Evaluation of Anisotropic Eddy‐Viscosity Model Based on History Effect in Calculation of Pipe Swirling Flow Using OpenFOAM

ABSTRACT Swirling flow refers to a type of fluid motion characterized by spatial rotational movement around the longitudinal axis. This flow pattern can occur in various environments, including natural phenomena such as tornadoes and whirlpools, as well as in engineering applications like combustion chambers and mixing tanks. Due to its ability to enhance mixing and increase heat transfer, swirling flow holds significant engineering importance. However, accurately simulating swirling flows is challenging. The flow is inherently three‐dimensional and involves swirl decay, making numerical calculations complex. Among various computational approaches, Reynolds‐averaged Navier–Stokes (RANS) turbulence modeling offers a good balance between accuracy and computational cost. Several types of turbulence models are available within the RANS framework to close the system of equations. In this study, we implement a new linear turbulence model with anisotropic eddy viscosity in the OpenFOAM software. This model is formulated in cylindrical coordinates and incorporates the assumption of time non‐locality in swirling flows, making it particularly suitable for such applications. To enable implementation in OpenFOAM, which is based on a Cartesian coordinate system, we developed a methodology for translating the model equations from cylindrical to Cartesian form. The model's performance was evaluated using a test case of swirling flow in a pipe. Numerical calculations were carried out using three existing turbulence models available in OpenFOAM, along with the newly implemented model. The effect of the time non‐locality term on swirl decay prediction was also investigated by varying this term during computations. Additionally, a detailed analysis of how the viscous coefficients influence the velocity profiles distribution was conducted. The results demonstrate that the new model outperforms the existing models in terms of accurately predicting both velocity profiles and swirl decay.