In this study, we investigate the intrinsic origin of unsteadiness in supersonic cavity–ramp flows under non-reacting conditions using an integrated framework that combines Reynolds-averaged Navier–Stokes (RANS) simulations, global stability analysis (GSA), and large-eddy simulations (LES). Steady two-dimensional RANS solutions serve as base states for GSA, from which unstable global modes and their characteristic frequencies and wavelengths are extracted. Unsteady three-dimensional RANS simulations then capture the evolution of perturbations and successfully reproduce the modal characteristics predicted by the GSA, including both the growth rates and spatial structures of the modes. To validate the existence of unsteadiness, LES is also conducted as a numerical experiment. The power spectral density (PSD) reveals low-frequency oscillations across different flow regions, with dominate frequencies nearly consistent with those predicted by GSA. Moreover, dynamic mode decomposition (DMD) captures spatial structures of these low-frequency motions, characterized by wavelengths in agreement with GSA predictions. Together, the PSD and DMD analyses confirm the presence of low-frequency unsteadiness and indicate that it originates from intrinsic global instabilities. This study provides new physical insight into the mechanisms governing cavity–ramp unsteadiness, highlighting the utility of GSA in clarifying the origins of low-frequency oscillations in supersonic non-reacting cavity–ramp flows.
Hu et al. (Thu,) studied this question.