A detailed comparative numerical study between the two-dimensional（2 D） and quasi-two-dimensional（quasi-2 D）turbulent Rayleigh–B＇enard（RB） convection on flow state, heat transfer, and thermal dissipation rate...A detailed comparative numerical study between the two-dimensional（2 D） and quasi-two-dimensional（quasi-2 D）turbulent Rayleigh–B＇enard（RB） convection on flow state, heat transfer, and thermal dissipation rate（TDR） is made. The Rayleigh number（Ra） in our simulations ranges up to 5×10^10 and Prandtl number（Pr） is fixed to be 0.7. Our simulations are conducted on the Tianhe-2 supercomputer. We use an in-house code with high parallelization efficiency, based on the extended PDM–DNS scheme. The comparison shows that after a certain Ra, plumes with round shape, which is called the ＂temperature islands＂, develop and gradually dominate the flow field in the 2 D case. On the other hand, in quasi-2 D cases, plumes remain mushroom-like. This difference in morphology becomes more significant as Ra increases, as with the motion of plumes near the top and bottom plates. The exponents of the power-law relation between the Nusselt number（Nu） and Ra are 0.3 for both two cases, and the fitting pre-factors are 0.099 and 0.133 for 2 D and quasi-2 D respectively,indicating a clear difference in magnitude of the heat transfer rate between two cases. To understand this difference in the magnitude of Nu, we compare the vertical profile of the horizontally averaged TDR for both two cases. It is found that the profiles of both cases are nearly the same in the bulk, but they vary near boundaries. Comparing the bifurcation height zb with the thermal boundary layer thickness dq, it shows that zb 〈 δθ（3 D） 〈 δθ（2 D） and all three heights obey a universal power-law relation z ~Ra^-0.30. In order to quantify the difference further, we separate the domain by zb, i.e., define the area between two zb（near top and bottom plates respectively） as the ＂mid region＂ and the rest as the ＂side region＂, and integrate TDR in corresponding regions. By comparing the integral it is found that most of the difference in TDR between two cases, which is connected to the heat transfer rate, occurs within the thermal boundary layers. We also compare the ratio of contributions to total heat transfer in BL–bulk separation and side–mid separation.展开更多
A direct numerical simulation(DNS) method is used to calculate the partitioned convection system with Ra number ranging from 10^7 to 2×10^9.Using the boundary layer thickness to normalize the height of gaps d, we...A direct numerical simulation(DNS) method is used to calculate the partitioned convection system with Ra number ranging from 10^7 to 2×10^9.Using the boundary layer thickness to normalize the height of gaps d, we find a strong consistency between the variation of the TD number(the average value of the temperature in each heat transfer channel is averaged after taking the absolute values) with the change of the height of gaps and the variation of the TD number with the change of Ra number in partitioned convection.For a given thickness of partition, heights of gaps are approximately equal to 0.5 or 1 time of the thermal boundary layer thickness λθ at different Ra numbers.TD number representing temperature characteristics is almost the constant value, which means that TD number is a function of d/λθ only.Analysis of local temperature field of area in gaps shows that the temperature distribution in the gaps are basically the same when d/λθ is certain.The heat transfer Nu number of the system at d/λθ≈ 0.5 is larger than that of d/λθ≈ 1, both of them have the same scaling law with Ra number and Nu^Ra^0.25.展开更多
We report a numerical study of the Prandtl-number(Pr)effects in two-dimensional turbulent Rayleigh-Bénard convection.The simulations were conducted in a square box over the Pr range from 0.25 to 100 and over the ...We report a numerical study of the Prandtl-number(Pr)effects in two-dimensional turbulent Rayleigh-Bénard convection.The simulations were conducted in a square box over the Pr range from 0.25 to 100 and over the Rayleigh number(Ra)range from 10^(7) to 10^(10).We find that both the strength and the stability of the large-scale flow decrease with the increasing of Pr,and the flow pattern becomes plume-dominated at high Pr.The evolution in flow pattern is quantified by the Reynolds number(Re),with the Ra and the Pr scaling exponents varying from 0.54 to 0.67 and-0.87 to-0.93,respectively.It is further found that the non-dimensional heat flux at small Ra diverges strongly for different Pr,but their difference becomes marginal as Ra increases.For the thermal boundary layer,the spatially averaged thicknesses for all the Pr numbers can be described byδθ~Ra^(-0.30) approximately,but the local values vary a lot for different Pr,which become more uniform with Pr increasing.展开更多
We study the characteristics of temperature fluctuation in two-dimensional turbulent Rayleigh–Benard convection in´a square cavity by direct numerical simulations.The Rayleigh number range is 1×10^(8)≤Ra≤...We study the characteristics of temperature fluctuation in two-dimensional turbulent Rayleigh–Benard convection in´a square cavity by direct numerical simulations.The Rayleigh number range is 1×10^(8)≤Ra≤1×10^(13),and the Prandtl number is selected as Pr=0.7 and Pr=4.3.It is found that the temperature fluctuation profiles with respect to Ra exhibit two different distribution patterns.In the thermal boundary layer,the normalized fluctuationθrms/θrms,max is independent of Ra and a power law relation is identified,i.e.,θrms/θrms,max∼(z/δ)0.99±0.01,where z/δis a dimensionless distance to the boundary(δis the thickness of thermal boundary layer).Out of the boundary layer,when Ra≤5×10^(9),the profiles ofθrms/θrms,max descend,then ascend,and finally drop dramatically as z/δincreases.While for Ra≥1×10^(10),the profiles continuously decrease and finally overlap with each other.The two different characteristics of temperature fluctuations are closely related to the formation of stable large-scale circulations and corner rolls.Besides,there is a critical value of Ra indicating the transition,beyond which the fluctuation hθrmsiV has a power law dependence on Ra,given by hθrmsiV∼Ra−0.14±0.01.展开更多
基金Project supported by the National Natural Science Foundation of China(Grant Nos.11772362 and 11452002)the Special Scientific Research Fund for Super Computing in the Joint Fund of the National Natural Science Foundation of Chinathe People’s Government of Guangdong Province(Phase Ⅱ,Grant No.nsfc2015 570)
文摘A detailed comparative numerical study between the two-dimensional（2 D） and quasi-two-dimensional（quasi-2 D）turbulent Rayleigh–B＇enard（RB） convection on flow state, heat transfer, and thermal dissipation rate（TDR） is made. The Rayleigh number（Ra） in our simulations ranges up to 5×10^10 and Prandtl number（Pr） is fixed to be 0.7. Our simulations are conducted on the Tianhe-2 supercomputer. We use an in-house code with high parallelization efficiency, based on the extended PDM–DNS scheme. The comparison shows that after a certain Ra, plumes with round shape, which is called the ＂temperature islands＂, develop and gradually dominate the flow field in the 2 D case. On the other hand, in quasi-2 D cases, plumes remain mushroom-like. This difference in morphology becomes more significant as Ra increases, as with the motion of plumes near the top and bottom plates. The exponents of the power-law relation between the Nusselt number（Nu） and Ra are 0.3 for both two cases, and the fitting pre-factors are 0.099 and 0.133 for 2 D and quasi-2 D respectively,indicating a clear difference in magnitude of the heat transfer rate between two cases. To understand this difference in the magnitude of Nu, we compare the vertical profile of the horizontally averaged TDR for both two cases. It is found that the profiles of both cases are nearly the same in the bulk, but they vary near boundaries. Comparing the bifurcation height zb with the thermal boundary layer thickness dq, it shows that zb 〈 δθ（3 D） 〈 δθ（2 D） and all three heights obey a universal power-law relation z ~Ra^-0.30. In order to quantify the difference further, we separate the domain by zb, i.e., define the area between two zb（near top and bottom plates respectively） as the ＂mid region＂ and the rest as the ＂side region＂, and integrate TDR in corresponding regions. By comparing the integral it is found that most of the difference in TDR between two cases, which is connected to the heat transfer rate, occurs within the thermal boundary layers. We also compare the ratio of contributions to total heat transfer in BL–bulk separation and side–mid separation.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.11772362 and 11452002)the Special Scientific Research Fund for Super Computing in the Joint Fund of the National Natural Science Foundation of China and the People’s Government of Guangdong Province,China(Phase Ⅱ,nsfc2015_570)
文摘A direct numerical simulation(DNS) method is used to calculate the partitioned convection system with Ra number ranging from 10^7 to 2×10^9.Using the boundary layer thickness to normalize the height of gaps d, we find a strong consistency between the variation of the TD number(the average value of the temperature in each heat transfer channel is averaged after taking the absolute values) with the change of the height of gaps and the variation of the TD number with the change of Ra number in partitioned convection.For a given thickness of partition, heights of gaps are approximately equal to 0.5 or 1 time of the thermal boundary layer thickness λθ at different Ra numbers.TD number representing temperature characteristics is almost the constant value, which means that TD number is a function of d/λθ only.Analysis of local temperature field of area in gaps shows that the temperature distribution in the gaps are basically the same when d/λθ is certain.The heat transfer Nu number of the system at d/λθ≈ 0.5 is larger than that of d/λθ≈ 1, both of them have the same scaling law with Ra number and Nu^Ra^0.25.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.11961160719,11702128,91752201,and 11772362)the Shenzhen Fundamental Research Program(Grant No.JCYJ20190807160413162)+1 种基金the Fundamental Research Funds for the Central Universities(Sun Yat-sen University under Grant No.19lgzd15)the Department of Science and Technology of Guangdong Province,China(Grant No.2019B21203001).
文摘We report a numerical study of the Prandtl-number(Pr)effects in two-dimensional turbulent Rayleigh-Bénard convection.The simulations were conducted in a square box over the Pr range from 0.25 to 100 and over the Rayleigh number(Ra)range from 10^(7) to 10^(10).We find that both the strength and the stability of the large-scale flow decrease with the increasing of Pr,and the flow pattern becomes plume-dominated at high Pr.The evolution in flow pattern is quantified by the Reynolds number(Re),with the Ra and the Pr scaling exponents varying from 0.54 to 0.67 and-0.87 to-0.93,respectively.It is further found that the non-dimensional heat flux at small Ra diverges strongly for different Pr,but their difference becomes marginal as Ra increases.For the thermal boundary layer,the spatially averaged thicknesses for all the Pr numbers can be described byδθ~Ra^(-0.30) approximately,but the local values vary a lot for different Pr,which become more uniform with Pr increasing.
基金the National Natural Science Foundation of China(Grant No.11772362)the Shenzhen Fundamental Research Program(Grant No.JCYJ20190807160413162)the Fundamental Research Funds for the Central Universities,Sun Yat-sen University,China(Grant No.19lgzd15).
文摘We study the characteristics of temperature fluctuation in two-dimensional turbulent Rayleigh–Benard convection in´a square cavity by direct numerical simulations.The Rayleigh number range is 1×10^(8)≤Ra≤1×10^(13),and the Prandtl number is selected as Pr=0.7 and Pr=4.3.It is found that the temperature fluctuation profiles with respect to Ra exhibit two different distribution patterns.In the thermal boundary layer,the normalized fluctuationθrms/θrms,max is independent of Ra and a power law relation is identified,i.e.,θrms/θrms,max∼(z/δ)0.99±0.01,where z/δis a dimensionless distance to the boundary(δis the thickness of thermal boundary layer).Out of the boundary layer,when Ra≤5×10^(9),the profiles ofθrms/θrms,max descend,then ascend,and finally drop dramatically as z/δincreases.While for Ra≥1×10^(10),the profiles continuously decrease and finally overlap with each other.The two different characteristics of temperature fluctuations are closely related to the formation of stable large-scale circulations and corner rolls.Besides,there is a critical value of Ra indicating the transition,beyond which the fluctuation hθrmsiV has a power law dependence on Ra,given by hθrmsiV∼Ra−0.14±0.01.
