Wind Turbine Large-Eddy Simulations on Very Coarse Grid Resolutions using an Actuator Line Model

TL;DR

This paper investigates the accuracy of the Actuator Line Model in large eddy simulations of wind turbines at very coarse grid resolutions, highlighting challenges and proposing a theoretical framework for improvement.

Contribution

It introduces a theoretical approach to compare filtered and unfiltered quantities in ALM, aiming to enhance simulation accuracy at coarse resolutions.

Findings

Flow predictions match wind tunnel data at coarse resolutions

ALM predictions of power output are affected by rotor-scale flow induction

Theoretical scaling relations help understand filtered versus unfiltered quantities

Abstract

In this work the accuracy of the Actuator Line Model (ALM) in Large Eddy Simulations of wind turbine flow is studied under the specific conditions of very coarse spatial resolutions. For finely-resolved conditions, it is known that ALM provides better accuracy compared to the standard Actuator Disk Model (ADM) without rotation. However, we show here that on very coarse resolutions, flow induction occurring at rotor scales can affect the predicted inflow angle and can adversely affect the ALM predictions. We first provide an illustration of coarse LES to reproduce wind tunnel measurements. The resulting flow predictions are good, but the challenges in predicting power outputs from the detailed ALM motivate more detailed analysis on a case with uniform inflow. We present a theoretical framework to compare the filtered quantities that enter the Large-Eddy Simulation equations as body forces with a scaling relation between the filtered and unfiltered quantities. The study aims to apply the theoretical derivation to the simulation framework and improve the current results for an ALM, especially in the near wake where the largest differences are observed.