Impact of Conjugate Heat Transfer on the Turbulence and Heat Transfer in an Upward Heated Pipe Flow at Supercritical Pressure

Jundi He,Wei Wang,Bing Xu,Shuisheng He
DOI: https://doi.org/10.1016/j.ijheatmasstransfer.2024.126004
IF: 5.2
2024-01-01
International Journal of Heat and Mass Transfer
Abstract:Direct numerical simulations (DNS) of an upward heated pipe flow of supercritical CO2 (at Re0=3600) were carried out with and without conjugate heat transfer (CHT). The aim is to identify the impact of CHT on this specific flow problem and provide information for relevant numerical studies. The configuration is referred to one of the experiments carried out by Luo (2014), with a stainless steel (SS 304) pipe and a thermal activity ratio of 0.061. When CHT is introduced, the predictions of heat transfer coefficient (HTC) and wall temperature obtained in DNS are slightly improved against the experiment measurements. Two primary factors arise when CHT is considered, which are the axial redistribution of wall heat flux and the damping of the enthalpy fluctuations. The former affects the heat transfer particularly within a short distance after the inlet, leading to a more rapid establishment of the thermal boundary layer and a more rapidly progressing ‘laminarising’ process. The latter introduces substantial differences in the fluctuations of enthalpy (∼30 times difference) and density (∼40 times difference), which is however confined within a small region close to the wall and diminishes rapidly further away, resulting in a minor effect on turbulence and heat transfer in the main flow. When CHT is omitted, turbulent kinetic energy is over-predicted up to 1.55 times at most, in particular when the flow is being ‘laminarised’ (z/D<30). This is attributed to the discrepancies in shear production rather than the buoyancy production of the turbulent kinetic energy. Additionally, CHT alters both the laminar (Nul) and turbulent contributions (Nut) to the Nusselt number, with Nul starting at a lower level close to the inlet and the peak of Nut being reduced by 12.3%. These findings provide an insight on the impact of CHT on turbulence and heat transfer at different stages and valuable information for a more accurate modelling method for supercritical fluid flows.
What problem does this paper attempt to address?