For the design of mixing and agitation facilities in process engineering it is of central importance to appraise the correct viscosity of fluids. This can be a challenging task when non-Newtonian and/or non-homogeneous fluids need to be processed. Since it is not always possible to analyze them in the classical ways, an propeller viscosimeter approach on the basis of the Rieger-Novak-Method is used. In recent years the Lattice Boltzmann Methods (LBM) are established as an alternative approach to classical computational fluid mechanics methods. The utilization of Cartesian grids avoids the need to discretize with boundary conform meshes. This makes the LBM suitable for complex geometries like a propeller in this case. Numerical simulations were carried out using a 3D in-house Lattice Boltzmann code called SAM-Lattice with our latest extension to non-Newtonian flow. We use a truncated form of the power-law approximation to accommodate the varying flow properties in non-Newtonian simulations, where the effective viscosity is a function of the shear rate. SAM-Lattice comprises the LBM solver and a highly automated grid generator for arbitrarily complex geometries. The code is capable of multi-domain grid refinement as well as multi reference frames and rotational boundaries. The post processing is done using an extension of the open source visualization tool Paraview. We compare results of experiments and LBM simulations for the Newtonian case (Glucose) to validate our Lattice Boltzmann solver. A study of the non-Newtonian, shear thinning case (Xanthan) is conducted to validate the generalized Newtonian model. The propeller viscosimeter is currently under development as a standalone solution for viscosity measurement. For calibration purposes the Metzner-Otto-constant of the propeller device has to be determined. While the constant is valid for the laminar region the numerical results for the agitator characteristics are presented. Different levels of grid refinement are tested to assure independence of the lattice resolutions.
Validation of Three-Dimensional Grid Refinement for Lattice Boltzmann Methods