Decoupling bi-directional fluid–structure interactions by the Koopman theory: Actualizing one-way subcases and the role of crosswind structure motion

Abstract

We propose a novel thinking of decoupling bi-directional fluid–structure interactions (bi-FSI) into simpler mono-directional components for analytical insights. The decoupling aims to overcome nonlinearity by the Koopman theory and transform bi-FSI into a linear superposition of the fluid-to-structure, structure-to-fluid, and interactive subcases. This first of a serial effort presents the wind tunnel experimental and computational fluid dynamics numerical actualizations of the fluid-to-structure and structure-to-fluid subcases via rigid and forced vibration models, which are indispensable requisites to the forthcoming Koopman analysis. The results have been analyzed with respect to flow field phenomenology, and the role of forced vibration, hence cross-structure motion alone, has been isolated and elucidated. Compared with the rigid case, crosswind motion weakens leading-edge separation, promotes shear layer curvature and the impingement of the asymmetric wall jets, and hastens reattachment. Consequently, it causes premature shedding of the roll substructure and delays the formation of the rib substructure, effectively altering the Kármán shedding frequency. It also reduces three-dimensional suppression of the Kármán shedding near the fix- and free-end boundary conditions, overarchingly devolumizing wake coherent structures and weakening the Kármán street's intensity. Results also suggest that increasing the wind speed from the characteristic speed of the vortex-induced vibration (VIV) to that of galloping intensifies vortical activities but causes no fundamental change in flow field phenomenology. Therefore, the underlying causes of VIV and galloping are not attributed to the flow field nor structure motion alone but to the interactive mechanisms unique to bi-FSI.