Politecnico di Torino (logo)

Machine Learning applications for Surface Roughness in Turning.

Vito Antonio Duca

Machine Learning applications for Surface Roughness in Turning.

Rel. Franco Lombardi, Giulia Bruno. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Gestionale (Engineering And Management), 2023

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (4MB) | Preview

Machining production relies heavily on the quality of the surface roughness, and researchers have spent years studying ways to predict it. Typically, three approaches are taken to model surface roughness: empirical methods, theoretical/simulation methods, and soft computing methods. This study involved using Machine Learning models to analyze two datasets generated from machining experiments. These experiments involved turning AISI H13 steel with cutting fluid. The first dataset, which contained 324 samples, was based on theoretically new-tool conditions. The second dataset, which contained 288 samples, varied cutting tool flank wear in three levels. To increase the available data, a strategy was employed to boost by six times the number of measurements without increasing the number of experiments. Machine Learning models were used to predict the output, which was the arithmetic mean deviation (Ra), for both datasets. Smaller datasets resulted in models that were prone to overfitting and performed worse than larger datasets. The models trained on the first dataset (without tool wear) were unable to generalize to the second dataset (with tool wear). This suggests that tool wear is a critical factor when using Machine Learning models to model surface roughness in turning processes. Additionally, Machine Learning models outperformed classical theoretical surface roughness equations. The study also covers techniques for data cleansing, data manipulation to extract and select features, the use of these features in training various Machine Learning models. Finally, ML models have been tested to determine how turning process parameters and cutting tool wear affect the surface roughness parameter "Ra”. Lastly, the study highlights various tools that can be used for the regression analysis of Machine Learning models, such as Linear Regression, Decision Tree Regression, Random Forest Regression, Bayesian Ridge Regression, KNN Regression, Kernel Ridge Regression, and Neural Network.

Relators: Franco Lombardi, Giulia Bruno
Academic year: 2022/23
Publication type: Electronic
Number of Pages: 117
Corso di laurea: Corso di laurea magistrale in Ingegneria Gestionale (Engineering And Management)
Classe di laurea: New organization > Master science > LM-31 - MANAGEMENT ENGINEERING
Aziende collaboratrici: UNSPECIFIED
URI: http://webthesis.biblio.polito.it/id/eprint/26588
Modify record (reserved for operators) Modify record (reserved for operators)