A robust multiaxial constitutive law is needed to predict stresses within parts formed from sheet metal to be able to compensate for such phenomena as elastic springback. Ideally, the constitutive law would not need to be "trained" using empirical mechanical property data alone, but would be able to compensate for changes in crystallographic texture of the incoming sheet.
To this end, this project seeks to develop constitutive laws based on the initial crystallographic texture and uniaxial stress-strain data, predicting the evolution of the yield surface in multiaxial tensile space. Crystal plasticity modeling treats the material as an assemblage of interacting single crystals with an orientation distribution recreated from the measured texture. As plastic yield evolves, the crystals slip according to the Schmid factor of each system, and each grain rotates to maintain boundary compatibility. The constitutive behavior is then calculated as an average mechanical response of the aggregate.