Impeller shape-optimization of stirred-tank reactor: CFD and fluid structure interaction analyses

Mixing is an important operation in the chemical industry. The present numerical study, validated by experimental data, offers a computational framework for minimizing the stirred-tank reactor power consumption and the impeller equivalent stress. The flow pattern, turbulence parameters, and power consumption of the three types of impellers, named as 6-blade Rushton turbine and U-shape and V-shape impeller designs, are numerically studied using the Computational Fluid Dynamic (CFD) analysis. The interaction between the fluid and the solid is then studied through a two-way Fluid Structural/Solid Interaction (FSI) analysis. Multi-objective Genetic Algorithm (MOGA) is used along with both Direct Optimization (DO) and Response Surface Optimization (RSO) to optimize the blade's geometrical parameters, including blade thickness (Tb), disk thickness (TD), and vertical angle of the blade (θ). The variation between experimental and numerical results, in terms of power consumption, is around 5%. The computational results show that the power number (Np) can decrease, respectively by 21% and 48%, for the V- and U-shape impellers compared to the 6-blade Rushton turbine. The maximum turbulent kinetic energy and the maximum dissipation rate of the turbulent kinetic energy of the U-shape (k = 0.065 m2/s2, ε = 65.14 m2/s3) are more than V-shape impeller (k = 0.044 m2/s2, ε = 25.16 m2/s3). The FSI results also show that the lowest von Mises stress occurs in the V-shape impeller.