Wind Field Simulation and Fluid-Structure Interaction in Wind Farms

Abstract

Wind energy as a renewable source has been a topic of immense interest over the past few decades. It is a well known fact that the wake structure influences the performance and operation of the wind turbines located further downstream. However, modelling the wind field as a whole along with the wind turbines involves complex geometrical models which require extensive computational resources, both in terms of memory and time. Thus, researchers have proposed different analytical models based on certain assumptions which, in some cases, have certain parameters involved which are difficult to obtain. This calls for an in-between approach such that the actual model can be built at a low cost and at the same time satisfying the theoretical requirements of the analytical models. There are certain mathematical approaches which can reduce this computational expense to a considerable extent and thus, certain special fluid-structure interaction (FSI) methods come into play. However, because of the size of a wind farm, the amount of data to be processed still remains substantial which can be easily handled computationally only if parallel numerical algorithms coupled with multi-core simulations are implemented.

A new model for laminar wind flow with vorticity and wake interaction has been proposed and its advantages over the potential flow model has been presented. This model has the ability to deal with both rotational and irrotational flow in 2D. This model coupled with a FSI method and parallel numerical algorithms has been used to simulate back-to-back actuator discs. The results were found to bear a nice resemblance with the analytical wind farm models with the added advantage of visualizing the velocity field under the light of vorticity. The added advantage is that unlike blade-resolved models which require modelling a complex mesh around the actuator disc, this model is simpler geometrically and as such is not expensive computationally. A linear spatial turbulence model has also been proposed based on Rapid Distortion Theory (RDT). The approach to develop the model uses Gaussian closure. The model has been simulated for a 2D steady fluid flow problem and the simulated results were found to be consistent. The laminar flow model with vorticity and wake interaction and the spatial turbulence model can form the basis of input to an individual wind turbine in a wind farm for subsequent analysis.