In many technical fields, e.g. heat exchangers, circular cylinders are involved in Fluid Structure Interaction (FSI) problems. Therefore correct frequency and magnitude of fluid forces, respectively Strouhal number, drag and lift coefficient are needed. If fluid forces are evaluated with Computational Fluid Dynamics (CFD), mostly flow around a rigid cylinder is used to verify model and numerical methods. Unfortunately experimental as well as numerical results show great variation, making verification and testing of models difficult. Reynolds number is regarded as main influencing parameter for a rigid cylinder in cross flow. Most of experimental deviations can be related to other parameters, which differ from experiment to experiment. In this paper such parameters are specified and it is shown, that a closer look is needed, if one really wants to verify a model. Besides experimental results, which can be found in literature, some parameters are investigated by numerical simulation. Like experiments CFD (Computational Fluid Dynamics) simulations show a huge bandwidth of results, even when the same turbulence model is used. Flow around cylinders separates over a wide range of Reynolds numbers. It will be demonstrated that, using CFD, large deviations in fluid forces can often be related to miscalculation of the point of separation.

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