On the effect of conjugate heat transfer on turbulence in supercritical fluids: Results from a LES application

- LES analyses are performed for a simple problem with supercritical pressure water.
- Cases with and without conjugate heat transfer at the heating wall are considered.
- Larger wall temperature oscilattions were observed without conjugate heat transfer.
- The need to perform resolved CFD analyses with conjugate heat transfer is suggested.

The present paper reports the results obtained from Large Eddy Simulation (LES) analyses concerning the effect of conjugate heat transfer with walls when dealing with heat transfer to supercritical pressure fluids. A specific operating condition was investigated with and without walls showing a clearly understandable effect on turbulence of the actual characteristics of the wall, something further complicating turbulence modelling, already quite difficult in this field.Unlike the experience of past studies, which considered conjugate heat transfer by LES or DNS when dealing with constant property fluids, and resulted in a limited influence on the observed phenomena, in the present work relevant effects are instead identified. In fact, in the case of supercritical pressure fluids, the strong changes in fluid properties close to the pseudo-critical threshold may provide a strong feedback on the velocity field and then on turbulence; in particular, the presence of a wall with realistic properties strongly damps the large temperature and fluid properties fluctuations obtained when imposing a constant heat flux.Consequently, unlike fluids in standard conditions, heat transfer to supercritical fluids seems to be depending on the actual fluid-and-wall coupling, thus adding a further challenging aspect in this already complicated topic. Though further analyses are underway for confirming the observed behaviour, the presented findings related to a simple example open new scenarios in the development of heat transfer correlations and CFD models to be used for supercritical fluids. In fact, the available data, both experimental and by DNS, can no more be considered independent from the imposed boundary conditions at the wall and the effect of the wall properties should be seriously taken into account.