Characterization of Random Composites Using Moving-Window Technique

Many critical damage phenomena in composite materials, including cracking, are related to local stresses that are linked to local variations in material properties associated with composite microstructure. Relatively little research, however, has been done on the effects of randomness in microstructural configuration on the material behavior of composites. For many engineering applications, it is assumed that smallscale fluctuations in material properties are averaged when evaluating macroscopic behavior. This assumption is generally valid when evaluating quantities such as displacements, average strain, or even average stress, but this approach does not provide any information about local stresses. The analysis of local stresses requires a method of characterizing material properties in terms of material microstructure. This characterization is made more difficult by the inherent randomness in composite microstructure. A methodology is presented whereby the micromechanics model known as the generalized method of cells is used in combination with a moving-window technique to produce material property fields, for elastic and inelastic material properties, associated with the random microstructure of a composite material. In this work, it is assumed that the properties of each constituent of the composite are deterministic and that the fields are the result of randomness in microstructural configuration. Subsequent statistical and probabilistic analysis of these fields will result in a probabilistic description of each property. In this work, the moving-window methodology is applied to a numerically generated micrograph and the real micrograph of a matrix-infiltrated fiber tow.