Supervisor’s Foreword
In gas–solid fluidised beds, solid particles in a vessel are set in motion by means of
a rising gas stream. Bubbles appear above a certain gas velocity. Such bubbling
fluidised beds are extensively used for the production of chemicals, solid fuel
combustion and gasification, waste conversion, coating, drying and other operations
of practical interest, especially when the process is exothermic and good gas–solid
contact is important. Despite their frequent use in industry, the complexity of their
hydrodynamics complicates fluidised bed scale-up and hampers uniform, reliable
operation. Gas channelling is another common problem.
Inspired by the regularly spaced ripples on beaches and dunes, induced by the
action of waves and wind, as well as the regular patterns that form when shallow
layers of particles are vibrated on a plate, we proposed about 20 years ago to apply
this nature-inspired principle of “dynamic self-organisation” in dissipative systems
to fluidised beds. A periodically varying gas flow, instead of the conventionally
used constant gas flow, was shown to impose a regular structure on the otherwise
chaotic dynamics of fluidised beds. We discovered that periodic inlet flows lead to
not only regular periodic wave patterns in 3D shallow fluidised beds, but also
triangular tessellations of rising bubbles in deep, quasi-2D beds. The latter are
unlike anything observed in vibrated layers: here, the chaotic bubble flows of a
typical fluidised bed (looking like a boiling liquid) turn into an orderly bubble
stream, neatly organised in space and in time. This occurs within a range of frequencies, amplitudes and offsets of the gas flow.
The Ph.D. thesis of Kaiqiao Wu investigates these dynamically structured flows
and how they form. What are the fundamental principles underpinning their formation? Can numerical simulations reproduce the experimental patterns? What are
the required ingredients of successful models? Qualitative exploration of conditions
under which patterns form is followed by rigorous, quantitative analysis using a
new pattern regularity or intensity index. This provides a parametric map for
dynamically self-organised bubble patterns. The thesis then shows that conventional two-fluid computational fluid dynamics (CFD) models, which represent the
solid phase as fluid-like, cannot reproduce these experimental patterns in deep beds.
This is because friction between the particles is too important for pattern
ix
In gas–solid fluidised beds, solid particles in a vessel are set in motion by means of
a rising gas stream. Bubbles appear above a certain gas velocity. Such bubbling
fluidised beds are extensively used for the production of chemicals, solid fuel
combustion and gasification, waste conversion, coating, drying and other operations
of practical interest, especially when the process is exothermic and good gas–solid
contact is important. Despite their frequent use in industry, the complexity of their
hydrodynamics complicates fluidised bed scale-up and hampers uniform, reliable
operation. Gas channelling is another common problem.
Inspired by the regularly spaced ripples on beaches and dunes, induced by the
action of waves and wind, as well as the regular patterns that form when shallow
layers of particles are vibrated on a plate, we proposed about 20 years ago to apply
this nature-inspired principle of “dynamic self-organisation” in dissipative systems
to fluidised beds. A periodically varying gas flow, instead of the conventionally
used constant gas flow, was shown to impose a regular structure on the otherwise
chaotic dynamics of fluidised beds. We discovered that periodic inlet flows lead to
not only regular periodic wave patterns in 3D shallow fluidised beds, but also
triangular tessellations of rising bubbles in deep, quasi-2D beds. The latter are
unlike anything observed in vibrated layers: here, the chaotic bubble flows of a
typical fluidised bed (looking like a boiling liquid) turn into an orderly bubble
stream, neatly organised in space and in time. This occurs within a range of frequencies, amplitudes and offsets of the gas flow.
The Ph.D. thesis of Kaiqiao Wu investigates these dynamically structured flows
and how they form. What are the fundamental principles underpinning their formation? Can numerical simulations reproduce the experimental patterns? What are
the required ingredients of successful models? Qualitative exploration of conditions
under which patterns form is followed by rigorous, quantitative analysis using a
new pattern regularity or intensity index. This provides a parametric map for
dynamically self-organised bubble patterns. The thesis then shows that conventional two-fluid computational fluid dynamics (CFD) models, which represent the
solid phase as fluid-like, cannot reproduce these experimental patterns in deep beds.
This is because friction between the particles is too important for pattern
ix