CFD prediction of heat transfer at supercritical pressure with rough walls: Parametric analyses and comparison with experimental data

In this paper, a low-Reynolds number turbulence model developed by the authors in past activities for rough walls is used for predicting heat transfer at supercritical pressure in the presence of different degrees of surface finishing. The model, able to reproduce the typical trends of friction factors from classical data reported by Nikuradse and summarised in the Moody diagram, is based on a simple-minded description of the effect of wall protrusions through the boundary layer on turbulence production. Though prior validation of the model only on the basis of friction factor data did not assure any basis for achieving accuracy in heat transfer prediction, a sensitivity analysis is firstly presented in order to characterise the obtained predictions at variable values of the roughness parameter, in particular concerning the possible suppression of deteriorated heat transfer by roughened surfaces. These analyses are extended to different fluids, making use of a fluid-to-fluid similarity theory recently proposed by the authors in order to establish similar boundary conditions and predicted phenomena. The results obtained by these analyses can be considered interesting, especially in view of the design of supercritical water-cooled nuclear reactors; however, an assessment against experimental data was obviously necessary. Experimental carbon dioxide data published in a very recent archival paper were thus addressed and were found useful in this regard. Considering these data allowed to extend the above analysis to provide confirmation of the promising features of the model in comparison with wall temperature values obtained with different boundary conditions. The model here described appears promising not only for its capability to predict experimentally measured effects, but also for the perspective to be used in the study of the behaviour of purposely roughened surfaces reducing the probability of occurrence of deteriorated heat transfer