Issue 47

L. Marsavina et al., Frattura ed Integrità Strutturale, 47 (2019) 266-276; DOI: 10.3221/IGF-ESIS.47.20 267 properly designed [2]. However, the deformation and fracture of such elements create malfunctions of structures made of PB. Different studies on bending properties of PB were published. In [3] design requirements for PB and Medium Density Fiberboard (MDF) plates under different loading conditions are presented. The performance of PB beams under four point bending are presented in [4]. The effect of different types of coatings on the strength and stiffness of PB are investigated in [5]. Kulman et al. [6] studied the effect of density and temperature on modulus of rupture and modulus of elasticity of PB and MDF. The fracture behavior of wood and its composites is reviewed by Stanzl-Tschegg and Navi [7]. However, only a few studies investigates the fracture toughness of PB [8-10]. The same like in the case of MDF, different values of fracture toughness were obtained: Matsumoto and Nairn [11] 2.57 MPa·m 1/2 for density of 609 kg/m 3 , respectively 3.77 MPa·m 1/2 for density of 769 kg/m 3 using Compact Tension (CT) specimens, while for wedge splitting specimens and a density of 710 kg/m 3 Niemz et al. [12] obtained 1.81 MPa·m 1/2 . Fewer investigations were carried out on mixed mode fracture toughness of PB and MDF, [13]. Today, several fracture approaches such as the Stress Intensity Factor (SIF) [14-16], the Crack Relative Displacement Factor (CRDF) [17-20] or the energy release rate [21-24] allow expressing fracture criteria. It should also be noted that usually the damage level could be evaluated from a local approach based on the mechanical fields assigned by the crack tip singularity or by a global approach using the mechanical fields far to the crack tip singularity. Starting from this analysis, in the present study, a formalism based on the SIF and the CRDF was applied to evaluate the fracture process. As will be shown latter the CRDF allows definition of the kinematic state around to the crack tip. As defined by Dubois et al. [17, 18], Pop et al. [19] and Jamaaoui et al. [25], the crack opening state represents the relative displacement between two points positioned on the upper and the lower crack flanks. Its evaluation can be performed directly from the experimental measurements. Associated more often to full fields techniques, the optical methods can be easily applied to observe and to analyze the fracture process. Today, several optical techniques and methods are developed in order to measure the different fracture properties. Among these methods, we remind here: interferometry, stereo correlation, moiré, photoelasticity, Digital Image Correlation (DIC) or mark- tracking methods [24, 26-32]. Nevertheless, their application to analyze the fracture process depends on the observation scale and the environmental boundary conditions (i.e. laboratory or in-situ). Concerning the characterization of mechanical and fracture properties of PB the DIC and the mark-tracking methods seem to be the better. For this purpose, the Crack Opening Displacement was measured by means the DIC. Associated with optical full field methods the DIC allows measure of the bi-dimensional displacement and strain fields. The interest of this method lies in its possibility to perform the measurements without contact. Moreover, the studied zone, sometimes called the zone of interest, can be easily adapted to the analyze scale (i.e. local or global). Today several algorithms to perform the DIC in order to evaluate the fracture parameters are proposed [29-37]. In the present study, the analysis was performed using Correla software’s, developed by PEM team of Pprim Institut of Poitiers [38-39]. The present paper presents the original results, obtained for two different densities and thicknesses of PB, for modulus of rupture, modulus of elasticity, the fracture toughness in mode I and predominantly mode II and the crack relative intensity factors. E XPERIMENTAL DETERMINATION OF MECHANICAL AND FRACTURE PROPERTIES Materials ests were carried out on medium density PB with thicknesses of 16 and 25 mm. The density was determined on each specimen resulting a mean density of 600 (±12) kg/m 3 for the PB with 16 mm thickness, respectively 587 (±15) kg/m 3 for the PB with 25 mm thickness. The specimens before testing were conditioned at 22±2°C room temperature and 65±5% relative air humidity. Bending tests The tests were performed on a Zwick Roell Z005 electromechanical universal testing machine under displacement control by setting the machine to 10 mm/min. During the test, the force versus deflection was measured by means of linear variable differential transformer (LVDT) position sensor (-/+ 0.01mm) and a load cell of 5 kN (±5%). Rectangular specimens, Fig. 1, were adopted for the Three Point Bending tests, with dimensions B (height)  B (width)  L (length). For 16 mm thickness the dimensions were B =16 mm, L =250 mm, and the span (distance between supports) S =192 mm respectively for 25 mm thickness B =25 mm, L =250 mm, and S =192 mm. The test program consisted of four test series (two different thicknesses of PB plates of 16 and 25 mm, respectively two orientations 1 and 2) with five tests in each T