Jonathan Chin

Large-scale lattice Boltzmann simulations of complex fluids: advances through the advent of computational Grids.

During the last 2.5 years, the RealityGrid project has allowed us to be one of the few scientific groups involved in the development of computational Grids. Since smoothly working production Grids are not yet available, we have been able to substantially influence the direction of software and Grid deployment within the project. In this paper, we review our results from large-scale three-dimensional lattice Boltzmann simulations performed over the last 2.5 years. We describe how the proactive use of computational steering, and advanced job migration and visualization techniques enabled us to do our scientific work more efficiently. The projects reported on in this paper are studies of complex fluid flows under shear or in porous media, as well as large-scale parameter searches, and studies of the self-organization of liquid cubic mesophases.

LATTICE BOLTZMANN SIMULATION OF THE FLOW OF NON-NEWTONIAN FLUIDS IN POROUS MEDIA

We present a LB study of the flow of single-phase non-Newtonian fluids, using a power law relationship between the effective viscosity and the local shear rate. Channel flow experiments were carried out to measure the velocity profiles. The simulation results are found to be in good agreement with theory. We also report simulations of the flow of non-Newtonian fluids in a 2-D porous medium.

Steering in computational science: mesoscale modelling and simulation

This paper outlines the benefits of computational steering for high performance computing applications. Lattice-Boltzmann mesoscale fluid simulations of binary and ternary amphiphilic fluids in two and three dimensions are used to illustrate the substantial improvements which computational steering offers in terms of resource efficiency and time to discover new physics. We discuss details of our current steering implementations and describe their future outlook with the advent of computational grids.

Lattice Boltzmann simulation of the flow of binary immiscible fluids with different viscosities using the Shan-Chen microscopic interaction model.

We present a lattice Boltzmann study of the flow of a binary fluid where the fluid components have different viscosities. For this purpose, a microscopic interaction model (due to Shan & Chen) is used. The model is validated for Poiseuille flow of layered immiscible binary fluids and the dispersion of a capillary wave. We then study the unstable displacement of a viscous fluid by a less viscous fluid in a two-dimensional channel. Although a finger-like structure was observed in many simulations, it is not clear if this structure was produced due to viscous fingering or due to other effects.

Simulations of amphiphilic fluids using mesoscale lattice-Boltzmann and lattice-gas methods

We compare two recently developed mesoscale models of binary immiscible and ternary amphiphilic fluids. We describe and compare the algorithms in detail and discuss their stability properties. The simulation results for the cases of self-assembly of ternary droplet phases and binary water-amphiphile sponge phases are compared and discussed. Both models require parallel implementation and deployment on large scale parallel computing resources in order to achieve reasonable simulation times for three-dimensional models. The parallelization strategies and performance on two distinct parallel architectures are compared and discussed. Large scale three-dimensional simulation of multiphase fluids requires the extensive use of high performance visualization techniques in order to enable the large quantities of complex data to be interpreted. We report on our experiences with two commercial visualization products: AVS and VTK. We also discuss the application and use of novel computational steering techniques for the more efficient utilization of high performance computing resources. We close the paper with some suggestions for the future development of both models.

Scientific grid computing: the first generation

The scientific users computing demands are becoming increasingly complex and can benefit from distributed resources, but effectively marshalling these distributed systems often introduces new challenges. The authors describe how researchers can exploit existing distributed grid infrastructure to get meaningful scientific results.

Structural and dynamical characterization of Hele-Shaw viscous fingering.

Viscous fingering occurs in the interfacial zone between two fluids confined between two plates with a narrow gap (Hele–Shaw geometry) when a highly viscous fluid is displaced by a fluid with relatively low viscosity. Using a mesoscopic approach—the lattice Boltzmann method—we investigate the dynamics of spatially extended Hele–Shaw flow under conditions corresponding to various experimental systems by tuning the ‘surface tension’ and the reactivity between the two fluids. We discuss the onset of the fingering instability (dispersion relation), analyse the structural properties (characterization of the interface) and the dynamical properties (growth of the mixing zone) of the Hele–Shaw systems, and show the effect of reactive processes on the structure of the interfacial zone.

Towards performance control on the Grid

Advances in computational Grid technologies are enabling the development of simulations of complex biological and physical systems. Such simulations can be assembled from separate components—separately deployable computation units of well-defined functionality. Such an assemblage can represent an application composed of interacting simulations or might comprise multiple instances of a simulation executing together, each running with different simulation parameters. However, such assemblages need the ability to cope with heterogeneous and dynamically changing execution environments, particularly where such changes can affect performance. This paper describes the design and implementation of a prototype performance control system (PerCo), which is capable of monitoring the progress of simulations and redeploying them so as to optimize performance. The ability to control performance by redeployment is demonstrated using an assemblage of lattice Boltzmann simulations running with and without control policies. The cost of using PerCo is evaluated and it is shown that PerCo is able to reduce overall execution time.

Chirality and domain growth in the gyroid mesophase

We describe the first dynamical simulations of domain growth during the self-assembly of the gyroid mesophase from a ternary amphiphilic mixture, using the lattice Boltzmann method. The gyroid is a chiral structure; we demonstrate that, for a symmetric amphiphile with no innate preference for left- or right-handed morphologies, the self-assembly process may give rise to a racemic mixture of domains. We use measurements of the averaged mean curvature to analyse the behaviour of domain walls, and suggest that diffusive domain growth may be present in this system.

Detection and tracking of defects in the gyroid mesophase

Certain systems, such as amphiphile solutions or diblock copolymer melts, may assemble into structures called “mesophases”, with properties intermediate between those of a solid and a liquid. These mesophases can be of very regular structure, but may contain defects and grain boundaries. Different visualization techniques such as volume rendering or isosurfacing of fluid density distributions allow the human eye to detect and track defects in liquid crystals because humans are easily capable of finding imperfections in repetitive spatial structures. However, manual data analysis becomes too time consuming and algorithmic approaches are needed when there are large amounts of data. We present and compare two different approaches we have developed to study defects in gyroid mesophases of amphiphilic ternary fl uids. While the first method is based on a pattern recognition algorithm, the second uses the particular structural properties of gyroid mesophases to detect defects.  2004 Elsevier B.V. All rights reserved.