Naslov
Hard Milling Process Based on Compressed Cold Air- Cooling Using Vortex Tube for Sustainable and Smart Manufacturing

Autori
Celent, Luka ; Bajić, Dražen ; Jozić, Sonja ; Mladineo, Marko

Izvornik
Machines (2075-1702) 11 (2023), 2; 264, 25

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
hard milling ; sustainability ; smart manufacturing ; vortex tube ; compressed cold air-cooling ; radial basis neural networks

Sažetak
Improving machining performance and meeting the requirements of sustainable production at the same time represents a major challenge for the metalworking industry and scientific community. One approach to satisfying the above challenge is to apply different types of cutting fluids or to optimise their usage during the machining process. The fact that cutting fluids are well known as significant environmental pollutants in the metalworking industry has encouraged researchers to discover new environmentally friendly ways of cooling and lubricating in the machining process. Therefore, the main goal is to investigate the influence of different machining conditions on the efficiency of hard machining and find a sustainable solution towards smart manufacturing. In the experimental part of the work, the influence of various machining parameters and conditions on the efficiency of the process was investigated and measured through the surface roughness, tool wear and cutting force components. Statistical data processing was carried out, and predictive mathematical models were developed. An important achievement is the knowledge of the efficiency of compressed cold air cooling for hard milling with the resulting lowest average flank wear of 0.05 mm, average surface roughness of 0.28 µm, which corresponds to grinding procedure roughness classes of N4 and N5, and average tool durability increase of 26% compared to dry cutting and conventional use of cutting fluids. Becoming a smart machining system was assured via technological improvement achieved through the reliable prediction of tool wear obtained by radial basis neural networks modelling, with a relative prediction error of 3.97%.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo, Temeljne tehničke znanosti

POVEZANOST RADA