Reconstruction and Prediction of Turbulence in the Atmospheric Boundary Layer Using Large-Eddy Simulations

Name: Pieter Bauweraerts

When: 04/09/2020 17:00
Language: en

Where: aula De Molen, 00.07, Kasteelpark Arenberg 50, 3001 Heverlee

Prof. dr. Johan Meyers (promotor)

Rapid development in simulation methodology, computational speed and measurements opens up possibilities for reconstruction, prediction and by extension control of turbulent flow fields in the atmospheric boundary layer (ABL). This can lead to applications such as monitoring the dispersion of pollutants, predicting the power of wind turbines, optimizing the power of a wind farm and reducing the structural loading on wind turbines. In this work we focus on the reconstruction and prediction of turbulence in a horizontally homogeneous ABL. For development and benchmarking purposes we use our in-house developed ABL turbulence simulation code SP-Wind as a virtual testing environment. SP-Wind uses the large-eddy simulation (LES) technique, which only simulates the large turbulent eddies, and models the effect of the small scales on the large scales with a subgrid-scale model.

We investigate the feasibility of using LES for real-time forecasting of instantaneous turbulent velocity fluctuations in an ABL. Although LES is generally considered computationally too expensive for real-time use, wall-clock time can be significantly reduced by using very coarse meshes. Thus, we focus on forecasting errors arising on such coarse grids, and investigate the trade-off between computational speed and accuracy. To this end, we set-up an idealized test scenario in which the forecasting error in an ABL are investigated based on a fine reference simulation, and a series of coarser LES grids. We find that errors only slowly increase with grid coarsening. A practical example, inspired by wind-energy applications, reveals that there is a range of forecasting horizons for which the variance of the forecasting error is significantly reduced, while at the same time, associated LES wall times are up to 300 times smaller than simulated time.

Thereafter, we investigate the reconstruction of a turbulent flow field in an ABL from a time series of lidar measurements, using LES and a state of the art data assimilation algorithm. Lidar sensors emit laser pulses, and can based on the travel time and frequency shift determine the wind speed simultaneously at around 100 locations along a line of a few kilometers.  Two lidar scanning modes are considered: a classical plan-position-indicator mode, which swipes the lidar beam in a horizontal plane, and a 3D pattern that is based on a Lissajous curve. To evaluate the approach, we construct a series of virtual lidar measurements from a fine-grid LES of an ABL. For both scanning modes we are succesfull in the reconstruction of turbulent velocity fields within the scanning area. For the 3D scanning pattern we are furthermore able to capture the large scale 3D turbulent flow structures.

https://lirias.kuleuven.be/handle/123456789/659089

• Prof. dr. Johan Meyers (promotor)
• Prof. dr. ir. Omer Van der Biest (chairman)
• Prof. dr. ir. Goele Pipeleers (secretary)
• Prof. dr. ir. Geert Lombaert
• Prof. dr. Jakob Mann , Technical University of Denmark
• Prof. dr. David Schlipf , Flensburg University of Applied Sciences