수직원관의 과냉각비등 현상에 대한 연구

Abstract

To enhance the multi-dimensional analysis capability of a subcooled boiling two-phase flow, the population balance equations and improved heat partitioning algorithm were coupled together.

Several models for the heat partitioning algorithm and the breakup & coalescence models for the population balance equation were reviewed and a detailed discussion was given to get a better understanding of the subcooled boiling flows.

To implement the models, two-fluid models of CFX.12.1 was developed with the improved heat partitioning algorithms and the latest breakup & coalescence models.

Improved heat partitioning algorithm included the bubble lift-off diameter model. it took into account the bubble's sliding on a heated wall after its departure from a nucleate site.

Lehr et al.(2002) model was adopted as the breakup model of the population balance model. It took into account bubble's capillary pressure. Due to bubbles's capillary pressure, A very small bubbles is unlikely to breakup and the formation of very small bubbles is also unlikely.

Coalescence effects due to the turbulent fluctuation, buoyancy and wake-entrainment were taken into account for bubbly flows. Among coalescence models, Luo(1993), Wang(2005), and Hibiki(2000) models were adopted for those coalescence effects.

There is no inter-phase turbulent models at CFX.12.1. Inter-phase turbulent models could do a significant role at the two-phase flows. The SST(Shear Stress Transport) turbulent model was developed with the Krepper(2010)'s inter-phase turbulent model.

With the improved two-fluid models and default models of CFX.12.1, the behavior of an upwards forced convective subcooled boiling flows in a vertical annular channel is simulated. Comparison of models predictions against local measurements is made for the void fraction, liquid velocity and vapour velocity covering a range of different mass, heat fluxes and inlet subcooling temperature.

A lot of difference was revealed between the default model's prediction and measurement. But better agreement is achieved between the improved model's predictions and measurements.