A machine learning aided interpretable model for rupture strength prediction in Fe-based martensitic and austenitic alloys

Osman Mamun,Jeffrey Hawk,Ram Devanathan,Madison Wenzlick

doi:10.1038/s41598-021-83694-z

Abstract

The class of 9–12% Cr ferritic-martensitic alloys (FMA) and austenitic stainless steels have received considerable attention due to their numerous applications in high temperature power generation industries. To design high strength steels with prolonged service life requires a thorough understanding of the long-term properties, e.g., creep rupture strength, rupture life, etc., as a function of the chemical composition and processing parameters that govern the microstructural characteristics. In this article, the creep rupture strength of both 9–12% Cr FMA and austenitic stainless steel has been parameterized using curated experimental datasets with a gradient boosting machine. The trained model has been cross validated against unseen test data and achieved high predictive performance in terms of correlation coefficient (R^{2} > 0.98 for 9–12% Cr FMA and R^{2} > 0.95 for austenitic stainless steel) thus bypassing the need for additional comprehensive tensile test campaigns or physical theoretical calculations. Furthermore, the feature importance has been computed using the Shapley value analysis to understand the complex interplay of different features.

Highlights

The class of 9–12% Cr ferritic-martensitic alloys (FMA) and austenitic stainless steels have received considerable attention due to their numerous applications in high temperature power generation industries
To boost the discovery of high strength ferrous materials, experimental trial and error combined with physicsbased constitutive equations and/or computational C ALPHAD13 evaluations have become the main approaches for developing predictive models for rupture life or rupture s trength[14]
The data used in this study were collected and compiled into a consistent and reliable set of data by the National Energy Technology Laboratory’s (NETL) effort on Extreme Environment Materials, eXtremeMAT (XMAT)

Summary

Introduction

The class of 9–12% Cr ferritic-martensitic alloys (FMA) and austenitic stainless steels have received considerable attention due to their numerous applications in high temperature power generation industries. A workflow combining ML with high quality experimental data has been developed to construct an accurate predictive model for rupture strength prediction in 9–12% Cr FMA and austenitic stainless steel. I.e., 9–12% Cr FMA and austenitic 347H stainless steel, are used to build the predictive models as both are very important for different components of a fossil energy power plant.

Results

Conclusion