Surface roughness monitoring by singular spectrum analysis of vibration signals. (1st February 2017)

• Record Type: Journal Article
• Authors: García Plaza, E.; Núñez López, P.J.
• Is Part Of: Mechanical systems and signal processing. Volume 84:Part A(2017)
• Page Start: 516
• Page End: 530
• Publication Date: 2017-02-01
• Subjects: Singular spectrum analysis (SSA) -- Vibrations signals -- Surface roughness monitoring -- CNC turning; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2016.06.039
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: This study assessed two methods for enhanced surface roughness ( Ra ) monitoring based on the application of singular spectrum analysis (SSA) to vibrations signals generated in workpiece-cutting tool interaction in CNC finish turning operations i.e., the individual analysis of principal components (I-SSA), and the grouping analysis of correlated principal components (G-SSA). Singular spectrum analysis is a non-parametric technique of time series analysis that decomposes a signal into a set of independent additive time series referred to as principal components. A number of experiments with different cutting conditions were performed to assess surface roughness monitoring using both of these methods. The results show that singular spectrum analysis of vibration signal processing discriminated the frequency ranges effective for predicting surface roughness. Grouping analysis of correlated principal components (G-SSA) proved to be the most efficient method for monitoring surface roughness, with optimum prediction and reliability results at a lower analytical-computational cost. Finally, the results show that singular spectrum analysis is an ideal method for analyzing vibration signals applied to the on-line monitoring of surface roughness. Highlights: Optimum prediction results were obtained for a window length L =10 (effective L =5). Feed acceleration ay was the primary data source in the monitoring of roughness. I-SSA and G-SSA methodologies provided excellent reliability and predictive power. G-SSA method is the most efficient method for monitoring surface roughness ( Ra ). SSA method permitted the grouping of effective principal components in monitoring.