Inverse modeling of inelastic properties of a two-phase microstructure

The prediction of material properties with the inclusion of morphology has been an area of increasing interest for material scientists in the past decades. A myriad of statistical continuum mechanics formulations have been developed to investigate the properties of a two-phase microstructure given its morphology. In this study, the structure-property model is inverted to create an inverse microstructure model for a two-phase Ti64 material to predict the microstructure required to achieve a desired property. For this purpose, an inverse formulation is developed using the two-point correlation function representation of the microstructure within the statistical continuum framework. Using this formulation the initial microstructure is then predicted by knowing a desired strength. This approach calculates the optimum values of the two-point probability functions which are associated with the minimum error in the predicted strength with respect to the desired strength. Finally, 2D microstructures are reconstructed using the predicted values of the two-point probability functions to represent the morphology of the initial microstructure at four different temperatures of Ti64 (850, 900, 950, 1000 degrees C).