Exploring the Effects of Low Amplitude Fatigue in Crack Growth Rates in High Temperature Aqueous Solution/Metal Systems

Abstract In this work, we explore the use of artificial neural networks (ANN, net) in sorting and interpreting the impact of mechanical variables [such as applied stress intensity factor (Kmax), amplitude and frequency of loading (ΔK, ω)] and environmental parameters [e.g., the corrosion potential (ECP)] on fatigue crack growth in steels in high temperature aqueous systems. In doing so, we reviewed and collected fatigue crack growth rate (FCGR) data from the open literature, we constructed a suitable database (mainly from data obtained from the Argonne National Laboratory) for use (as inputs) with the Artificial Neural Network (ANN), we designed an ANN and trained it on the data base, and we used the ANN to extrapolate the range of input variables. We discuss the predictions of the ANN, and we compare and contrast our findings with known and expected trends.