Thermal mixing of flows in horizontal T-junctions with low branch velocities

Abstract

High cycle thermal fatigue (HCTF) damage in the vicinity of T-junction piping, where mixing of coolant streams at significant temperature differences occur, is considered to be an important problem in the context of operational safety in nuclear power plants (NPPs). The present study deals with the experimental and numerical investigations of T-junction flow mixing at realistic temperature differences encountered in NPPs. Experiments were carried out at the fluid-structure interaction (FSI) T-junction test facility at the University of Stuttgart. Near-wall thermocouples were used in the mixing zone to record temperature information inside the flow and the structure. A conductivity based wire mesh sensor (WMS) instrumentation designed at the Laboratory of Nuclear Energy Systems at the ETH Zurich for high-pressure high-temperature applications was also used during the measurements to characterize the cross-sectional thermal field in the mixing zone. Aside from experiments, numerical studies were carried out using the large-eddy simulation (LES) turbulence model to gain further insights into the flow mixing behavior and its predictions were validated with measurement data.With a constant mass flow rate ratio (main/branch) of 4:1, thermal mixing tests were carried out at temperature differences (ΔT) between the mixing fluids of 65°C (Case 1) and 143°C (Case 2), respectively. Two distinct flow trends were identified in the mixing region: (i) an unstable stratified flow being subjected to extreme oscillations in case 1 and (ii) a stable stratified flow being subjected to significant buoyancy effects in case 2. Near-wall thermal fluctuations (a factor in thermal fatigue analysis) with highest amplitudes were recorded by the thermocouples located in the vicinity of the stratification layer in both the cases. Frequency analyses of thermal fluctuations using the power spectral density (PSD) method indicates no dominant frequency (spectral peak) in the frequency range identified to be of relevance for HCTF (0.1–10Hz). Instead, the energy of thermal fluctuations were mainly contained in the frequency range of 0.1–2Hz. LES predictions of parameters, viz., mean temperature, thermal fluctuations and its frequency distribution showed good agreement with measurement data.