Questions of generalization of statistical modeling results on the example of face milling of titanium alloys

Abstract

The article presents the results of the experiment on face milling of titanium alloy Ti-6Al-4V in coarse-grained and ultra-fine-grained states. The nanostructured state of the alloy was obtained by equal-channel angular pressing that leads to higher strength characteristics of material. The experiment was carried out on a universal milling machine PROMA model FHV-50PD. During the three-factor experiment, the cutting modes were varied: cutting depth (t = 0.4; 0.6; 0.8 mm), feed (s = 65; 100; 185 mm/min) and the rotation speed of the mill (n = 580; 720; 875 rpm). The cutting force was determined indirectly by measuring the displacement of the mill holder using a vibration sensor. The simulation of the dependence of the main component of the cutting force on the processing modes was performed by linear regression for exponential dependence. The model adequately describes the experimental data. A comparison of the results of regression analysis obtained during full and short experiments is presented. The questions of generalization of the results of the obtained force dependencies for processing conditions that differ from the basic ones are considered. The analysis is carried out using a modern method of statistical analysis – bootstrap.