A new efficient and simple algorithm based on two-point probability functions was developed to construct isotropic and anisotropic heterogeneous mi- crostructures with desired effective physical property. To take into account the complex heterogeneities geometry, the developed microstructure design tool uses the strong-contrast formulations of the statistical continuum the- ory to compute the effective properties. This approach uses n-point correla- tion functions to take into account the details of the microstructure. The mi- crostructure construction starts by randomly distributing unit-cells that repre- sent heterogeneities within the matrix. These initial unit-cells are represented by one pixel each that grows to become multi-pixel heterogeneities using an optimization algorithm. The construction is achieved by minimizing an ob- jective function (OF) defined as the difference between the effective desired property and the one calculated at each iteration.

The presentation could be downloaded here.