PSI - Issue 39

Georg Schnalzger et al. / Procedia Structural Integrity 39 (2022) 313–326 Author name / Structural Integrity Procedia 00 (2019) 000–000

319

7

= ∆CTOD - ∆CTSD ∆ CTOD

(3)

with ΔCTOD being the cyclic crack tip opening displacement and ΔCTSD the cyclic crack tip sliding displacement. ΔCTOD and ΔCTSD are evaluated from the nodal displacements available in the FEA output. 3. Results This section divides into two parts. At first, the results from numerical simulations are presented. In the second part, the experimental results from FCG tests with undeformed and pre-deformed R260 are compared in terms of the crack propagation characteristics and macroscopic crack path. 3.1. Numerical predictions In a first step, the correlation between the crack inclination angle and the mode-mixity ratio is studied numerically to determine the Mode-II dominated region. Fig. 6 depicts the mode-mixity parameter P mix computed from the FEA for initial crack inclination angles β between 5 and 60 degrees. The mean shear stress Δ τ mean , initial crack length a 0 and stress ratio R amount 82 MPa, 0.8 mm and 0.3, respectively. Ten load cycles considering the elastic-plastic Chaboche model for undeformed R260 are simulated and the mode-mixity is evaluated from Eq.(3). Mode-II dominated crack propagation is predicted for inclination angles smaller than approximately 33 degrees. Consequently, cracks should not exceed this critical value in the experiment to ensure Mode-II dominated propagation conditions.

Fig. 6. Computed mode-mixity ratio P mix for different crack inclination angles β using elastic-plastic material behaviour of undeformed R260. The mean shear stress, initial crack length and stress ratio amount 82 MPa, 0.8 mm and 0.3, respectively.

A second parameter study investigates the effect of different multiaxial loading combinations on the crack propagation. The model assumes an idealized crack geometry with 0.8 mm initial length a 0 as observed typically after pre-fatigue. At first ten cycles of pre-fatigue considering a stress amplitude of 105 MPa at a stress ratio of 0.1 are simulated. Afterwards the model is loaded multiaxially for a further ten cycles to simulate the FCG test. Finally, the stress intensity factor range Δ K and crack growth direction is evaluated for the crack tip node based on the configurational forces using the procedure from Daves et al. (2019). Fig. 7 (a) presen ts the computed Δ K values and Fig. 7 (b) the crack growth angle determined for the maximum computed stress intensity factor K max . The shear stress amplitude amounts between 45 and 180 MPa at a stress ratio R of 0.3. Additionally, three different static axial stress levels of -70, 0 and 70 MPa are superimposed. The results from these numerical simulations allow a first estimation of optimum loading conditions for the FCG experiment to measure Mode-II dominated crack propagation data.