Phase field fracture simulations of the cyclic fatigue behavior of gray cast iron: Effect of microstructure, porosity and graphite content

Gray cast iron has been used structurally since the 19th century due to its low cost, reliability, and durability. However, these structures can become fatigued over time and structural safety can be compromised. This study examines, for the first time, the impact of microstructural features, including pores, graphite content, and surface roughness, on the cyclic fatigue life of gray cast iron by using a holistic approach comprising modern modeling techniques, small-scale testing, ultrasonic measurements, and optical and electron microscopy characterization. Overall, this leads to the development of a predictive framework for the fatigue life of gray cast iron components across a wide range of microstructural defects. In order to reach this target, the phase field fracture approach, traditionally used to model crack nucleation and propagation during monotonic load of brittle materials, is extended to high cycle fatigue and applied to assess the remaining life of gray cast iron components extracted from civil engineering infrastructures. The phase field fracture model is complemented with a degradation function that pre-multiplies the critical energy release rate and represents fracture nucleation during cyclic load. This approach is validated empirically by predicting the lifetime of gray cast iron samples under different cyclic loading amplitudes and for different microstructures. Specifically, cyclic fatigue experiments, chemical analysis, and optical and scanning electron microscopy experiments are used to measure the cyclic fatigue life as a function of the porosity and graphite content. First, the model is calibrated to match the fatigue life SN curves of non-porous gray cast iron from the literature. Second, a multiscale approach is adopted, in which a small region of the larger sample is simulated that contains a single pore or graphite flake at which crack nucleation takes place. Different pore sizes and elastic constants are used, which reflect the sample-to-sample variability observed by ultrasonic measurements and optical and electron microscopy. The simulation results showed that the analyzed specimens fractured significantly earlier than previously conducted experiments on non-porous gray cast iron specimens researched in literature. The newly developed phase-field fracture model proved to be very accurate in predicting the decrease in fatigue life depending on microstructural variations and loading conditions. The simulation, including the pore or graphite flake, demonstrated a decrease in its fatigue life for both cases when compared to the defect-free model simulations. The fatigue life of the specimen, including different sizes of a pore or a graphite flake, was recorded, and a relationship was found. It was concluded that porosity has a much higher negative effect than graphite on fatigue life reduction and that the developed model can predict the fatigue life of samples with a high density of defects without needing a specific calibration. This is further validated by microscopy images of the fractured surfaces, which reveal an accumulation of graphite flakes and the presence of porosity. Overall, this research presents a virtual qualification framework for assessing the remaining life of heritage infrastructure, using modern microscopy techniques and adapting to changing loading conditions. This is useful for infrastructure owners because they can evaluate structural integrity, prioritize maintenance and extend service life by assessing the porosity and graphite flakes in small scale specimens extracted from larger infrastructures, while using the proposed computational technique to avoid large scale testing and save costs.