Download files
Abstract
The spouted bed is a reactor designed for the gas-solid contacting purpose and can deal with particles of a wide range of particle density, complex shape, and different sizes. Thus, it has been widely applied to a variety of processes, such as drying, mixing, granulation, coating, gasification, pyrolysis, and combustion. Its superior gas-solid mixing performance can be mainly attributed to the symmetrical particle circulation movement brought by the strong vertical spouting. Particles experience the dilute region (i.e., spout region and fountain region) and dense region (i.e., annulus region around the spout region) in a routine process, thereby ensuring the homogeneity and efficiency of the gas-solid contacting.
However, the spout deflection phenomenon keeps disturbing the normal spouting, in which spout channel deflects to the annulus region in the lateral direction, rather than keeping vertically upward. This phenomenon is easily triggered by hydrodynamics disturbance and small operating parameter changes, and thus has been reported in both spouted bed and spout fluidised bed. The spout deflection phenomenon is usually regarded as a challenge to the normal spouting because it destroys the symmetrical particle circulation movement and narrows the operation range for stable spouting. Thus, many studies have been conducted aiming to eliminate the spout deflection and enlarge the operating range for stable operation. However, these interventions on the gas-solid hydrodynamics bring new problems and are not satisfying. This can be owed to the lack of fundamental understanding of the spout deflection. On the other side, no valid proof shows that vertical spouting is necessary for high gas-solid mixing efficiency in spouted beds. The gas-solid mixing performance under the spout deflection flow pattern remains unknown. Thus, either the elimination or application of spout deflection all requires an in-depth understanding of the spouted deflection phenomenon.
This thesis focuses on the understanding of the spout deflection phenomenon in terms of its quantification, mechanism, and possible application by means of the Computer Fluid Dynamic-Discrete Element Method (CFD-DEM) coupling approach as well as PIV experiments. It is noted that the rectangular spout fluidised bed, as a kind of spouted bed, is widely used in this study because the spout deflection phenomenon in it is most obvious.
In the past, the understanding of the spout deflection remains on qualitative description because there lacks a tool to study this mesoscale spout movement. Therefore, the first task of this thesis is to develop a tool to quantify the spout deflection movement. Therefore, spout deflection angle was defined to characterise the spout horizontal movement in the reactor. At first, the CFD-DEM simulation was conducted to obtain detailed information of both the solid and gas phase in the spout deflection process. As the spout channel often shows the polyline shape, and the bottom part (termed as the original spout) is chosen as the indicator of the spout deflection. It is noted that the spout channel has two biggest features, high porosity and large gas velocity. But the former feature is sensitive to background gas. Therefore, the angles of the gas velocity vectors in the original spout core are averaged to represent the spout deflection degree. The amplitude, regularity, and speed of the alternating spout deflection movement are quantified for the first time. The spout deflection angle makes it possible to compare the intensity of alternating spout deflection under different operating conditions.
Then, the spout deflection angle is applied to systematically investigate how the operating conditions, including static bed height (H0), background velocity (Ubg), and reactor column width (W), affect the alternating spout deflection in a spout fluidised bed. Some interesting results were found. For example, there is an upper limit for the deflection amplitude because of the self-locking phenomenon in the bottom corner. Large W promotes the regularity of the alternating spout deflection while the effects of H0 on its regularity in the investigated range is neglectable. In addition, the main frequency of the alternating spout deflection reduces with increasing H0 and increasing Ubg, and finally stabilises at around 1.7 Hz in the investigated cases.
It is noted that the previous hydrodynamics study about spout deflection is all based on the simulation method. To further validates the findings about the spout deflection, experimental work is conducted. In this experiment, a flat-bottomed rectangular spouted bed is built. The spout deflection angle definition is extended to the experimental data obtained by Particle Image Velocimetry (PIV). The results show that the amplitude of the alternating spout deflection is stable with increasing spouting velocity, which can be owed to the self-locking phenomenon. The regularity of the alternating spout deflection keeps good at first and then becomes chaotic with increasing H0. These findings are similar to the previous result in numerical simulation.
After that, the mechanism of the spout deflection is explored in order to control or eliminate the spout deflection better. In this part, two hypotheses are proposed for the spout deflection mechanisms in spout fluidised beds, and two virtual experiments are designed to verify the corresponding hypotheses using the CFD-DEM. To verify Hypothesis I that the unsymmetrical particle fountaining, including particle spurting and particle falling, is one main reason for the spout deflection, the virtual experiment I is designed where the particles are manually collected from the spurting surface and then symmetrically released and fall to the free space above the two annulus regions. Three symmetrical release cases are compared, and the results confirm that the spout deflection disappears when the symmetrical particle fountaining is deployed. Moreover, two univariate tests are conducted to check the role of an unsymmetrical particle falling and unsymmetrical particle spurting, respectively, confirming the Hypothesis I again. To verify Hypothesis II that the rheological properties of the annulus region also play a key role in determining whether the spout deflection happens, virtual experiment II is designed where the rheological properties of the annulus region are changed in terms of particle restitution coefficient, particle friction coefficient, background gas velocity, and particle diameter. The simulation results show that spout deflection tends to happen when the annulus regions have better rheological properties (i.e., larger particle restitution coefficient, smaller particle friction coefficient, larger particle diameter, and larger background velocity of the gas phase). This work unveils the mechanism of spout deflection in spout fluidised beds.
The last work is to explore the application possibility of the alternating spout deflection, that is, to get the alternating spout deflection under control and get it independent on the operating conditions. Thus, an umbrella-like baffle is added on the route of the spout channel to obtain the spout deflection. As a result, the spout channel is divided into two parts by the baffle, the first spout and the second spout, respectively. Both first spout and second spout keep deflecting alternatively with time, and their deflection is always in opposite directions. This dynamic balance of the two spouts enables the multistage spouting higher robustness. The efficiency of gas-solid interactions also improves because of two reasons. The first one is that the horizontal spout movement enhance the gas-solid momentum transformation, and the second reason is that the four extra particle circulations occur on both sides of the first spout and second spout due to the obstruction of the baffle. In addition, the control of the alternating spout deflection below and above the baffle can be achieved by adjustments of baffle opening angle. In summary, the alternating spout deflection shows significant advantages in terms of robustness and gas-solid interaction efficiency and is of fascinating engineering and industrial implications.