| Summary: | Within the course of this master thesis project, subcooled nucleate boiling in a vertical pipe has been modeled using CFD. The modeling has been carried out within the OpenFOAM framework and a two-phase Eulerian approach has been chosen. The code can be used to predict the distribution of the local ow parameters, i.e.the void fraction, the bubble diameter, the velocity of both liquid and gas, the turbulent intensity as well as the liquid temperature. Special attention has been devoted to the phenomena which govern the void fraction distribution in the radial direction. Two di erent solvers have been implemented and the simulations have been performed in two dimensions. Firstly, isothermal turbulent bubbly ow is mechanistically modeled in a solver named myTwoPhaseEulerFoa-mAdiabatic. The conservation equations of mass and momentum are solved for the two phases, taking special care in the modeling of the interfacial forces. The turbulence phenomena are described by a classical k- model in combination with standard wall functions for the near-wall treatment. Furthermore, an interfacial area concentration equation is solved and two di erent models for its sink- and source terms (corresponding to bubble coalesence and bubble breakup) have been investigated. Secondly, a solver named myTwoPhaseEulerFoamBoiling has been developed based on the rst solver in order to model a heated wall leading to subcooled nucleate boiling and subsequent condensation in the subcooled liquid. Additional terms accounting for the phase change have been included in the mass and momentum conservation equations as well as in the interfacial area equation. Assuming the gas phase being at saturation conditions, only one energy equation for the liquid phase needs to be solved. The adiabatic solver has been validated against the DEDALE experiment and the simulation results showed satisfactory agreement with the measured data. The predictions obtained from myTwoPhaseEulerFoam-Boiling have been compared to the DEBORA experimental data base. They are qualitatively similar but rather high quantitative discrepancies exist. Grid dependence tests revealed that the latter solver depends on the near-wall grid resolution, a yet unresolved issue related to the application of the wall heat ux as the boundary condition. However, the results were shown to be insensitive to small variations in the applied inlet conditions.
|
|---|
