Flow-induced vibration (FIV) caused by reactor coolant flow is a critical factor leading to fretting wear of fuel assemblies and threatening the safe operation of nuclear power plants, with turbulence-induced excitation being a primary driver. This paper investigates the turbulent fluctuation characteristics of axial flow in pressurized water reactor (PWR) fuel assemblies through numerical simulations using Large Eddy Simulation (LES) and Fluid-Structure Interaction methods. Turbulence excitation spectra at various locations were obtained via signal processing to analyze the characteristics and spatial correlation of turbulent fluctuations. Furthermore, the cross-power spectral density (CPSD) and coherence between monitoring points were quantitatively evaluated. On this basis, Empirical Mode Decomposition (EMD) and the Hilbert-Huang Transform (HHT) were introduced to perform adaptive time-frequency analysis of the pressure fluctuation signals, further revealing their multi-scale composition and intermittent characteristics. The results indicate that, compared to time-averaged turbulence models, LES more accurately captures the turbulent fluctuations and power-law decay characteristics in the inertial subrange within fuel assemblies. While the coherence of monitoring points on the same cross-section is relatively low, monitoring points separated by one hydraulic diameter along the axial direction exhibit highly consistent trends in both CPSD and power spectral density (PSD), reflecting the evolutionary relationship of coherent structures along the axial direction. EMD adaptively decomposes the pressure fluctuations into 7-8 intrinsic mode functions (IMFs), whose Hilbert marginal spectra agree closely with the PSD in the range of 10-2000 Hz (log-scale correlation coefficient of 0.946-0.984), and the Hilbert spectrum further reveals the temporal intermittency of energy bursts. These findings can be utilized to identify the homology of pressure fluctuations and provide a basis for constructing turbulence excitation envelope spectra required for vibration analysis.

Flow-induced vibration (FIV); Turbulence-induced excitation; Large Eddy Simulation (LES); Power spectral density (PSD); Empirical Mode Decomposition (EMD)