Flow and residence time characteristics of Taylor-Couette flow device

Abstract

In this study, characteristics of Taylor-Couette (T.C.) flow was investigated using experiment and CFD simulation. T.C. flow is generated between two co-axial cylinders by the rotating inner cylinder and stationary outer cylinder. The T.C. flow device has been applied in various areas. To apply T.C. flow reactor in a variety of fields, characterization of reactor is essential. Previous researches have been limited to batch type T.C. flow reactor. Characterization of continuous T.C. flow reactor has not been reported extensively in operating condition. Flow regime and flow motion in a T.C. flow were studied. Flow regime changed depending on Taylor number. We confirmed flow regime using rheoscopic fluid. From simulation of fluid flow, velocity vector was displayed for confirmation of flow regime and streamline used to estimate flow motion and number of vortex revolution. The number of vortex revolutions was decreasing with increasing inner cylinder rotation speed during rotation of a big circle. Number of vortex revolution was from 5.77 to 0.65 during a big circle. Residence time distribution (RTD) in a T.C. flow reactor was investigated. We conducted experiment and simulation in a wide range of operating condition. Residence time distribution is an essential information for the characterization of continuous device. From residence time distribution, we can confirm mixing property of reactor. Effect of inner cylinder rotation speed was investigated on residence time distribution. In vortex flow regime, RTD became narrower than RTD without rotation. In turbulent flow regime, variance of RTD increased by increasing rotation speed. Peak time of RTD came earlier with an increase of rotation speed. CFD simulation was conducted to understand the shift of peak time with rotation speed. Axial velocity data was extracted from simulation results. Due to increased average axial velocity, RTD was peak time of RTD came earlier by increasing rotation speed. Effect of inlet flow rate was also studied. RTD became broader with lower intensity by decreasing flow rate. Peak time of RTD came faster with a decrease of inlet flow rate. Effect of size and density of particle were studied on the particle residence time in a T.C. flow reactor. We prepared particle injection and separation apparatus for particle RTD experiment. In the system, we simulated and described particle trajectory and particle residence time by CFD software. Residence time was analyzed more than 1,000 particles for each particle size and particle density from simulation results by discrete phase model. Particle RTD became broader and peak time of particle RTD appeared later with an increase of particle size at 1.5 g/mL. Influence of particle density was also investigated for 1.06 and 1.5 g/mL. When particle density is close to fluid density, particle RTD follows fluid RTD. Particle position in a T.C. flow reactor was analyzed by CFD simulation. Particle coordinate information was extracted and analyzed from particle trajectory depending on time for different particle size and density. Particle density was higher in outer region with increasing size and density. Based on this research, we described characteristics of a T.C. flow reactor in a wide range of operating condition and a T.C. reactor including particle. Results of RTD and particle RTD will provide a basis to control reaction time and degree of mixing in a T.C. reactor. These flow characteristics and RTD information would be useful to apply continuous T.C. flow reactor in industry.