DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Adversarial Ensemble Modeling of Multi-modal Mechanical Properties for Iron-Based Alloys

Journal Article · · JOM. Journal of the Minerals, Metals & Materials Society

Mechanical properties of alloys are controlled by their microstructure; and microstructure evolution is controlled by internal and external stressors. Chemical complexity during the alloy processing may result in heterogeneity and observation of multimodal performance patterns. Thus, to ensure the desired performance of stressed components it is important to understand the origin and mechanisms of such behavior. Adversarial ensemble modeling was introduced here to explain multimodal mechanical properties of the iron-based alloys. The modeling results showed that the areas of a single mechanism predominance were contiguous across the alloy compositions clustered by similarity, with sharp and persistent boundaries separating the single-mechanism domains. The marginal compositions resulted in increased competition between the adversarial models but only led to the multimodal behavior when the competing models diverged. Transparency of the clustering method allowed explicit interpretation of the chemistries leading to such competition. The adversarial ensemble was interpreted through transition from ductile to brittle microstructure.

Research Organization:
National Energy Technology Laboratory (NETL), Pittsburgh, PA, Morgantown, WV, and Albany, OR (United States)
Sponsoring Organization:
USDOE Office of Fossil Energy (FE)
OSTI ID:
1865279
Journal Information:
JOM. Journal of the Minerals, Metals & Materials Society, Journal Name: JOM. Journal of the Minerals, Metals & Materials Society Journal Issue: 4 Vol. 74; ISSN 1047-4838
Publisher:
SpringerCopyright Statement
Country of Publication:
United States
Language:
English

References (14)

Materials data analytics for 9% Cr family steel
  • Romanov, Vyacheslav N.; Krishnamurthy, Narayanan; Verma, Amit K.
  • Statistical Analysis and Data Mining: The ASA Data Science Journal, Vol. 12, Issue 4 https://doi.org/10.1002/sam.11406
journal February 2019
Mapping Multivariate Influence of Alloying Elements on Creep Behavior for Design of New Martensitic Steels journal April 2019
Statistical distributions of toughness and fracture stress for homogeneous and inhomogeneous materials journal January 1986
Yield stress of fine grained materials journal August 1998
What is behind the inverse Hall–Petch effect in nanocrystalline materials? journal June 2007
9Cr steel visualization and predictive modeling journal October 2019
Deep-freeze graph training for latent learning journal November 2021
Predictions of long-term creep life for the family of 9–12 wt% Cr martensitic steels journal January 2020
Screening of heritage data for improving toughness of creep-resistant martensitic steels journal August 2019
Mechanical properties of nanocrystalline materials journal May 2006
A machine learning aided interpretable model for rupture strength prediction in Fe-based martensitic and austenitic alloys journal March 2021
Mapping Non-Linear Influence of Alloying Elements on Tensile Strength of Martensitic Steel conference December 2018
Latent Learning with pyroMind.2020 conference December 2021
Application and Characteristics of Low-Carbon Martensitic Stainless Steels on Turbine Blades journal January 2015