13th International Conference on Fracture June 16–21, 2013, Beijing, China -1- Experimental and numerical analysis of mixed mode fracture Octavian POP1,*, Mamadou Méité1, Frédéric Dubois1, Joseph Absi2 1 GEMH, Université de Limoges, Centre Universitaire de Génie Civil, 19300 Egletons 2 GEMH, Université de Limoges, IUT du Limousin, Département GM, 87000 Limoges * Corresponding author: ion-octavian.pop@unilim.fr Abstract Based on the global approach and the experimental measurements by digital image correlation, the presented study proposes a coupling between these two approaches in order to evaluate energy release rate. The proposed formalism allows calculating the energy release rate fracture parameters without considering the elastic parameters of material. The experimental analysis is realized using the specimens made in PVC isotropic material and Douglas fir (orthotropic material) under different mixed mode loadings. The loading under displacement control is applied using the Arcan system. From experimental data optimized by an adjustment procedure, the kinematic state in the crack tip vicinity is evaluated through the crack opening relative displacement factors. In parallel the stress state in vicinity of the crack tip is evaluated by a numerical analysis. This analysis is performed using the finite elements method and the integral invariant Mtheta, in order to evaluate the stress intensity factors. The finite element analysis is based on the reproduction of experimental test in terms of specimen geometry, experimental boundaries conditions and loading configurations. Then the energy release rate can be estimated by coupling of these two factors calculated without considering the material elastic parameters. Moreover, this method allows defining the local mechanical behavior. Keywords Fracture mechanics, Mixed Mode, Experimental, Numerical, Digital Image Correlation 1. Introduction Cracked structures are most often subjected to complex loadings in mixed-mode configurations that can lead to a catastrophic collapse of the structure and modify their mechanical behavior. In this case in order to avoid structural behavior, it is necessary to evaluate fracture process parameters and local mechanical behavior. Within this field of study, several numerical investigations have been carried out in the literature for the purpose of characterizing crack tip parameters through use of the energy method for mixed-mode configurations [1-4], for isotropic and orthotropic media. At present, these efficient techniques require an explicit knowledge of material properties; for orthotropic cases in particular, the complete compliance tensor is needed. In this context, our study proposes a new formalism that allows uncoupling the fracture parameter identification relative to material elastic properties. Based on Digital Image Correlation (DIC) and a Finite Element Method (FEM), the fracture parameter identification can be performed from the kinematic and stress distributions in the crack tip vicinity. The complementarity of these two approaches distinguishes the calculation of energy release rates relative to opening and shear modes from the calculation of local material elastic proprieties. According to our approach, DIC is employed to measure the displacement field evolution for the specimens made from a rigid Polyvinyl Chloride polymer and Douglas fir loaded under mixed-mode configurations. The application of such an experimental technique enables capturing both strong and weak kinematic discontinuities in the crack tip vicinity so as to characterize Crack Relative Displacement Factor (CRDF) [5-9]. The iterative Newton-Raphson method is then coupled with DIC to provide not only the real crack tip position needed for an accurate CRDF determination, but also the comprehensive raw dataset optimized through Williams' asymptotic series expansion solution. The evaluation of stress distribution is realized using a finite element model. The model is based on experimental sample geometry and experimental boundary conditions. According to the Mq
RkJQdWJsaXNoZXIy MjM0NDE=