Abstract
The aim of this work is to predict numerically the turbulent flow through a straight square duct using Reynolds Average Navier-Stokes equations (RANS) by the widely used k – ε and a near wall turbulence k – ε − fμ models. To handle wall proximity and no-equilibrium effects, the first model is modified by incorporating damping functions fμ via the eddy viscosity relation. The predicted results for the streamwise, spanwise velocities and the Reynolds stress components are compared to those given by the k – ε model and by the direct numerical simulation (DNS) data of Gavrilakis (J. Fluid Mech., 1992). In light of these results, the proposed k – ε − fμ model is found to be generally satisfactory for predicting the considered flow.
Highlights
The prediction of turbulent flows involving secondary motions and using numerical simulations of the Reynolds Average Navier-Stokes (RANS) equations has great practical value in fluid mechanics and many applications can be found in centrifugal machinery design
10:17 am Numerical simulation of turbulent flow through a straight square duct using a near wall linear k – ε model based on the linear stress-strain relation initially proposed by Boussinesq
We show the comparisons between the linear k – ε model, the k – ε – f μ model, direct numerical simulation (DNS) of Gavrilakis [10] and the experimental data of Cheesewright et al [27]
Summary
The prediction of turbulent flows involving secondary motions and using numerical simulations of the Reynolds Average Navier-Stokes (RANS) equations has great practical value in fluid mechanics and many applications can be found in centrifugal machinery design. By far the most popular turbulence model is the standard k − ε model which is Numerical simulation of turbulent flow through a straight square duct using a near wall linear k – ε model based on the linear stress-strain relation initially proposed by Boussinesq. This type of closure has been revealed robust and efficient with respect to CPU time than more highorder models [1,2].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have