Numerical Simulation of Flow and Heat Transfer of Fluids at Supercritical Pressure

Simple geometries such as circular pipes or annular channels have been used to investigate the heat transfer experimentally in the vicinity of the pseudo-critical temperature using water or CO2. A large database exists worldwide. In these experiments large deviations from the ‘normal’ heat transfer behaviour were documented, in particular a strong dependency of the heat transfer coefficient on high wall heat fluxes and a sharp, sometimes unexpected drastic rise of the wall temperature, known as the ‘heat transfer deterioration’. This body of data is also useful for model testing. Our own efforts were directed towards an analytic modelling of the near-wall turbulent layer and the laminar sub-layer in order to derive RANS turbulence model parameters and CFD wall functions, which are reliable for a large range of the flow and wall heat flux parameters. A close investigation on the flow mechanisms under deterioration conditions reveals that the thickness of the laminar, heat-conducting sub-layer must be modelled accurately as a function of the local Prandtl number. Furthermore it is necessary to take account of the drastic property changes, in particular of the heat capacity, with temperature or enthalpy in the turbulent near-wall layer. An attempt to model this effect has been made by using a probability-density function approach (pdf model). Buoyancy effects must also be modelled. Various three-dimensional numerical examples of the flow and heat transfer within the reactor core cooling channels of the European High-Performance Light-Water Reactor (HPLWR) are presented. An important geometrical element of its core, are the wires wrapped around each rod as spacers. The wires enhance mixing and mass exchange between the sub-channels of an assembly, but they also may cause a non-uniform wall temperature distribution with local high-temperature regions. In order to validate our models, an experiment of a single rod with wrapped wire has been performed in the supercritical water flow loop of Xi’an Jiaotong University, China. The comparison of these and other experiments with CFD models shows, that great achievements have been made but the reliability of our models for supercritical flow conditions must be further improved.