scholarly journals Influence of the computational domain on DNS of turbulent heat transfer up to Reτ=2000 for Pr=0.71

Author(s):  
F. Lluesma-Rodríguez ◽  
S. Hoyas ◽  
M.J. Perez-Quiles
1994 ◽  
Vol 116 (3) ◽  
pp. 577-587 ◽  
Author(s):  
S. H. Kim ◽  
N. K. Anand

Two-dimensional turbulent heat transfer between a series of parallel plates with surface mounted discrete block heat sources was studied numerically. The computational domain was subjected to periodic conditions in the streamwise direction and repeated conditions in the cross-stream direction (Double Cyclic). The second source term was included in the energy equation to facilitate the correct prediction of a periodically fully developed temperature field. These channels resemble cooling passages in electronic equipment. The k–ε model was used for turbulent closure and calculations were made for a wide range of independent parameters (Re, Ks/Kf, s/w, d/w, and h/w). The governing equations were solved by using a finite volume technique. The numerical procedure and implementation of the k–ε model was validated by comparing numerical predictions with published experimental data (Wirtz and Chen, 1991; Sparrow et al., 1982) for a single channel with several surface mounted blocks. Computations were performed for a wide range of Reynolds numbers (5 × 104–4 × 105) and geometric parameters and for Pr = 0.7. Substrate conduction was found to reduce the block temperature by redistributing the heat flux and to reduce the overall thermal resistance of the module. It was also found that the increase in the Reynolds number decreased the thermal resistance. The study showed that the substrate conduction can be an important parameter in the design and analysis of cooling channels of electronic equipment. Finally, correlations for the friction factor (f) and average thermal resistance (R) in terms of independent parameters were developed.


Author(s):  
Robert Bergant ◽  
Iztok Tiselj

Direct Numerical Simulation (DNS) of fully developed turbulent flow in a flume and channel was used to study the heat transfer near the wall. Two different geometries for numerical simulations of turbulent heat transfer in infinite channel and flume were used. Reynolds number based on friction velocity, channel half height (or flume height) and kinematic viscosity was set to Reτ = 170.8, whereas the Prandtl number was set to unity, Pr = 1. The computational domain of 2146×341.6×537 wall units and 2146×170.8×537 wall units was used for channel geometry and flume geometry, respectively. Comparison of these two different geometries with the same physics, were highlighted and discussed.


2020 ◽  
Vol 401 ◽  
pp. 117-130
Author(s):  
Younes Menni ◽  
Ali J. Chamkha ◽  
Oluwole Daniel Makinde

In this work, the thermal behavior of a turbulent forced-convection flow of air in a rectangular cross section channel with attached W-shaped obstacles is investigated. The continuity, momentum and energy equations employed to control the heat and velocity in the computational domain. The turbulence model of k-ε is employed to simulate the turbulence effects. The finite volume method with SIMPLE algorithm is employed as the solution method. The results are reported temperature, local and average Nusselt numbers, and mean velocity contours. The subject is relevant and important for industrial applications.


2006 ◽  
Vol 129 (4) ◽  
pp. 449-463 ◽  
Author(s):  
Arun K. Saha ◽  
Sumanta Acharya

Turbulent flow and heat transfer in rotating ribbed ducts of different aspect ratios (AR) are studied numerically using an unsteady Reynolds averaged Navier–Stokes procedure. Results for three ARs (1:1, 1:4, and 4:1) and staggered ribs with constant pitch (P∕e=10) in the periodically developed region are presented and compared. To achieve periodic flow behavior in successive inter-rib modules calculations are performed in a computational domain that extends to two or three inter-rib modules. The computations are carried out for an extended parameter set with a Reynolds number range of 25,000–150,000, density ratio range of 0–0.5, and rotation number range of 0–0.50. Under rotational conditions, the highest heat transfer along the leading and side walls are obtained with the 4:1 AR, while the 1:4 AR has the highest trailing wall Nu ratio and the lowest leading wall Nu ratio. The 1:4 AR duct shows flow reversal near the leading wall (leading to low Nu) at high rotation numbers and density ratios. For certain critical parameter values (low Re, high Ro, and/or DR), the leading wall flow is expected to become nearly stagnant, due to the action of centrifugal buoyancy, leading to conduction-limited heat transfer. The 4:1 AR duct shows evidence of multiple rolls in the secondary flow that direct the core flow to both the leading and trailing surfaces which reduces the difference between the leading and trailing wall heat transfer relative to the other two AR ducts.


Sign in / Sign up

Export Citation Format

Share Document