(601d) Solid Suspension in Stirred Tank Reactor: Hysteresis in Cloud Height, Measurement of Solid Velocity Using Uvp and Cfd Simulations

Solid suspension in stirred tank reactor (STR) is commonly used in process industries for catalytic reactions, dissolution of a solid, crystallization, and so on. In such slurry systems, the quality of suspension is a critical parameter for reliable design, optimum performance and scale-up. Suspension quality in stirred tank reactor depends upon complex interactions of impeller generated flow, turbulence and solid loading. Recent advances in computational fluid dynamics (CFD) and experimental techniques for characterizing solid distribution, cloud height and solid velocity profiles provide new ways of understanding solids suspension in stirred tanks. Several attempts have been made to simulate solid suspension in stirred tanks (see for example, Gosman et al., 1992; Micale et al., 2000; Khopkar et al., 2006). However, many of these modeling efforts were evaluated by comparing results on averaged solids volume fraction profiles. Unfortunately comparison of averaged solid volume fraction is not an effective way to discriminate between closure models of inter-phase drag force. Thus, there is still no consensus on closures of inter-phase drag force terms for solid-liquid suspension in these published studies. In order to provide more stringent tests for evaluating CFD models, in this work we report new data on hysteresis of cloud height and local solid phase velocities measured using ultra-sound velocity profiler (UVP). There are relatively few studies available on measurements of local solid velocity profiles (Guiraud et al., 1997; Fishwick et al., 2003) and cloud height of suspended solids (Hicks et al., 993, 1997; Bujalski et al., 1999). When solid particles are not uniformly suspended, one can distinguish clearly two regions in the vessel: the bottom region in which solids are suspended called as cloud region and a distinct layer of clear liquid on top of solid cloud. Fundamentally, lifting of particles from bottom of the vessel requires more input energy than energy required for particles to keep them in suspended condition. This manifest into a hysteresis in observed cloud heights when impeller rotational speed increases and decreases (see Figure 1). When impeller rotational speed is zero, all solids are settled at the bottom. As impeller speed increases, the cloud height starts increasing. When impeller rotational speed is reduced after reaching uniform suspension of solids at high rotational speed, it exhibits hysteresis in cloud height as shown in Figure 1. The experimental data on hysteresis in cloud height is expected to provide a stringent test for the CFD models since it crucially depends on inter-phase drag force and its dependence on solid volume fraction and prevailing turbulence.  In this work we provide data on hysteresis in cloud heights for two different particles and for two different impellers. In addition to this data, measurements of local velocity of solid phase using UVP (Ultrasound Velocity profiler) were also carried out. These local velocity measurements and hysteresis data will be used for evaluating the CFD models. The CFD models were developed to simulate the experiments reported here. Influence of various drag force formulations was studied. The details of work done are discussed in the following.

In this work, a cylindrical, flat-bottomed, acrylic tank of diameter 0.7m with standard baffles of width T/10 was used for experiments. Pictorial view of experimental set up is shown in Figure 2a. Experiments were conducted with 6-bladed pitched blade turbine (PBTD-down-pumping) and hydro-foil impeller with solid loadings such as 1%, 3%, 5%, and 7% v/v of glass particles (d_p = 250 mm and 50 mm) for impeller speeds ranging from 150-600rpm. Local solid velocity profiles were obtained by UVP. Time averaged radial velocity and axial velocity data was acquired at three axial locations (z/R = 0.3, 0.53, 0.8) and seven radial locations respectively. Cloud height measurements were done with visual observation (shown in Figure 2b). For the study of hysteresis effect in cloud height, impeller speed was increased gradually from 150rpm to 600rpm and after attaining uniform suspension of particles the impeller speed was decreased gradually from 600rpm to 150rpm. A sample of experimentally observed hysteresis data is shown in Figure 1. CFD model was developed using the Eulerian-Eulerian approach including the standard k-ε mixture turbulence model. Multiple Reference Frame (MRF) approach was used to simulate impeller rotation with the steady-state condition. Computational geometry developed and schematic presentation of solution domain shown in Figure 3a. Different inter-phase drag force closures were used and tested in this work (Brucato et al., 1998; Khopkar et al., 2006). Contour plots of solid phase volume fraction at different impeller speed shown in Figure 3b. The model was used to simulate suspension of solid particles from bottom of vessel as well as when particles were uniformly distributed to simulate experiments on hysteresis effect. Preliminary results of this are shown in Figure 1. The reported experimental data, modeling approach and presented results in this work will be useful for enhancing understanding of solid suspension and quality of suspension in stirred reactors. The results will also be useful for designing and scaling up of solid-liquid stirred reactors.  

References

Brucato A., Grisafi F. and Montante G., 1998, Chem. Eng. Sci., 53 (18), 3295-3314.

Bujalski W.K. 1999, Suspension and liquid homogenisation in high solid concentration stirred chemical reactors, Trans. Inst. Chem. Eng., 77, 241-247.

Fishwick R. P., Winterbottom J.M., Stitt E.H., 2003, Catalysis Today, 79-80, 195-202.

Gosman A. D., Lekakou C., Polithis S., Issa, R., Looney, M.K., 1992, AIChE J, 1992, 38(12), 1947

Guiard P., Costes J., and Bertrand J., 1997, Chem. Eng. J., 68, 75-86

Hicks M.T.,1993, Cloud height, fillet volume, and the effect of multiple impellers in solid suspension, presented at Mixing XIV, Santa Barbera, CA, June 20-25.

Hicks M.T., K.J. Myers, and A. Bakker., 1997, Cloud height in solid suspension agitation, Chem. Eng.Commun., 160, 137-155.

Khopkar A. R., Kasat G., and Ranade V.V., 2006, Ind. Eng. Chem. Res. 45, 4416 - 4428

Micale G., Montante G., Grisafi F., Brucato A., Godfrey J., 2000, Trans. Inst. Chem. Eng. Part A, 78, 435        

             

Figure 1: Hysteresis in height of suspended solids (cloud height) with impeller rotational speed  

       

                        Figure 2a                                              Figure 2b

Figure 2: Experimental set-up and measurement of cloud height for a case with average volume fraction of 7%

         

                        Figure 3a                                                          Figure 3b

Figure 3: Solution domain and predicted contour plots of solid volume fraction for a case with average volume fraction of 7%