Cutting tool condition monitoring using surface texture via neural network

Hülya KAÇAR; Hacı SAĞLAM

Cutting tool condition monitoring using surface texture via neural network

Date

2003

Authors

Hülya KAÇAR

Hacı SAĞLAM

Abstract

For defining surface finish and monitoring tool wear is essential for optimisation of machining parameters and performing automated manufacturing systems. There is very close relationship between tool wear and surface finish parameters as surface roughness $(R_a)$ and maximum depth of profile $(R_t)$. The machined surface reflects the rate of tool wear and the plot of surface provides reliable information about tool condition. In this paper an approach for estimating $R_a$ and $R_t$ in milling process using the artificial neural networks is proposed. Feed-forward multi-layered neural networks, trained by the back-propagation algorithm are used. In training phase seven input parameters $(v, f, d, F_{x}, F-{y}, F_{z} and Vb)$ and two output parameters are used and the network architecture is as 7x6x6x6x2. It was found that the ANN results are very close to the experimental results. The developed model can be used to define the quality of surface finish in tool condition monitoring systems.