Degree Name

Master of Engineering (Hons.)


Department of Mechanical Engineering


In-process tool wear monitoring has become an important aspect of modem automated manufacturing system. One of the important area of research in the tool wear monitoring system is the application of artificial intelligent techniques for sensor fusion strategy and rational decision making for tool and process conditions. Considerable work has been attempted by many previous authors to develop this technique. However many of these monitoring systems did not adequately give the desired information regarding to the cutting process condition, due to insufficient sensor information and the complex nature of the cutting process itself In this research, a multisensor or sensor fusion approach has been used to integrate the acoustic emission (AE) and forces signals from the cutting process, in order to estimate the tool and cutting condition. Signal processing techniques including spectrum density analysis (FFT) and time series were employed to extract additional information from the signals. These signal features along with cutting parameters were then used to estimate the tool flank wear, crater wear and workpiece surface roughness by using an artificial intelligent neural network.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.