This study investigates the complete stall process of a counter-rotating axial compressor through multi-passage unsteady numerical simulations, integrating the Liutex vortex identification method and reduced-order variational mode decomposition (RVMD) technique. The analysis reveals that the tip secondary vortex (TSV) plays a pivotal role in stall inception. During stall initiation, two TSVs emerge within the tip passage: the upstream TSV generates backflow and undergoes periodic vortex shedding, while the downstream TSV interacts with adjacent blades, triggering leading-edge spillage that produces stall spikes. These TSVs collectively form an axially fluctuating structure with periodic and selfsustained unsteadiness, propagating circumferentially. The accumulation of fluctuation energy disrupts the stability of tip leakage vortex, ultimately causing complete blockage at the rotor 2 tip and initiating compressor stall. Furthermore, the RVMD technique proves effective in identifying non-periodic and transitional features during stall evolution, enabling efficient extraction of dominant flow structures from large-scale unsteady data. To address the high computational complexity of RVMD, singular value decomposition is employed for spatial dimensionality reduction. However, challenges persist in parameter optimization, often leading to over- or under-decomposition of flow characteristics. The findings highlight the dynamic interplay of TSV-driven mechanisms in stall development and demonstrate the potential of RVMD for complex flow diagnostics, while underscoring the need for improved adaptive parameter selection to enhance its practicality. This work provides insights into compressor stall dynamics and advances data-driven methods for turbomachinery flow analysis.
Chen et al. (Sat,) studied this question.