Bluff bodies and vortex-induced vibrations


    • Prof Richard Willden

Bluff body Flows

Paul_bluffThe fluid flow past a bluff body is an important consideration in many engineering applications and can have a significant impact on design. However, to date, many aspects of the wake dynamics of such a flow are still to be fully understood. This is particularly true when the flow direction is not perpendicular to the axis of the bluff body and relatively little attention is paid to such cases in the literature. The so called Independence Principle has been proposed to relate the forces on a yawed cylinder to those on an unyawed cylinder with the same perpendicular inflow velocity component. For the case of a circular cylinder this has been investigated both experimentally and computationally in the literature; as yet no firm conclusion as to the validity of the Independence Principle has been reached.

Experimental studies have encountered obstacles in negating the impact of end effects and the intrinsic nature of vortex shedding not parallel to the cylinder axis is the topic of much debate. Full three-dimensional computational studies using periodic boundary conditions at low yaw angles have yet to show anything other than parallel shedding, however no thorough examination of the effect of spanwise cylinder length for the case of yawed cylinders has been undertaken.

The aim of this report is to improve the understanding of the flow over yawed bluff bodies and to establish whether significantly increasing the spanwise length produces slantwise shedding for low yaw angles. The results of this investigation will go some way to providing a firm conclusion as to whether slantwise shedding is intrinsic to the wake of a yawed circular cylinder and hence confirm or disprove the validity of the Independence Principle.

Spectral/hp element methods have been used to simulate the flow at various yaw angles and cylinder lengths. These simulations allow the full 3D flow field to be visualised, and various parameters to be calculated to allow comparisons with previous studies.

Vortex-Induced Vibrations

Bluff_IA fluid flow past a bluff body, such as a circular cylinder, will result in the periodic shedding of vortices into the body’s wake for all but the lowest flow speeds. This process gives rise to oscillatory lift and drag forces which, if the body is compliant or elastically supported, can result in Vortex-Induced Vibrations (VIV). VIV can in turn lead to fatigue damage in vibrating structures, which makes it an important issue in the design of bridges, chimney stacks and marine riser pipes.

The work of the group is divided into two main areas: fundamental research into VIV of circular cylinders in low Reynolds number flows, and VIV of long riser pipes in flows with much higher Reynolds numbers.

The principal aim of the low Reynolds number VIV research is to use numerical models to further the understanding of the types of synchronisation that can exist between the vortex shedding form a cylinder submerged in a flow and the motion of the cylinder. A high order spectral element Computational Fluid Dynamics (CFD) code, Nektar, is employed to compute fluid flow past a circular cylinder that is undergoing vibrations transverse to the direction of the flow. Presently the mode of vortex shedding, which describes the manner in which the vortices are formed at the surface of the cylinder and is an indicator of the level of synchronisation that exists between the cylinder and the shedding, has been carefully analysed over a wide parameter space and has exposed some interesting trends. The lift component of the lift force acting in-phase with the cylinder’s velocity contains information relating to the direction of power transfer between the fluid and cylinder: whether the fluid is exciting or damping the motion of the cylinder; the present research has shown good agreement with experimental data in this respect, and has uncovered the hitherto unknown reason for which hysteretic responses are often observed in laboratory experiments.

Fatigue damage due to VIV in marine riser pipes represents Riser_Ia significant problem for the offshore oil industry. Current industry-standard VIV prediction strategies involve the use of very large safety factors, which is unsatisfactory from an engineering perspective. A means by which VIV response can be accurately and reliably predicted is therefore extremely desirable. The work of the group in this area focuses on the use of a two –dimensional CFD code, VIVIC, to predict such riser pipe vibrations. The code takes a strip theory approach, whereby fluid flow is computed on multiple two-dimensional planes that are positioned at intervals along the axis of a long vibrating pipe, modelled by a Finite Element assemblage. The model permits motions in both the in-line, x, and cross-flow, y, directions. By numerically simulating the experimental set-up of Trim et al (Journal of Fluids and Structures, 2005), the research has shown that VIVIC is capable of accurate simulation of in-line and cross-flow VIV, and has indicated that travelling waves are predominant along the length of the pipe, even when the flow profile is uniform. Super-harmonic responses have been also been observed, and they have been shown to be dominant with respect to riser curvature in a number of instances. Furthermore, fatigue analysis has revealed that the super-harmonic responses cab increase fatigue damage rate by an order of magnitude in the in-line direction and by two orders in the cross-flow direction.