Abstract

A numerical study has been carried out to investigate the combined forced and natural convection heat transfer in a differentially heated 3D obstructed cavity with a thermally insulated rotating circular cylinder. The cavity has a hot stationary bottom wall and a cold top lid-driven wall, and all the other walls completing the domain are motionless and adiabatic. The simulations are performed for different Reynolds numbers, Re=5000, 10,000, 15,000 and 30,000, and for dimensionless rotational speeds of the cylinder, 0≤Ω≤10. The performance of two turbulence methods, Large Eddy Simulation (LES) and Unsteady Reynolds-Averaged Navier-Stokes (URANS), has been evaluated in this research. The flow and thermal fields are studied through flow vectors, isotherm contours and iso-surfaces temperature, as well as through the average Nusselt number (Nuav) and velocity components. The results demonstrate clearly that the flow patterns and the thermal fields are influenced strongly by increasing either the rotating cylinder speed or the Reynolds number. Furthermore, both LES and URANS solutions can capture the essential feature of the primary eddies in the cavity. But this study has shown convincing evidence that only the LES method can predict the structure details of the secondary eddies that have profound effects on the heat transfer behaviour within the enclosure.

Highlights

  • Over the last few decades, the numerical simulations of turbulent flows and heat transfer have become one of the essential attentions of engineering applications in the industrial and engineering fields

  • The commercial code ANSYS©FLUENT [39] was adapted to complete the simulations. Both steady and unsteady Reynolds182 averaged Navier–Stokes equations were solved besides the large eddy simulation method

  • The final chosen number of grid points in the current study was 929160, which was proved satisfactory by different indicators that are important in order to obtain high quality results: the non-dimensional time step is 0.004, the dimensionless wall distance y+ ≈ 1 and the Courant-Friedrichs-Lewy number CFL = 0.3

Read more

Summary

Introduction

Over the last few decades, the numerical simulations of turbulent flows and heat transfer have become one of the essential attentions of engineering applications in the industrial and engineering fields. Large Eddy Simulation (LES) method is increasingly involved in predicting the detailed turbulent flow scales [1, 2], though the Reynolds Averaged Navier-Stokes (RANS) approach is still a useful tool in evaluating the time averaged features of the turbulent flows [3,4,5,6]. Courant–Friedrichs–Lewy number width of the cavity on z-axis (m) cylinder diameter (m) finite volume method Grashof number (gβmΔTW3/νm2) convective heat transfer coefficient (W/m2K) turbulent kinetic energy (m2/s2) width of the cavity on x-axis (m) Nusselt number Prandtl number (νm/αm) Rayleigh number (Gr Pr) Reynolds number (U0,mW/νm) Richardson number (Gr/Re2) large-scale strain rate tensor for grid-filter temperature of the fluid (K) time velocity component at x-direction (m/s) dimensionless velocity component at x-direction lid velocity (m/s) velocity component at y-direction (m/s) dimensionless velocity component at y-direction dimensionless velocity component at z-direction distance along the x-coordinate distance along the nondimensional x-coordinate (x/L).

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call