Estimating Effective Moduli of Composites Using Quantitative Image Analysis

The statistical data required for quantitative analysis of the physical properties of composite materials may be obtained using digital image processing techniques. This approach allows accurate quantitative determination of volume fraction data, of interface area per unit volume, and of more sophisticated measures of the microgeometry of material mixtures including the n-point spatial correlation functions. The twoand three-point correlation functions can be measured quite accurately both for synthetic materials and for real materials such as porous sandstones. To extract the desired parameters of the microgeometry from the three-point correlation functions, methods of interpolating and integrating these lattice-based empirical values have been developed. Methods of nessing the theory to give estimates of physical properties when the available data are not adequate to permit use of the full theory have also been developed.