Bridging Diagnostic Condition Monitoring and NVH Tonal Excitation Through Frequency-Domain Structural Mapping

In general, condition monitoring (CM) and noise, vibration and harshness (NVH) are often treated as separate disciplines, despite the fact that both rely on vibration measurements. CM relies on broadband statistical metrics such as RMS, kurtosis, and envelope analysis to detect faults. Meanwhile, NVH investigates tonal excitation mechanisms related to gear mesh frequency (GMF) and its modulation components. In this study, we investigate whether a numerical relationship can be established between classical CM indicators and physically based tonal excitation indicators derived from frequency-domain analysis. Using a controlled gearbox degradation dataset, Spearman correlation analysis was performed between broadband metrics and GMF-related tonal features, including GMF-band energy and absolute sideband energy. Results show moderate but statistically significant correlations between RMS, envelope peak amplitude, and tonal indicators, whereas kurtosis exhibits no meaningful association. Additionally, tonal amplification due to degradation is shown to be structurally localized rather than uniformly distributed across sensor locations. These findings demonstrate that broadband CM indicators partially encode tonal excitation growth, establishing a reproducible data-driven bridge between diagnostic condition monitoring and NVH excitation analysis.