Ti6Al4V alloy has superior fatigue and corrosion properties, hence extensively used in aerospace and biomedical applications. Considering the intricate shapes of aerospace and biomedical components and the ability of additive manufacturing to create parts that are stronger and lighter than the parts made using traditional manufacturing, additive manufacturing has acquired noteworthiness lately. The components in the aerospace and biomedical industry have to undergo severe cyclic loading, which requires analysis of the fatigue behavior of the material.
The fatigue behavior of additively manufactured Ti6Al4V alloy is investigated, and predictive models were developed using Machine learning in the present work. The relationship between the Fatigue crack growth rate and Stress intensity factor was expressed using the Paris law model; subsequently, Fatigue Life and Stress Amplitude were expressed using S-N curves. The data that has been used is assembled from the reported literature. The experimentations were conducted by varying specific processing, post-processing, and operating conditions given as input features. The effect of particular Processing parameters, Post Processing parameters, and testing conditions was investigated using statistical methods. Various Machine Learning algorithms, including Decision trees, ensemble learning, and K-nearest neighbor, were employed to analyze their effect on the fatigue data. Parameter optimization is done on hyperparameters to improve the performance of the model. It is observed that the high cycle fatigue region in the S-N curve and two nonlinear regions in the Paris Law curve were predicted with the model explaining 90% of the variance present in the data.