The availability of high-performance computational resources has increased steadily, but we are still far from the capacity to perform high-fidelity simulations for turbulent flows in real-world applications. Thus, we still rely on computationally cheaper surrogates like Reynolds-Averaged Navier-Stokes (RANS) turbulence modeling. The most commonly used RANS models are the linear eddy viscosity models (LEVM), which rely on the turbulent vis- cosity hypothesis for their Reynolds stress closure, a known source of structural uncertainty. Despite the development of theoretically superior turbulence models such as algebraic mod- els or Reynolds stress transport models, the LEVMs remain the most widely used class of turbulence models due to their efficiency and stability. This work combined a nonlinear eddy viscosity model with a deep neural network to yield improved predictions of the anisotropy tensor on flow cases with surface curvature and flow separation traditionally challenging to LEVMs. The neural network used an extensive set of rotationally invariant local flow fea- tures for predictions and incorporated realizability constraints in the training process. Using visualizations based on the barycentric map, our results indicate that the proposed machine learning method’s anisotropy tensor predictions offer a significant improvement over the best performing LEVM (baseline) and compare very well with the DNS/LES (ground truth). The predicted anisotropy tensor could uncover secondary flow information in some flow cases, which the baseline model completely missed. However, in a significant number of cases, the improved predictions did not translate into an improvement of the mean velocity and pressure fields when measured against the best-performing LEVM.     «

The availability of high-performance computational resources has increased steadily, but we are still far from the capacity to perform high-fidelity simulations for turbulent flows in real-world applications. Thus, we still rely on computationally cheaper surrogates like Reynolds-Averaged Navier-Stokes (RANS) turbulence modeling. The most commonly used RANS models are the linear eddy viscosity models (LEVM), which rely on the turbulent vis- cosity hypothesis for their Reynolds stress closur...     »

The availability of high-performance computational resources has increased steadily, but we are still far from the capacity to perform high-fidelity simulations for turbulent flows in real-world applications. Thus, we still rely on computationally cheaper surrogates like Reynolds-Averaged Navier-Stokes (RANS) turbulence modeling. The most commonly used RANS models are the linear eddy viscosity models (LEVM), which rely on the turbulent vis- cosity hypothesis for their Reynolds stress closure, a known source of structural uncertainty. Despite the development of theoretically superior turbulence models such as algebraic mod- els or Reynolds stress transport models, the LEVMs remain the most widely used class of turbulence models due to their efficiency and stability. This work combined a nonlinear eddy viscosity model with a deep neural network to yield improved predictions of the anisotropy tensor on flow cases with surface curvature and flow separation traditionally challenging to LEVMs. The neural network used an extensive set of rotationally invariant local flow fea- tures for predictions and incorporated realizability constraints in the training process. Using visualizations based on the barycentric map, our results indicate that the proposed machine learning method’s anisotropy tensor predictions offer a significant improvement over the best performing LEVM (baseline) and compare very well with the DNS/LES (ground truth). The predicted anisotropy tensor could uncover secondary flow information in some flow cases, which the baseline model completely missed. However, in a significant number of cases, the improved predictions did not translate into an improvement of the mean velocity and pressure fields when measured against the best-performing LEVM.     «

The availability of high-performance computational resources has increased steadily, but we are still far from the capacity to perform high-fidelity simulations for turbulent flows in real-world applications. Thus, we still rely on computationally cheaper surrogates like Reynolds-Averaged Navier-Stokes (RANS) turbulence modeling. The most commonly used RANS models are the linear eddy viscosity models (LEVM), which rely on the turbulent vis- cosity hypothesis for their Reynolds stress closur...     »