| dc.description.abstract | The state-of-the-art actuator line modeling (ALM) method has been used extensively in the past decade to perform large-eddy simulations (LES) of wind turbines as fully resolved blade simulations require complex rotating meshes and are computationally expensive. The ALM method represents rotor blades as the line of the elements. Each element actuates forces that are projected onto the flow field as a body force using volumetric projection. Thus, the thesis mainly focuses on smoothing the radius of actuator force projection along the blade span because the rotor power and blade tip loading are susceptible to this ALM parameter. The commonly used constant chord and grid sizes method for the volumetric force projection radius overpredicts the blade tip loading on the wind turbine. As a result, the researcher developed the elliptic distribution method to improve the prediction of blade tip loading, but it has not been used extensively. Therefore, these three methods are applied to the NTNU model wind turbine and the MEXICO rotor to prove their validity. The NTNU model wind turbine has a smaller size and aspect ratio than the MEXICO rotor. The results showed that the elliptic chord distribution method on the NTNU wind turbine gave a better power and blade tip loading prediction than the constant chord and grid-based Gaussian radius method. However, the three Gaussian radius methods gave the same results for the power and blade tip loadings prediction on the MEXICO rotor. The time step and grid size selection for LES-ALM simulation have also been discussed. The recent shift towards the multi-rotor concept has been observed in the offshore wind turbine industry as it can potentially decrease the levelized cost of energy due to its smaller components which are easier to handle. Hence, the second part of the thesis measures the aerodynamic performance of the multi-rotor system in comparison to the single rotor. The ALM results showed an increase in individual rotor power and blade loading by 4% and 2%, respectively, for multi-rotor compared to the single rotor. In addition, less velocity deficit is observed for the two rotors compared to the single rotor. | |
