Comparison of empirical models with an experimental database for condensation on banks of tubes

Empirical models for calculating the heat-transfer coefficients for condensation on banks of tubes were compared with the experimental data set obtained by the various previous investigators consisting more than 4000 data points for 6 different condensing fluids and 13 different tube bank configurations. All the banks considered in this study, involved the condensation of the pure vapours, the exceptions are that of the Briggs and Sabaratnam (2003) and Shah (1978, 1981), since their pure vapours consisted of incondensable air. For the forced convection flow region (F < 3.5), it was observed that some of the data were underpredicted by the recent models, recommending that the effect of the shear stress due to high vapor velocity overcomes the effect of the inundation on the heat transfer rate while vice versa is the case for the models overpredicting results. Similarly, for the free convection flow region (F > 3.5), it is suggested that the data overprediction by some of the models was due to the boundary layer separation and inundation effects, whereas the data underprediction was due to the generation of the turbulence within the condensate film due to high velocity on the condensate film. The inclusion of the inundation effect to a pure forced convection model of Shekriladze and Gomelauri (1966) as recommended by Cipollone et al. (1983) lead the model of Cavallini et al. (1985) to be the most accurate model compared to the other models for steam only. It was found that the Fujii and Oda model (1986) is the most accurate model among the empirical models been demonstrated in this paper, giving an agreement with the experimental data base to within an average absolute of the errors of about 21.5%. It accounts for the effects of the shear stress on the surface of the film condensate and the inundation within the bank.