Fluid-structure interaction ( fsi ) modeling of thin plates

The fluid-structure interaction (FSI) of both flat and curved plates in parallel flow has been successfully simulated using popular numeric codes. The FSI models are carried out using a prototypic geometry of one 1.016 mm (40 mils) plate bounded by fluid channels of differing thickness. The modeled plate uses isotropic structural properties of Al 6061-T6 while the fluid is assumed to be water. High velocity flow combined with the dissimilar fluid channel gaps leads to a pressure differential across the plate that causes plate deflection. The plate’s structural behavior is modeled using Abaqus, a finite element based code, and the fluid flow in the channels is modeled using Star-CCM+, a control volume based code. A loosely coupled (explicit) scheme is utilized in the FSI models in an effort to decrease runtime. However, utilizing a loosely coupled scheme amplifies solution-destabilizing behavior of the FSI models for the complex problem evaluated. In particular, incompressible turbulent flow around a slender geometry and comparable densities of the fluid and the structure contribute to solution instability. The iterative data exchange between the computational domains therefore needs to be managed very carefully depending on the spatial mesh used and the time step employed in the models. It was discovered that the range of stable time steps diminishes with increasing flow velocity. Furthermore, both an upper and lower limit of stable time steps was found through a range of flow velocities. These upper and lower limits on the time step are likely dependent on the spatial mesh used. The flat plate FSI models were compared to both curved plate models and experiment data collected using the Hydro-Mechanical Flow Loop at the University of Missouri. It was found that the curved plate models deflected less than the flat plate models by more than an order of magnitude. This is due to the increased stiffness of the curved plates. Comparison to the experiment shows the flat plate models deviating from the experiment data. It is hypothesized this is a result of assuming idealized geometry in the models when in reality the experiment has small geometric deviations. Thus, future FSI models should account for these geometric deviations in the experiment to help facilitate convergence of the experimental data with the model’s predicted solutions.