RADIALINĖS KRYPTIES PEAK-TO-PEAK POSLINKIO INTERVALAI KAIP PROCESINIS PAVIRŠIAUS ŠIURKŠTUMO PRIEŽIŪROS KRITERIJUS BAIGIAMAJAME EN AW-2011 ALIUMINIO LYDINIO TEKINIME

Abstract

This article presents a practical process-supervision interpretation of a broader calibration analysis that relates low-frequency radial (i.e., normal-direction) peak-to-peak displacement to surface roughness in finish turning of EN AW-2011 aluminum alloy. The study uses only a 190-record synchronized displacement-roughness analysis subset obtained in one controlled machine-tool-sensor setup. The regression calibration showed very strong within-dataset relations:          R_a = 14,204 + 24,191V (R² = 0,9929, RMSE = 0,052 µm) and R_z = 63,207 + 105,253V (R² = 0,9905, RMSE = 0,264 µm). For operational interpretation, displacement values were divided into three empirical bands: low, medium, and high. Mean roughness increased monotonically across these bands, from R_a = 14,512 µm and R_z = 64,547 µm in the low band to R_a = 16,076 µm and R_z = 71,362 µm in the high band. An additional one-way ANOVA confirmed statistically significant differences between the bands (p < 0,001). The results support the use of displacement bands as a setup-specific warning logic for roughness-related supervision. However, the bands should not be treated as universal limits, because the data represent one time-ordered calibration sequence, and no independent validation set was used.