Analysis of Linear Regression to Predict Abrasive Water Jet Marble Drilling Quality

Marble material is characterised by its high
fragility and heterogeneousness, which directly depend on
two main aspects: the origin and nature. Marble
manufacturing for commercial and industrial uses
involves many processes, such as the machining process,
which requires detailed knowledge of marble
characteristics and features. One of these important
characteristics is the stone cutter point, an essential input
to identify the marble machining parameters. In this
study, an examination was achieved to optimise the
marble drilling industry quality process. The main
objective is to achieve high-quality marble drilling
without fracturing. White Carrara marble was used to
examine its drilling process. The utilised drilling method
was the Abrasive-Water-Jet (AWJ); unlike conventional
drilling methods, such as rotating abrasive can achieve
better drilling quality. The performance and quality of
the AWJ process depend on several considerations,
including jet pressure, the nozzle Traverse-Speed (TS),
hole diameter, Stand-Off-Distance (SOD) as well as
Abrasive-Flow-Rate (AFR). All these parameters were
studied and modelled, besides the most common defect
types. The studied defect types involve hole location error
and circularity, hole location error, surface roughness,
and hole cylindricity. Variance analysis of artificial
neural networks, as well as linear regression were utilised
for modelling these types of defects. Accordingly, marble
stone cutters can be effectively determined and utilised in
configuring marble machining parameters, hence
optimising its manufacturing process.