Prediction of Cutting Forces in Milling Using Machine Learning Algorithms and Finite Element Analysis

• Published in: Journal of materials engineering and performance Vol. 30; no. 3; pp. 2002 - 2013
• Main Author: Charalampous, Paschalis
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2021
• Subjects: Characterization and Evaluation of Materials; Chemistry and Materials Science; Corrosion and Coatings; Engineering Design; Materials Science; Quality Control; Reliability; Safety and Risk; Tribology; finite element method; machine learning; cutting forces; milling process; simulation
• ISSN: 1059-9495; 1544-1024
• DOI: 10.1007/s11665-021-05507-8

Cutting force estimation is an essential factor for the stability and the optimization of the machining process. The present article deals with the prediction of the cutting forces during the face milling of AISI 4140 with coated cemented carbide inserts under dry cutting conditions. The required data for the development of the predictive models were obtained through milling investigations conducted in various cutting conditions. Two modeling approaches, particularly machine learning algorithms and finite element analysis, have been employed in the study. A series of machine learning algorithms were applied such as support vector regression, k-nearest neighbor, polynomial regression and random forest. The machine learning and FE models evaluate the dependence of the cutting force magnitude on milling parameters like cutting speed, radial depth of cut and feed per tooth. Results acquired by the predictive models have been compared with the experimental ones. Taking into account the sufficient convergence between them, the introduced machine learning models could be effectively applied in order to estimate the effect of cutting conditions on the developed milling forces. Innovative machine learning models could be implemented into CAM software to increase the productivity of manufacturing procedures.