Variability of large-crack fatigue-crack-growth thresholds in structural alloys

Abstract

The fatigue threshold of large cracks is known to show substantial variations due to microstructural variability in structural alloys. The ⌬K th variations are dependent on the stress ratio (R); they are extremely large at low R ratios, e.g., R Ͻ 0.5, but are drastically reduced at high R ratios (R Ͼ 0.8). The origins of these large variations due to intrinsic and extrinsic mechanisms are examined by theoretical analyses. First, an intrinsic fatigue crack growth (FCG) threshold model is developed for structural alloys by considering the cyclic slip process at the crack tip. Second, the effects of extrinsic mechanisms such as residual plastic stretch, crack deflection, fracture-surface roughness, and oxide wedging are considered both individually and concurrently in order to delineate their relative contributions to threshold variability. The theoretical results indicate that the intrinsic threshold depends on the elastic properties, magnitude of the Burgers vector, yield stress, and Taylor factor (i.e., texture), but is independent of the R ratio or the maximum applied stress intensity factor, K max . The large variability of ⌬K th at low R ratios and their corresponding dependence on K max appear to arise from various crack closure mechanisms. Applications of the threshold models to structural alloys show good agreement between theory and experimental data from the literature for steels, Ti, Al, Ni, Cu, Nb, and Mo alloys.

Development of a Physically-Based Methodology for Predicting Material Variability in Fatigue Crack Initiation and Growth Response

“…Comparison of model predictions against measured AKth for Ti-6AI-4V as a function of Kmax. From Chan [9].…”

“…Analytical expressions for computing individual U, terms, which are summarized in a recent paper [9], include Newman's formulation for treating plasticity-induced crack closure [10], Suresh's formulation for crack deflection and roughness-induced crack closure [ 11 ], as well as an oxide-wedging model.…”

“…Once intrinsic material properties were obtained, the FCG model was utilized to predict deterministic da/dN curves for various R ratios by applying the crack closure model and using average values of grain size, yield stress, and fatigue ductility coefficient. The crack closure model [9] includes plasticity [10], deflection-, and roughness-induced [11] closure mechanisms, but not oxide-induced crack closure. For crack growth in Ti-6AI-4V, grain size was the dominant microstructural variable, and was modeled as a lognormal random variable (mean = 11.7 pim, standard deviation = 3.2 Pn) based on grain size measurements shown in Figure 3.…”

“…with the intrinsic threshold, AKth,in given by [9] AK,• =.0 1 ME1 2 EbV2 (9) where b is the magnitude of the Burgers vector. The proposed model was applied to compute the AKth for Ti-6AI-4V for R values ranging from -1 to 1 and the results are presented in Figure 7.…”

Probabilistic micromechanical modeling of fatigue-life variability in an α+β Ti alloy

“…Analytical expressions for computing individual U i terms, which have been summarized in a recent article, [20] include Newman's formulation for treating plasticity-induced crack closure, [21] Suresh's formulation for crack deflection and roughness-induced crack closure, [22] as well as an oxide-wedging model. [23,24] The FCG threshold model in MicroFaVa contains an intrinsic term, which originates from the crack-tip cyclic-slip process, and an extrinsic term, which arises from a combination of plasticity-induced closure with crack deflection,…”