Stirred Vessels are used in a variety of industries to keep mixtures of fluids/solids (i.e. a slurry) in a homogenous state. Preventing solids from settling and keeping the slurry uniform is a key requirement for a variety of processes (e.g. leaching of mineral ores, feeding a filter/thickener, feeding a slurry pipeline). Round tanks are used for these applications with mixers installed on center. Wall baffles are installed in a stirred vessel to re-direct the tangential flow created by the impeller and promote a top to bottom flow pattern, which is necessary to keep the solids in suspension . Otherwise, the solids settle near the bottom of the vessel while the liquid remains in motion, which leads to liquid/solid separation.
The vessel manufacture typically provides the baffles with the tank and takes responsibility for the mechanical design of the baffles. Each baffle is held in place to the tank structure by a series of supports. There is a peak load experienced on each baffle during operation due to the fluid impact from the impeller. Historically these loads have been estimated using empirical data with large safety factors.
Computational fluid dynamics (CFD) has been used for a couple of decades by mixer manufacturers and industrialists to predict flow fields in stirred vessels. This output can be used to optimize impeller placement within the vessel, evaluate whether an incoming flow on a continuously operated tank will short circuit and/or assist in process troubleshooting. In order to rely on the CFD results and make sound decisions based on them, experimental validation of the CFD output data is required. Laser Doppler Velocimetry is an experimental method used by mixer manufactures to measure the flow performance of their impeller designs within a mixing vessel .
CFD can also be used to predict fluid pressure acting upon tank internals such as baffles. By using a steady state approach, the average fluid pressures can be determined on a baffle surface using CFD. When using a multiple reference frame approach to simulate the impeller rotation, the relative location of the blade versus the baffle in the model affects the velocity field near the baffle and thus the estimated baffle force. Thus it is crucial to consider modelling several conditions under steady state to determine the maximum load experienced at each baffle. Considering a sliding mesh model is another approach, which requires considerably more computational time to complete. The computational grid style, density and quality all have an effect on the results as well. Thus multiple computational grids should be used for the modeling to confirm that the results are grid independent .
For this presentation, steady state CFD predictions on baffle forces are presented for two impeller geometries; a Rushton impeller (SPX LIGHTNIN R100) and a pitched blade turbine (SPX LIGHTNIN A200). The tank modeled is a 48” diameter tank with four (4) 4” wide wall baffles filled with 48” of water. The impeller diameters and off bottom distances were the same for both impeller styles. A grid independence study, focusing on grid density near the baffles and impeller blades was conducted. A Realizable K-Epsilon turbulent model was used to solve the steady state flow field in each case. The impeller rotation was modeled using a multiple reference frame approach. Multiple impeller positions were considered and baffle force results are presented for each position modeled. Experimental validation using LDV measurements was conducted to validate the flow field. Baffle forces were experimentally estimated by measuring baffle deflection while the mixer was turning over the entire height of the baffle. This data was used to validate the computational approach. Future work will include comparing these results to CFD results using a sliding mesh time dependent model if time permits prior to the conference. The results using both methods will then be compared.
 Atiemo-Obeng, V.A., Penney, W. R., Armenante, P. Solid Liquid Mixing, in Paul, E.L, V. Atiemo-Obeng, and S.M. Kresta, “Handbook of Industrial Mixing”, Wiley, New York, 2004.
 Marshall, E.M., Bakker, A., Computational Fluid Mixing, in Paul, E.L., V. Atiemo-Obeng and S.M. Kresta, “Handbook of Industrial Mixing”, Wiley, New York, 2004.
 Oldshue, J., “Fluid Mixing Technology”, McGraw-Hill Publications Co., New York, New York, 1983.