Coupling between the DualSPHysics solver and multiphysics librarieimplementation, validation and real engineering applications

  1. Martínez Estévez, Iván
Supervised by:
  1. Alejandro Jacobo Cabrera Crespo Director
  2. José Manuel Domínguez Alonso Director

Defence university: Universidade de Vigo

Fecha de defensa: 23 May 2024

Committee:
  1. María Teresa de Castro Rodríguez Chair
  2. Corrado Altomare Secretary
  3. Lorenzo Cappietti Committee member

Type: Thesis

Abstract

Wave Energy Converters (WECs) are devices designed to harness wave energy from both coast and offshore. These devices include a power take-off (PTO) system to convert the wave energy into electricity. Probably the biggest challenge for wave energy is to ensure the efficiency and survivability of WECs by making the most of the energy potential of waves. For this reason, there is the need of using tools that facilitate the tasks of design and analysis of the behaviour of WECs. Computational Fluid Dynamics (CFD) is a fundamental tool in that allows numerical tests to be carried out without the need to build scale models or real prototypes, thus saving time and money. Numerical models cannot replace physical tests, but they can significantly reduce their number and at the same time provide information that is difficult or impossible to measure in real tests. CFD models can be divided into mesh-based and mesh-free models. Among the mesh-free models, the Smoothed Particle Hydrodynamics (SPH) method is worth mentioning. The SPH method can perform simulations of fluid-structure interactions with high accuracy, where the free surface does not require special treatment and high deformations of the free surface can be solved without the problems that would appear in mesh models. However, one of the main limitations of these models is the high computational cost needed to solve the calculations. DualSPHysics model is a CFD based on the SPH method, oriented to the simulation of free surface fluids and their interaction with fixed and floating structures. It is a reference in the field of coastal engineering. Its main fields of application are: the study of wave energy generating devices, the evaluation and study of floating devices, the design of dykes and coastal protection and the interaction of waves with coastal structures. Normally, the use of a single model is not sufficient to simulate complex real-world problems because different physical processes or mechanisms not solved by CFD are involved. Therefore, coupling DualSPHysics with other models is necessary. MoorDyn+ is a mooring library based on the MoorDyn code. The development of MoorDyn+ arose from the need to couple DualSPHysics with a library that solves the mooring dynamics needed to simulate floating devices that are moored to the seabed. Project Chrono is a model that provides a multi-physics simulation engine. The application areas in which Project Chrono is most frequently used are vehicle dynamics, robotics and machine design. This model is able to simulate a large number of mechanical problems of varying complexity with high accuracy and efficiency, such as: rigid and deformable objects, collision detection, friction support, springs, joints, etc. Despite the existence of the DualSPHysics couplings with Project Chrono and MoorDyn+, there is a need to extend the existing functionalities in order to simulate more complex scenarios, such as reproducing WEC devices, PTO systems, simulation of flexible objects and simulation of mooring lines called tensors. Additionally, these models are computationally very slow by their nature, which means that they need to be accelerated by applying high-performance computing (HPC) techniques. The research work during the thesis period will focus on two main lines: i) the development of SPH computational code ideal for studying wave-floating structure interaction, which allows to evaluate and study the efficiency and survivability of WECs and ii) an optimisation and acceleration of the models to make them more computationally efficient by applying HPC techniques.