High performance computational marine hydrodynamics for renewable energy

Significant progress has been made in renewable energy (RE) innovation in the past decades, for which numerical simulation tools continue to play significant roles, especially for fast prototyping and innovative technology development. Today, to meet the fast-growing demand for productivity and solve the ever-complex problems emerging in engineering practice, high-performance simulation tools are in great demand to support those above and empower RE research and development. Most of these tools require a substantial long research and development (R&D) cycle along with a solid background in both fluid mechanics and programming skills. Efforts made to develop efficient simulation tools hence are relatively rare and specialized expert simulation software, are also in great demand in the high-performance software community. Therefore, the motivation of this PhD work is to develop both marine RE technology and simulation tool by carrying out highly efficient computational marine hydrodynamics (CMH) study, which covers numerical tool development, research in RE capability enhancement, and a cloud-aided platform to accelerate the execution of numerical hydrodynamic codes. CMH simulation tools, such as propeller software, are usually developed to design and analyse marine propulsion, which has complex geometry generation component, run data processing component and computational kernel (i.e., the hydrodynamic kernel). The purpose of this study is to investigate whether the CMH codes used for propellers can be enhanced to predict the performance of turbine blades efficiency and their hydrodynamics characteristics. The theory of similarity between propeller and turbine is established. A 4- quadrant computational model was created in terms of forwarding, reverse shaft revolution and the direction of inflow speeds to produce four operation quadrants for a propeller-rotor. The effective angle of attack concept was used to analyses the instantaneous flow condition for an unsteady 3D foil with both pitch and heave along with translation motion. For these models, validation and verification against measurements were carried out, ranging from the induced velocity at a designated location downstream, velocity profile downstream, to the effect of the number of blades on hydrodynamic efficiency. The panel method code used as example in this study has been proven accurate and robust for many propellers tested in the past. The numerical tool then was used to address two emerged issues in tidal and wind turbine problems. Firstly, as propulsion system is one of the primary pollution sources on board a vessel, in order to reduce environmental impact in shipping, an innovative dual-mode propeller technology was developed. The results obtained from the dual-mode propeller-tidal turbine indicate that the optimized fixed pitch propeller can be performed as both propulsion and tidal/current turbine with a balanced efficiency in both modes for low-speed ships, especially for yachts. The balanced, relatively high-power productivity and propulsive performance are achievable for low-speed ships anchored in current or regular sailboat for which a propeller is used as a towed turbine. Secondly, because the existing wind turbine blades are prone to damage due to severe conditions and the costs associated with repairing them is high, a new capability is enhanced, which fully coupled aero-hydrodynamic performance and the structural strength as a new numerical model. Validations were also carried out, which show capability and reliability on aerodynamic prediction of wind turbine for optimization with coupled aerodynamics and structural strength optimization target. A 10-meter diameter wind turbine from NREL was used as a base design example and was optimized under an extreme wind speed of 100 km/hr. The results achieved a total savings of 18.72 percent in blade material. The introduction of advanced simulation technologies requires large amounts of performance data hence produces a large run matrix. These new challenges require reliable and fast computational technology deployment in the local infrastructure or cloud. In this project, a tailored container-based private cloud platform with two key strategies is designed and implemented. The results show that compared with the computing scheme of a traditional HPC workstation, this solution can accelerate the speed and shorten makespan by six times. The new method is scalable to address above two computing cases up to a run matrix size 10\\(^7\\). Overall, it is concluded that this research has developed a reliable and accurate renewable energy expert numerical simulation system by enhancing existing CMH codes, solved some key problems in turbine design and optimization area, including software tool enhancement, zero-emission shipping technology development, structure fracture/failure of offshore wind turbines. This research improved numerical simulation productivity by using containerization technology. These solutions greatly shorten the R&D cycle of marine RE research and development and are expected applicable in other science and engineering disciplines.