Acoustics-driven vortex dynamics in channel branches with round intersections: Flow mode transition and three-dimensionality

Abstract

A combined experimental and large eddy simulation study was conducted to investigate acoustics-driven vortex dynamics inside channel branches with round intersections. The underlying flow mode transition and intensified flow three-dimensionality, which are closely related to the Coanda effect at round intersections, were comprehensively demonstrated. A dynamic pressure transducer array was first used to establish the relationship of the excited acoustic pressure pulsations to the channel-branch intersection radius (r) and the mainstream Reynolds number. In complementary simulations, three configurations with r/D = 0, 0.2, and 0.4 (where D is the short edge of the side-branch) were selected for demonstration. First, the simulated results were well validated in terms of acoustic pressure pulsations and phase-dependent flow fields. Subsequent analysis of the time-averaged and statistical flow characteristics revealed the existence of significantly intensified flow fluctuations inside the round channel branches having r/D = 0.2 and 0.4. Next, the proper orthogonal decomposition analysis was conducted to extract the dominant flow modes and to identify the energy transition from the streamwise vortex-shedding mode to vertical flow-oscillation mode. To this end, the influence of flow-mode transition on the phase-dependent flow fields was further investigated. The intensified branch-flow streaks resulted in a channel flow transition from synchronous convection of co-rotating vortex pairs into alternating convection of a single large-scale vortex, yielding a stronger flapping motion of the mainstream flow. Finally, the intensified flow three-dimensionality, presented by the essential spanwise Reynolds shear stresses inside the round channel branches, was found to relate to the strong turbulent mixing process caused by the flapping mainstream flow and the vertical branch flow oscillation. These findings are of great significance for industrial pipeline design and optimization.