Publication: Experimental study of a bubbling fluidized bed with a rotating distributor
Loading...
Identifiers
Publication date
2008-06
Defense date
2008-06-06
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Esta tesis consiste en la caracterización experimental de la hidrodinámica de un nuevo
lecho fluido de distribuidor rotatorio.
Existen numerosas referencias en la literatura en las que se analizan los factores que influyen
en la calidad de la fluidización en lechos burbujeantes, como son la tasa de mezcla,
el tamaño de burbuja y la heterogeneidad en el lecho. Entre estos factores se encuentran la
geometría del lecho, el caudal de gas empleado en la fluidización y el tipo de distribuidor.
Las heterogeneidades que se producen con frecuencia en lechos industriales han hecho que
numerosos investigadores hayan incorporado modificaciones de distinta índole sobre los
lechos convencionales, como por ejemplo alterar el sistema de suministro de aire o usar
diseños innovadores. El nuevo diseño de distribuidor que se estudia en esta tesis intenta,
mediante la introducción del giro del distribuidor, lograr mayores tasas de mezcla del gas y aumentar la dispersión de las partículas, al tiempo que se consigue una fluidización más
uniforme. La posibilidad de controlar la velocidad de giro del distribuidor permite operar
en un amplio rango de condiciones de operación sin perder la calidad de fluidización.
Los experimentos fueron realizados en un lecho constituido por un cilindro transparente
de diámetro 192 mm y altura 0.8 m lleno de partículas de arena del tipo B de acuerdo con
la clasificación de Geldart. El distribuidor rotatorio es una placa perforada acoplada en su
eje al eje de un motor eléctrico. La velocidad de giro se controla mediante un inversor de
frecuencia que permite trabajar con un rango de velocidades que en los experimentos se
varía de 0 a 100 rpm. La descripción completa de la instalación experimental se encuentra
en el Capítulo 2.
La caracterización experimental de la hidrodinámica del lecho realizada en la tesis incluye, por un lado, la descripción global del lecho sin giro y con giro en el distribuidor, mediante medidas absolutas de presión (Capítulo 4) y por otro, el estudio de las propiedades de
las burbujas que se forman en el lecho usando sondas ópticas específicamente diseñadas
y construidas para esta tesis (Capítulo 6).
Con el fin de interpretar de manera adecuada las señales de presión que se utilizan para la
caracterización del lecho, el Capítulo 3 contiene un estudio de los valores de la desviación
estándar de las señales de presión en lechos fluidos burbujeantes. Se ha obtenido una
función semi-empírica, que depende de la velocidad del gas, que permite predecir dichas
fluctuaciones de presión en lechos con partículas del tipo B. Este modelo permite explicar
las diferencias en las medidas cuando se emplean sensores de presión en modo diferencial o
absoluto, obteniéndose una buena correspondencia entre los valores teóricos y las medidas
experimentales para diferentes tamaños de lechos, posición de los sensores y propiedades
de las partículas.
En el Capítulo 4 se estudia el efecto del giro del distribuidor en el comportamiento
hidrodinámico global del lecho, analizándose el cambio en la mínima velocidad de fluidización y en las fluctuaciones de presión. Se ha observado una disminución en el valor de la mínima velocidad de fluidización a medida que aumenta la velocidad de giro. Además se analizaron los espectros de frecuencia y la desviación estándar de las fluctuaciones de presión. Las medidas se repitieron a distintas alturas iniciales del lecho para ver como afectaba esta altura a la magnitud del efecto provocado por el giro. Se ha comprobado que la rotación del distribuidor permite fluidizar lechos con poca altura, que en ausencia de giro presentan una estructura de chorros y no se consiguen fluidizar. Por otro lado, conforme aumenta la altura inicial del lecho, el efecto de la rotación sobre la velocidad de mínima fluidización tiende a disminuir. Se demuestra por tanto que mediante el ajuste de la velocidad de giro en el distribuidor, se puede cambiar la velocidad del aire necesario para fluidizar el lecho, lo que permite mantener unas condiciones uniformes de fluidización en un rango mayor de caudales.
Una vez realizado el análisis global, se estudiaron las características locales del lecho. Para ello, se usaron sensores de presión diferencial y sondas ópticas con las que se midieron las cuerdas de las burbujas que se forman en el lecho y su velocidad. Los resultados obtenidos usando las dos sondas se encuentran en el Capítulo 5. Las funciones de densidad de probabilidad de la cuerda y de la velocidad se calcularon aplicando el Método de la Máxima
Entropía. Existe un tamaño mínimo de cuerda que es posible medir usando sondas intrusivas
para que el error sea tolerable. Este límite inferior se ha tenido en cuenta en la formulación de las ecuaciones para la obtención de las funciones de distribución de probabilidad.
La función de densidad de probabilidad de los diámetros se ha deducido a partir
de las medidas experimentales de las cuerdas, aplicando herramientas estadísticas. Los
resultados de las sondas de presión y de las sondas ópticas son bastante parecidos, aunque
las sondas ópticas proporcionan información más local, y pueden utilizarse en posiciones
muy próximas al distribuidor. Se ha comprobado que el método de la Máxima Entropía
es un método simple que ofrece varias ventajas frente a otros métodos aplicados hasta la
fecha para la obtención de las distribuciones de tamaño en lechos fluidos: no es necesario
suponer a priori la forma de la distribución, el número de muestras requeridas es menor
que en otros métodos y se evita la transformación inversa, que es un cálculo complejo.
Una vez desarrollado y validado el método de transformación de cuerdas en diámetros se
estudió el efecto del giro del distribuidor en el tamaño de las burbujas y su frecuencia
de paso a distintas posiciones en el lecho. Los resultados se presentan en el capítulo 6.
Primero se ha obtenido un modelo simple para analizar la influencia de la aceleración
centrifuga que actúa sobre la burbuja en el momento en que se desprende del distribuidor
una vez formada. Este análisis indica que el giro hace que el diámetro inicial de la burbuja
sea menor que si el distribuidor estuviera parado. Los resultados experimentales
muestran que, a igualdad en el exceso de aire, el tamaño de las burbujas es menor cuando
el distribuidor gira. Los tamaños medidos de burbuja en distintas posiciones radiales
confirman la tendencia puesta de manifiesto por el modelo: para el distribuidor rotatorio
el diámetro medio disminuye a distancias mayores del eje del lecho, donde la aceleración
centrífuga es mayor. La rotación del distribuidor también hace que la distribución de
burbujas en la sección radial del lecho sea más homogénea. Además, para el distribuidor
rotatorio se observa que el aumento del tamaño de las burbujas a medida que aumenta la
altura es menos acusado que con ausencia de giro. Esto puede deberse a una disminución
de la coalescencia lograda por la ruptura de los caminos preferenciales de ascensión de las
burbujas gracias al giro.
_________________________________________________
This thesis presents the experimental fluid dynamic characterization of a new fluidized bed with a rotating distributor. Many works in the literature analyze the factors that influence the quality of fluidization in bubbling beds, e.g. the rate of solids mixing, the size of the bubbles and the extent of heterogeneity in the bed. These factors include among other, the bed geometry, the gas flow rate and the type of gas distributor. The non heterogenous structures often found in industrial fluidization processes have led many investigators to modify the conventional fluidized bed devices, alter the air supply system or try innovative designs to avoid these heterogeneities. The novel distributor design studied in this thesis tries to solve some of these difficulties; the aim of the distributor rotation being to overcome low radial gas mixing and particle dispersion, and to achieve a more uniform fluidization. The possibility to control and adjust the rotational speed of the distributor plate offers a wide range of operating conditions while maintaining the quality of fluidization. The fluidized bed is a transparent cylinder with 192 mm ID and a height of 0.8 m filled with Geldart B silica particles. The distributor is a perforated plate that is coupled to the shaft of an AC electric motor. It can rotates around the bed axis and the rotational speed can be varied using a frequency inverter. In the experiments this speed was varied between 0 and 100 rpm. A complete description of the experimental set-up can be found in Chapter 2. The experimental fluid dynamic characterization presented in this thesis includes a global description of the bed behavior with and without rotation of the distributor, using pressure measurements (Chapter 4). In addition, the differences in the characteristics of the generated bubbles are studied by means of in house made optical probes (Chapter 6). To better understand the pressure signal recorded for the bed characterization, the behavior of the standard deviation of pressure fluctuations in fluidized beds for group B particles in the bubbling regime is studied in Chapter 3. An empirical-theoretical function, which depends on the gas velocity, is proposed for predicting the pressure signal fluctuations. The differences in the standard deviation of pressure fluctuations obtained from absolute or differential sensors are analyzed and compared to experimental values corresponding to different bed sizes, pressure probe positions and particle properties. In chapter 4 the effect of the rotational speed of the distributor plate on the global hydrodynamic behavior of the bed is studied. Minimum fluidization velocity and pressure fluctuations were first analyzed. A decrease in the minimum fluidization velocity is observed when the rotational speed increases. The standard deviation and the power spectra of the pressure fluctuations are also discussed. Measurements with several initial static bed heights were taken in order to analyze the influence of the initial bed mass inventory over the effect of the distributor rotation on the bed hydrodynamics. The rotation of the distributor allows to fluidize very shallow beds which had a jet structure when the static distributor is used; the effect of the rotation becomes globally less important for deeper beds. These characteristics show that adjusting the rotational speed it is possible to change the gas velocity needed to fluidize the bed, facilitating the fluidization and maintaining a uniform fluidization. Once the global fluid dynamic behavior is known, local characteristics are analyzed. Differential pressure and optical probe measurements were carried out in order to obtain the size and velocity of the bubbles rising in the bed. Results obtained from the two types of probes were compared in Chapter 5. The probability distributions of bubble pierced length and velocity were obtained applying the Maximum Entropy Method. The minimum bubble pierced length that it is possible to measure using intrusive probes, due to their finite size, has been introduced as a constraint in the derivation of the size distribution equations. The probability density function of bubble diameter was inferred applying statistical tools to the pierced length experimental data. Results on bubble size obtained from pressure and optical probes have been found to be very similar, although optical probes provide more local measurements and can be used even at very low heights in the bed, near the distributor. The Maximum Entropy Method has been found to be a simple method that offers many advantages over other methods applied before for size distribution modeling in fluidized beds: the distribution shape does not have to be pre-established, the number of samples required is lower than in other methods and the backward transformation procedure is avoided. The effect of the distributor rotation on the bubble size, bubble passage frequency and bubbles distribution at different radial and axial positions in the bed was studied with the optical probes and the results are presented in Chapter 6. A simple theoretical expression is obtained in order to analyze the centrifugal acceleration influence on the bubble when it detaches from the distributor. This analysis points out that the centrifugal acceleration imparted by the distributor rotation causes the decrease of the initial bubble radius. The experimental results show that the bubbles are smaller when the rotating distributor is used if the excess gas for the static and rotating configuration is similar. The bubble size radial profile indicates that when the distributor rotates, the diameter of the bubbles close to the bed walls is smaller due to the effect of a higher centrifugal acceleration. The distributor rotation also promotes the more homogenous distribution of the bubbles over the bed surface. At higher axial positions even smaller bubbles are found for the rotating case. This may be due to a lower coalescence rate of the bubbles when the distributor rotates as the rotation may break the channeling in the bed.
This thesis presents the experimental fluid dynamic characterization of a new fluidized bed with a rotating distributor. Many works in the literature analyze the factors that influence the quality of fluidization in bubbling beds, e.g. the rate of solids mixing, the size of the bubbles and the extent of heterogeneity in the bed. These factors include among other, the bed geometry, the gas flow rate and the type of gas distributor. The non heterogenous structures often found in industrial fluidization processes have led many investigators to modify the conventional fluidized bed devices, alter the air supply system or try innovative designs to avoid these heterogeneities. The novel distributor design studied in this thesis tries to solve some of these difficulties; the aim of the distributor rotation being to overcome low radial gas mixing and particle dispersion, and to achieve a more uniform fluidization. The possibility to control and adjust the rotational speed of the distributor plate offers a wide range of operating conditions while maintaining the quality of fluidization. The fluidized bed is a transparent cylinder with 192 mm ID and a height of 0.8 m filled with Geldart B silica particles. The distributor is a perforated plate that is coupled to the shaft of an AC electric motor. It can rotates around the bed axis and the rotational speed can be varied using a frequency inverter. In the experiments this speed was varied between 0 and 100 rpm. A complete description of the experimental set-up can be found in Chapter 2. The experimental fluid dynamic characterization presented in this thesis includes a global description of the bed behavior with and without rotation of the distributor, using pressure measurements (Chapter 4). In addition, the differences in the characteristics of the generated bubbles are studied by means of in house made optical probes (Chapter 6). To better understand the pressure signal recorded for the bed characterization, the behavior of the standard deviation of pressure fluctuations in fluidized beds for group B particles in the bubbling regime is studied in Chapter 3. An empirical-theoretical function, which depends on the gas velocity, is proposed for predicting the pressure signal fluctuations. The differences in the standard deviation of pressure fluctuations obtained from absolute or differential sensors are analyzed and compared to experimental values corresponding to different bed sizes, pressure probe positions and particle properties. In chapter 4 the effect of the rotational speed of the distributor plate on the global hydrodynamic behavior of the bed is studied. Minimum fluidization velocity and pressure fluctuations were first analyzed. A decrease in the minimum fluidization velocity is observed when the rotational speed increases. The standard deviation and the power spectra of the pressure fluctuations are also discussed. Measurements with several initial static bed heights were taken in order to analyze the influence of the initial bed mass inventory over the effect of the distributor rotation on the bed hydrodynamics. The rotation of the distributor allows to fluidize very shallow beds which had a jet structure when the static distributor is used; the effect of the rotation becomes globally less important for deeper beds. These characteristics show that adjusting the rotational speed it is possible to change the gas velocity needed to fluidize the bed, facilitating the fluidization and maintaining a uniform fluidization. Once the global fluid dynamic behavior is known, local characteristics are analyzed. Differential pressure and optical probe measurements were carried out in order to obtain the size and velocity of the bubbles rising in the bed. Results obtained from the two types of probes were compared in Chapter 5. The probability distributions of bubble pierced length and velocity were obtained applying the Maximum Entropy Method. The minimum bubble pierced length that it is possible to measure using intrusive probes, due to their finite size, has been introduced as a constraint in the derivation of the size distribution equations. The probability density function of bubble diameter was inferred applying statistical tools to the pierced length experimental data. Results on bubble size obtained from pressure and optical probes have been found to be very similar, although optical probes provide more local measurements and can be used even at very low heights in the bed, near the distributor. The Maximum Entropy Method has been found to be a simple method that offers many advantages over other methods applied before for size distribution modeling in fluidized beds: the distribution shape does not have to be pre-established, the number of samples required is lower than in other methods and the backward transformation procedure is avoided. The effect of the distributor rotation on the bubble size, bubble passage frequency and bubbles distribution at different radial and axial positions in the bed was studied with the optical probes and the results are presented in Chapter 6. A simple theoretical expression is obtained in order to analyze the centrifugal acceleration influence on the bubble when it detaches from the distributor. This analysis points out that the centrifugal acceleration imparted by the distributor rotation causes the decrease of the initial bubble radius. The experimental results show that the bubbles are smaller when the rotating distributor is used if the excess gas for the static and rotating configuration is similar. The bubble size radial profile indicates that when the distributor rotates, the diameter of the bubbles close to the bed walls is smaller due to the effect of a higher centrifugal acceleration. The distributor rotation also promotes the more homogenous distribution of the bubbles over the bed surface. At higher axial positions even smaller bubbles are found for the rotating case. This may be due to a lower coalescence rate of the bubbles when the distributor rotates as the rotation may break the channeling in the bed.