Chemical Engineering Science 63 (2008) 2559 – 2575 www.elsevier.com/locate/ces Mass transfer across the falling film: Simulations and experiments Z.F. Xu, B.C. Khoo ∗ , N.E. Wijeysundera Department of Mechanical Engineering, National University of Singapore, Singapore 119260, Singapore Received 16 August 2007; received in revised form 31 January 2008; accepted 5 February 2008 Available online 15 February 2008 Abstract Mass transfer across the thin falling film gas–liquid interface is a very important process as in chemical engineering and other fields, and yet there is still a lack of general predictability of the transfer quantity based on basic hydrodynamic parameters and independent of the geometrical setup. In this work, a numerical simulation is carried out for a vertical falling film arrangement. The wave dynamics and the associated mass transfer phenomena are discussed and compared with previous experimental empirical relationships. Based on the validity of the simulated results for wave parameters, numerical experiments for mass transfer were carried out with the aim of comparing to the empirical relation based on a single hydrodynamic parameter  (the gradient of the vertical fluctuating velocity at the interface) established previously by Law and Khoo [2002. Transport across a turbulent gas–liquid interface. A.I.Ch.E. Journal 48(9), 1856–1868.] and Xu et al. [2006. Mass transfer across the turbulence gas–water interface. A.I.Ch.E. Journal 52, 3363–3374] with various non-falling film experiments. Separately, experiments in an inclined plate thin falling film apparatus were carried out to determine the  distribution and associated mass transfer. It is found that there is reasonable concurrence with the mentioned empirical relation, hence suggesting the general applicability of  characterizing the scalar transport across the gas–liquid interface independent of the means of turbulence generation.  2008 Elsevier Ltd. All rights reserved. Keywords: Falling films; Mass transfer; Hydrodynamics; Interface; Hanratty’s ; Turbulence 1. Introduction Thin liquid films falling under the influence of gravity along vertical or inclined solid surface are encountered in a wide range of industrial process equipments, including wetted-wall absorbers, falling film chemical reactors, condensers, and verti- cal tube evaporators. Reliable design of these processes depends on the ability to accurately predict the transport rates of heat and mass to the flowing film. However, the interfacial and wall- to-liquid heat and mass transport processes for wavy falling films are somehow affected by the unsteady hydrodynamic characteristics of the wavy films such as the film thickness, velocity distribution, wall shear stress variations and others. The hydrodynamics of falling film have been studied by numer- ous researchers, and yet there is no consensus on a universally accepted general correlation between the scalar transport and the (dominant) thin film hydrodynamic parameter(s) applicable over a wide range of Reynolds numbers. ∗ Corresponding author. Tel.: +65 68742889; fax: +65 67791459. E-mail address: mpekbc@nus.edu.sg (B.C. Khoo). 0009-2509/$ - see front matter  2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ces.2008.02.014 Invariably, several researchers have focused on the free sur- face wave characteristics and used various measurement meth- ods to investigate the wave patterns of liquid films. However, making reliable experimental measurements of the local flow structure in the film is exceedingly difficult due to the extremely thin film dimension (≈ 1 mm), the short passage time of each wave (≈ 60 ms) and the random location of the wave height. Previous experimental methods can be divided into two cat- egories: (1) intrusive method, such as time-recording of film thickness at one or more location(s) using various probes; (2) non-intrusive methods, such as shadowgraph method and light absorption and reflection method. Karapantsios et al. (1989) evaluated the statistical char- acteristics of a falling film within a vertical pipe by means of the “parallel-wire conductance probe” technique. Lyu and Mudawar (1991) developed an alternative technique based on the principle of hot-wire anemometry. Nosoko et al. (1996) have used a needle contact technique to measure the wavy peak thick- ness. The obvious disadvantage of this class of methods is that no matter the small size of the probes, it will affect the flow field