The Mechanism Investigation Of Turbulent Heat Transfer Of Supercritical Fluids Using DNS Method

Supercritical fluid technology has been widely applied to the energy cycle and chemical engineering due to its superiority.In the energy cycle,turbulent heat trans-fer plays a important role for the heat transfer of supercritical fluid.Due to the sharp property variation,flow acceleration and buoyancy effect,the heat transfer mode of supercritical fluid shows three different mode:normal heat transfer,heat transfer de-terioration and heat transfer enhancement.Especially,heat transfer deterioration has a obvious hazard for the system of energy cycle.Due to the limitations of the experimental means of measurement under supercritical conditions,it is difficult to study the complex processes and mechanisms involved by traditional experimental methods.The conven-tional CFD method based on conventional Reynolds-averaged Navier-Stokes(RANS),in addition to failing to obtain detailed turbulent field characteristics,lacks turbulent models and turbulent heat flux models under variable properties conditions,resulting in large computational errors.The Direct Numerical Simulation(DNS)method,which does not require any model,can accurately simulate turbulent flow and heat transfer pro-cesses.It can not only analyze the complex mechanism of supercritical fluid turbulent heat transfer process in depth,but also obtain a large amount of detailed turbulent flow and heat transfer data,which can provide a basis for building a turbulent heat trans-fer model under supercritical conditions.Therefore,based on the DNS method,this paper develops a parallel DNS calculation procedure for the turbulent heat transfer of supercritical fluids to numerically simulate the turbulent flow and heat transfer process of supercritical water in a circular tube and study the complex turbulent heat transfer mechanism in supercritical fluid thermal convection.Direct numerical simulation using a thermophysical property table was conducted to study the characteristics of strongly heated air flows and heat transfer.The predicted mean velocity and wall temperature have an excellent agreement with the experimen-tal data.The analysis of turbulent statistics shows that the flow acceleration is the major cause of turbulence attenuation in the strongly heated air flow.Also the buoyancy against flow direction has a suppressive contribution to the turbulence and heat transfer.The decomposition of the skin friction and Nusselt number illustrates that the turbulent con-tribution to the skin friction persistently decreases and laminar contribution increases along the streamwise direction,while the laminar contribution to the Nusselt number is dominant except in the entrance where the inhomogeneous contribution is dominant due to the quick development of the thermal boundary layer.Direct numerical simulations have been conducted for two cases of forced flow at different supercritical pressures in a heated pipe with constant heat flux and a bulk Reynolds number of Re0=5400.The present study aims to reveal the mechanisms of turbulent heat transfer of superciritical fluids at different pressures in a spatial-developing flow.The results show that at supercritical pressure closer to the critical pressure,where the property variations become more drastic,both the skin friction coefficient and Nusselt number become smaller.The decompositions of skin friction and Nusselt number show it is mainly due to the large turbulence reduction along the streamwise direction.The analyses of turbulent kinetic energy(TKE),the turbulent shear stress,the production of TKE also confirm this point.Moreover,it was found that the thermophysical property fluctuations are very large and significantly influence the turbulent statistics in the supercritical fluid flows.Due to the large property fluc-tuations,it is found that the density-fluctuation-related terms are significant and their values are even comparable to the mean-density-related terms.Due to their negative contributions to turbulent shear stress and turbulent heat flux,the turbulence and heat transfer are severely attenuated by the large thermophysical property fluctuations.For near-wall scaling in spatial-developing flows at supercritical pressures,the semi-local velocity transformation with a semi-local coordinate shows a better agreement in the logarithmic region.However,a clear deviation still exists especially for mean tempera-ture because all the transformations only incorporate the local mean property variations and cannot consider their fluctuations.Direct numerical simulation was adopted to study the turbulent heat transfer of su-percritical water in a heated pipe with different heat flux.Due to the strong heat flux,when acceleration parameter Kv closed to 1.5*10-6,flow begins to change from tur-bulence to laminar state,turbulent shear stress experiences a sharp reduction,therefore,turbulence attentuation rises that suppress heat transfer.From the turbulence to the lam-inar state,friction factor will experience a prosess from decrement to increment due to the wall shear stress increases.Also,the present velocity transformation method has faced with a severe challenges due to the mismatching of velocity law under turbulent state.Due to the flow acceleration,the production of turbulent statistics becomes more important to total turbulent quantities.Direct numerical simulation was performed to investigate the turbulent heat trans-fer of supercritical water in upward pipe flows with and without buoyancy force.A small buoyancy first impairs the heat transfer and then improves the heat transfer due to the turbulence attenuation and recovery.With buoyancy increasing,the turbulence recovery gets earlier and stronger,and the wall temperature is significantly reduced due to the large buoyancy production of turbulent kinetic energy(TKE)in supercritical wa-ter flows.The instantaneous vortical structures based on Q criterion also clearly show the process of turbulence attenuation and recovery.In addition,the simulation using two eddy-viscosity turbulent models is compared to DNS data.The results show that the failed prediction of wall temperature and Nusselt number by those turbulent mod-els is possibly due to the bad results of TKE,production of TKE and turbulent Prandtl number Prt.Direct numerical simulation was applied to reveal the turbulent heat transfer of supercritical water in a plate.From the results,the increasing of Reynolds number lead to the averaged enthalpy increase in the turbulent core region due to the increasing of turbulence.Conditional velocity and temperature law show a bad agreement in the logarithmic law region.