This paper compares invasive and non-invasive voidage measurement techniques from five different ... more This paper compares invasive and non-invasive voidage measurement techniques from five different Canadian laboratories for the determination of local voidage profiles in gas-fluidized beds of 104 μm Fluid Cracking Catalyst (FCC) and 332 μm silica sand in a column of 133 mm i.d. A novel modular system and auxiliary components were designed and constructed to facilitate transportation to, and deployment at, different research centers. Profiles of time-averaged voidage were determined by optical fiber probes, dynamic pressure measurements, electrical capacitance tomography, X-ray computed tomography and radioactive particle tracking by researchers from different research groups. These results are compared with average measurements from pressure gradients and overall bed expansion. Results were kept separate from each other in a double-blind method. There is general, but imperfect, agreement among the different methods for average voidage for both the FCC and silica sand. Invasive probes gave similar results and scatter to non-invasive techniques.
Journal of The American Oil Chemists Society, 2011
In order to improve the oxidative stability and cold flow properties of oleic acid or methyl olea... more In order to improve the oxidative stability and cold flow properties of oleic acid or methyl oleate, branch chain isomerization was conducted using a beta zeolite catalyst. Reaction conditions of temperature (200–300 °C), pressure (0.1–3.0 MPa), and co-catalyst (0–2 wt%) were optimized based on branch chain conversion and the cloud point of the ester following the isomerization reaction of oleic acid or methyl oleate. Fourier transform infrared spectroscopy (FTIR) and Gas Chromatograph equipped with Mass Spectrometry (GC/MS) analyses were used to analyze and quantify the isomerization product samples, while the cloud point of each sample was tested. The lowest and therefore, best cloud point measured was −15.2 °C at conditions of 200 °C, 3 MPa, and 2% co-catalyst using methyl oleate as a starting material. The highest branch chain conversion achieved was 50% under conditions of 300 °C, 1.5 MPa and 0% co-catalyst using oleic acid as a starting material. The use of oleic acid and methyl oleate is based on whether it is optimal to carry out the skeletal isomerization before or after the esterification reaction. Performing the isomerization reaction on the ester was preferred over the fatty acid based on the trans isomerization and cloud point results. Reducing the unbranched trans isomers was desirable in obtaining a low cloud point.
The hydrodynamics of fluidized beds of FCC particles in a column of diameter 0.29 m were investig... more The hydrodynamics of fluidized beds of FCC particles in a column of diameter 0.29 m were investigated based on gauge and differential pressure signals, as well as optical voidage probe data for conditions approaching and beyond the onset of the turbulent fluidization flow regime. It is shown that any treatment of the system as a dense phase/dilute phase binary is oversimplified given the broad spectrum of properties and the lack of clear delineation between two separate phases. On the other hand, chaos analysis indicates complex bifractal behaviour, with two separate values of the Hurst exponent corresponding to different scales of motion, while wavelet analysis is successful in again decomposing signals into scales of motion associated with voids and separate particles.
Hydrodynamics of turbulent fluidized beds of different diameters
Powder Technology, 2004
... A lack of standardized terminology and characterization methodology may have contributed to t... more ... A lack of standardized terminology and characterization methodology may have contributed to the controversy regarding ... Two external cyclones in parallel at the top of the column separategas and solids ... a distributor at the bottom of the loop seal controls the solid circulation rate ...
Turbulent #uidization has only been widely recognized as a distinct #ow regime for the past two d... more Turbulent #uidization has only been widely recognized as a distinct #ow regime for the past two decades, even though it is commonly utilized in industrial #uidized-bed reactors due to vigorous gas}solids contacting, favourable bed-to-surface heat transfer, high solids hold-ups (typically 25}35% by volume), and limited axial mixing of gas. Despite its practical importance, turbulent #uidization has received much less attention than the adjacent #ow regimes of bubbling, slugging and fast #uidization, due to the challenges of experimental and theoretical work related to this #ow regime. However, recent years have seen an upsurge in interest in turbulent #uidization. Various methods } pressure #uctuations, visual observations, capacitance signals, optical "bre probes and bed expansion } have been used to determine the transition velocity, usually denoted ;
Chemical reaction engineering is one of the most important courses in any chemical engineering pr... more Chemical reaction engineering is one of the most important courses in any chemical engineering program, dealing with many fundamental concepts that can be applied in a variety of fields related to chemical engineering. Given the broad variety of backgrounds of the students in graduate programs, teaching complex reaction engineering applications is challenging. In this paper, we report on our experience in including a robust and flexible research component in a graduate reaction engineering course. It is suggested that open research questions allow students to become familiar with course content, as well as with other tools required in their graduate programs, thesis writing and later careers. Steps in implementing the research component are presented together with details found to be helpful in designing a successful reaction engineering course. In addition, a representative research project in which a batch reactor was designed for the isomerization of unsaturated fatty acids is presented as a case study.
A multifluid Eulerian computational fluid dynamics (CFD) model with granular flow extension is us... more A multifluid Eulerian computational fluid dynamics (CFD) model with granular flow extension is used to simulate a liquid-solid fluidized bed. The numerical simulations are evaluated qualitatively by experimental data from the literature and quantitatively by comparison with new experimental data. The effects of mesh size, time step and convergence criteria are investigated. Varying the coefficient of restitution did not alter the results significantly. The Gidaspow drag relationship predicted a higher voidage than the Wen and Yu drag law. Two different liquid distributors (uniform and non-uniform) were simulated and compared, but a better representation of the geometry of the distributor plate did not greatly influence the results. Qualitatively, the simulations show trends similar to experimental trends reported by various authors. The predictions are also compared with new experimental results for 1.13 mm glass spheres at a wide variety of superficial liquid velocities (0.0085-0.110 m/s) and two different temperatures (12 and 33 • C) significantly affecting the liquid viscosity. The CFD model predictions are within 5% of the steady-state experimental data and show the correct trend with variation in viscosity. ᭧
The hydrodynamics of fluidized beds strongly influence their operation, but are complicated and c... more The hydrodynamics of fluidized beds strongly influence their operation, but are complicated and chaotic. There are many measurement techniques, but none fully characterizes gas-solid fluidized beds. Acoustic signals from fluidized beds cover a wide frequency spectrum and can be correlated to bed characteristics. Experiments were conducted to study the acoustic signals from ultrasonic transducers mounted on the outer wall of a two-dimensional fluidization column. The acoustic signals were related to bubble behavior in 550 μm glass beads. Simultaneous acoustic and pressure measurements allowed direct comparison of these signals for single bubbles, pairs and chains of bubbles. The envelope of acoustic signals, generated by particle collisions and particle-wall impacts, provided information on the behavior of bubbles. Significant peaks appeared as the top portions of the bubble wakes approached the acoustic sensor. Pressure waves propagated considerably in the horizontal direction, whereas acoustic signals propagated little in the lateral direction, but transmitted forward in the wall in the direction of bubble motion, maintaining the wave profile invariant during transmission. The strong lateral localization of acoustic signals is promising for determining the lateral bubble position in the bed. Acoustic signals provide a potential means of determining such bubble properties as velocity, frequency and volume, with some advantages relative to pressure signals.
The multi-scale properties of gas-uidized bed hydrodynamics are examined using three di erent siz... more The multi-scale properties of gas-uidized bed hydrodynamics are examined using three di erent sized optical probes in a two-dimensional uidized bed of uid cracking catalyst particles. It is shown that the probe scale determines the scale of structures re ected in the signals analyzed using Fourier and wavelet transformation methods. The conventional fast Fourier transform method failed to remove the noise inevitably present in experimental measurements, while a soft thresholding wavelet transform method was successful in minimizing the noise. The level of noise at superÿcial gas velocities between 2.5 and 6:2 cm=s is shown to be comparable, and formed a basis for setting the threshold values used in the soft threshold de-noising method. The probe scale, as well as the analysis method, in uences the results, and thus the interpretation of experimental hydrodynamic data. ?
In computational fluid dynamics modelling of gas–solid two phase flow, drag force is one of the d... more In computational fluid dynamics modelling of gas–solid two phase flow, drag force is one of the dominant mechanisms for interphase momentum transfer. Despite the profusion of drag models, none of the available drag functions gives accurate results in their own original form. In this work the drag correlations of Syamlal and O'Brien (Syamlal and O'Brien, Int. J. Multiphase Flow. 1988; 14(4):473–481), Gidaspow (Gidaspow, Appl. Mech. Rev. 1986; 39:1–23), Wen and Yu (Wen and Yu, Chem. Eng. Prog. Symp. Ser. 1966; 62(2):100–111), Arastoopour et al. (Arastoopour et al., Powder Technol. 1990; 62(2): 163–170), Gibilaro et al. (Gibilaro et al., Chem. Eng. Sci. 1985; 40:1817–1823), Di Felice (Di Felice, Int. J. Multiphase Flow. 1994; 20(1):153–159), Zhang-Reese (Zhang and Reese, Chem. Eng. Sci. 2003; 58(8):1641–1644) and Hill et al. (Hill et al., J. Fluid Mech. 2001; 448:243–278) are reviewed using a multi-fluid model of FLUENT V6.3.26 (FLUENT, 2007. Fluent 6.3 User's Guide, 23.5 Eulerian Model, Fluent, Inc.) software with the resulting hydrodynamics parameters being compared with experimental data. The main contribution of this work is to propose an easy to implement and efficient method for adjustment of Di Felice drag law which is more efficient compared to the one proposed by Syamlal-O'Brien. The new method adopted in this work showed a quantitative improvement compared to the adjusted drag model of Syamlal-O'Brien. Prediction of bed expansion and pressure drop showed excellent agreement with results of experiments conducted in a Plexiglas fluidized bed. A mesh size sensitivity analysis with varied interval spacing showed that mesh interval spacing with 18 times the particle diameter and using higher order discretization methods produces acceptable results.Dans la modélisation par la dynamique des fluides par ordinateur de l'écoulement diphasique gaz-solide, la force de traînée est l'un des mécanismes dominants dans le transfert de quantité de mouvement interphase. Malgré la profusion des modèles de traînée, aucune des fonctions de traînée disponibles ne donnent de résultats précis dans leur forme originale. Dans cet article, les corrélations de traînée de Syamlal and O'Brien (Syamlal and O'Brien, Int. J. Multiphase Flow. 1988; 14(4):473–481), Gidaspow (Gidaspow, Appl. Mech. Rev. 1986; 39:1–23), Wen and Yu (Wen and Yu, Chem. Eng. Prog. Symp. Ser. 1966; 62(2):100–111), Arastoopour et al. (Arastoopour et al., Powder Technol. 1990; 62(2):163–170), Gibilaro et al. (Gibilaro et al., Chem. Eng. Sci. 1985; 40:1817–1823), Di Felice (Di Felice, Int. J. Multiphase Flow. 1994; 20(1):153–159), Zhang-Reese (Zhang and Reese, Chem. Eng. Sci. 2003; 58(8):1641–1644) et Hill et al. (Hill et al., J. Fluid Mech. 2001; 448:243–278) sont examinées à l'aide du modèle multi-fluides du logiciel FLUENT V6.3.26 (FLUENT, 2007. Fluent 6.3 User's Guide, 23.5 Eulerian Model, Fluent, Inc.), les paramètres hydrodynamiques résultants étant comparés aux données expérimentales. La principale contribution de ce travail est de proposer une méthode efficace et facile à mettre en œuvre pour l'ajustement de la loi de traînée de Di Felice qui est plus efficace comparativement à celle proposée par Syamlal-O'Brien. La nouvelle méthode adoptée dans ce travail montre une amélioration quantitative par rapport au modèle de traînée ajusté de Syamlal-O'Brien. La prédiction de l'expansion de lit et de la perte de charge montre un excellent accord avec les résultats des expériences menées dans un lit fluidisé en plexiglass. Une analyse de sensibilité de la taille des mailles avec des mailles de taille variable variés montre qu'une taille de maille égale à 18 fois le diamètre des particules et l'utilisation de méthodes de discrétisation d'ordre supérieur donnent des résultats acceptables.
Uploads
Papers by Naoko Ellis