A primitive machine learning tool for the mechanical property prediction of multiple principal element alloys

Aug 5, 2026

Upload by Patrick

Multi-principal element alloys (MPEAs) combine metallic elements across varied proportions and have shown strong mechanical-property performance. This study builds machine-learning models using a curated dataset containing six mechanical properties of MPEAs. A parser converts alloy chemical composition into ML-ready input features, and multiple models are trained and compared. Gradio then enables interactive visualization and user-driven prediction, providing an initial primitive workflow despite not incorporating production or processing-route factors.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Introduction

Motivation for ML in MPEAs

Dataset and workflow overview

Model development and user interface

A primitive machine learning tool for the mechanical property prediction of multiple principal element alloys R. Tan 1, Z. Li 1, S. Zhao 1, N. Birbilis 1,2 * 1College of Engineering, Computing and Cybernetics, The Australian National University, Acton, ACT, 2601, Australia. 2Faculty of Science, Engineering, and Built Environment, Deakin University, Waurn Ponds, VIC, 3216, Australia *[nick.birbilis@deakin.edu.au](nick.birbilis@deakin.edu.au) Abstract Multi-principal element alloys (MPEAs) are produced by combining metallic elements in what is a diverse range of proportions. MPEAs reported to date have revealed promising performance due to their exceptional mechanical properties. Training a machine learning (ML) model on known performance data is a reasonable method to rationalise the complexity of composition dependent mechanical properties of MPEAs. This study utilises data from a specifically curated dataset, that contains information regarding six mechanical properties of MPEAs. A parser tool was introduced to convert chemical composition of alloys into the input format of the ML models, and a number of ML models were applied. Finally, Gradio was used to visualise the ML model predictions and to create a user-interactive interface. The ML model presented is an initial primitive model (as it does not factor in aspects such as MPEA production and processing route), however serves as a an initial user tool, whilst also providing a workflow for other researchers. Keywords: multi principal element alloys, high entropy alloys, machine learning, property prediction, alloys. 1. Introduction Multi principal element alloys (MPEAs) have recently gained significant research attention, with this category of alloys also being inclusive of the so-called high-entropy alloys (HEAs)[1-6] . Whilst research into such alloy systems is entering its second decade as a widespread domain in materials and metallurgy, a number of fundamental aspects related to the composition-performance relationship in MPEAs is still emerging [7-11] . As a data science tool, machine learning (ML) has been demonstrated to simplify several cumbersome steps in data usage and interpretation in material science research, whilst also avoiding errors to a certain extent [12, 13] . ML utilises computing power to improve the generalisation ability and efficiency in materials science [14] . Today, ML is increasingly used in many aspects of material science, such as prediction of materials, material optimisation, crystal structure prediction, and material property prediction [15-20] . The traditional (empirical) production of alloy materials can be an expensive and difficult manner to approach the development of new materials [21] . Consequently, machine learning has an important role to play in helping inform engineers and scientists with respect to possible down-selection in alloy synthesis. There exists a desire for approaches, including highthroughput methods, relevant to the accelerated exploration of MPEAs [22, 23] . To date, based on the exploitation of experimental data, ML models have shown the ability to find relationships between alloy composition and alloy mechanical properties through model training. At the same time, the ML models may therefore predict the mechanical properties of alloy materials on demand [24] . Some examples from the domain includes the recent work of [25] who used linear regression (LR), gradient boosted regression (GBR), and random forest regression (RFR) models to analyse MPEAs. Such work was able to successfully predict the elasticity in composite multi-principal element alloys (MPEAs) by training the density functional theory (DFT) data set constant. The present paper is not intended to be a comprehensive or exhaustive treatise of the topic of MPEAs or MPEA design. Instead, this work is the concise compilation of a study that explored the use of ML models upon a recently compiled dataset of MPEA properties – which th

FAQs

What does the proposed tool predict?

It predicts mechanical properties of multiple principal element alloys using machine-learning models trained on curated property data.

How are alloy compositions prepared for the ML models?

A parser tool converts the chemical composition of alloys into the input format required by the ML models.

How can users interact with the model predictions?

Gradio is used to visualize model predictions and provide a user-interactive web interface for requesting property predictions.

Aug 5, 2026

Upload by Patrick