PSI - Issue 21

Mehmet F. Yaren et al. / Procedia Structural Integrity 21 (2019) 31–37 Yaren M. F. et al / Structural Integrity Procedia 00 (2019) 000 – 000

35

5

A piecewise function is given in Eq. 9 to determine the effective stress intensity factor range for any cycle. It’s noted that because of mathematical error Willenborg model cannot be used if the overload ratio is greater than two. To solve this problem, Gallagher J.P. proposed a modification multiplying K r with parameter Ø, which is a function of shut-off ratio. The Generalized Modified Willenborg model was obtained by adding underload capability to the Generalized Willenborg model; Gallegher (1974). Calculation of Ø is defined as a piecewise function in Modified Generalized Willenborg model. Another crack growth model is proposed by Forman and Mettu (1990). Some other studies in the literature by Sander and Richard (2006) used Forman-Mettu model in the Willenborg model and obtained good results. This model requires more material constants than Paris-Erdogan or Forman models. 3. Experimental Procedure In this section, details of crack growth experiments are explained. Tests are performed on an MTS 809 axial torsional fatigue test machine at Sakarya University by using standard compact tension (CT) specimens. The distance between the center of pin hole and back of the specimen is 50 mm and thickness is 25 mm. Other dimensions of the CT specimen are determined according to ASTM E-647 (2016). The specimens are machined from rolled aluminum 7075-T651 plates in the L-T rolling direction.

Fig. 2. Experimental setup.

Fatigue crack growth tests are done under constant amplitude mode-I loading to obtain material constants in Paris Erdogan or Forman Equations. Half-milimetric scales are pasted on both sides of the specimen and crack growths on both sides are observed with high-zoom cameras during the tests. Fatigue pre-cracking is performed such that the crack length is 25 mm according to the view from the high zoom cameras and then fatigue crack growth tests are started. The views of crack growth on both sides of the specimen at certain cycle intervals are recorded simultaneously. After the tests, crack sizes from the recorded pictures are measured by pixel sizing method and crack growth rate curves are plotted. The experimental setup can be seen in Fig. 2. Single overloads are applied to investigate retardation effects. During the test, the loading process is applied as follows. Initially, 7 kN load is applied for 7000 cycles. Then just one cycle overload is applied and the test is continued with 7 kN load until the specimen is broken. Different magnitudes of overloads (9-11-12-13-14 kN) are applied for different tests.