The vortex induced vibration is an important problem for the offshore industry, due to the frequent use of structures, such as risers and umbilicals. In the last decade, the use of numerical methods to study the vortex induced vibration has increased. Some disagreement between numerical and experimental results has been verified in the literature. In the present work, the numerical model solves the Reynolds average Navier-Stokes equations using an upwind and a Total Variation Diminishing (TVD) conservative scheme, written in curvilinear coordinates. The turbulent flow in the wake of the cylinder has been modelled using the k–ε model. The cylinder is supported by a spring and a damper and free to vibrate in the transverse direction. Results obtained for phase angle, amplitude, frequency, and lift coefficient are compared to experimental data from Morse et al. [1]. The results indicate that the two dimensional simulation represents better the behaviour of cylinder displacement obtained experimentally using the unattached or attached end plate setup used by Morse et al. [1]. These results suggest that some difference between experimental and numerical data could be associated with the end condition used in the experiments. For the case using the unattached or attached end plate the difference in the peak amplitude between experimental and present results seems to be related with the initial condition adopted.

This content is only available via PDF.
You do not currently have access to this content.