Session: ASME Undergraduate Student Design Expo

Paper Number: 174637

Machine Learning-Driven Prediction of High Entropy Alloy Properties for Optimized Material Design 

High-entropy alloys (HEAs) have emerged as a revolutionary class of materials, offering exceptional mechanical properties such as high strength, corrosion resistance, and thermal stability. These characteristics make HEAs particularly valuable for advanced applications in the military and automotive industries, where materials must withstand extreme conditions while maintaining durability and performance. High-entropy alloys (HEAs) are a class of materials that are a result of an alloy-designed approach where multiple elements, generally more than five, are mixed in equal or near-equal concentrations. The three phases seen in HEAs are solid solution (SS), intermetallic compound(IM), and mixed SS and IM. It is because of their unique composition that HEAs exhibit superior and desirable properties, which makes them highly sought after for various applications.  However, designing HEAs with tailored properties remains challenging due to the vast compositional space and the complexity of predicting their behavior.

Predicting phase stability, mechanical strength, and other performance characteristics is complex, making the development of HEAs a time-intensive and costly endeavor. Traditional experimental approaches often rely on extensive trial and error, requiring numerous iterations to identify optimal compositions. While computational models have been introduced to accelerate the alloy design process, many existing methods suffer from limitations in accuracy, efficiency, or both, making it difficult to achieve reliable predictions on a large scale.

To address these challenges, machine learning (ML) has emerged as a powerful tool for optimizing HEA development. By leveraging ML algorithms such as artificial neural networks, researchers can analyze vast datasets, uncover hidden patterns, and predict alloy properties with greater speed and precision. These data-driven approaches enable scientists to systematically explore compositional possibilities, significantly reducing computational and experimental costs while enhancing predictive accuracy. The integration of ML-driven techniques into HEA research streamlines the alloy design process by allowing for faster identification of promising material candidates with desired properties.

This research explores ML-driven methodologies to predict the properties of high entropy alloys (HEA) in order to optimize HEA design. Through simulation-based approaches, the study aims to streamline the alloy design process, reducing development time while improving performance predictions. By integrating AI-driven techniques into HEA research, this work contributes to advancing material science, enabling the efficient development of next-generation alloys that offer superior durability and resilience. Ultimately, these advancements have the potential to revolutionize the development of high-performance materials in defense, aerospace, and transportation industries, paving the way for cost-effective, high-strength alloys that meet the demands of extreme environments.

Presenting Author: Paola Mora San Diego State University

Presenting Author Biography: I am a Computer Science student at San Diego State University with a focus on artificial intelligence, automation, and cybersecurity. I currently serve as Co-Team Lead of the Controls Software subteam for Aztec Electric Racing and work as an AI/Data Intern developing intelligent tools and automation systems for a commercial real estate firm. In addition, I have conducted undergraduate research applying machine learning to materials science, where I built models to predict thermodynamic properties of high-entropy alloys.

My interests lie at the intersection of data, systems, and real-world impact, and I am passionate about leveraging technology to address complex challenges.

Authors:

Paola Mora San Diego State University
Sara Adibi San Diego State