Validation Criteria for DNS of Turbulent Heat Transfer in Pipe Flow

Abstract

Direct numerical simulation (DNS) of turbulent flow and heat transfer involves directly solving the unsteady Navier-Stokes and thermal energy equations without considering any assumptions about the physics and resolve all the scales of the flow, including the energy and dissipation spectral peaks. Data from DNS of fully developed turbulent pipe flows subjected to a constant surface heat-flux are used to validate with the existing DNS database. The present validation process involves matching of first and second order turbulence statistics. However, accurate comparison of DNS data sets requires satisfying the conformity of necessary and sufficient computational parameters. This paper will address the influence of the computational conditions on the thermal statistics, which affect the comparison of the matching statistics from different DNS data sets. It was observed that apart from the governing parameters (Kármán and Prandtl numbers), computational mesh and domain size significantly influence the resulting thermal statistics, which must be considered while performing validation of DNS data.