This thesis describes the development of a Lattice Boltzmann (LB) model for a binary gas mixture. Specifically, channel flow driven by a density gradient with diffusion slip occurring at the wall is studied in depth. The first part of this thesis sets the foundation for the multi-component model used in the subsequent chapters. Commonly used single component LB methods use a non-physical equation of state, in which the relationship between pressure and density varies according to the scaling used. This is fundamentally unsuitable for extension to multi-component systems containing gases of differing molecular masses that are modelled with the ideal gas equation of state. Also, existing methods for implementing boundary conditions are unsuitable for extending to novel boundary conditions, such as diffusion slip. Therefore, a new single component LB derivation and a new method for implementing boundary conditions are developed, and validated against Poiseuille flow. However, including a physical equation of state reduces stability and time accuracy, leading to longer computational times, compared with 'incompressible' LB methods. The new method of analysing LB boundary conditions successfully explains observations from other commonly used schemes, such as the slip velocity associated with 'bounce-back'.The new model developed for multi-component gases avoids the pitfalls of some other LB models, a single computational grid is shared by all the species and the diffusivity is independent of the viscosity. The Navier-Stokes equation for the mixture and the Stefan-Maxwell diffusion equation are both recovered by the model. However, the species momentum equations are not recovered correctly and this can lead to instability. Diffusion slip, the non-zero velocity of a gas mixture at a wall parallel to a concentration gradient, is successfully modelled and validated against a simple one-dimensional model for channel flow. To increase the accuracy of the scheme a second order numerical implementation is needed. This can be achieved using a variable transformation method which does not result in an increase in computational time. Simulations were carried out on hydrogen and water diffusion through a narrow channel, with varying total pressure and concentration gradients. For a given value of the species mass flux ratio, the total pressure gradient was dependent on the species concentration gradients. These results may be applicable to fuel cells where the species mass flux ratio is determined by a chemical reaction and the species have opposing velocities. In this case the total pressure gradient is low and the cross-channel average mass flux of hydrogen is independent of the channel width. Finally, solutions for a binary Stefan tube problem were investigated, in which the boundary at one end of a channel is permeable to hydrogen but not water. The water has no total mass flux along the channel but circulates due to the slip velocity at the wall. The cross-channel average mass flux of the hydrogen along the channel increases with larger channel widths. A fuel cell using a mixture of gases, one being inert, will experience similar circulation phenomena and, importantly, the width of the pores will affect performance. This thesis essentially proves the viability of LB models to simulate multi-component gases with diffusion slip boundaries, and identifies the many areas in which improvements could be made.

The use of accelerator technology, particularly Graphics Processing Units (GPUs), for scientific computing has increased greatly over the last decade. While this technology allows larger and more complicated problems to be solved faster than before it also presents another opportunity: the real-time and interactive solution of problems. This work aims to investigate the progress that GPU technology has made towards allowing fluid-structure interaction (FSI) problems to be solved in real-time, and to facilitate user interaction with such a solver. A mesoscopic scale fluid flow solver is implemented on third generation nVidia ‘Kepler’ GPUs in two and three dimensions, and its performance studied and compared with existing literature. Following careful optimisation the solvers are found to be at least as efficient as existing work, reaching peak efficiencies of 93% compared with theoretical values. These solvers are then coupled with a novel immersed boundary method, allowing boundaries defined at arbitrary coordinates to interact with the structured fluid domain through a set of singular forces. The limiting factor of the performance of this method is found to be the integration of forces and velocities over the fluid and boundaries; the arbitrary location of boundary markers makes the memory accesses during these integrations largely random, leading to poor utilisation of the available memory bandwidth. In sample cases, the efficiency of the method is found to be as low as 2.7%, although in most scenarios this inefficiency is masked by the fact that the time taken to evolve the fluid flow dominates the overall execution time of the solver. Finally, techniques to visualise the fluid flow in-situ are implemented, and used to allow user interaction with the solvers. Initially this is achieved via keyboard and mouse to control the fluid properties and create boundaries within the fluid, and later by using an image based depth sensor to import real world geometry into the fluid. The work concludes that, for 2D problems, real-time interactive FSI solvers can be implemented on a single laptop-based GPU. In 3D the memory (both size and bandwidth) of the GPU limits the solver to relatively simple cases. Recommendations for future work to allow larger and more complicated test cases to be solved in real-time are then made to complete the work.

A novel modeling approach is applied to karst hydrology. Long-standing problems in karst hydrology and solute transport are addressed using Lattice Boltzmann methods (LBMs). These methods contrast with other modeling approaches that have been applied to karst hydrology. The motivation of this dissertation is to develop new computational models for solving ground water hydraulics and transport problems in karst aquifers, which are widespread around the globe. This research tests the viability of the LBM as a robust alternative numerical technique for solving large-scale hydrological problems. The LB models applied in this research are briefly reviewed and there is a discussion of implementation issues. The dissertation focuses on testing the LB models. The LBM is tested for two different types of inlet boundary conditions for solute transport in finite and effectively semi-infinite domains. The LBM solutions are verified against analytical solutions. Zero-diffusion transport and Taylor dispersion in slits are also simulated and compared against analytical solutions. These results demonstrate the LBM’s flexibility as a solute transport solver. The LBM is applied to simulate solute transport and fluid flow in porous media traversed by larger conduits. A LBM-based macroscopic flow solver (Darcy’s law-based) is linked with an anisotropic dispersion solver. Spatial breakthrough curves in one and two dimensions are fitted against the available analytical solutions. This provides a steady flow model with capabilities routinely found in ground water flow and transport models (e.g., the combination of MODFLOW and MT3D). However the new LBM-based model retains the ability to solve inertial flows that are characteristic of karst aquifer conduits. Transient flows in a confined aquifer are solved using two different LBM approaches. The analogy between Fick’s second law (diffusion equation) and the transient ground water flow equation is used to solve the transient head distribution. An altered-velocity flow solver with source/sink term is applied to simulate a drawdown curve. Hydraulic parameters like transmissivity and storage coefficient are linked with LB parameters. These capabilities complete the LBM’s effective treatment of the types of processes that are simulated by standard ground water models. The LB model is verified against field data for drawdown in a confined aquifer.

The Pleistocene carbonate rock Biscayne Aquifer of south Florida contains laterally-extensive bioturbated ooltic zones characterized by interconnected touching-vug megapores that channelize most flow and make the aquifer extremely permeable. Standard petrophysical laboratory techniques may not be capable of accurately measuring such high permeabilities. Instead, innovative procedures that can measure high permeabilities were applied. These fragile rocks cannot easily be cored or cut to shapes convenient for conducting permeability measurements. For the laboratory measurement, a 3D epoxy-resin printed rock core was produced from computed tomography data obtained from an outcrop sample. Permeability measurements were conducted using a viscous fluid to permit easily observable head gradients (~2 cm over 1 m) simultaneously with low Reynolds number flow. For a second permeability measurement, Lattice Boltzmann Method flow simulations were computed on the 3D core renderings. Agreement between the two estimates indicates an accurate permeability was obtained that can be applied to future studies.

No description available.

This dissertation presents the development of numerical models based on lattice Boltzmann (LB) and cellular automaton (CA) methods for solving phase change and microstructural evolution problems. First, a new variation of the LB method is discussed for solving the heat conduction problem with phase change. In contrast to previous explicit algorithms, the latent heat source term is treated implicitly in the energy equation, avoiding iteration steps and improving the formulation stability and efficiency. The results showed that the model can deal with phase change problems more accurately and efficiently than explicit LB models. Furthermore, a new numerical technique is introduced for simulating dendrite growth in three dimensions. The LB method is used to calculate the transport phenomena and the CA is employed to capture the solid/liquid interface. It is assumed that the dendritic growth is driven by the difference between the local actual and local equilibrium composition of the liquid in the interface. The evolution of a threedimensional (3D) dendrite is discussed. In addition, the effect of undercooling and degree of anisotropy on the kinetics of dendrite growth is studied. Moreover, effect of melt convection on dendritic solidification is investigated using 3D simulations. It is shown that convection can change the kinetics of growth by affecting the solute distribution around the dendrite. The growth features of twodimensional (2D) and 3D dendrites are compared. Furthermore, the change in growth kinetics and morphology of Al-Cu dendrites is studied by altering melt undercooling, alloy composition and inlet flow velocity. The local-type nature of LB and CA methods enables efficient scaling of the model in petaflops supercomputers, allowing the simulation of large domains in 3D. The model capabilities with large scale simulations of dendritic solidification are discussed and the parallel performance of the algorithm is assessed. Excellent strong scaling up to thousands of computing cores is obtained across the nodes of a computer cluster, along with near-perfect weak scaling. Considering the advantages offered by the presented model, it can be used as a new tool for simulating 3D dendritic solidification under convection.

Yes / Traditional Lattice-Boltzmann modelling of advection–diffusion flow is affected by numerical instability if the advective term becomes dominant over the diffusive (i.e., high-Péclet flow). To overcome the problem, two 3D one-way coupled models are proposed. In a traditional model, a Lattice-Boltzmann Navier–Stokes solver is coupled to a Lattice-Boltzmann advection–diffusion model. In a novel model, the Lattice-Boltzmann Navier–Stokes solver is coupled to an explicit finite-difference algorithm for advection–diffusion. The finite-difference algorithm also includes a novel approach to mitigate the numerical diffusivity connected with the upwind differentiation scheme.

Millions of years of evolution have led to a wealth of highly adapted functional surfaces in nature. Among the most fascinating are superhydrophobic surfaces which are highly water-repellent and shed drops very easily owing to their chemical hydrophobicity combined with micropatterning. Superhydrophobic materials have attracted a lot of attention due to their practical applications as ultra-low friction surfaces for ships and pipes, water harvesters, de-humidifiers and cooling systems. At small length scales, where surface tension dominates over gravity, these surfaces show a wealth of phenomena interesting to physicists, such as directional flow, rolling, and drop bouncing. This thesis focuses on two examples of dynamic drop interactions with micropatterned surfaces and studies them by means of a lattice Boltzmann simulation approach. Inspired by recent experiments, we investigate the phenomenon of the self-propelled bouncing of coalescing droplets. On highly hydrophobic patterned surfaces drop coalescence can lead to an out-of-plane jump of the composite drop. We discuss the importance of energy dissipation to the jumping process and identify an anisotropy of the jumping ability with respect to surface features. We show that Gibbs' pinning is the source of this anisotropy and explain how it leads to the inhibition of coalescence-induced jumping. The second example we study is the novel phenomenon of pancake bouncing. Conventionally, a drop falling onto a superhydrophobic surface spreads due to its inertia, retracts due to its surface tension, and bounces off the surface. Here we explain a different pathway to bouncing that has been observed in recent experiments: A drop may spread upon impact, but leave the surface whilst still in an elongated shape. This new behaviour, which occurs transiently for certain impact and surface parameters, is due to reversible liquid imbibition into the superhydrophobic substrate. We develop a theoretical model and test it on data from experiments and simulations. The theoretical model is used to explain pancake bouncing in detail.

This thesis deals with the lattice-Boltzmann method (LBM) in combination with other methods to solve thermal flow problems. The three primary, current approaches for thermal lattice-Boltzmann method (TLBM) will be introduced. The three approaches are the multispeed approach by McNamara and Alder , the passive scalar approach by Shan, and the thermal distribution model proposed by He et al. Shi et al. simplified the thermal distribution model for incompressible thermal flows. The model proposed by Shi et al. was simulated and compared to a hybrid LBM and energy equation model proposed by Khiabani et al. The thermal lattice-Boltzmann method will be compared to the temperature fields generated by the energy equation of the hybrid method. To determine which method is better suited from computer simulations the two will be compared for computational demands, and the speed of both convergence and computation.

Particle suspensions are common both in nature and in various technological applications. The complex nature of hydrodynamic interactions between particles and the solvent makes such analysis difficult that often requires numerical modeling to understand the behavior of particle suspensions. In this dissertation, we employ a hybrid computational model that integrates a lattice spring model for solid mechanics and a lattice Boltzmann model for fluid dynamics. We use this model to study several practical problems in which the dynamics of spherical and spheroidal particles and deformable capsules in dilute suspensions plays an important role. The results of our studies yield new information regarding the dynamics of solid particle in pressure-driven channel flows and disclose the nonlinear effects associated with fluid inertia leading to particle cross-stream migration. This information not only give us a fundamental insight into the dynamics of dilute suspensions, but also yield engineering guidelines for designing high throughput microfluidic devices for sorting and separation of synthetic particles and biological cells. We first demonstrate that spherical particles can be size-separated in ridged microchannels. Specifically, particles with different sizes follow distinct trajectories as a result of the nonlinear inertial effects and secondary flows created by diagonal ridges in the channel. Then, separation of biological cells by their differential stiffness is studied and compared with experimental results. Cells with different stiffness squeeze through narrow gaps between solid diagonal ridges and channel wall, and migrate across the microchannel with different rates depending on their stiffness. This deformability-based microfluidic platform may be valuable for separating diseased cells from healthy cells, as a variety of cell pathologies manifest through the change in mechanical cell stiffness. Finally, the dynamics of spheroid particles in simple shear and Poiseuille flows are studied. Stable rotational motion, cross-stream migration, and equilibrium trajectories of non-spherical particles in flow are investigated. Effects of particle and fluid inertia on dynamics of particles are disclosed. The dependence of equilibrium trajectory on particle shape reveals a potential application for shape based particle separation.