A numerical wave flume based on the Consistent Particle Method (CPM) is presented to model extreme wave impact on a fixed platform in a wave flume. One distinct advantage of CPM is that it computes spatial derivatives based on Taylor series expansion, achieving better accuracy and hence largely reducing spurious pressure fluctuation. A wave absorbing layer is used to reduce the computational domain while maintaining the non-reflection boundary. For validation purpose, an experimental study is conducted. The experimental results show that the developed numerical wave flume can simulate highly deformed plunging waves and wave impact pressure with good accuracy.
Extreme waves possess tremendous destructive power and may cause serious damages to ocean platforms such as semi-submersible and tension-leg platform (Rudman and Cleary, 2013). Therefore, accurate simulation of extreme wave motion and impact force is necessary for safe and cost-effective design of platform structures. Many numerical methods have been developed to simulate wave hydrodynamic problems. Compared to mesh-based methods (Lin and Liu, 1998; Xue and Lin, 2011; Wang et al., 2015), the particle methods (Shao and Lo, 2003; Khayyer et al., 2009; Liu et al., 2013) are better in simulating water waves of large deformation because of the meshless nature. In addition, by adopting the governing equation of Lagrangian type, particle methods can avoid the numerical dissipation induced by the discretization of the convection term in the governing equations.
In many research works on particle methods, however, the approximation of partial differential operators requires pre-defined kernel functions. When particle distribution is highly non-uniform in violent fluid motion, this approach induces significant errors in the computed spatial derivatives. The errors cause unphysical pressure fluctuation, which is one of the key issues in particle methods. To address this problem, the Consistent Particle Method (CPM) has been developed by adopting the Taylor series expansion to compute the spatial derivatives in governing equation (Koh et al., 2012). This way of computing derivatives has been demonstrated to be of good accuracy and hence CPM has largely overcome the problem of spurious pressure fluctuation.
