Coupled Fluid-Structure Interaction to model Three-Dimensional Dynamic Behaviour of Ships in Waves
An accurate prediction of global wave-induced motions and loads is crucial in the design of ships and offshore platforms. Increase in the size of ships and offshore platforms has resulted in ‘softer’ or flexible hulls which require their deformations to be taken into account when predicting motions and loads. In such cases, the deformation of the structure can significantly affect the flow field and the fluid loading, resulting in a fully coupled system. The study of this fully coupled fluid-structure interaction is termed as hydroelasticity. Hull girder flexibility can amplify the hull girder stress and contribute to significant fatigue damage and extreme hull girder loading.
Application of CFD (Computational Fluid Dynamics) to study wave-body interactions of ships and offshore structures using RANS (Reynolds-averaged Navier-Stokes equations) have increased over the years due to the increase in computational power. RANS solvers take into account the nonlinearities associated with free surface flows, viscous effects, as well as local flow features. The vast majority of investigations using RANS solvers for predicting wave-induced loads is carried out using one-way coupling, thus omitting hydroelastic effects. However, a fully coupled CFD/FE analysis has to be established to accurately model the responses of flexible floating bodies in waves. UoS continues to be the leader in hydroelasticity analysis through this and similar projects.
The computational time required to investigate a number of different wave conditions is now practicable due to supercomputers such as the Iridis cluster. For example, in the coupled simulations of a flexible barge, the CPU hour for 1 sec of real time simulations varied from 35 hours to 126 hours, depending on the total mesh size. Numerical simulations by applying the two-way coupling using commercial CFD software, Star-CCM+, and FE software, Abaqus is made feasible by the computational power offered by the Iridis Cluster.
Programming languages and libraries: Matlab