Direct-contact condensation heat transfer with noncondensable gases and interfacial shear for co-current stratified wavy flow in nearly-horizontal channels

The calibrating method for an electrochemical probe, neglecting the effect of the normal velocity on the mass transport, can cause large errors when applied to the measurement of wall shear rates in thin wavy flow with large amplitude waves. An extended calibrating method is developed to consider the contributions of the normal velocity. The turbulence-induced normal velocity component is included into the 2-D mass transport equation by means of its root mean square value multiplied by a random function. The wave-induced normal velocity component is postulated to be proportional to the normal interfacial velocity in thin wavy flow. The inclusion of the turbulence-induced normal velocity term is found to have a negligible effect on the mass transfer coefficient. The contribution of the wave-induced normal velocity can be classified on the dimensionless parameter, V. If V is above a critical value of V, V_crit, the effects of the wave-induced normal velocity become larger with an increase in V. While its effects negligible for V < V_crit. The unknown shear rate is numerically determined by solving the 2-D mass transport equation inversely. The present inverse method can predict the unknown shear rate more accurately in thin wavy flow with large amplitude waves than the previous method. If the normal velocity is neglected, the error that the predicted shear rate is lower than the real value is obtained. The interfacial shear stress is experimentally investigated for co-current air-water stratified flow inclined rectangular channels having a length of 1854mm, width of 120mm and height of 40mm at almost atmospheric pressure. Experiments are carried out in several inclinations from 0 .deg. up to 10 .deg. .The local film thickness and the wave height are measured at three locations, i.e., L/H = 8, 23, and 40. According to the inclination angle, the experimental data are categorized into two groups: nearly horizontal data group (0 .deg. ≤ θ ≤ 0.7 .deg. ), and inclined channel data group (0.7 .deg. ≤ θ ≤10 .deg. ). Experimental observations show that the super-critical flow only exists in the channel for the inclined channel data group while the water level gradient and hydraulic jump do for the nearly horizontal data group. For the inclined channel data group, a dimensionless wave height, Δh/h, is empirically correlated in terms of Re_G and h/H. A modified root-mean-square wave height is proposed to consider the effects of the interfacial and wave propagation velocities. It is found that an equivalent roughness has a linear relationship with the modified root-mean-square wave height. Direct-contact condensation experiments of atmospheric steam and steam/air mixture on subcooled water flowing co-currently in nearly horizontal channels are carried out and the logarithmic and local heat transfer coefficients are obtained. In the case of pure steam, the heat transfer coefficient increases as an increase in the inlet steam flow rate and inclination. The heat transfer coefficient due to the increased inlet water flow rate is slightly increased. In the presence of noncondensable gas, the heat transfer coefficient is much reduced as an increase in the inlet air mass fraction. The heat transfer coefficient increases as an increase in the inlet mixture flow rate. However, it decreases as an increase in the inlet water flow rate due to the change of the heat transfer mechanism from water layer to the mixture layer. The local heat transfer coefficient is enhanced in the large amplitude region even though the noncondensable gas exists. The heat transfer resistance also depends on the water side for the large amplitude region. For all cases it is found that the heat transfer coefficient increases as an increase in inclination. Direct-contact condensation database is constructed. Assessment and improvement of the direct-condensation model in RELAP5/MOD3.2 are performed. The horizontally stratified condensation model of RELAP5/MOD3.2 overpredicts both co-current and counter-current experimental data. The correlation proposed by Kim. predicts the database relatively well compared with that of RELAP5/MOD3.2. In the presence of noncondensable gases, RELAP5/MOD3.2 overpredicts the database with a percentage error of 103.0% and 65.0% for θ = 2.1 .deg. and θ = 5.0 .deg. , respectively. However, when Kim's correlation is substituted with the Dittus-Boelter type correlation in RELAP5/MOD3.2, the prediction errors are much reduced to 25.9% and 30.0% for θ = 2.1 .deg. and θ =5.0 .deg. , respectively. Therefore, it is concluded that Kim's correlation is superior to the correlation used in RELAP5/MOD3.2 both with and without noncondensable gases