Experimental study on the transition modes of falling film between horizontal 3D finned tubes and their transitional Reynolds number prediction model

To investigate the effect of different falling film modes on the heat transfer performance of three-dimensional (3D) finned tubes in a falling film heat exchanger, the falling film transition modes are experimentally investigated by observing the flow modes on 3D finned tubes and determining the Reynolds numbers of flow transition modes. A test facility, which contains an array of three horizontal test tubes, is constructed to study the effect of tube spacing and fin structure on the falling film Reynolds number (Re). The results show that tube spacing and fin structure significantly affect the Re and observed mode. With the increase in tube spacing, the Re overall shows an increasing trend for the four transition modes, especially for the transition between the column and the column–sheet mode. With the increase in the ratio for fin structure parameters, the Re overall shows a downward trend, and this phenomenon is more evident with the increase in the tube spacing. Machine learning methods are utilized to predict the Re, considering the effects of tube spacing and fin structure. Both this method and the linear regression method are used to predict the Re of the literature and this experiment, and the results indicate that machine learning has a lower prediction deviation.