This work reports on an empirical study to account for the effects of the drilling parameters of cutting speed: V, m/s (max at the outer periphery), tool feed (f, mm/rev)), and drill diameter (D, mm) on the generated thrust (T, N) and torque (M, N.m). A total of 50 drilling tests were conducted on a vertical machining center based on design of experiment (DOE) test matrix using JMP-SAS/STAT® software. The test matrix employed 2 different aluminum alloys: one wrought (6061-T6) and another cast (A356-T6) with both tempered to the T6 condition. Two classical chisel drills of different diameters (10mm and 12.7mm) were used. Drilling parameter variables employed varied from very mild to very aggressive combinations including 5 values of spindle speeds (796, 1592, 3183, 6366, and 9868 rpm) and 5 values of drilling feeds (0.04, 0.08, 0.16, 0.32, 0.64 mm/rev). Drilling forces and torques were recorded using a 4-component dynamometer (Kistler model 9123) at sampling rate of 200 Hz. At full drill engagement, cutting torque and thrust status were identified for each drilling test case.
By fitting to the measured drilling forces, power law equations with parametric variables D, f, and V are developed. Initial values of the model’s power coefficients were initially identified using Matlab® using a least square optimization function with target to minimize the difference between the predicting model and the experimental data. These values were fed as initially-guessed values to the nonlinear modeling Gauss method in JMP-SAS/STAT® with 50 observations for each set of experiments. For both torque and thrust and for both aluminum alloys employed, the model’s coefficient values were identified setting the convergence criteria to 10−15 with total of 60 iterations. For both thrust and torque, it was found that the power coefficients for feed and cutting speed are statistically significant (better than p-values < 0.05) with values of about 0.7 and −0.1, respectively.