Roughness prediction using machine learning models in hard turning: an approach to avoid rework and scrap
Luiz Gustavo Paes de Souza,Guilherme Augusto Vilas Boas Vasconcelos,Lucas Alves Ribeiro Costa,Matheus Brendon Francisco,Anderson Paulo de Paiva,João Roberto Ferreira
DOI: https://doi.org/10.1007/s00170-024-13951-8
IF: 3.563
2024-07-19
The International Journal of Advanced Manufacturing Technology
Abstract:In recent years, the effectiveness of integrating machine learning methods to predict surface roughness in machining processes has become well-established. However, there is a noticeable gap in the literature concerning the inclusion of tool wear as an input variable. In this context, this study proposed an innovative approach by including tool flank wear as an input variable along with cutting speed, feed rate, and depth of cut to train three machine learning models: Decision Tree Regression, Random Forest, and Support Vector Regression. These models aim to predict surface roughness during the dry turning of hardened AISI 52100 steel. As a result, all three models exhibited high prediction accuracy, with R -squared values exceeding 90% and lower values for both the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). However, the Random Forest model outperformed the others, boasting the lowest RMSE and MAE values of 0.05 and 0.038, respectively, alongside the highest R -squared value of 0.91. The confirmation runs demonstrated the accuracy of the Random Forest model, with actual roughness values very close to those predicted (variation of ± 0.01 μm). The correlation analysis revealed that roughness is correlated with feed rate and flank wear. This outcome underscores the importance of including tool wear as an input variable for roughness modeling. Since roughness prediction depends on the tool wear level, it is feasible to forecast the roughness of parts machined using the same cutting edge until flank wear reaches 0.30 mm (end of life). By anticipating and understanding roughness behavior as wear progresses, decision-makers can choose cutting configurations that ensure parts meet specifications, thus avoiding rework and scrap.
engineering, manufacturing,automation & control systems