Numerical simulation of a flexible net in currents with the smoothed particle hydrodynamics method

Abstract

The aim of this study is to evaluate the applicability and suitability of the Smoothed Particle Hydrodynamics (SPH) method in modelling a flexible flat net in currents. SPH is a meshless Lagrangian method that discretises the system into particles, making it particularly suitable for modelling turbulent and highly nonlinear flows with the presence of a free surface. For the modelling of the flexible net, a methodology is proposed that uses fluid-driven elements (spheres) and dynamic links to discretise the geometry of the net within the SPH framework. This numerical approach allows the determination of the deformation of the net elements in the numerical model, the analysis of the fluid velocity field through the discrete elements and the calculation of the drag forces on the system under different flow regimes. The proposed method for the numerical modelling of the net is based on the coupling of DualSPHysics and MoorDyn+. DualSPHysics is a fully Lagrangian and meshless open-source code based on weakly compressible SPH that solves the fluid-structure interaction. MoorDyn+ is a dynamic library based on a lumped-mass numerical model, which solves tensions in the lines and allows the establishment of connections between the floating elements. In addition, an in-house algorithm is used to generate the configuration of the net in the numerical model, preserving the hydrodynamic and structural properties of the studied net. The proposed method for the discretisation of the net is validated by a comparative analysis of the drag forces obtained in the numerical model for a flat net fixed in a rigid frame, with different solidity values and those obtained by analytical solutions. Then, the method is applied to simulate a flexible flat net in steady currents with different velocities. The results are compared with those obtained from physical modelling tests and show a satisfactory agreement of the proposed method in the majority of the cases analysed, with a difference of less than 11%.