PSI - Issue 42

Andrea Zanichelli et al. / Procedia Structural Integrity 42 (2022) 118–124 Author name / Structural Integrity Procedia 00 (2022) 000 – 000

121

4

The fracture tests have been performed according to the Modified Two-Parameter Model. In particular, during the test the specimen is loaded monotonically up to the peak load, under crack mouth opening displacement (CMOD) control. Then, the post-peak stage follows and, when the force is equal to about 95% of the peak load, the specimen is fully unloaded in order to determine the unloading compliance. Finally, the specimen is re-loaded up to failure. From the flexural tests, a mean value of the peak load equal to 5.30 kN (with a standard deviation of 0.94 kN) has been obtained, and the corresponding mean flexural strength was equal to 1.60 MPa (with a standard deviation of 0.25 MPa). Moreover, from the fracture tests, a mean value of the peak load equal to 0.397 kN (with a standard deviation of 0.050 kN) has been obtained, together with the mean values of both elastic modulus and fracture toughness equal to 6919 MPa (with a standard deviation of 213 MPa) and 0.261 MPa . m 0.5 (with a standard deviation of 0.075 MPa . m 0.5 ), respectively. Note that, only five specimens have been considered since one specimen was not characterised by an ‘acceptable’ failure (i.e., the crack started far from the notch tip). 4. Numerical simulations and discussion In the present work, the numerical model proposed by some of the present authors and summarized in Section 2 is applied to simulate the experimental tests carried out on the shot-earth 772 (see Section 3). The finite element analyses of the prismatic specimens subjected to three-point bending loading are performed by using the commercial finite element software Strauss for the mesh generation. Note that, the averaged geometrical sizes of the specimens are employed for the bi-dimensional model of both flexural and fracture tests. As far as the flexural test simulation is concerned, the mesh discretisation is shown in Figure 2 (a) and it is characterised by a symmetric distribution of 995 four-node plate elements. Three values are considered for the ultimate tensile strength: the experimental mean value and both an upper and lower value, obtained respectively by adding and subtracting from the mean value its standard deviation. Accordingly, three numerical load-deflection curves, named lower, mean and upper curves, have been obtained, which are characterized by three different values of peak load, that is, 5.08 kN, 5.84 kN, and 6.80 kN, respectively. By comparing the numerical results in terms of peak load with the corresponding experimental value, it can be observed that a satisfactory agreement has been obtained, since both the mean and the lower numerical values fall within the experimental scatter band.

(a)

(b)

Fig. 2. Mesh discretisation for: (a) flexural test simulations; (b) fracture test simulations.