Fatigue is one of the most prevalent mechanisms of failure. Thus, the evaluation of the fatigue crack growth process is fundamental in engineering applications subjected to cyclic loads. The fatigue crack growth rate is usually accessed through the da/dN-ΔK curves, which have some well-known limitations. In this study a numerical model that uses the cyclic plastic strain at the crack tip to predict da/dN was coupled with the Gurson–Tvergaard–Needleman (GTN) damage model. The crack propagation process occurs, by node release, when the cumulative plastic strain reaches a critical value. The GTN model is used to account for the material degradation due to the growth of micro-voids process, which affects fatigue crack growth. Predictions with GTN are compared with the ones obtained without this ductile fracture model. Crack closure was studied in order to justify the lower values of da/dN obtained in the model with GTN, when compared with the results without GTN, for lower ΔK values. Finally, the accuracy of both variants of the numerical model is accessed through the comparison with experimental results.

No abstract

The fatigue crack growth (FCG) process is usually accessed through the stress intensity factor range, ΔK, which has some limitations. The cumulative plastic strain at the crack tip has provided results in good agreement with the experimental observations. Also, it allows understanding the crack tip phenomena leading to FCG. Plastic deformation inevitably leads to micro-porosity occurrence and damage accumulation, which can be evaluated with a damage model, such as Gurson–Tvergaard–Needleman (GTN). This study aims to access the influence of the GTN parameters, related to growth and nucleation of micro-voids, on the predicted crack growth rate. The results show the connection between the porosity values and the crack closure level. Although the effect of the porosity on the plastic strain, the predicted effect of the initial porosity on the predicted crack growth rate is small. The sensitivity analysis identified the nucleation amplitude and Tvergaard’s loss of strength parameter as the main factors, whose variation leads to larger changes in the crack growth rate.

In this work, the boundaries of small‐scale yielding (SSY) and large‐scale yielding (LSY) have been experimentally evaluated from the analysis of crack tip opening displacement (CTOD) measured by digital image correlation (DIC). According to a previous numerical work, the ratio between the elastic CTOD range and the total CTOD range (Δδe/Δδt) defines the boundaries of SSY conditions and LSY. Three materials have been studied, commercially pure titanium and 2024‐T3 and 7050‐T6 aluminum alloys, tested at different stress ratio values. SSY conditions are shown to dominate when Δδe/Δδt ≥ 79% and ≥78% for titanium and the two aluminum alloys, respectively. In addition, LSY can be established when Δδe/Δδt ≤ 66.3% and ≤67.2% for titanium and for 2024‐T3 and 7050‐T6 aluminum alloys, respectively. Transition or LSY conditions are more probable in fatigue tests conducted at low R ratio and for crack lengths above 40% with respect to the width of the specimen.

Residual stresses affect the fatigue behavior, given that compressive stresses delay the phenomenon, while tensile stresses accelerate it. However, the mechanisms behind the effect of residual stresses are not totally understood. A numerical study is developed here to understand the effect of thermal residual stresses (TRSs) on fatigue crack growth (FCG). The crack driving force was assumed to be the cumulative plastic strain at the crack tip. The heating of a region ahead of the crack tip produced elastic compressive TRS, which were 69% of material’s yield stress. Alternatively, plastic deformation was produced by severe cooling followed by heating to generate compressive residual stresses. The crack propagation in the compressive residual stress field produced a decrease in the FCG rate. On the other hand, without the contact of crack flanks, the TRS showed no effect on FCG. Therefore, the TRSs only affect FCG by changing the crack closure level.