Laboratory for Flow Imaging & Diagnostics

Dynamic in situ measurement of axial segregation of E-CAT particles in a bubbling fluidized bed

Abstract

The present paper summarizes comprehensive in situ measurements of the temporal segregation/mixing behavior of the non-spherical Geldart B and D E-CAT particles mixture in a bubbling fluidized bed. The experiments are performed for both unsieved and sieved E-CAT particles with varying particle size distribution width and at the fluidization velocity ranging from 1.3 to 5Umf. The high-speed pressure and image data were used to provide time-resolved quantification of dynamic expanded bed height and pressure drop during the experiment. In parallel, a hybrid machine learning-aided image processing algorithm has been developed for quantifying the instantaneous particle size distribution at varying axial locations. Results reveal that the segregation exists only at low fluidization velocities (1.3 and 1.6Umf) for unsieved cases with a wider size distribution. In contrast, particles with a narrower distribution exhibited well-mixed behavior for the full range of tested gas velocities. The segregation intensity and spatial extent were found to correlate with particle size distribution width and fluidization velocity, respectively. In addition, by analyzing pressure, expanded bed height, particle size and motion, and bubble characteristics, we provided a comprehensive multi-scale dynamic view of the segregation process. This in-depth understanding of the process, coupled with the time-resolved datasets obtained, can be instrumental for benchmarking numerical models in solid–gas multiphase flows and for optimizing the design of fluidized beds.

Experimental setup

The experimental setup collected data on bed height, pressure drop, and axial particle distribution using high-speed imaging and pressure sensors, to analyze the segregation behavior of ECAT particles.

Figure 1. The (a) original and (b) cumulative particle size distribution for both sieved and unsieved cases measured using QICPIC®, and (c) ECAT particle samples along with micrographs showing particle non-sphericity.
Figure 1. The (a) original and (b) cumulative particle size distribution for both sieved and unsieved cases measured using QICPIC®, and (c) ECAT particle samples along with micrographs showing particle non-sphericity.
Figure 2. The schematics of the 25.4 mm inner diameter cylindrical fluidized bed and the fluidization visualization and measurement systems.
Figure 2. The schematics of the 25.4 mm inner diameter cylindrical fluidized bed and the fluidization visualization and measurement systems.

Methodology

Bed height data were analyzed through a five-step procedure (Fig. 3), pressure-drop data were obtained directly from the sensors, and particle size distributions were evaluated at three axial positions—top, middle, and bottom—as illustrated in Fig. 4.

Figure 3. Image-processing pipeline of bed height analysis.
Figure 3. Image-processing pipeline of bed height analysis.
Figure 4. Image processing flow chart for non-intrusive extracting particle size information.
Figure 4. Image processing flow chart for non-intrusive extracting particle size information.

Data process and analysis

The particle size analysis indicates clear segregation in the Unsieved 1.3 Umf case: over time, larger particles accumulate in the bottom and middle regions while smaller particles concentrate at the top, and a defluidization region gradually develops near the bed surface.

Figure 5. Temporal evolution of volumetric mean particle size (dp) at 1.3, 1.6, and 2Umf for (a,b,c) unsieved and (d,e,f) sieved cases. The blue, red, and green dots represent the experimental data for the top, middle and bottom region of the bed, respectively. The solid color lines correspond to an exponential fit.
Figure 5. Temporal evolution of volumetric mean particle size (dp) at 1.3, 1.6, and 2Umf for (a,b,c) unsieved and (d,e,f) sieved cases. The blue, red, and green dots represent the experimental data for the top, middle and bottom region of the bed, respectively. The solid color lines correspond to an exponential fit.
Figure 6. Flow region detection at four different moments at 1.3Umf for unsieved case.
Figure 6. Flow region detection at four different moments at 1.3Umf for unsieved case.

Publication

Luo, B., Lu, B., Gao, X., & Li, C. (2024). Dynamic in situ measurement of axial segregation of E-CAT particles in a bubbling fluidized bed. Chemical Engineering Journal, 485, 149861.

Address: 241 Daxue Road, Jinping District, Shantou, Guangdong Province, China

©Guangdong Technion-Israel Institute of Technology 2018

 

ABOUT

ACADEMIC

QUICK LINKS

粤ICP备17036470号，版权所有©广东以色列理工学院