Applications of a novel many-component Lattice Boltzmann simulation

There are many important systems which involve the flow of dense suspensions of deformable particles; blood is an example. However, the modeling of such materials represents a considerable challenge to established methods such as computational fluid dynamics (CFD). In this PhD project, work has been undertaken to extend the lattice Boltzmann method (LBM), and enable it to represent a large number of mutually immiscible, deformable, droplets (separated by narrow interfaces). The new method imposes a relatively small computational overhead and has been validated against experimental observations. The work has also led to very promising developments in the simulation of micro-fluidic systems, allowing much quicker simulations than traditional CFD methods.

The principal target application of this project is the mesoscale modeling of blood flow, where the typical length is about 10 red blood cells' diameters. We address in here the identified gap in models capable of modeling efficiently, explicitly, many deformable bodies within a surrounding incompressible fluid. We generalised, improved, and extended an existing LBM model for binary fluids. Our N>>2 noncoalescing fluids (droplets) are defined to represent the different deformable particles of the suspension. Their interactions with the walls as well as their deform ability are controlled by local fluid-wall wetting and fluid-fluid surface tensions methods, which we have also been developed and validated.

All interfacial methods suffer from small but spurious flows that disturb the solution. Our model is, unfortunately, no exception when used in a high surface tension and low Reynold's number regime. We describe several steps taken to address this problem which yield a significant reduction in these 'micro-currents' and important improvements in stability and flow field noise reduction. This enabled our model to access successfully the computationally non-trivial problems of binary fluid microfluidics.

Using our model, we also recover the expected behaviour of deformable and solid particle suspensions with respect to experimental observations on flow of solid and deformable spheres in pressure-driven straight pipe flow. In order to serve as calibration, we measured the macroscopic effect of the droplets' effective deformability against their microscopic properties (surface tension, internal viscosity). This new model opens very promising and unique grounds of research, by the new capability its offers.