Fluid-Structure Interaction in Naval Engineering
By Robert A. Sielski
In traditional ship structural analyses, the ship is statically balanced on a wave. In quasi-static loading by long waves, fluid-structure interaction is normally handled by a frequency-domain panel method hydrodynamic analysis code linked to a finite element analysis. Vibration calculations treat fluid-structure interaction through the concept of added mass effects. Dynamic wave impact calculations use computational fluid dynamics methods as well as the results of model and full-scale experiments. Fundamental challenges to the field of fluid-structure interaction simulation includ a theoretical understanding of the dominant and emerging coupling methods such as convergence properties, error bounds, computational scale and complexity, visualization and data transfer of huge datasets, and software engineering across legacy codes and pairing these codes to new high-performance computing architectures. There are two general approaches to fluid-structure interaction coupling, loose and tight coupling. Loose coupling is used for where the density of the fluid is much less than the density of the structure. Tight coupling is used for ship hydrodynamic loading, and it will involve either monolithic or partitioned coupling strategies. This paper, which is based upon a workshop that was held in July, 2016 in Ann Arbor, Michigan, reviews the current state-of-the art in fluid-structure interaction as seen from the differing perspectives of the workshop participants. A taxonomy of fluid-structure interaction is given and then a review is made of the application to three use cases, global ship loading and response, local ship loading and response, and propellers and appendages. Research needs are explored in the areas of improving generation and understanding of large data sets, limitations of current fluid simulation models, limitations of current structural simulation models, limitations of current coupling models, requirements for two-way coupling, needs for verification and validation and uncertainty quantification, future experimental needs, and training and human interaction challenges.