Optimization of roller burnishing process parameters using Taguchi and ANN methods

Abstract

The paper presents the results of research on the optimization of the roller burnishing process of brushing used in aeronautical structures by means of Taguchi method and ANN method. Spindle speed (S), feed rate (F) and burnishing depth (D) are chosen to be input parameters, and surface roughness (Ra) is chosen to be the objective of the optimization. Results of analyzing the data about used process parameters and measured surface roughness show that: 1 - Burnishing depth has the biggest influence on surface roughness, then spindle speed and feed rate; 2 - In the considered range of values of process parameters, optimal (minimum) surface roughness is reached as S and D are of largest values whilt F is of smallest one; 3 - Value of optimal surface roughness obtained by both methods is consistent with the experimental data. However, ANN method provides unstable results, the reason may be the fact that the data amount used is not big.