13th International Conference on Fracture June 16–21, 2013, Beijing, China -9- The calibrated computational model is now employed to predict the crack extension behavior of M(T) and MSD specimens. Three M(T) specimens with W = 300 mm and a/W ratios of 0.33, 0.42 and 0.5 are analyzed. The element size and arrangement in the region near the crack front are kept the same as used in the C(T) specimen. The nominal remote stress, Rσ , characterizes the loading for these specimens. Figure 7(a) compares the computed load versus crack extension responses with experimental measurements, showing very good agreement for all three cases. Figure 7(b) compares the computed load versus crack extension responses with experimental measurements for a MSD specimen containing three cracks as shown in Figure 4(d). This specimen has the same width as the M(T) specimens. The center-crack length is 2a2 = 100 mm. The two lead cracks have the same length of a1 = 12.5 mm. The tip-to-tip distance between the lead crack and the center crack is b = 12.5 mm. The model prediction captures accurately the load versus crack extension curve. The cusp on the predicted load versus crack extension curve corresponds to the point when the lead crack and the center crack link up. Δa (mm) 0 10 20 30 40 50 σR (MPa) 0 50 100 150 200 250 Exp.(a/W=0.50) Exp.(a/W=0.42) Exp.(a/W=0.33) Num.(a/W=0.50) Num.(a/W=0.42) Num.(a/W=0.33) 0 10 20 30 40 0 40 80 120 160 Prediction Experiment Δa (mm) σR (MPa) Figure 7. Comparison of the model predicted load versus crack extension responses (lines) with experimental measurements (symbols): (a) M(T) specimens, (b) MSD specimen containing three cracks. 4. Concluding Remarks Based on the mechanism of ductile fracture in metallic alloys, this paper describes a method to predict crack growth in engineering structures. To model extensive crack extension, homogenized porous plasticity models need to be adopted to describe the material behavior in the fracture process zone and these models must be calibrated such that the material behavior is accurately captured. Unit cell analysis of the representative material volume reveals the strong effect of the stress state on the void growth and coalescence. Calibration of the porous plasticity models requires the predicted macroscopic stress-strain response and void growth behavior of the representative material volume to match the results obtained from detailed unit cell analysis. As an application, a numerical procedure is proposed to predict ductile crack growth in thin panels of a 2024-T3 aluminum alloy. The material specific GLD porous plasticity model is used to describe the void growth process and the failure criterion is calibrated using experimental data. The calibrated computational model is then applied to predict crack extension in fracture specimens having various initial crack configurations. The numerical predictions show good agreement with experimental (a) (b)
RkJQdWJsaXNoZXIy MjM0NDE=