Drilling Studies on MWCNT- and Zirconia-Reinforced Aluminium Alloy 8011 Hybrid Composite: A Machine Learning Approach

Emerging novel composites fabricated from aluminium alloys incorporating high strength and fracture-resistant ceramics with multi-walled carbon nanotubes (MWCNTs) are used frequently in aeronautical and structural engineering fields. This research aims to examine the machinability of composites made from AA8011, consisting of 3 wt.% MWCNTs and varying weight proportions of zirconium dioxide (ZrO₂) (5, 10, and 15 wt.%) fabricated via ultrasonic-assisted stir casting method. Drilling experiments are designed based on the mixed-level design (L₁₈ 2¹,3³) of Taguchi's approach with two levels of drill point angle and three levels of % ZrO₂, rotational speed (RS), and feed rate (FR). The casted micrographs show near uniform dispersal of reinforcements present in the grain boundaries, improving the dislocation density and material strength. Observation shows that with a higher addition of ZrO₂, tool wear (TW) and surface roughness (SR) increase with a reduction in material removal rate (MRR), attributed to the hard ceramic reinforcements in the soft matrix. MWCNTs act as a solid lubricant, which lowers the friction between the workpiece and carbide tool. The desirability approach produces an ideal condition of 118°-point angle, 5% addition of ZrO₂, 1439 rpm of RS, and 0.174 mm/rev of FR. With optimum conditions, the MRR is increased by 9.95%, and TW and SR are reduced by 6.37 and 0.69%. A machine learning approach random forest algorithm (RFA) implemented produces a near-predicted value to the experimental outputs with a correlation coefficient of 99.5, 99.56, and 98.95 for MRR, TW, and SR.