Heat Transfer in Vapour-Liquid Flow at High Reduced Pressures

During the last decade a number of studies of boiling heat transfer in carbon dioxide notably increase. As a field of CO₂ practical using corresponds to high reduced pressures, and a majority of available experimental data on CO₂ flow boiling even in submillimetric channels relate to turbulent liquid flow regimes, a possibility arises to develop sufficiently general method for HTC predicting. Under the above conditions nucleate boiling occurs up to rather high flow quality, even in annular flow regime due to extremely small size of an equilibrium vapour bubble. This conclusion is in agreement with the available experimental data. The predicting equation for nucleate boiling heat transfer developed by one of the present authors in 1988 is valid for any nonmetallic liquid. A contribution of forced convection in heat transfer is calculated according to the Petukhov et al. equation with correction factor, which accounted for an effect of velocity increase due to evaporation. This effect can be essential at relatively small heat fluxes and rather high mass flow rates. The Reynolds analogy and homogeneous model are used in order to account for the convective heat transfer augmentation in two-phase flow. Due to low ratio of liquid and vapour densities at high reduced pressures the homogeneous approximation of two-phase flow seems to be warranted. A total heat transfer coefficient is calculated as an interpolated value of boiling and convective HTCs. The experimental data on CO₂ flow boiling related to regimes before heated wall dryout incipience are in rather good agreement with the calculations. Besides the data on carbon dioxide flow boiling, the results on water, helium, nitrogen and some refrigerants were used for comparison; at rather high reduced pressures the computed and the measured values of HTCs are in a good agreement. The data include results obtained in the channels of a diameter from 0.6mm up to 18mm. It is clear that at high reduced pressures there is no strong variation in boiling heat transfer with channel size decrease, it means that a classification on channel size has no sense if it does not consider liquid/vapour densities ratio.