Estimation of surface roughness and dimensional accuracy using process parameters in wire cut EDM by artificial neural network

H. V. Ravindra, B. B. Manjunatha, N. Kuruvila

Abstract

Wire cut EDM is a widely accepted non-traditional material removal process to manufacture components with intricate shapes and profiles irrespective of hardness. Due to complicated stochastic process mechanisms in wire-EDM, the relationships between the cutting parameters and cutting performance are hard to model accurately. Experiments were carried out machining the SKD11/D2A2, Tungsten Carbide and Mild steel material using brass wire of diameter 0.25 mm as tool. The input process parameters considered during experiments were servo voltage, offset distance, machining speed, pulse-on and pulse-off. Data’s were taken for different thickness and for different materials of same thickness and corresponding dimensional accuracy and roughness were measured. Artificial neural network is used for the estimation of the dependent parameters (dimensional accuracy and roughness) of WEDM. Finally, from the comparison it was observed that at 90% data in training, data estimated using Artificial Neural Network correlates well with measured value.

Keywords

WEDM; artificial neural network; roughness; dimensional accuracy

Download

IMEKO-TC14-2007-71.pdf

DOI

-

IMEKO TC

TC14 - Measurement of Geometrical Quantities

Back to the proceedings