Data-driven optical method for full-field stress measurements

We demonstrate an optical method to directly determine full-field stress by combining a data-driven method with digital image correlation (DIC) strain data. In order to obtain the full-field stress distribution, the full-field strain measured by 2D-DIC and an isotropic ideal elastoplastic material stress-strain dataset are used. The proposed data-driven full-field stress measurement method determines the full-field stress components by performing iterative calculations using the dataset and a distance-functional-based data-driven algorithm without assuming any constitutive equations. The proposed method is implemented in an experiment-in-loop simulation and a practical application. The stress fields of uniaxial tensile and compact tensile processes are determined by the developed algorithm. The results show that the proposed data-driven method can obtain accurate full-field stress using an already established material stress-strain dataset. Furthermore, the proposed method satisfies the equilibrium and compatibility constraints and therefore allows the correction of erroneous strain results from DIC calculations, resulting in more accurately calculated full-field strain and stress values.