10
1 Introduction
Applications of the KTGF to study granular flows and fluidised beds at lab
scales have had tremendous success, especially in predicting bubble properties. For
example, Hulme et al. [48] simulated a 2D bed of Geldart B glass beads fluidised
at 2U mf , and demonstrated that the computed flows agreed quantitatively with the
experimental measurements, such as axial profiles of the mean bubble size and the
bubble distribution, measured by X-ray fluoroscopy. Moreover, many comprehensive comparisons, including size distributions, aspect ratio, number, rise velocity of
bubbles, also showed reasonable agreement [16, 67]. Circulation patterns and expansion of solids in bubbling beds of Geldart A and B particles have been also captured
with TFM approaches [67]. Owing to its relatively low computational cost, TFM
is especially favoured in tackling large-scale applications. Li et al. [65] provided
comparisons between 2D and 3D TFM simulations of three differently configured
CFB risers with varying geometries and powder feeding. They observed significant
quantitative differences in terms of flow properties, but acceptable agreement in the
axial profiles of pressure gradient and solids fraction. Nonetheless, the validations
extended to properties of the emulsion phase give rise to more concerning differences. Hernández-Jiménez et al. [46] conducted particle image velocimetry (PIV)
and digital image analysis (DIA) in a 5 mm thick quasi-2D fluidised bed at 1.75U mf .
They observed out that the TFM model overestimates the rising velocity of particles
by nearly one order of magnitude, despite a good agreement observed in several
bubble properties, such as size and rising velocity. Subsequently, researchers recognised the critical role of the correlations for solid frictional stress in simulating flow
behaviour in a dense granular flow system, such as solid circulation and bubbling
[38, 86].
It is obvious that numerical models can successfully predict apparent characteristics of flow patterns with a decent degree of agreement, such as bubble size, rising
velocity, solid flux, and function as reliable tools for engineering design. However,
due to the complexity and chaotic nature of multiphase flows, accurate spatiotemporal prediction of the dynamics is hardly achieved. To overcome these disadvantages
induced by complex flow patterns, it is desirable to seek the assistance of an additional actuator to shape the chaotic hydrodynamics into a determined form that is
predictable.
1.4 Methods for Structuring Bed Hydrodynamics
One possibility of gaining control over the system hydrodynamics is to “structure” the
gas-solid flows in a fluidised bed by imposing additional degrees of freedom. If one
suppresses, or at least controls, the effects of hydrodynamic instability, it is possible
to decouple conflicting design objectives, such as promoting the solid mixing through
a vigorous bubble flow, but also reducing the size of bubbles to maximise interphase
contact time. In the best scenario, gas-solid suspensions can be fully structured to
simplify scale-up practice, enhance interphase heat and mass transport and, therefore,
manipulate contact time and conversion rate. A decade ago, van Ommen et al. [112]
1 Introduction
Applications of the KTGF to study granular flows and fluidised beds at lab
scales have had tremendous success, especially in predicting bubble properties. For
example, Hulme et al. [48] simulated a 2D bed of Geldart B glass beads fluidised
at 2U mf , and demonstrated that the computed flows agreed quantitatively with the
experimental measurements, such as axial profiles of the mean bubble size and the
bubble distribution, measured by X-ray fluoroscopy. Moreover, many comprehensive comparisons, including size distributions, aspect ratio, number, rise velocity of
bubbles, also showed reasonable agreement [16, 67]. Circulation patterns and expansion of solids in bubbling beds of Geldart A and B particles have been also captured
with TFM approaches [67]. Owing to its relatively low computational cost, TFM
is especially favoured in tackling large-scale applications. Li et al. [65] provided
comparisons between 2D and 3D TFM simulations of three differently configured
CFB risers with varying geometries and powder feeding. They observed significant
quantitative differences in terms of flow properties, but acceptable agreement in the
axial profiles of pressure gradient and solids fraction. Nonetheless, the validations
extended to properties of the emulsion phase give rise to more concerning differences. Hernández-Jiménez et al. [46] conducted particle image velocimetry (PIV)
and digital image analysis (DIA) in a 5 mm thick quasi-2D fluidised bed at 1.75U mf .
They observed out that the TFM model overestimates the rising velocity of particles
by nearly one order of magnitude, despite a good agreement observed in several
bubble properties, such as size and rising velocity. Subsequently, researchers recognised the critical role of the correlations for solid frictional stress in simulating flow
behaviour in a dense granular flow system, such as solid circulation and bubbling
[38, 86].
It is obvious that numerical models can successfully predict apparent characteristics of flow patterns with a decent degree of agreement, such as bubble size, rising
velocity, solid flux, and function as reliable tools for engineering design. However,
due to the complexity and chaotic nature of multiphase flows, accurate spatiotemporal prediction of the dynamics is hardly achieved. To overcome these disadvantages
induced by complex flow patterns, it is desirable to seek the assistance of an additional actuator to shape the chaotic hydrodynamics into a determined form that is
predictable.
1.4 Methods for Structuring Bed Hydrodynamics
One possibility of gaining control over the system hydrodynamics is to “structure” the
gas-solid flows in a fluidised bed by imposing additional degrees of freedom. If one
suppresses, or at least controls, the effects of hydrodynamic instability, it is possible
to decouple conflicting design objectives, such as promoting the solid mixing through
a vigorous bubble flow, but also reducing the size of bubbles to maximise interphase
contact time. In the best scenario, gas-solid suspensions can be fully structured to
simplify scale-up practice, enhance interphase heat and mass transport and, therefore,
manipulate contact time and conversion rate. A decade ago, van Ommen et al. [112]