While it is well known that Energy-Dispersive X-ray Diffraction (EDXRD) can measure elastic strains along a specific direction—depending on the positioning of the sample, source, and detector—such setups are complex and typically unavailable in standard laboratory environments. A primary motivation for using EDXRD is to obtain elastic strain measurements necessary for regularizing the problem of constitutive modeling.
A more accessible alternative is Angular-Dispersive X-ray Diffraction (XRD), which is commonly available in laboratory settings. In our work, we are combining XRD (to measure elastic strains) with Digital Image Correlation (DIC, to measure deformation gradients) to develop a data-driven approach for learning the constitutive behavior of metals.
In this newly developed XRD-DIC system, specimens are mechanically loaded while both surface elastic strain and deformation gradients are measured. These measurements are then used to infer the material’s constitutive behavior through machine learning and other data-driven techniques.
