Simulating Wave-induced Whipping Responses of Ships

With the rapid growth of global trading and transportation by ships, the modern ship size has increased continuously with the largest container ships exceeding 400 m. This makes the large steel ships respond in ocean waves as flexible beams rather than rigid bodies. Such flexibility of large ships may lead to hull girder vibrations known as whipping response which makes the ships vulnerable to damage and even severe structural failure. The main challenge of modelling whipping of ships are related to the difficulties to simulate the highly nonlinear hydrodynamic process and to obtain the short-term and long-term extreme responses in random ocean waves. Previous related studies in the literature [1] have shown that using the inverse First Order Reliability Method (FORM) can significantly reduce the required computational efforts in CFD to achieve the targeted extreme whipping responses. This CFD+FORM framework uses efficient potential-flow solvers as a predictor to suggest critical wave episode, and a high-fidelity corrector based on free-surface CFD analysis. It has been demonstrated that the required three-hour real sea state time domain simulation can be decreased to 50-second sea state with most likely response wave [1]. However, some unexpected slamming events [2]have been observed in the earlier framework. We are working on the development of a more accurate and robust corrector using Star-CCM+ to simulate slamming induced hydroelastic responses of ships. Star-CCM+ is validated in the literature [3] to simulate the violent free surface accurately by the Volume of Fluid (VOF) method. Meanwhile, the overset mesh and morphing technology makes it more capable to deal with deformable boundaries. At present, we are developing a CFD model to solve the structural response based on ship beam theory. We build a beam model coupled with the flow solver to account for the flexibility of the hull girder using modal superposition method inside Star-CCM+ instead of interacting with another external CAE (computer-aided engineering) software, such as Abaqus. The deformation of the body surface is applied by morphing both the inside and outside boundaries of the overset mesh around the body. By this means, the efficiency of flow solver can be improved significantly since the present model avoids data exchange and mapping between the flow solver and CAE software at each time step. We will apply this approach to investigate the hydroelastical response of a flexible barge and a container ship.