| Accurate prediction of the occurrence of dryout is very important for the estimation of reactor safety margin, because it is needed for calculating the DNBR and the maximum cladding temperature in the reactor safety analysis. Based on CIAE dryout experimental data, an evaluation of the typical CHF prediction methods was carried out, and a modified model was proposed. With this model, the effects of various parameters and non-uniform heating on the dryout ware studied.There are many prediction methods on the dryout, such as CHF empirical correlations, physical models and Look-up Table, but they are generally derived from respective data with big differences, and no one can be used for a wide range of parameters with confidence.Under the hypothesis of Annular Film Dryout (AFD), a modified model was proposed based on CIAE and abroad dryout experimental data. It gives better prediction under pressure of 0.1--13.78MPa, mass flux of 100--2000 kg/(m~2·~PS), tube diameter of 4—14mm and heating length of 1.0--2.1m.The dryout of the liquid film is the result of upstream heating, exhibiting the memory effect. The total power is more suitable for representing the main feature of dryout, and critical quality is the characteristic parameters. Critical quality increases with flow rate decreasing and pressure increasing, which is especially obvious under lower pressure and lower mass flow. For lower pressure and / or higher mass flow, dryout could occur at very low quality conditions. The effecting length is well beyond that commonly accepted for full development of T-H condition (L/D=60).The influence of non-uniform heating is closely related to pressure and flow rate. |