Parameter identification of a phase-field fracture model using integrated digital image correlation

Phase-field fracture (PFF) modeling is a popular approach to model and simulate fracture processes in solids. Accurate material parameters and boundary conditions are of utmost importance to ensure a good prediction quality of numerical simulations. In this work, an Integrated Digital Image Correlation (IDIC) algorithm is proposed to calibrate boundary conditions, Poisson’s ratio, fracture energy and internal length, all at once, by using the phase-field model itself and images of a deforming sample. The presented approach is applied to virtual experiments mimicking a single edge notched shear test and implemented in the opensource deal.II-based software pfm-cracks and the digital image correlation library Correli 3.2. The reliability of the results is investigated for different levels of acquisition noise, thereby demonstrating high robustness and accuracy for a wide range of noise levels. The conditioning of the problem is analyzed via sensitivity fields for all parameters and the eigendecomposition of the Hessian matrix used in the IDIC algorithm.