Blockage-driven turbulent vortex dynamics and modal evolution in high-Reynolds-number cavity flows
Abstract
This study presents two-dimensional particle image velocimetry measurements acquired at the mid-plane of the cavity, together with higher-order dynamic mode decomposition (HODMD) analysis, for lid-driven square-cavity flows containing cubical obstacles at high Reynolds numbers (Re = 1.161 × 105–3.483 × 105). Obstacle sizes of 3, 6, and 9 cm, corresponding to h/D = 0.10, 0.20, and 0.30, were systematically examined to clarify blockage-driven variations in vortex dynamics, velocity distribution, planar turbulent kinetic energy (TKE2D), and modal characteristics on the cavity middle plane. The results show that increasing Reynolds number drives the primary vortex toward the cavity center and promotes a transition from relatively fragmented, high-rotation structures to smoother large-scale recirculation. Meanwhile, increasing obstacle size strengthens geometric confinement, shifts the primary vortex toward the left wall, and induces vortex splitting at the largest blockage ratio. The low-velocity region expands markedly, reaching 43.29% for h/D = 0.30, while the high-velocity region is reduced to 0.8% at the highest Reynolds number. At Re = 2.322 × 105, increasing h/D from 0.10 to 0.30 increases the area proportion of the high normalized-TKE region from 0.83% to 4.74%, while decreasing that of the low normalized-TKE region from 50.72% to 36.52%, indicating a blockage-induced redistribution of the resolved in-plane fluctuation energy. HODMD further identifies a dominant zero-frequency mean-flow mode and higher-order modes with clear harmonic relations, corresponding to the multi-scale evolution of the cavity flow. The reconstructed flow fields show a root mean square error as low as 2.08%, substantially lower than that of standard DMD in strongly nonlinear conditions. These results provide experimental evidence for blockage-controlled turbulent vortex dynamics in high-Reynolds-number cavity flows and offer reference data for the validation of numerical simulations in confined flow systems.