consisted of E-glass fibers and isophthalic polyester resin. The laminates with 6 mm in thickness consisted of a thin polyester veil in the outer surfaces, two combined mat layers on each side and a roving layer in the middle. Each mat layer comprised of 0/90 woven fabric stitched to a chopped strand material. The fiber architecture of the laminate is schematically shown in Fig. 1. Based on burn-off test, according to ASTM D3171-11 [15] and fiber density of 2560 kg/m3, the fiber content was 43.3 vol. %. A two-component epoxy adhesive system Sikadur 330, Sika AG, Switzerland, was used for the bonding of the GFRP Figure 1 : Scheme and micrograph of the laminate architecture. Figure 2 : Scheme of the mixed-mode bending test. T