Abstract
In recent years, cutting-edge preparation has emerged as a topic of great relevance in the manufacturing industry, given its influential role in the performance of cutting tools. This article details the use of the brushing-polishing BP method of cutting-edge preparation on tungsten carbide tools for the broaching process. The innovation of this approach lies in the use of stiff ceramic bristle brushes, which exhibit exceptional impact resistance, a crucial requirement for brushing and polishing operations. The main process parameters are addressed and analysed, together with their influence on aspects such as cutting-edge rounding, material removal rate (MRR) and resulting surface quality/roughness. In parallel, a repeatability and reproducibility (R&R) analysis is carried out, cutting force estimation by FEA, as well as a prediction of the development and growth of the cutting-edge radius using machine learning (ML) algorithms. The results obtained reveal that the required radius and surface roughness were achieved in very short times, being less than 10 min to reach the maximum radius and less than 2 min to obtain optimum roughness (Ra = 0.12, Rz = 0.8). The accuracy of the reproducibility of the cutting-edge radius is comparable to other preparation methods, such as drag finishing and microblasting, with an edge preparation system affectation rate of 14%. The cutting force by the filament obtained by FEA is 0.4 N, like similar processes studied and sufficient for the abrasion process. The effectiveness of the prediction method is evidenced by the low training and cutting-edge radius prediction errors of 2.6% and 1.7%, respectively, demonstrating the effectiveness of the ML prediction approach. Ultimately, it is verified that the planning-polishing process is feasible and reliable for cutting-edge preparation on broaching tools under controlled conditions, allowing the required cutting-edge radius to be obtained with the necessary surface quality for subsequent anti-wear coatings.
Keywords
Get full access to this article
View all access options for this article.
