Prediction model for surface quality during 2.5D milling of Inconel 718 using Artificial Neural Network

Resumen

Inconel 718 is indeed a difficult to cut material to manufacture, and its products actually
require a high level of surface quality in aerospace engineering. Thus, it is critical for the
assessment and forecast of surface roughness, as it affects the deflection of the product and
the performance of machining of the machined Inconel 718 workpiece to be created. Speed
(vc), depth of cut (da), feed per tooth (ft), and nose radius (nt) are few of the most influencing
machining parameters to the surface quality during milling. Therefore, the Box–Behnken
design (BBD) with three levels for each parameter has been explored in this investigation.
Artificial Neural Network has been applied to predict the surface quality of the machined
component. The prediction model has been compared with the regression analysis prediction
modal obtained by Response Surface Methodology (RSM). The surface quality of the
components has been observed with the help of Mitutoyo surftest S-310. ANOVA was
performed for the significance and applicability of the proposed model, as well as the impacts
of process parameters on surface quality. It has been observed that the ANN provides more
accurate results as compared to Regression RSM.