Abstract Ultrasonic machining is a process used to machine brittle and hard materials (glass, quartz, ceramic). The material is removed by the abrasion and erosion caused by particle agitation and pressure variation within the abrasive fluid. Much work has already been done on the process: the effects of parameters on performance have been described, but restricted to sinking or drilling. This paper concerns the study of contour machining using a 3 axis numerical control machine. In order to understand its principles and to be able to predict the material removal rate for modelling and simulation of the process, an experimental and analytical work is being carried out. Plans of experiments have made it possible to get the performance of the process, characterized by the removal rate and surface quality. The effects of the main parameters have been studied using two statistical methods. They show an advantage for the neural network one, which gives a more precise prediction of the removal rate compared with the Taguchi method. Moreover the proposed approach seems to be well adapted to the modelling of multi parameters processes.
[1]
George Chryssolouris,et al.
An Experimental Study of Strategies for Integrating Sensor Information in Machining
,
1989
.
[2]
Toshimichi Moriwaki,et al.
Recognition of Cutting State Based on Neural Network Sensor Fusion.
,
1993
.
[3]
Tetsutaro Uematsu,et al.
A New Grinding Method for Ceramics Using a Biaxially Vibrated Nonrotational Ultrasonic Tool
,
1993
.
[4]
M. F. DeVries,et al.
Neural Network Sensor Fusion for Tool Condition Monitoring
,
1990
.
[5]
Amitabha Ghosh,et al.
A fundamental approach to the study of the mechanics of ultrasonic machining
,
1985
.
[6]
Tae Jo Ko,et al.
Cutting state monitoring in milling by a neural network
,
1994
.
[7]
Kuldeep Singh,et al.
On the mechanics of material removal in ultrasonic machining
,
1979
.