Numerical Simulation of Mixing Time in Stirred Reactors Based on Mean Age Theory

PAN Ao,XIE Minghui,XIA Jianye,ZHUANG Yingping

DOI: https://doi.org/10.14135/j.cnki.1006-3080.20180716001

2019-01-01

Abstract:A mixing time simulation method based on mean age theory is established. This method is suitable for industrial stirred reactors for its computing resource saving.The simulation process is as follows: 1) velocity field of bioreactor simulated by computational fluid dynamics (CFD); 2) mean age distribution simulated by CFD steady-state calculation; 3) volume integration process performed on the histogram of the mean age distribution to simulate mixing time. To get accurate mixing time, three conditions should be satisfied: 1) unstructured mesh element number should guarantee that the power calculated by energy dissipation rate should exceed 80% of the power calculated by the torque; 2) inlet flux should be two orders lower in magnitude than the impeller pumping flow rates; 3) number of mean age groups should be more than 12500 for Matlab post-processing.When ratio of height to diameter of reactor is lower than 1.5:1, simulation results are in good agreement with the experimental values, and average error is within 10%. When reactor has a height-to-diameter ratio of more than 1.5: 1, Reynolds number less than 2.1×10 5 will cause difference between simulation result and experimental value to become larger, and error will reach 12%. Comparing with pitch blade turbine pumping down (PBTD) and pitch blade turbine pumping up (PBTU), the simulation results of PBTD are all within the experimental data range, and the simulation results of PBTU are higher than the experimental values. The maximum error is up to 14%. In addition, many mixing time empirical formulas do not consider the volume factor. Through the investigation of this thesis, it is found that N·θ number obtained by simulation is not affected by scales of reactors.