Understanding the relationship between machining processes and surface properties is crucial for optimizing tribological characteristics of mechanical components. Different
machining operations create specific surface structures, however, a systematic methodology for their identification and characterization has not been developed.
We analyzed surface topography of samples machined with four basic processes (turning, milling, grinding, polishing) in three intensities. Using contact stylus profilometer, we measured 19 roughness parameters on 60 samples. We applied multivariate statistical analysis, principal component analysis, power spectral density analysis, and anisotropy analysis.
Results show that each machining operation creates unique topographical properties with statistically significant differences (p < 0.001). Profile asymmetry parameter (Rsk) achieves the highest discriminatory power, while traditional parameters (Ra, Rq) are not the best discriminators. We developed a 3-level classification model for automatic identification of machining processes. Principal component analysis showed that 19 parameters can be reduced to four components explaining 97% of variability.
|
