Theoretical and experimental investigations of CHF in round tubes and rod bundles

A knowledge of the condition leading to critical heat flux (CHF) is of great importance in the design of nuclear reactors. Although many efforts have been devoted to the subject of CHF during the last few decades, information on the burnout phenomenon at low velocity condition is very limited. Furthermore, in most cases, the applicable range of a bundle CHF correlation is restricted to a narrow region mainly due to the limitation of the CHF data base used in the correlation development. In view of these points, theoretical and experimental investigations are performed in this study for round tubes and rod bundles. A CHF prediction model for low velocity conditions is proposed throughout the assessment of CHF data from various sources with mass velocities less than 500 kg/m2s. The CHF data base is classified into seven groups with respect to the flow pattern characteristics at CHF conditions. CHF data for each group is analyzed by several CHF prediction models including; the flooding correlations, the flow regime transition criteria, the complete evaporation model, and the empirical correlations. At zero inlet flow or extremely low mass velocity conditions, the flooding correlation can be used for predicting CHF employing appropriate constant. In the slug or churn-turbulent flow regime, CHF seems to occur at the annular flow transition conditions. When CHF occurs at the annular flow region, the empirical correlation such as AECL CHF lookup table gives accurate predictions except for the ranges where density-wave instability is expected. A phenomenological model for the prediction of dryout locations under flooding-limited CHF condition is developed based on the liquid film dryout model and the two-phase mixture level theory. The mass and energy conservation equations are applicable to the liquid film considering no entrainment of liquid droplets from the film region. The variation of the two-phase mixture level after the onset of flooding is calculated based on the homogeneous mixture model. In view of the result comparing with Katto's experimental data, it is revealed that the present model gives upper and lower bounds of the dryout locations measured in the boiling system with closed bottom end. The model accuracy is improved as the L/D ratio increases. CHF experiment at low velocities including zero inlet flow conditions are conducted for uniformly heated round tube with various flow obstructors positioned at the top of the heated tube. At very low mass velocities or zero inlet flow conditions, where CHF is expected to be governed by the counter-current flow limitation of liquid and vapor, CHF values tend to decrease as the blockage area by the obstructor increases. This effect cannot be distinguished at higher velocities approximately greater than 50 kg/m2s. Furthermore it reveals that the effect of the obstruction shape is not significant for low velocity CHF. A bundle correction method, based on the conservation laws of mass, energy, and momentum in an open subchannel, is proposed for the prediction of the CHF in rod bundles from round tube CHF correlations without detailed subchannel analysis. It takes into account the effects of the enthalpy and mass velocity distributions at subchannel level using the first derivatives of CHF with respect to the independent parameters. Three different CHF correlations for tubes (AECL CHF lookup table, Katto correlation, and Biasi correlation) have been examined with uniformly heated bundle CHF data collected from various sources. A limited number of CHF data from a non-uniformly heated rod bundle are also evaluated with the aid of Tong's F-factor. The proposed method shows satisfactory CHF predictions for rod bundles, both uniform and non-uniform power distributions