View in EDS

Optimization of Cutting Parameters and Result Predictions with Response Surface Methodology, Individual and Ensemble Machine Learning Algorithms in End Milling of AISI 321.

Saved in:
Bibliographic Details
Title: Optimization of Cutting Parameters and Result Predictions with Response Surface Methodology, Individual and Ensemble Machine Learning Algorithms in End Milling of AISI 321.
Authors: Demircioglu Diren, Deniz, Ozsoy, Neslihan, Ozsoy, Murat, Pehlivan, Huseyin
Source: Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ); Sep2023, Vol. 48 Issue 9, p12075-12089, 15p
Subject Terms: RESPONSE surfaces (Statistics), MACHINE learning, K-nearest neighbor classification, CUTTING force, SURFACE roughness, ANALYSIS of variance, DECISION trees
Abstract: Optimizing the parameters in the milling method is important in terms of cost, energy, and time. The forces that arise during milling cause undesirable results, such as tool wear and energy loss. In this study, cutting parameters were optimized during the milling of AISI 321 material. Cutting speed (60, 70, 80 m/min), feed per tooth (0.04, 0.05, 0.06 mm/tooth), and depth of cut (0.25, 0.5, 0.75 mm) were selected as input parameters. Cutting force in the X and Y axes and the surface roughness were selected as the output parameters. Optimum parameters (60.80 m/min for cutting speed, 0.04 mm/tooth for feed per tooth, and 0.25 mm for depth of cut) were found using response surface methodology. The effect of cutting parameters was calculated by analysis of variance. The most influential parameters were found, depth of cut as 87.49% for cutting force on the X-axis, 86.48% on the Y-axis, and for surface roughness, the cutting speed with 36.48%. Prediction models are compared to choose the best model. Individual (Neural network, decision tree, and k-nearest neighbor algorithms) and ensemble methods (vote) from machine learning and response surface methodology from statistical methods were used for models. The error rates of the models were compared according to the mean absolute percentage error performance criterion. The lowest MAPE values were obtained with the vote method 11.163% in the X-axis force, the artificial neural network algorithm with 7.749% in the Y-axis force, and RSM with 0.93% in the surface roughness. [ABSTRACT FROM AUTHOR]
Copyright of Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ) is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Complementary Index
Description
Abstract:Optimizing the parameters in the milling method is important in terms of cost, energy, and time. The forces that arise during milling cause undesirable results, such as tool wear and energy loss. In this study, cutting parameters were optimized during the milling of AISI 321 material. Cutting speed (60, 70, 80 m/min), feed per tooth (0.04, 0.05, 0.06 mm/tooth), and depth of cut (0.25, 0.5, 0.75 mm) were selected as input parameters. Cutting force in the X and Y axes and the surface roughness were selected as the output parameters. Optimum parameters (60.80 m/min for cutting speed, 0.04 mm/tooth for feed per tooth, and 0.25 mm for depth of cut) were found using response surface methodology. The effect of cutting parameters was calculated by analysis of variance. The most influential parameters were found, depth of cut as 87.49% for cutting force on the X-axis, 86.48% on the Y-axis, and for surface roughness, the cutting speed with 36.48%. Prediction models are compared to choose the best model. Individual (Neural network, decision tree, and k-nearest neighbor algorithms) and ensemble methods (vote) from machine learning and response surface methodology from statistical methods were used for models. The error rates of the models were compared according to the mean absolute percentage error performance criterion. The lowest MAPE values were obtained with the vote method 11.163% in the X-axis force, the artificial neural network algorithm with 7.749% in the Y-axis force, and RSM with 0.93% in the surface roughness. [ABSTRACT FROM AUTHOR]
ISSN:2193567X
DOI:10.1007/s13369-023-07642-x