Numerical investigation and parametric analysis of an attached eddy model applied to inlet condition

Generating a realistic turbulent field at the inflow is of great importance as well as a complex challenge for large-eddy simulation. As a new synthetic turbulence method, the attached eddy model (AEM) was initially proposed by Townsend, where the velocity field is induced by a hierarchy of randomly distributed lambda-shape eddies by using the Biot-Savart law. Although extensive research has theoretically proved the existence and effectiveness of AEM, there have been a few numerical investigations on its practical applications. In this paper, the AEM method is implemented in an open-source software code_saturne to generate inlet conditions. The AEM generation process is detailed and described by defining various parameters. The new model is then applied to turbulent channel flows with Re-tau = 180, 395, and 590, respectively. The results are compared with the direct numerical simulation to validate its ability to accurately predict the velocity and turbulent kinetic energy profiles. It is also compared with the simulation by using the synthetic eddy method to assess its potential to improve the drop process of the wall shear stress. AEM is shown to be efficient in developing turbulent kinetic energy in the near-wall region. Furthermore, a parametric analysis of the present model is carried out to discuss in detail the specific effect of each factor. This parametric analysis shows the impact of different model settings on the velocity fluctuation. These results are expected to inspire future practical applications of AEM. Published under an exclusive license by the AVS.

Automated summary

This paper introduces a new synthetic turbulence method called AEM and applies it to generate inlet conditions in the open-source software code_saturne. By comparing with direct numerical simulation and synthetic eddy method, the ability of AEM to accurately predict turbulent kinetic energy profiles is validated. Parametric analysis is also conducted to discuss the specific effect of each factor.