Eventos Anais de eventos
EPTT 2020
12th Spring School on Transition and Turbulence
Evaluation of an Anisotropic Turbulence Model for Simulation of Three-dimensional Cases
Submission Author:
Leonardo Machado da Rosa , SC
Co-Authors:
Leonardo Machado da Rosa, Daniela Koerich, Tommaso Oggian, Henry França Meier
Presenter: Leonardo Machado da Rosa
doi://10.26678/ABCM.EPTT2020.EPT20-0077
Abstract
Turbulence is a flow regime that can be observed in the most diverse situations. However, even though itis commonly found in nature and in industrial applications, its understanding is still challenging due to the fact that it involves a broad spectrum of scales. There are different ways to calculate the Reynolds tensor. In the Reynolds Stress Model (RSM), each component of this tensor is obtained by solving a separate transport equation, which implies greater computational cost when compared to simpler models such as those based on the Boussinesq hypothesis, which postulates that there is a linear relationship between the tensor itself and the strain tensor. The present study evaluates ananisotropic model to calculate the turbulent tensor. In the simulation of the cyclonic separator, all components associatedwith the tangential direction were obtained through the mixture length model, whereas the others through the k-Є model toguarantee the Galilean invariance of the Reynolds stress. This approach was implemented as a library in an open source CFD code. Tests were performed using an axisymmetric representation of a cyclone, adopting a transient formulation for the solution of the single-phase transport equations. The hybrid model predicted values closer to the experimental ones for the tangential velocity when compared against those obtained using the standard k-Є model, at similar computational cost. The hybrid model was also used to simulate the turbulent flow in a three-dimensional representation of the cyclone, hence demonstrating the applicability of the model to complex geometries.
Keywords
Modeling and Simulation, Turbulence, k-epsilon model, Mixing length model, Reynolds Tensor
