Surface quality in turning is still evaluated mainly by post-process profilometry, which limits the use of sensor feedback during machining. This article examines whether perpendicular vibration displacement can be used as a practical indirect indicator of surface roughness in the turning of aluminum alloys. The study is based on 204 synchronized segment-level vibration–roughness observation pairs collected during 408 s of turning. The vibration meter operated in displacement mode, continuously measuring vibration while the SD logger stored one perpendicular displacement p-p reading every 2 s; Ra and Rz were then associated with the corresponding machined segment. The analysis combined descriptive time-domain statistics, low-frequency FFT/STFT descriptors of process-state evolution, phase segmentation, correlation analysis, and linear regression. Very strong within-dataset relationships were obtained between perpendicular vibration displacement and surface roughness, with R2 = 0.992 for Ra and R2 = 0.988 for Rz. Entry, steady-state, and exit phases showed different variability levels, and the steady-state segment provided the most stable basis for roughness estimation. Because the logger sampling interval was 2 s, the spectral results should be interpreted as low-frequency process-state descriptors rather than as direct chatter measurements. Within this scope, the results support the use of perpendicular displacement sensing as a low-cost feasibility approach for in-process roughness indication. Broader transfer to CNC production, other alloys, and higher-bandwidth monitoring requires additional validation.
Karpavičius et al. (Sat,) studied this question.