ARTIFICIAL NEURAL NETWORK-BASED OPTIMIZATION OF THE CRYOGENIC-INTERNAL DIAMOND BURNISHING PROCESS IN TERMS OF SURFACE QUALITY

Abstract

Internal diamond burnishing is a prominent solution to produce surface
finishing for interior holes. This work aims to propose a novel diamond burnishing
process, in which an integrative lubrication using the Vortex tube and liquid CO2 is
applied. Three key process parameters, including the spindle speed (S), feed rate
(f), and burnishing depth (D) are optimized to decrease the surface roughness (SR)
and improve the Vickers hardness (VH). The Box-Behnken method is applied to
conduct the burnishing experiments. The artificial neural network (ANN) is used
to develop burnishing response models, while the entropy method is utilized to
compute the weights. The optimal solution is determined using the multiple-
objective particle swarm optimization (MOPSO) algorithm. The results indicated
that the optimal outcomes of the S, D, and f were 630rpm, 0.12mm, and
0.04mm/rev., respectively. The SR was decreased by 60.9%, while the VH was
increased by 10.2% at the optimal solution. The outcomes could be applied to
practical diamond burnishing to enhance the surface quality of the internal holes.
The optimizing technique could be used to present the non-linear data and obtain
optimal global results.