Seminar - Raphael Blumenfeld - Systematic derivation of structure-property relations in porous and cellular materials

Wednesday 30 September 2009, 1.10 pm - 1.55 pm
Civil Engineering Lecture Theatre 3

The School of Civil Engineering is pleased to welcome Raphael Blumenfeld, from Imperial College, London and the Cavendish Laboratory, Cambridge, UK.

Abstract:
A method has been developed for morphological characterization of porous media and systematic derivation of macroscopic properties. It is illustrated for structure-permeability relation. The method involves several stages.

1. A novel mathematical description of the disordered porous microstructure, using a locally-defined shape tensor.
2. Analysis of the statistics of the microstructure, based on a recently developed statistical-mechanical approach.
3. Computation of structural characteristics and their distributions as expectation values of the statistics, e.g. the throat size distribution which relates to bulk permeability to fluid flow.
4. Extraction of statistically equivalent networks, whose nodes and edges represent, respectively, the pores and the connections between them.
5. Computation of the permeability to flow in the network.
6. If required, upscaling of the permeability from the network scale to large-scale systems, using an effective medium approach.

The method paves the way to several significant results:
(i) it makes possible quantitative comparisons between morphologies of different materials;
(ii) it enables characterization of any structural property and their distributions through expectation values over a certain statistics;
(iii) it makes it possible to relate the structure to bulk transport properties (e.g. permeability, electrical conductivity, heat transport, heat exchange, radiation heating, and mechanical properties);
(iv) it provides specifications for generating synthetic networks of designed statistical characteristics, alleviating the burden of time and cost of producing real samples, visualizing them in 3d and analysing the visual data;
(v) it makes it possible to predict changes in macroscopic behavior from experimenting with changes in the statistics. This paves the way to better understanding of different properties in different solids.

A key advantage of this approach is its generality - it can be used to compute a range of transport and mechanical properties.