Studies on heat-transfer film coefficients inside a horizontal tube in falling film evaporators

The effects of liquid load, tube diameter, evaporation boiling point, condensation temperature difference and steam flow velocity on the heat-transfer film coefficients inside horizontal tubes were studied. The experiments were conducted with various operation parameters. In fact, if there is little shear stress on the film surface, it affects film thickness and blocks the condensate flowing inside the tube, but it is not strong enough to make the condensate to be ring flow. We defined the shearing Reynolds number in order to describe the effect of this small shear on the film of condensate. Using a regression of the experimental data, the models of condensation heat-transfer coefficient were built. According to the literature, some theoretical models cannot be used to explain why condensation heat-transfer coefficients increase gently with steam flow velocity. Our models of condensation heat-transfer coefficient can be used to explain this problem. These results provide a basis for further developing the high efficiency heat-transfer process and novel evaporators.

A study on the condensation heat-transfer coefficient h(")1 inside a smooth red copper tube was made. We conclude that: 1. With the change of liquid load outside the tubes, the variety of h(")1 inside the heat-transfer tubes is very small. With an increment of the evaporation boiling point t2, h(")1 increases greatly. With an increment of )t1, h(")1 is greatly reduced. With an increment of vs, h(")1 slightly increases. 2. Considering there is less shear stress on the film surface, we defined the shearing Reynolds number, which reflects the shear effect of steam inside the tube on the condensation film. The values of all the parameters can be calculated through a multivariable regression analysis of the experimental data. Thus, we can get the values of h(")1 and Eq. (5). The value of the complex dependent coefficient and biased correlation coefficient indicates that the linear regression result is very good. The effects of two variables — eigen function of measurement and shearing Reynolds number — are significant. Therefore, they cannot be ignored.