Steady-State Natural Convection in Empty and Partitioned Enclosures at High Rayleigh Numbers

1990 ◽  
Vol 112 (3) ◽  
pp. 640-647 ◽  
Author(s):  
D. A. Olson ◽  
L. R. Glicksman ◽  
H. M. Ferm

Steady-state natural convection, which occurs in building enclosures (Rayleigh numbers of 1010), was studied experimentally in a full-scale room and in a 1:5.5 small-scale physical model containing R114 gas. The model was geometrically similar, had the same Rayleigh number, and had the same dimensionless end wall temperatures as the full-scale room. Configurations were tested with the enclosure empty, with a vertical partition extending from the floor to midheight, and with the vertical partition raised slightly off the floor. For isothermal opposing end walls at different temperatures, excellent agreement was found between the full-scale room and the scale model in flow patterns, velocity levels, temperature distributions, and heat transfer, even though the radiation heat transfer was not scaled between the two models.

1991 ◽  
Vol 113 (3) ◽  
pp. 635-642 ◽  
Author(s):  
D. A. Olson ◽  
L. R. Glicksman

Transient natural convection at Rayleigh numbers of 1010 was studied experimentally in two enclosures of aspect ratio 1/3, one a 1:5.5 physical scale model containing the dense refrigerant gas R114, and the second a full-scale room containing air. In one type of transient the vertical endwall temperature was suddenly changed, while in a second type of transient the isothermal, well-mixed gas was suddenly exposed to hot and cold vertical endwalls. The experiments indicated that the dominant time constant was a convective one. Comparisons between the scale model and full scale show that R114 gas can simulate the transient behavior of air-filled enclosures.


Author(s):  
Behnam Moghadassian ◽  
Farshad Kowsary ◽  
Hamed Gholamian

The problem of natural convection radiation with the presence of participating fluid in a tilted square cavity has been investigated numerically. Two vertical walls are at uniform different temperatures while the others are adiabatic. The working fluid is taken as grey, absorbing, emitting and non-scattering. The finite volume method is used to solve the dimensionless governing equations and SIMPLE algorithm is applied for pressure velocity coupling. The radiative heat flux gradient is estimated by finding radiative intensities from the radiative transfer equation (RTE). A very recent method, called the QL method, is utilized to solve RTE. In this study the effects of the inclination angle, Rayleigh number and optical thickness on the heat transfer and flow characteristics are studied. A great accuracy in the results was observed in the prediction of flow contours and average radiative and convective Nusselt numbers at walls.


1982 ◽  
Vol 104 (1) ◽  
pp. 96-102 ◽  
Author(s):  
S. M. ElSherbiny ◽  
G. D. Raithby ◽  
K. G. T. Hollands

Measurements of the heat transfer by natural convection across vertical and inclined air layers are reported. The air layer is bounded by flat isothermal plates at different temperatures and around the edges by a perfectly conducting boundary (i.e., one that takes on a linear temperature distribution between the two plates). Measurements are reported for six aspect ratios between 5 and 110, covering a portion of the range of practical interest for windows, solar collectors, etc. Rayleigh numbers were in the range 102 to 2 × 107. The present measurements permitted the role of aspect ratio to be clearly defined. In addition, correlation equations are presented which allow the heat trnasfer across a vertical or inclined air layer to be calculated.


1981 ◽  
Vol 103 (4) ◽  
pp. 623-629 ◽  
Author(s):  
M. W. Nansteel ◽  
R. Greif

Heat transfer by natural convection in a two-dimensional rectangular enclosure fitted with partial vertical divisions is investigated experimentally. The horizontal walls of the enclosure are adiabatic while the vertical walls are maintained at different temperatures. The experiments are carried out with water, Pr ≃ 3.5, for Rayleigh numbers in the range, 2.3 × 1010 ⩽ RaL ⩽ 1.1 × 1011, and an aspect ratio, A = H/L = 1/2. The effect of the partial vertical divisions on the fluid flow and temperature fields is investigated by dye-injection flow visualization and by thermocouple probes, respectively. The effect of the partitions on the heat transfer across the enclosure is also studied and correlations for the Nusselt number as a function of RaL and partition length are generated for both conducting and non-conducting partition materials. Partial divisions are found to have a significant effect on the heat transfer; especially when the divisions are adiabatic. The results also indicate that the partial divisions may have a stabilizing effect on the laminar-transitional flow on the heated vertical walls of the enclosure.


2019 ◽  
Vol 7 (1) ◽  
pp. 43-53
Author(s):  
Abbas Jassem Jubear ◽  
Ali Hameed Abd

The heat sink with vertically rectangular interrupted fins was investigated numerically in a natural convection field, with steady-state heat transfer. A numerical study has been conducted using ANSYS Fluent software (R16.1) in order to develop a 3-D numerical model.  The dimensions of the fins are (305 mm length, 100 mm width, 17 mm height, and 9.5 mm space between fins. The number of fins used on the surface is eight. In this study, the heat input was used as follows: 20, 40, 60, 80, 100, and 120 watts. This study focused on interrupted rectangular fins with a different arrangement and angle of the fins. Results show that the addition of interruption in fins in various arrangements will improve the thermal performance of the heat sink, and through the results, a better interruption rate as an equation can be obtained.


Computation ◽  
2021 ◽  
Vol 9 (6) ◽  
pp. 65
Author(s):  
Aditya Dewanto Hartono ◽  
Kyuro Sasaki ◽  
Yuichi Sugai ◽  
Ronald Nguele

The present work highlights the capacity of disparate lattice Boltzmann strategies in simulating natural convection and heat transfer phenomena during the unsteady period of the flow. Within the framework of Bhatnagar-Gross-Krook collision operator, diverse lattice Boltzmann schemes emerged from two different embodiments of discrete Boltzmann expression and three distinct forcing models. Subsequently, computational performance of disparate lattice Boltzmann strategies was tested upon two different thermo-hydrodynamics configurations, namely the natural convection in a differentially-heated cavity and the Rayleigh-Bènard convection. For the purposes of exhibition and validation, the steady-state conditions of both physical systems were compared with the established numerical results from the classical computational techniques. Excellent agreements were observed for both thermo-hydrodynamics cases. Numerical results of both physical systems demonstrate the existence of considerable discrepancy in the computational characteristics of different lattice Boltzmann strategies during the unsteady period of the simulation. The corresponding disparity diminished gradually as the simulation proceeded towards a steady-state condition, where the computational profiles became almost equivalent. Variation in the discrete lattice Boltzmann expressions was identified as the primary factor that engenders the prevailed heterogeneity in the computational behaviour. Meanwhile, the contribution of distinct forcing models to the emergence of such diversity was found to be inconsequential. The findings of the present study contribute to the ventures to alleviate contemporary issues regarding proper selection of lattice Boltzmann schemes in modelling fluid flow and heat transfer phenomena.


1987 ◽  
Vol 109 (2) ◽  
pp. 388-391 ◽  
Author(s):  
E. M. Sparrow ◽  
M. A. Ansari

Measurements were made of the combined natural convection and radiation heat transfer from a horizontal finned tube situated in a vertical channel open at the top and bottom. In one set of experiments, both walls of the channel were heavily insulated, while in a second set of experiments, one of the insulated walls was replaced by an uninsulated metallic sheet. In general, the heat transfer coefficients were found to be lower with the metal wall in place, but only moderately. With the finned tube situated at the bottom of the channel, the differences in the heat transfer coefficients corresponding to the two types of walls were only a few percent. When the tube was positioned at the mid-height of the channel, larger differences were encountered, but in the practical range of Rayleigh numbers, the differences did not exceed 5 percent.


Sign in / Sign up

Export Citation Format

Share Document