Forced-convective heat transfer and hydrodynamic aspects of co-current, two-phase, two-component flow r¡ere studied experimentally in a 0.46 in. (1.168 cm) i.d. vertical tube with essentiaÌIy no evaporation. The effect of reducing the surface tension on the heat-transfer coefficients, both loca1 and length mean, the frictional pressure drop, and the flow patterns were investigated. Three Iiquids (water, glycerine & water solution of 58% glycerine by weight, and silicone tiguid, 5 CS viscosity grade) were used in the study. The glycerine & water solution was chosen so that it would have essentially the same Prandtl number as that for the silicone liquid, while its surface tension was 3.4 times that of the latter.

Àt low-to-medium liquid flow rates, the mean heattransfer coefficient for silicone-air increased with the increase of the gas flow rate untit it reached a maximum, after which there nas a substantial drop as the mixture was approaching the annular-mist transition. A further increase in m_G was followed by a monotonic increase in h_TP with the attendant condition of the gas phase becoming more important in the heat-transfer process.

The behaviour and shape of the loca] heat-transfer coefficients along the test section length remained unchanged in most cases with trends that varied according to the combination of the liquid and gas flow rates.

The effect of lowering the surface tension on the flow pattern vras most profound on the bubble-slug boundary, which fell considerably, expanding the bubble flow region for the silicone-air system. smaller changes were observed for the other flow Pattern transitions.

The frictional-pressure-drop data for the silicone-air were slightly lower than those for the glycerine & water-air data in terms of the ratio ϕ² vs v_SG/v_SL, tt the same liquid Reynolds numbers.

À comprehensive study and comparisons Ytere done with most of the existing correlations that predict the mean heattransfer coefficient in two-phaSe, two-component, Cocurrent, upltard flow in vertical tubes. These correlations were tested against a large set of data which included a very wide range of variables and fluid properties.

Two correlations are presented for determining the mean heat-transfer coefficient in the laminar and turbulent flow regions. These two correlations give the best combination of predictive capabilities together with minimum restrictions.