13th International Conference on Fracture June 16–21, 2013, Beijing, China -3- iii) by finally introducing a seed crack (~2mm-long) with a razor blade. The test is conducted at room temperature (20 °C), under displacement control at 1 µm/s using a testing machine of 1 kN capacity (Instron 5882). The specimens exhibited a linear load–displacement diagram prior to fracture, confirming the predominantly linear elastic behavior of the material (Figure 1c). During the test, the surface of the sample is observed, with a CCD camera, (pixeLINK®, definition: 2500 x 1600 pixels, digitization: 8 bits) operating at 0.1 Hz and equipped with a telecentric objectif (GO Edmund, Techspec® gold series, Max distorsion: 0.35%, telecentricity: < 0.2°). The telecentric is used to minimize artifacts related to out-ofplane motions. The physical pixel size corresponds to 44.2 µm. Prior to the experiment, surface of the sample is painted in white and speckled with black paint. Two halogen lights are used for illuminating. For a realistic simulation, we need to obtain the typical wedge displacement-crack length curve from mechanical test and the crack path trajectory. These informations are directly measured from the location of crack tip via the camera using Digital Image Correlation (Figure 2a and b). a) b) Figure 2. a) Displacement-crack length curve from PMMA fracture test, b) Crack path trajectory determined by DIC. 3. Finite element analysis (Abaqus) Two-dimensional linear elastic fracture mechanics is retained to model the Wedge Splitting Test (WST) using a finite element method (Abaqus 6.6 software). To be in agreement with the experimental conditions, the choice of plane-stress state conditions was assumed. The computation of the stress intensity factor (K) and T-stress based on the domain integral “J” is carried out using six contours [10,12]. Apart from the first contour, the J-integral is path independent for the remaining 5 contours surrounding the crack tip. The material model was linear elastic, with the same properties for the experimental tests. The loading wedge is modeled as rigid bodies. The specimen is loaded by applying a displacement to the wedge in the vertical direction using an experimental displacement versus crack length data in the loading boundary conditions and taking under consideration the crack path trajectory (see Figure 2a and b. All other motions of the wedge are restrained. Surface-to-surface contact with a finite-sliding formulation is defined between the wedge and the rolling specimens, on the one side, and the specimen and the support, on the other. We assume that the contact is frictionless. Two analysis steps are used. In the first step, contact is established between the wedge and the rolling specimen by applying a small displacement (1×10-3 mm) in the vertical direction. In the second step controlled displacement loading of the wedge is applied. The virtual crack extension direction is specified with the q-vector. In the present model, it is defined with the starting point at the crack tip and the end point at the red dot, as shown in Figure 3.
RkJQdWJsaXNoZXIy MjM0NDE=