Zhi Jian Wang

High-order methods have demonstrated orders of magnitude reduction in computational cost for large eddy simulation (LES) over low-order methods in the past decade. Most such simulations are wall-resolved implicit LES (ILES) without an explicit sub-grid scale (SGS) model. The use of high-order ILES for severely under-resolved LES such as wall-modeled LES (WMLES) often runs into robustness and accuracy issues due to the low dissipation embedded in these methods. In the present study, we investigate the performance of several popular SGS models, the static Smagorinsky model, the wall-adapting local eddy-viscosity (WALE) model and the Vreman model, to improve the robustness and accuracy of under-resolved LES using high-order methods. The models are implemented in the high-order unstructured grid LES solver called hpMusic based on the discontinuous flux reconstruction method. The length scales in these SGS models are calibrated using the direct numerical simulation (DNS) database for the turbulent channel flow problem. The Vreman model has been found to produce the most accurate and consistent results with a proper choice of the length scale for WMLES.

A passive flow control technique of utilizing an equilateral triangular trip close to the leading edge was developed and tested for a micro-scale Eppler E423 airfoil at a Reynolds number based on the chord of 40 000. The analysis was carried out via high-order wall-resolved large eddy simulation using the computational solver HpMusic. Angles of attack of 5° and 20° were tested. It was shown that at an angle of attack of 5°, the trip height of two times the local boundary layer thickness outperformed existing passive flow control techniques from the literature by almost a factor of five in terms of the lift-to-drag ratio. To understand the underlying physics which allowed the trip to provide this very significant performance benefit, metrics such as pressure coefficient profiles, oil flows, iso-surfaces of Q-criteria, and leading-edge flow behavior were examined. It was found that this trip configuration simultaneously removed the flow separation regions on both the suction and pressure sides of the wing.

In the present study, an algebraic equilibrium wall model previously developed for hexahedral elements is extended to handle mixed meshes including prismatic, tetrahedral, and pyramidal elements in the context of a discontinuous high-order method. This extension is needed for complex geometries, for which high-order mixed elements (e.g., tetrahedral and pyramidal elements) are often necessary near solid walls to avoid meshing challenges. Various design decisions are discussed to achieve the best performance on massively parallel CPU/GPU architectures for a production-level high-order large-eddy simulation solver based on the flux reconstruction/correction procedure via reconstruction method, hpMusic. The extension to other elements is first evaluated using a benchmark channel flow problem at various Reynolds numbers. After that, flow over the NASA high-lift Common Research Model (CRM-HL) from the 4th AIAA High-lift Prediction Workshop is computed to further test the new implementation. Computational results at the third- and fourth-order accuracies are compared with experimental data.

View Video Presentation: https://doi.org/10.2514/6.2022-1201.vid In this paper, we use a discontinuous high-order method to evaluate three error indicators specifically designed for mesh adaptation in large eddy simulations. The indicators are based on different physical and numerical arguments. The first indicator is based on the unsteady residual. The second indicator is based on a local smoothness indicator. The third indicator is based on an estimate of the small scale turbulent kinetic energy. We test the three error indicators to drive autonomous mesh refinement on implicit large eddy simulations, deploying the error indicators on the widely used benchmark problem T106c. We compare the adaptively meshed results against a DNS simulation to evaluate the effectiveness of the error indicators in reducing solution errors in LES.

View Video Presentation: https://doi.org/10.2514/6.2022-1202.vid Artificial turbulent inflow is often generated to reduce the computational cost of large eddy simulation while maintaining certain turbulent flow characteristics. As one of the popular turbulent inflow generators, the digital filter approach is known for its efficiency and flexibility. It has mostly been used with low-order or structured-mesh flow solvers. In this paper, we extend it to unstructured meshes with the high-order flux reconstruction method. Efficient solution interpolations are employed to transfer data between the structured and unstructured meshes in a distributed-memory parallel environment. Test cases are used to demonstrate the capability of the digital filter approach with the high-order flux reconstruction method.