Using artificial neural network techniques to predict volume removal rate of ultrahigh pressure water jet cutting

As a potential machining tool, water jet cutting has considerable advantages over traditional techniques. These advantages include its particular ability to cut special shapes, and to cut hard materials that are problematic to machine, such as foam and composites. However, the water jet cutting process is sometimes a combination of water flow and the flow of abrasives, and these two processes have particular properties. When they operate together, the combined two phase flow becomes difficult to predict. Thus it is more difficult to establish an appropriate model for the water jet cutting process compared with cutting using traditional tools. Several models have been established to model cutting by abrasive–water jets. However, there are still several problems associated with the existing models. In the present paper, the search for a sufficiently accurate and reliable water jet cutting model is continued. A model of the cutting rate of an ultrahigh pressure water jet has been investigated. An artificial neural network system is established to correlate the volume removal rate with the cutting parameters. The results are compared with experimental data and an existing model. Finally, several conclusions are drawn.