Machinability study on abrasive waterjet drilling of YSZ-coated inconel 718 superalloy: optimization and modelling by machine learning

Abstract

The purpose of this research is to ascertain the optimal abrasive waterjet drilling parameters for making holes on Inconel 718 superalloy coated with yttrium-stabilized zirconia, namely waterjet pressure (JP), standoff distance (SD), abrasive flow rate (FR), and angle of impact (AI). The study explores the predictive modeling of the entry diameter (DN) and exit diameter (DX) of the drilled holes using an amalgamation of experimental analysis, response optimization and machine learning approaches. Eight different ML techniques are used to predict DN and DX. Better values of DN=1.31352 mm and DX=1.00515 mm are obtained through Random Forest for the setting of JP=175 MPa, FR=250 g/min, SD=1.45455 mm, and AI=0.909091 deg. Further, the tuning of hyperparameters of Random Forest algorithm is performed to study the improvement in measuring DN and DX. The least value of total absolute error=0.468 is observed while finding the DN and DX using Random Forest algorithm than the traditional response optimization method with reference to the confirmation test results. The work provides important insights for enhancing the machinability of YSZ-coated Inconel 718 superalloy utilizing the AWJ drilling process by bridging the gap between manufacturing research, machine learning, and real-world applications.