A sensitivity analysis of bubble departure behavior in vertical channel nucleate boiling

Boiling is a very efficient heat transfer mechanism and had been exploited in numerous industrial and engineering sectors. However, limited by the experimental techniques and measuring tools, the experimental data related to the bubble dynamics especially in high heat fluxes on the wall surface is still rare and random. To contribute to this knowledge gap, the present paper proposes a statistical analysis method, attempting to identify the effect of input variables to bubble departure diameter on superheated wall by sensitivity analysis. More than 400 experimental data are collected by literature reviewing and a group of simulation data generated by selected correlation are included to supplementary and enlarge the database. Finite Mixture Model (FMM) is employed to conduct the sensitivity analysis study, where the hydraulic diameter of channel, water flow rate, flow subcooling and wall superheat are selected as input variables and the bubble departure diameter is output variable. With adopting two definitions of general distance (Kullback-Leibler divergence and chi(2) distance), the sensitivity degree order of the input variables is evaluated. Those two definitions of general distance have almost the same results concerning the order. The innovative sensitivity analysis approach is compared with the standard variance decomposition method (Sobol). The result indicates that FMM is more accurate than Sobol based on present database. The preliminary results of sensitivity analysis approach show that the flow condition has a more significant influence on bubble departure behavior than temperature does.