Predicting Thermophysical Property of Aluminum Oxide/Ethylene Glycol-Water Nanofluid: A Machine Learning Approach

Articles in Press

Document Type : Full Length Research Article

Authors

1 Department of Engineering College of Engineering and Technology University of Technology and Applied Sciences Muscat, Sultanate of Oman

2 Department of Computer Science. NMAM Institute of Technology Nitte Deemed to be University Nitte, India

Abstract

Nanofluids are used in industrial thermal applications because of their significant thermal characteristics. Machine learning algorithms have recently advanced to the point that they can properly anticipate the thermophysical properties of nanofluids. The literature study provides the data needed to train the models. The gathered data will be separated into groups for testing and training according to 20% and 80% ratios. The thermophysical characteristics of the water-EG base fluid at various percentages mixed with Al2O3 nanoparticles are analyzed in this work. The thermophysical properties were predicted using several machine-learning algorithms. The mean square error and coefficient of determination (R2) were used to compare the models' accuracy. According to the study's findings, machine learning models are the most accurate and quick ways to forecast thermophysical parameters. The accuracy of the model was found to be 99%. The MSE and R2 value of the XGBoost algorithm was found to be 0.0001 and 0.99 respectively. An XGBoost machine learning model was proposed in this study to forecast the thermophysical characteristics of the Al2O3/water_EG nanofluid. This work's novelty lies in the powerful, data-driven alternative that machine learning techniques offer, enabling real-time, high-accuracy predictions of thermal conductivity based on simulation or experimental datasets. This method improves the design and optimization of nanofluids for specific thermal applications, fills in data gaps through substitute modeling, and drastically lowers experimental effort and expense.

Keywords

Main Subjects

Journal of Heat and Mass Transfer Research

Articles in Press, Accepted Manuscript
Available Online from 09 June 2025

Files

History

  • Receive Date: 04 February 2025
  • Revise Date: 15 April 2025
  • Accept Date: 09 June 2025

Share

How to cite

Statistics