Important Factors in Bubble Coalescence Modeling in Stirred Tank Reactors

Abstract

B ubble size distribution is among the important factors controlling interfacial mass transfer rate in gas liquid stirred tank reactors. This distribution is determined by a balance of coalescence and breakage rates. Despite a large number of articles reporting bubble coalescence, most of the previous investigations, such as those by Ueyema et al. (1993), Kim and Lee (1987), and Oolman and Blanch (1986), were done for a pair of bubbles under non-turbulent flow conditions.Bubble coalescence in a stirred tank reactor (STR) has been shown photographically for the first time by Rennie and Valentin (1968). Tse et al. (1998) have tried to visualise bubble coalescence in a coalescence cell, a STR and a bubble column. Unfortunately, the use of low record rate (50 frames per second, fps) in visualising bubble coalescence in the STR made it impossible to observe the coalescence occurring in time scale shorter than the record rate.Several researchers (Takahashi and Nienow, 1993;Bakker, 1992;Barigou and Greaves, 1992) show that Sauter mean diameter, d 32 , becomes larger as the radial distance from the impeller increases. They suggest that coalescence of bubbles takes place as the bubbles move away from the high shear rate region near the impeller blades toward the tank wall. Takahashi and Nienow (1993) have tried to quantify the coalescence rate solely based on measurement of bubble size during passage of bubbles from the impeller to the tank wall. In their approach, they did not differentiate the break-up from coalescence rate. Therefore their result cannot be used in developing population balance(s) coupled with equations describing hydrodynamics of gas-liquid flows.Near the wall, d 32 generally increases as the axial distance from the impeller plane toward the liquid surface increases indicating the coalescence takes place near the tank wall (Takahashi and Nienow, 1993;Bakker, 1992;Barigou and Greaves, 1992). The increase is more pronounced in the leeward or downstream side of the baffle compared to the windward or upstream side, especially in a short distance above and below the impeller plane (Takahashi and Nienow, 1993;Barigou and Greaves, 1992). Recently, Sudiyo et al. (2002) have observed consecutive frames using a fast CCD camera (up to 2000 fps), to quantify the coalescence rate and have shown that bubble coalescence occurs much more frequent on the leeward side than that in the windward side. This phenomenon is not clearly understood due to the lack of measurement data of liquid velocities and turbulent properties. Bubble coalescence has been studied in a 2.6 L stirred tank. Instantaneous velocity fields were measured using PIV and corresponding turbulent kinetic energy, dissipation rate, various length and timescales were estimated. All of these data, combined with data of local gas hold-up, bubble size and coalescence rate obtained with shadowgraph were used to assess bubble coalescence at different positions. Results show that bubble coalescence takes place mostly near the tank wall, especially on the leeward...

Bubble trapping and coalescence at the baffles in stirred tank reactors

“…The stationary vortices rotate fast, resulting in a low‐pressure region within the vortices. Bubbles tend to be captured by the vortex core, leading to higher gas hold‐up within the vortex, as shown experimentally by Sudiyo et al15 and Sudiyo16 in the case of a stationary vortex on the leeward side of the baffle, and by Van't Riet and Smith17 in the case of trailing vortices behind the impeller blade.…”

“…Instantaneous velocity fields obtained from PIV measurements15 were then analyzed in terms of turbulent structures. A detailed description of the PIV measurements, including the experimental set‐up and data handling, has been given by Sudiyo et al15 Turbulent structure is identified on the basis of Reynolds‐decomposed fluctuations and vorticity analysis, as only the large‐scale structures are of interest 23. The small structures do not contain sufficient energy to significantly affect the bubble motions.…”

“…In turbulent flows of a stirred tank, a stationary vortex is found on the leeward side of each baffle at the impeller plane 15, 16. Additionally, two trailing vortices originated from behind each blade can be seen as stationary vortices, since they are generated constantly at more or less the same position.…”

Mass transfer rates from bubbles in stirred tanks operating with viscous fluids

“…In case of stirred tanks, even though it may be considered that buoyancy collision rate, y B ij , is not important, this mechanism has an important contribution according to recent studies (Sudiyo et al, 2003). On the other hand, the laminar stress collision rate, y LS i (Friedlander, 1977), will be considered negligible for this model.…”

Particle-image-velocimetry measurements in organic liquid multiphase systems for an optimal reactor design and operation

“…There are uncountable publications describing ParticleImageVelocimetry (PIV) measurements. Most consider aerodynamic applications, but many PIV measurements have been documented for chemical engineering applications involving a liquid bulk phase in stirred tank reactors; for instance in Kim et al 2001;Pan et al 2008;Sharp and Adrian 2001;Yoon et al 2005), to analyze the velocity field near Rushton turbines under different conditions; or in (Baldi and Yianneskis 2003;Kilander and Rasmuson 2005;Sharp et al 1999;Sheng et al 2000;Sheng et al 1998;Sudiyo et al 2003), where turbulence features have been analysed in stirred tank reactors and used in particular to validate numerical modelling approaches (Sheng et al 1998;Yoon et al 2001). The impact of a secondary gas flow on hydrodynamics has been considered, e.g., in Hall et al 2005;Montante et al 2013).…”