LAUSR

Abstract Recent papers by the Authors proposed a successful rationale to establish a similarity theory for heat transfer at supercritical pressures, making use of original definitions that seemingly solved the problem of finding a closely similar phenomenological behaviour with very different fluids. As promised in the conclusions of the latest publication, this paper reports on the result of the work performed by further exploring the features of the theory, confirming its choices in front of additional evidence. In particular, several further RANS calculations, made by the same improved turbulence model adopted in recent works, have been performed, assessing various additional features of the theory and finding excellent confirmations of its capabilities. While waiting for experimental confirmations, RANS analyses are here used as a suitable workbench to compare data obtained by four different fluids (water, carbon dioxide, ammonia and R23) for “similar” conditions. The predicted behaviour, in addition to be conditioned to the limited capabilities of the adopted turbulence model, can be attributed to the different thermodynamic and thermo-physical properties, i.e., to intrinsic features of supercritical fluids. The obtained results further clarify relevant issues, as the most appropriate selection of supercritical pressure for each fluid in front of a given reference condition, together with the level of accuracy achieved in establishing similarity and the role of the most important dimensionless numbers. It is shown that the reported results depict a very clear picture of the role played by the different boundary conditions in determining the relevant heat transfer regimes, with specific reference to deterioration and recovery, opening the possibility for a thorough and physically based revision of the only partially successful engineering correlations for heat transfer proposed up to now.