Study of fluid-to-fluid scaling for upward pipe flows of supercritical fluids using direct numerical simulation

Fluid-to-fluid scaling is widely applied in experiments of supercritical fluid flows to reduce the cost and technical difficulties. Different scaling schemes, in most cases, formulated using a set of non-dimensional parameter groups that characterise the heat transfer of a supercritical fluid flow, have been developed and validated in the literature. Validations were previously mostly based on experimental data, empirical correlations and Reynolds-averaged Navier–Stokes (RANS) simulations. In the present study, direct numerical simulations (DNS) are used to assess one of the scaling schemes Ambrosini and De Rosa [1]. Simulations of upward pipe flows of four different supercritical fluids were carried out with their boundary conditions scaled accordingly. Excellent similarity in turbulence and heat transfer is achieved among the chosen fluids when the parameter group recommended by Ambrosini and De Rosa [1] together with the inlet Reynolds number (Re0) is used. The similarity becomes weaker when Re0 is replaced by the inlet Peclet number Pe0. Further understandings in terms of the conditions to achieve similarity and the parameters that determine the turbulence and heat transfer in upward heated flows of supercritical fluids are established.