TED-AJ03-582 NUMERICAL AND EXPERIMENTAL STUDY ON TURBULENT THERMAL MIXING IN A T-JUNCTION FLOW

概要

論文の詳細を見る
Thermal fatigue of the structure around a T-junction is a technically important issue for the safety of nuclear power plants. The cause for this is related to mixing of hot and cold fluids, though detailed mechanisms and effects of various parameters on those are still unclear. In the present study, direct numerical simulation (DNS) and experiments of flow in two square ducts connected via a T-junction, as shown in Fig.A1,are carried out. The Reynolds number of the main stream based on the bulk mean inlet velocity and the hydraulic diameter is about 4485. The hydraulic diameter of the branch duct is a half of that of the main duct. The velocity ratio of the branch jet to the main stream, V_r, is 2. Due to computational cost, a lower Prandtl number (Pr=0.71) is used in DNS as compared to those in experiments (Pr=2.6∿8.8). Five different Richardson numbers (Ri=gαΔTδ/U_<b1>^2=0,±0.93,±9.3; + : stable, - : unstable) are examined to investigate effects of buoyancy on the turbulent velocity and temperature fields, especially on temperature fluctuations on the walls. Three types of grid systems are used to investigate dependency on the mesh size and the domain size. Nine corresponding cases of experiments are carried out in order to obtain verification data for DNS. Laser Doppler Velocimeter (LDV) and thermocouples are used to measure velocity and temperature, respectively. Mean and RMS velocities in the main-flow and vertical directions and temperature from the simulation with the fine grid system agree quantitatively well with the experiments in the neutral case. Flow visualization reveals that the skewed temperature fluctuations on the duct walls are caused by various kinds of large-scale coherent structures near the T-junction, such as a groove on the branch jet, upper and lower kidney vortices, separation in the branch duct and roll-ups shed from the leading and trailing edges of the duct. The buoyancy dramatically alters the mean velocity and temperature distribution in the strongly stable/unstable case (Ri=±9.3), whilst its effects are localized at Ri=±0.93. As shown in Fig.A2,the magnitude of temperature fluctuations downstream on the bottom wall is the largest in the case without buoyancy. The reason is found to be as follows. In the stable cases, relatively small penetration rates of the branch jet to the bottom wall lead to stratification. In the unstable cases, the condition changes into a stable state near the bottom wall once the cold fluid from the branch jet goes under the hot main stream. Those stabilizing mechanisms are missing in the neutral case.[figure]
一般社団法人日本機械学会の論文

TED-AJ03-582 NUMERICAL AND EXPERIMENTAL STUDY ON TURBULENT THERMAL MIXING IN A T-JUNCTION FLOW

スポンサーリンク

概要

著者

関連論文

スポンサーリンク