The EPSRC-funded RealityGrid e-Science project has simulated and visualised how complex fluids, such as blood, milk, polymers and oil, flow. One outcome of this work could be a better understanding of the factors affecting blood flow and clotting in arteries. Other possible outcomes include a better choice of lubricant for bearings and improved methods for extracting oil from porous rocks. Several simulations are described in a paper published on 15 August in a special Theme Issue of Philosophical Transactions of the Royal Society A on scientific grid computing.
Density and speed are all you need to build a large-scale model that describes how simple fluids, such as water, flow. The flow of complex fluids, however, cannot be described in terms of such large-scale features alone. In theory, it could be derived from the movement of individual molecules, but that would be too big a computational task in practice. So the RealityGrid project, led by Professor Peter Coveney at UCL, has adopted a third way that describes the flow in terms of properties of the fluid that are not too small and not too large, using what is called the lattice-Boltzmann method.
In many complex fluids, molecules form aggregates, perhaps because one end of the molecule is attracted to water and the other to oil. These aggregates influence strongly how the fluid behaves. Professor Coveney and colleagues used the properties of these aggregates to simulate and visualise how a number of different complex fluids flow. These include oil undergoing shear forces inside a bearing and a fluid being forced into a rock that is already saturated with another fluid.
In one particular type of fluid, the aggregates self assemble into a pattern called a gyroid: single layers of a third phase, such as detergent molecules, come between the oil and water phases, creating interesting patterns. Gyroid phases have been known experimentally for about 20 years and some are thought to influence the movement of other molecules across cell membranes. One of RealityGrid’s most significant achievements has been to simulate and visualise how such a gyroid phase develops and changes as the fluid flows.
It takes a lot of computing power to simulate and visualise such processes, watch how they develop and even see how their development responds to a change in conditions, for example temperature. Only a Grid of distributed high performance computers is up to the task. RealityGrid routinely uses high performance computers at the UK’s national supercomputing facilities. In the TeraGyroid project, high performance computers in the US and UK were linked to perform the largest set of simulations of a gyroid ever attempted.
Notes
Please note: links open in a new browser window.
Note 1: ReaityGrid is one of eight major pilot projects funded by EPSRC to explore how new grid technologies can be developed and used in some key scientific applications. By modelling and simulating very large or complex condensed matter structures, it is able to address challenging scientific problems that would otherwise remain out of reach, or even be impossible to study.
http://www.realitygrid.org
Note 2: Large-scale lattice-Boltzmann simulations of complex fluids: advances through the advent of computational grids by J. Harting, J. Chin, M. Venturoli and P. V. Coveney Philosophical Transactions of the Royal Society A 363 1833 (15 August 2005)