Dynamic Analysis of Submerged Structures

Abstract

There are a wide range of situations where a structural component is in contact with or submerged in water. Offshore wind turbines, ship hulls, or components in nuclear reactor pools are examples of such structures. Being able to understand and model the behavior of these structures is of great interest. The behavior of a submerged structure can be drastically different from its behavior in air. When comparing the dynamic response of a structure in air and in water, the main effects of submersion are increased damping and decreased eigenfrequency. Added mass and added damping (AMAD) are well-known concepts that are widely used in the analysis of submerged structures. These concepts are used to emulate the interaction between a structure and surrounding water. Analytical expressions for the added mass of simple geometrical shapes undergoing rigid body translation in a fluid have been derived and can be found in design norms e.g. ASME N-1311. These analytical expressions are often applied in an approximate manner to more complex structures for simplicity. Approximate values of damping, based on experiments, can be found in design norms for different types of general structures. The AMAD model is a computationally effective way of emulating the effects of submersion. Other methods for simulating fluid-structure interaction effects include the use of acoustic finite elements and coupled CFD (computational fluid dynamics) and FE (finite element) solvers. The viability and performance of these three methods have been evaluated in this thesis. The purpose was to investigate alternative and more accurate modeling techniques for the dynamic analysis of submerged structures. The dynamic behavior of a simple geometry was analyzed using the different methods. The results were compared to experimental data for validation. Acoustic finite elements and coupled CFD-FEM have proven to be accurate alternatives to AMAD that could be used more in future work. However, the increased computational cost and complexity of the alternative models call for careful consideration when deciding to use them.