Back to Top
Skip to main content
NETL Study that Advances Data Science for Durable Materials Earns Journal Recognition
A photo of molten durable material.

NETL data analysis research is expanding the work of the Lab’s eXtremeMAT program to enable improved materials property prediction that could support the design of cutting-edge clean energy systems.

In their 2021 study, “Data Science Techniques, Assumptions, and Challenges in Alloy Clustering and Property Prediction,” which was cited for excellence by a prestigious academic journal, the researchers explain steps taken to use data from eXtremeMAT to accelerate the development of materials for extreme environments, including the high-temperature, high-pressure conditions in advanced power plants.

The study manuscript was named an Editor’s Choice Article by the Journal of Materials Engineering and Performance. In a Feb. 22 letter to the research team, Editor-in-Chief Rajiv Asthana noted, “This selection is reflective of the comprehensive nature of your paper and its overall excellence.” The journal selects only six papers annually for this recognition.

“This designation validates the importance of the work we are completing to harness the unique capabilities of data science and high-performance computing to make faster and more accurate predictions and develop cost-effective materials needed in flexible energy systems,” said Madison Wenzlick, NETL researcher and lead author of the study.

A thermal power plant’s internal environment is unforgiving. Advanced power generation systems with operating temperatures of more than 650 degrees Celsius and stresses exceeding 50 megapascals will put a plant’s metal components to the test. “But these high temperatures and pressures, along with reliable components, are critical to achieve thermodynamic efficiency that results in reduced carbon emissions and increased cost-effectiveness,” Wenzlick said.

Data analytics methods have been increasingly applied to understand materials chemistry. Wenzlick and her co-authors, which include NETL’s Kelly Rose and Jeffrey Hawk along with Pacific Northwest National Laboratory researchers, noted in the paper that the challenges in interpreting the results of large datasets and assumptions made during data analysis are a barrier to widespread adoption and application of data science methods and tools.

The researchers drew data from the U.S. Department of Energy’s Office of Fossil Energy and Carbon Management eXtremeMAT program, a database of physical and mechanical properties of materials for fossil energy power generation (but applicable to many engineering fields), which was launched in 2018.

The paper describes the limitations of existing studies and provides a clear and methodical example of how advanced machine learning (ML) methods can be used to improve materials property predictions while also ensuring the explainability and reproducibility of those results.

Researchers found their work indicates that strong correlations identified with small datasets can become weaker when more data are collected. Therefore, a certain degree of restraint is called for in drawing conclusions from small datasets. New techniques were applied to the updated and expanded eXtremeMAT dataset. For instance, in order to investigate trends and underlying patterns in the dataset, visualization tools were applied to the data and a correlation matrix was created.

Applying these techniques and others resulted in the identification of new steel groupings that were not present in previous iterations of database analysis and will enable improved property predictions.

Wenzlick noted that the Editor’s Choice honor means the manuscript will be posted as a free-access article on the journal website. “Open access to this article will increase the visibility of this work and will help promote the work of eXtremeMAT and NETL in the area of materials data science.”

This work undertaken in the study is an example of the advanced research supported by NETL’s Science-based Artificial Intelligence and Machine Learning Institute (SAMI). Established in 2020, SAMI combines the strengths of NETL’s energy computational scientists, data scientists and subject matter experts with strategic partners to drive solutions to today’s energy challenges. The Institute has a vision to leverage science-based models, artificial intelligence and ML methods, data analytics and high-performance computing to accelerate applied technology development for clean, efficient and affordable energy production and utilization.

NETL is a U.S. Department of Energy national laboratory that drives innovation and delivers technological solutions for an environmentally sustainable and prosperous energy future. By leveraging its world-class talent and research facilities, NETL is ensuring affordable, abundant and reliable energy that drives a robust economy and national security, while developing technologies to manage carbon across the full life cycle, enabling environmental sustainability for all Americans.