Optimization of Machining Parameters for Nimonic PE16 Using Machine Learning Models
Matthew Jansen1, Ibrahim Deiab2
1Matthew Jansen, College of Engineering and Physical Sciences, University of Guelph, Guelph, Canada.
2Ibrahim Deiab, College of Engineering and Physical Sciences, University of Guelph, Guelph, Canada.
Manuscript received on 04 July 2024 | Revised Manuscript received on 11 July 2024 | Manuscript Accepted on 15 September 2024 | Manuscript published 30 September 2024 | PP: 1-6 | Volume-13 Issue-2, September 2024 | Retrieval Number: 100.1/ijrte.C812413030924 | DOI: 10.35940/ijrte.C8124.13030924
Open Access | Editorial and Publishing Policies | Cite | Zenodo | OJS | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Machining high-temperature alloys such as Nimonic PE16 demands precise control of machining parameters to achieve desired outcomes while minimizing tool wear and optimizing surface finish. In this study, we propose using machine learning regression models combined with synthetic data and response surface methodology strategies to optimize machining parameters for PE16. We aim to develop a predictive model that accurately estimates optimal cutting speeds and feed rates based on key output parameters, including cutting forces and surface roughness. Our methodology involves collecting experimental data from controlled machining tests conducted on PE16 samples under varying conditions. We used the datasets to train and validate regression models to establish correlations between input parameters and machining outcomes. The performance of each model is evaluated based on metrics such as mean absolute error and coefficient of determination. These metrics show relationships within the data and can determine a model’s success. The proposed machine learning framework offers a data-driven approach to optimize machining processes for PE16, facilitating enhanced efficiency, productivity, and quality in nuclear and other high-performance applications. Our findings contribute to understanding machining dynamics in challenging materials and provide valuable insights for intelligent machining systems.
Keywords: Nimonic PE16, Machine Learning, Machinability, Regression, Response Surface Methodology, Synthetic Data.
Scope of the Article: Computer Science and Applications