Critical heat flux and associated phenomena in forced convective boiling in nuclear systems

Abstract

In evaporation of a liquid flowing in a tube or nuclear fuel element, there exists a transition (known as "dryout", "burnout", "boiling crisis" or "critical heat flux", CHF) from a high heat transfer coefficient regime to one of greatly reduced heat transfer coefficient. The conditions leading to dryout or CHF and the behaviour of wall temperatures in the ("post dryout or post CHF") region beyond it are of immense importance in nuclear reactor safety. In a nuclear reactor, the clad temperature excursion in the post-dryout region may be unacceptably high and the prediction of the location of dryout and the magnitude of the temperature excursion into the post-dryout region is of great importance. Moreover, the dryout transition and its effects are important not only in nuclear plant but also in many other types of heat transfer equipment. The main focus of work described in this thesis was the improvement and validation of phenomenological models for the prediction of CHF and of heat transfer beyond CHF ("post CHF" or "post dryout" heat transfer). The main focus has been on the process of annular film dryout. In phenomenological modelling of this process the dryout location prediction is sensitive to the boundary value of entrained fraction at churn annular transition, especially at high flow rates. The model was extended to churn flow so that integration of entrainment, deposition and evaporation processes could be started from onset of churn flow. A new correlation for the prediction of entrainment rate in churn flow was presented. The application of the new methodology to experimental data leads to improved predictions of CHF. Another long-standing problem, i.e. effect of heat flux on droplet entrainment, is addressed by analysing the contradictory results of previous experiments by using the annular film dryout model. The capability of phenomenological models to cover the whole range of CHF scenarios, i.e. from subcooled or very low quality to very high quality CHF, was demonstrated by using a possible transition criterion from bubble crowding model (an improved version of the Weisman Pie model) to annular film dryout model. These improved phenomenological models captured trends of CHF data very well (including the Look Up Table data of Groeneveld et al. 2007) and produced improved results over a wide range of system parameters such as pressure, mass flux and critical quality. The implementation of the phenomenological models was pursued by modifying and developing an Imperial College computer code GRAMP. In addition to its application in modelling CHF, the GRAMP code was extended to the post dryout region and predictions for this region compared to a range of data and the results were found to be satisfactory.