| Popis: |
After very many rather naive attempts to understand and model Critical Heat Flux (CHF) on the basis of unique mechanisms that resulted in a myriad models and of correlations, we came “back” to Look Up Tables that implicitly recognize that the data are not correlatable in general ways and then, with the advent of Computational Multi-Fluid Dynamics (CMFD) methods back again to predicting CHF mechanistically. The situation is clearly much more complicated in the fuel rod bundle assemblies used in nuclear reactors than in tubes or other simple channel geometries. For rod bundles, subchannel analysis methods have been used but the challenge of correlating CHF to the local conditions remains, in addition to the difficulties in computing the inter-subchannel fluxes. Recently, CMFD computations of the critical heat flux became the grand challenge and efforts were undertaken on both sides of the Atlantic to produce predictions of CHF for both the Departure from Nucleate Boiling (DNB) and the Dryout situations in PWRs and BWRs, respectively. CMFD methods come closer to tackling in a “natural” way the fuel-bundle and transient problems also, as the same basic, microscopic-level models and numerical techniques can be applied to any geometry also under transient conditions if sufficient computational resources are available. Much more advanced instrumentation is in the meantime producing the much more detailed, microscopic data needed to validate the CMFD methods. The paper reviews briefly the historical developments and then focuses on the recent CMFD work, its promises and still existing restrictions; the phenomena are too complex to be fully resolved by CMFD. Although we had set this as a goal over ten years ago, we are not there yet, but the spinoffs from this effort have been beneficial. |
Full Text from ScienceDirect