Aerodynamics

Being driven by the wind, the study of wind turbines and wind farms is inherently highly dependent on aerodynamic modeling capability. This topic area is highly coupled to much of the research and development work within the wind energy industry. Aerodynamic design is used to optimize the performance of wind turbines, balance the loads in high wind or high turbulence events to minimize damage to the whole system and improve reliability, and improve the energy recovery downstream of a wind turbine to minimize the wake loss throughout the wind farm. In general, aerodynamics models choose between fidelity and computational cost, and practitioners should also find the balance that fits their needs.

Software Stack

The study of aerodynamics spans various technical areas and applications.

%%{init: { 'logLevel': 'debug', 'theme': 'forest' } }%% mindmap root("Aerodynamics") ps(Engineering-fidelity Physics Models) OpenFAST FAST.Farm OWENS hfm(High Fidelity Models) AMR Wind Nalu Wind ERF WindSE do(" ") ctrl(Controls) FLORIS ra(" ") cost(" ") wild(" ") om(" ")

In addition to separating by application, this software stack can be divided by fidelity. The following graphic roughly orders this tool suite from high fidelity at the top to low fidelity at the bottom.

flowchart TD A["Nalu-Wind LES, time accurate, blade resolved"] B["AMR-Wind LES, time accurate, actuator line"] C["WindSE RANS, LES, time accurate, actuator disk"] D["OpenFAST/FAST.Farm/OWENS Engineering models, time accurate, BEM-derived models + FVW for HAWTs and VAWTs"] E["FLORIS Analytical, steady state"] subgraph ExaWind A B end A ~~~ C B ~~~ C C ~~~ D D ~~~ E

ExaWind, part of the DOE Exascale Computing Project, has developed wind energy simulation software capable of leveraging state of the art supercomputers achieves exascale performance (10^18 floating point operations per second). This software is made up primarily of AMR-Wind and Nalu-Wind. AMR-Wind is a large eddy simulation (LES) code that uses adaptive mesh refinement through AMReX to resolve the wind farm flow field. Nalu-Wind is another LES code that uses a fixed, unstructured mesh to resolve the flow field around the blade. ExaWind brings the two codes together through advanced numerical methods to efficiently model the large spread in spatial and temporal scales in the full wind farm. ExaWind couples with OpenFAST to simulate fluid-structure interactions (FSI).

WindSE is a wind farm systems engineering tool that uses a variety of computational fluid dynamics (CFD) models to simulate the wind farm flow field including Reynolds-averaged Navier-Stokes (RANS), LES, and an actuator disk model. It is coupled to optimization algorithms for use in wind farm layout design.

OpenFAST includes actuator disk-based aerodynamics models including Blade Element Momentum theory (BEMT) and Dynamic BEM (DBEMT).

FAST.Farm is a farm-level simulator building on OpenFAST, and it includes a free vortex wake (FVW) model to simulate the wake of wind turbines within a farm. OpenFAST also includes TurbSim, a stochastic turbulence model, and InflowWind, a tool for generating wind inflow conditions or OpenFAST simulations.

OWENS includes Actuator Cylinder (AC) and Double Multiple Streamtube (DMS) aerodynamics models including 3-D and unsteady approximations, with corrections for curved and deforming blades. OWENS is also coupled to the OpenFAST free vortex wake model for higher fidelity and multi-turbine capabilities.

FLORIS is a low fidelity simulator that includes steady state analytical wind turbine wake models. While this is explicitly not aerodynamics, this tool is useful for rapidly estimating and optimizing wind farm layout and controls.