Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Modeling Turbulence in the Atmospheric Boundary Layer with Spectral Element and Finite Volume Methods

Ananias Tomboulides Matthew Churchfield, Paul Fischer, Michael Sprague, Misun Min

TL;DR

This work develops and validates LES-based turbulence models for the atmospheric boundary layer with direct relevance to wind-energy applications. It juxtaposes a high-order spectral-element code (Nek5000/RS) and a block-structured finite-volume code (AMR-Wind) on exascale-capable platforms, implementing multiple SGS strategies (MFEV with HPF, SMG, and TKE) and traction boundary conditions derived from Monin–Obukhov theory. The study demonstrates grid-convergence of bulk ABL parameters, cross-code consistency, and alignment with literature benchmarks, highlighting how high-order methods can achieve finer-scale resolution at reduced grid counts compared to lower-order schemes. These results support robust, exascale-ready simulations of stable ABL turbulence for wind-energy design and analysis, including LLJ dynamics and surface flux characteristics.

Abstract

We present large-eddy-simulation (LES) modeling approaches for the simulation of atmospheric boundary layer turbulence that are of direct relevance to wind energy production. In this paper, we study a GABLS benchmark problem using high-order spectral element code Nek5000/RS and a block-structured second-order finite-volume code AMR-Wind which are supported under the DOE's Exascale Computing Project (ECP) Center for Efficient Exascale Discretizations (CEED) and ExaWind projects, respectively, targeting application simulations on various acceleration-device based exascale computing platforms. As for Nek5000/RS we demonstrate our newly developed subgrid-scale (SGS) models based on mean-field eddy viscosity (MFEV), high-pass filter (HPF), and Smagorinsky (SMG) with traction boundary conditions. For the traction boundary conditions, a novel analytical approach is presented that solves for the surface friction velocity and surface kinematic temperature flux. For AMR-Wind, standard SMG is used and discussed in detail the traction boundary conditions for convergence. We provide low-order statistics, convergence and turbulent structure analysis. Verification and convergence studies were performed for both codes at various resolutions and it was found that Nek5000/RS demonstrate convergence with resolution for all ABL bulk parameters, including boundary layer and low level jet (LLJ) height. Extensive comparisons are presented with simulation data from the literature.

Modeling Turbulence in the Atmospheric Boundary Layer with Spectral Element and Finite Volume Methods

TL;DR

This work develops and validates LES-based turbulence models for the atmospheric boundary layer with direct relevance to wind-energy applications. It juxtaposes a high-order spectral-element code (Nek5000/RS) and a block-structured finite-volume code (AMR-Wind) on exascale-capable platforms, implementing multiple SGS strategies (MFEV with HPF, SMG, and TKE) and traction boundary conditions derived from Monin–Obukhov theory. The study demonstrates grid-convergence of bulk ABL parameters, cross-code consistency, and alignment with literature benchmarks, highlighting how high-order methods can achieve finer-scale resolution at reduced grid counts compared to lower-order schemes. These results support robust, exascale-ready simulations of stable ABL turbulence for wind-energy design and analysis, including LLJ dynamics and surface flux characteristics.

Abstract

We present large-eddy-simulation (LES) modeling approaches for the simulation of atmospheric boundary layer turbulence that are of direct relevance to wind energy production. In this paper, we study a GABLS benchmark problem using high-order spectral element code Nek5000/RS and a block-structured second-order finite-volume code AMR-Wind which are supported under the DOE's Exascale Computing Project (ECP) Center for Efficient Exascale Discretizations (CEED) and ExaWind projects, respectively, targeting application simulations on various acceleration-device based exascale computing platforms. As for Nek5000/RS we demonstrate our newly developed subgrid-scale (SGS) models based on mean-field eddy viscosity (MFEV), high-pass filter (HPF), and Smagorinsky (SMG) with traction boundary conditions. For the traction boundary conditions, a novel analytical approach is presented that solves for the surface friction velocity and surface kinematic temperature flux. For AMR-Wind, standard SMG is used and discussed in detail the traction boundary conditions for convergence. We provide low-order statistics, convergence and turbulent structure analysis. Verification and convergence studies were performed for both codes at various resolutions and it was found that Nek5000/RS demonstrate convergence with resolution for all ABL bulk parameters, including boundary layer and low level jet (LLJ) height. Extensive comparisons are presented with simulation data from the literature.
Paper Structure (8 sections, 39 equations, 21 figures, 1 table)

This paper contains 8 sections, 39 equations, 21 figures, 1 table.

Table of Contents

  1. Introduction
  2. Large Eddy Simulation Model
  3. ABL GABLS Benchmark
  4. SGS Modeling Approaches in Nek5000/NekRS
  5. Traction Boundary Condition
  6. Convergence and Verification Tests
  7. GABLS results from Nek5000/NekRS and AMR-Wind
  8. Conclusion

Figures (21)

  • Figure 1: NekRS simulation for the atmospheric boundary layer flows with particle tracer (Simulation by Lidquist lindquist21.
  • Figure 2: Nek5000/RS Convergence in horizontally averaged velocity with (a) MEFV/HPF and (b) MFEV/SMG and potential temperature profiles at $t=7h$ with (c) MEFV/HPF and (d) MFEV/SMG.
  • Figure 3: Nek5000/RS Comparison between mean profiles obtained with MEFV/HPF and MFEV/SMG with resolution (a) $n=128^3$, (b) $n=256^3$, (c) $n=512^3$, and (d) $n=1024^3$ .
  • Figure 4: Nek5000/RS horizontally averaged (a) streamwise, spanwise velocities and (b) potential temperature at $t=6h$ using MFEV/SMG and MFEV/HPF with traction boundary conditions, compard with AMR-Wind for $512^3$.
  • Figure 5: Nek5000/RS Convergence for MEFV/SMG and MFEV/HPF.
  • ...and 16 more figures