Napredna pretraga

Pregled bibliografske jedinice broj: 486108

Flank wear prediction in end milling using regression analysis and radial basis function neural networks


Bajić, Dražen; Jozić, Sonja; Celent Luka
Flank wear prediction in end milling using regression analysis and radial basis function neural networks // IN-TECH 2010 / Jan Kudlaček, Branimir Barišić, Xavier Velay, Kazuhiro Ohkura (ur.).
Prag: Tisk AS s.r.o., Jaromer, 2010. str. 250-254 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Flank wear prediction in end milling using regression analysis and radial basis function neural networks

Autori
Bajić, Dražen ; Jozić, Sonja ; Celent Luka

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
IN-TECH 2010 / Jan Kudlaček, Branimir Barišić, Xavier Velay, Kazuhiro Ohkura - Prag : Tisk AS s.r.o., Jaromer, 2010, 250-254

ISBN
978-80-904502-2-6

Skup
International Conference on Innovative Technologies

Mjesto i datum
Prag, Češka Republika, 14.09.2010. do 16.09.2010

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Flank wear; end milling; regression analysis; radial basis function neural network

Sažetak
End milling is commonly used machining process for the manufacturing of dies and molds, as well as numerous very high precision machine components. Flank wear develops due to abrasion of the cutting tool edge against the machined workpiece surface and is measured by the average width of wear land on the primary clearance face. This study presents the prediction of flank wear in end milling process. Machining parameters (cutting speed, vc, feed per tooth, ft, radial depth of cut, ae) and machining time, t, have been used as input variables. Since the flank wear has an influence on surface quality, the surface roughness has also been observed in this study. Regression analysis and radial basis function neural networks have been applied to data experimentally determined by means of the design of experiment and the effective mathematical models have been developed. The results obtained by the models have been compared. Both models have the relative prediction error below 7.66 %. The best prediction of flank wear and surface roughness shows radial basis function neural network model with the average relative prediction error of 4.48 % and 4.47 %, respectively.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo



POVEZANOST RADA


Projekt / tema
023-0692976-1742 - Istraživanje visokobrzinske obrade materijala (Dražen Bajić, )

Ustanove
Fakultet elektrotehnike, strojarstva i brodogradnje, Split