Influences on the accuracy of fluid force in fluid structure interaction simulation

In the wake of computer hardware and CAE software technology rapid progress, it becomes possible to consider FSI in practice to simulate more complex problems. There are many researches on the force acting on the body in a fluid flow. A typical example is 'Flow Past a Circular Cylinder (Re=100; Reynolds Number) induced vortex and vibration'. When Re is in a range of 30 ∼ 104, drag coefficient CD can be 4/Re0.5. When Re>103, fluid pressure on the cylinder accounts for 87% of CD. Roshko also showed the pressure distribution around a circular cylinder. Achenbach showed the experiment data of CD near the critical Re of cylinders with different course surfaces. Therefore, most of the researches are based on measuring the steady state fluid force, but very less on the approaches of how to calculate the transient fluid force itself between fluid and structure body as well as the sensitivity of these forces with fluid parameters. In this paper, the difficulties of getting an accurate fluid force are examined in detail with a well-known example of a cylinder under different flow rates in the fluid field. The variation of the flow rate, pressure and shear force against different fluid parameters and different numerical methods have shown that it is necessary to do certain comparison when analysts use FSI in engineering practice. Further comparison on the fluid forces for the boundary conditions, mesh sizes and laminar/turbulent models gives other factors that may affect the results significantly. An example of this is quasi-stationary phenomenon, which is the most difficult to approach as it induces a transient vibration analysis that will cost much more resource than a steady-state analysis, such as CPU time and memories. In this study, we have summarized certain characters of fluid force in steady state, quasi-steady state and transient flow, which must be noticed if a steady state approach is used. (author)