What is it about?
18 % Ni Maraging steel is known and used for its high strength-toughness combination. However, to fully realize this superiority, the steel is to be processed under closely controlled conditions so as to regulate the volume fraction of reverted austenite. This study examined how machine learning can be used to predict the formation of reverted austenite in this steel.
Featured Image
Photo by Markus Winkler on Unsplash
Why is it important?
Experimental work to determine the volume fraction of reverted austenite in this steel is time consuming and also expensive. The machine learning approach reported here spares such experimental work. This work highlights the value of data-driven approaches to monitor the kinetics of a critical phase transformation in 18 wt % Ni maraging steel. The work demonstrates that machine learning, with traditional experiments complimenting, can be exploited to monitor and control volume fraction of a critical phase, the reverted austenite in this steel.
Perspectives
The work was carried out as an extension of the doctoral research carried out by Dr Rohit Benjamin, a scholar who pursued his PhD program under my supervision. He brought out that different machine learning approaches are available to satisfactorily predict the response of 18 wt % Ni steel to thermal treatments, with particular reference to formation of reverted austenite. The work highlights the value of data-driven approaches in materials science. It serves to exemplify how machine learning, in conjunction with traditional experimentation, can provide rapid, reliable insights into processing, structure-property relationships, heat treatment optimization, thereby playing a major role in alloy design. I do hope that the work kindles much interest among researchers and practicing engineers working with materials.
Nageswara Rao Muktinutalapati
Read the Original
This page is a summary of: Machine Learning Prediction of Reverted Austenite in 18% Ni Maraging Steel, Journal of Phase Equilibria and Diffusion, April 2026, Springer Science + Business Media,
DOI: 10.1007/s11669-026-01242-6.
You can read the full text:
Contributors
The following have contributed to this page
