Principal component idealizations of the dominant modes of variability in the mechanics of the cutting process in metal turning

Quantitative information about the contributions of individual cutting phenomena to linear roughness profiles may aid in optimizing processes with fewer expensive successive trial parts. Linear roughness profiles of metallic hard turned parts contain feed marks, each mark representing a snapshot in time of the state of the cut. This suggests that roughness measuring machines may be an attractive avenue for offline, inexpensive, non-destructive, quantitative evaluation of the time-dependent mechanisms active during the cut. Principal component analysis of feed marks reveals theoretically expected feed mark deformations without coercing the data by fitting. Novel in this paper, we show that those components of feed mark variability appear to correspond to radial and axial displacement of the cutting tool, ploughing, and side flow. Those components are sufficient to explain nearly all the variability between feed marks. The components are easily idealized in a general manner, and their influences on experimental profiles are quantified as percentage contributions to ordinary roughness parameters.