the local stress triaxiality near the crack tip and the plane-stress behavior remote from the crack. Threedimensional analyses with a flat, straight crack-front maintain a straightforward modeling approach, yet capture the complex constraint behavior missing from typical 2D analysis. The methodology is applied by finding Yc such that the analysis matches the average maximum load for coupon tests (typically, compact tension, W = 152 mm, specimen). This angle is used in subsequent analyses for predictions of crack extension and fracture. Figure 1 shows a schematic of a typical fracture surface. Initially, the material fails in tension, most likely due to micro-void coalescence. Tunneling occurs on the interior, and shear bands start on the surface and eventually join to form a single shear dominated fracture surface. Recently, a wide variety of fracture tests were conducted on 6.35 mm thick 2024-T351 aluminum alloy [4]. Both C(T) (compact tension) and M(T) (middle-crack tension) specimens were tested. Figure 2 shows selected load-crack extension results, including both test and analysis results. The 152 mm wide C(T) was used to find Yc, the critical value of CTOA, by matching the average maximum load for the tests. This value of Yc was used to predict the behavior of the other C(T) and M(T) configurations. The results show that the constant critical CTOA fracture criterion is transferable between C(T) specimens and M(T) specimens and show that the analysis was able to accurately predict the maximum load for C(T) specimens ranging in size from 50 mm to 152 mm and for M(T) specimens ranging in size from 75 mm to 1016 mm. A consistent observation made of these load-crack extension curves (and similar curves for other thin sheet and plate materials) is that the straight crack-front analysis generally over-predicts crack extension, before and after maximum load. A number of factors could be contributing to the discrepancies in crack extension, including crack-front tunneling, transition from initially flat fracture to slant fracture after maximum load, and the simplicity of the constant value of CTOA. CTOA is a local fracture criterion that is always measured a fixed distance from the current crack tip. An alternative fracture criterion is d5, which is the displacement measured across the original crack tip location using a 5 mm gauge length. d5 is initially a local parameter, but after crack extension behaves more like a remote parameter. Figures 3 (a) and (b) show load-d5 plots for the 152 mm wide C(T) and 1016 mm wide M(T), respectively. In contrast to the load-crack extension comparisons, shown in Figure 2, the analysis matches the d5 behavior of the tests very well. The d5 results compared well before and after maximum load, corresponding respectively to the local and remote stages of crack extension. The results in Figure 3 (b) are somewhat abbreviated because the clip gage measuring d5 went out of range near the end of the test. However, data was collected past maximum load. The results in Figure 3 strongly suggest that the flat, straight crack front in the 3D analysis represents the “average” crack length at any given load. Returning to Figure 2, the amount that the analysis over-predicts crack extension is on the order of the plate thickness. The crack extension was measured on the surface of the specimen during the test using a traveling-stage optical microscope. The surface measured value is the shortest crack length if the specimen is experiencing crack-front tunneling. A more appropriate comparison metric may be to use an average crack length measure, such as unloading compliance or an area average. TUNNELING TESTS Experiments and analyses were performed to characterize the crack tunneling for the 6.35 mm thick 2024T351 aluminum alloy. The typical approach to characterize crack tunneling is to perform a multiple specimen test to obtain one crack-front per specimen [2]. In an effort to obtain more data per specimen, a combined approach was taken here. Three specimens are single crack-front tests, while the final specimen was used as a multiple crack-front specimen test. A number of methods may be available to mark the crack fronts and measure the tunneling. Dye penetrants could be used while the crack is held open to take advantage of the capillary action at the crack. Radiographic measuring methodologies may be able to characterize the crack extension. Herein the crack front was marked with fatigue cycles at a high stress ratio (R = 0.75 or 0.8) and a relatively high load (80%
RkJQdWJsaXNoZXIy MjM0NDE=