EXPERIMENTAL INVESTIGATION OF THE VORTEX-INDUCED VIBRATION RESPONSE OF A FLEXIBLY-MOUNTED RIGID CYLINDER IN THE SHEAR-THINNING AND INERTIAL-VISCOELASTIC FLOW REGIMES

Flexible or flexibly-mounted structures with bluff cross-sections in flow can shed vortices at frequencies that increase with increasing flow velocity. When this shedding frequency is equal to the structure\u27s natural frequency, the structure can oscillate. This is called vortex-induced vibrations (VIV). VIV is present in multiple fluid-structure interaction (FSI) systems which can be found in industrial, medical, and engineering applications. These oscillations can be desirable or undesirable, so understanding the physics behind this phenomenon is important. This work seeks to investigate experimentally the VIV response in the inertial-viscoelastic regime where fluid inertia and elasticity influence the system. The subcritical Newtonian VIV response is first investigated. In fixed cylinders, vortex shedding begins at Reynolds numbers of Re\u3e47. When the cylinder is flexibly mounted, a subcritical instability exists. Previous numerical VIV work has found that vortex shedding and cylinder oscillations can occur at Reynolds numbers as low as Re=18. A rotating water channel is used to generate flow and a novel experimental setup is employed that allows the Reynolds number to remain fixed while the independent variable, the reduced velocity is varied over the entire lock-in range (i.e., the range for which oscillations are observed). The subcritical VIV response is investigated experimentally for the first time and VIV is found to occur at Reynolds numbers as low as Re=19. When the fluid is non-Newtonian, the subcritical response can be shifted or entirely suppressed. In inelastic shear-thinning fluids, the critical Reynolds number required for the subcritical response to be observed is increased when the fluid is made more shear-thinning and less extensionally-thinning. The maximum oscillation amplitude is reduced along with the lock-in width and the critical reduced velocity at which lock-in begins. Viscoelasticity further influences the VIV results. Similar to the shear-thinning case, the critical Reynolds number is increased and the lock-in range is shifted. The oscillation amplitude is greatly reduced, even at high Reynolds number and the overall response changes significantly when compared to Newtonian VIV. In the higher Reynolds number cases, the amplitude increases with increasing reduced velocity until it sharply drops to zero