Fatigue-induced cracking and fracture in Francis turbine runner blades are regarded as serious threats to the operational stability and safety of hydropower units. Due to the geometrically complex twisted profiles and enclosed configurations of Francis turbine blades, significant challenges were encountered in conducting crack propagation and fracture tests on prototype or scaled models. To address this limitation, a combined approach involving fluid-structure interaction (FSI) numerical simulations and experimental methodologies was adopted, focusing on ZG06Cr13Ni4Mo martensitic stainless steel blades. Acoustic emission (AE) signals were processed using wavelet threshold denoising, fast Fourier transform (FFT), and short-time Fourier transform (STFT) to isolate signatures associated with fracture events. It was determined that the dominant frequency band corresponding to fracture phenomena in ZG06Cr13Ni4Mo blades lies within the range of 200~320 kHz. These results provide a quantifiable frequency-domain criterion for the identification of fracture events, offering a theoretical foundation for the development of predictive maintenance strategies and data-driven shutdown protocols to prevent catastrophic failures and reduce economic losses in hydropower systems.