Prediction Model for BTA Deep-Hole Machining using Fuzzy Clustering Approach: Experimental Study

Abstract

In this article, a subtractive clustering-based fuzzy identification method and a Sugeno-type fuzzy inference system are used for modeling in metal cutting. This approach is considered with its application on the experimental study of Boring and Trepanning Association (BTA) deep-hole drilling. The model for the surface roughness is identified by using the cutting speed and feed as input data and roughness as the output data. Using subtractive clustering in both input and output spaces performs the model-building process. Minimum error model is obtained through enumerative search of clustering parameters. The fuzzy model obtained is capable of predicting the surface roughness for a given set of inputs (speed and feed). Therefore, the operator can predict the quality of the surface for a given set of working parameters and will then be able to set the machining parameters to achieve a certain surface quality. The fuzzy model is verified experimentally by further experimentation using different sets of inputs. The tool life is also investigated using the same approach. The fuzzy inference system obtained is capable of predicting the tool life for a given set of cutting parameters. Therefore, the operator will be able to predict how many minutes the cutting tool is going to last and will set the time for the next tool change.

Keywords:

• Deep-hole machining
• Subtractive clustering
• Sugeno fuzzy system
• Surface roughness
• Tool life
• Factorial design of experiments

Acknowledgment

We acknowledge the cooperation and help provided by Mr. J. Seeger in Deep Hole Machining Lab at Concordia University. The experiments were conducted during the first author visit in July 2001.