A dry-spot model of critical heat flux and transition boiling in pool and subcooled forced convection boiling

A new dry-spot model for critical heat flux (CHF) is proposed. The new concept for dry area formation based on Poisson distribution of active nucleation sites and the critical active site number is introduced. The model is based on the boiling phenomena observed in nucleate boiling such as Poisson distribution of active nucleation sites and formation of dry spots on the heating surface. It is hypothesized that when the number of bubbles surrounding one bubble exceeds a critical number, the surrounding bubbles restrict the feed of liquid to the microlayer under the bubble. Then a dry spot of vapor will form on the heated surface. As the surface temperature is raised, more and more bubbles will have a population of surrounding active sites over the critical number. Consequently, the number of the spots will increase and the size of dry areas will increase due to merger of several dry spots. If this trend continues, the number of effective sites for heat transport through the wall will diminish, and CHF and transition boiling occur. The model is applicable to pool and subcooled forced convection boiling conditions, based on the common mechanism that CHF and transition boiling are caused by the accumulation and coalescences of dry spots. It is shown that CHF and heat flux in transition boiling can be determined without any empirical parameter based on information on the boiling parameters such as active site density and bubble diameter, etc., in nucleate boiling. It is also shown that the present model well represents actual phenomena on CHF and transition boiling and explains the mechanism on how parameters such as flow modes (pool or flow) and surface wettability influence CHF and transition boiling. Validation of the present model for CHF and transition boiling is achieved without any tuning parameter always present in earlier models. It is achieved by comparing the predictions of CHF and heat flux in transition boiling using measured boiling parameters in nucleate boiling from given boiling conditions with the pool CHF data measured by Dhir and Liaw and Paul and Abdel-Khalik and the subcooled flow CHF data measured by Del Valle M. and Kenning and with the heat flux data in transition boiling measured by Dhir and Liaw. The predictions show good agreement with the existing data. To use the present phenomenological model as a prediction tool, a study has been performed to predict CHF in pool and subcooled forced convection boiling using existing correlations for active site density, maximum bubble diameter, and heat transfer coefficients in nucleate boiling. Comparison of the model predictions with experimental data for pool boiling of water and upward flow boiling of water in vertical, uniformly-heated round tubes is performed. The data set (2438 data points) for CHF in subcooled forced convection boiling covers wide ranges of operating conditions (0.1≤P≤14.0 MPa, 0.00033≤D≤0.0375 m: 0.002≤L≤2 m: 660 ≤G≤90000 kg/m²s: 70≤Δh,≤1456 kJ/kg). Without any tuning factor, 1492 data points out of 2438 (61.2%) are calculated with a r.m.s. error of 41.3% and about 80% of the calculated data points are predicted within ±50%. It is also shown that by a modification of suppression factor in subcooled boiling, the predictive capability of the present model can be improved, i.e., 2421 data points (99.3%) are calculated with a r.m.s. error of 20.5% and 82.3% of the calculated data points are predicted within ±25%. In addition, the parametric trends of CHF in subcooled forced convection boiling have been investigated under local conditions hypothesis