Prediction of start and finish transformation temperatures for different types of weld metals via machine learning

The interactions of different alloying elements in metallic alloys, combined with the application of thermal cycles and/or welding operations, have a significant impact on the microstructural characteristics and mechanical properties of materials.Continuous cooling transformation (CCT) diagrams, typically obtained through dilatometric tests, describe phase transformations in specific alloys during thermal cycling.However, this methodology requires long processing times, specialized equipment, and expertise, making it unsuitable for fast decision-making.This work presents a Machine Learning-based model, developed and validated to predict austenite decomposition during continuous cooling, using a dataset of experimental CCT diagrams of different weld metals available in the literature.The results demonstrate that the CCT diagrams predicted by the Machine Learning approach can be used as a promising tool to assist in studying microstructural changes occurring during the continuous cooling of weld metals.Moreover, this method can complement dilatometric analysis, reducing experimental time and costs while providing fast and accurate responses that could support the development of welding procedures.