Parameter Estimation for Phase-field Crack Propagation Simulation Using Nonsequential Data Assimilation

Reliable fracture simulations are expected to contribute to the longer life cycle of structures. The phase-field fracture (PFF) model is anticipated to be a robust approach to reproducing crack propagation and branching. Measurement techniques for detecting cracks and fractures have also advanced and are widely used for health monitoring. In the field of structural analysis, a data assimilation method using the full-field measurement data obtained by digital image correlation has been proposed. The data assimilation method is attracting attention as a method of improving the simulation accuracy by utilizing the time-series measurement data obtained from such measurement techniques. In this study, we apply for the first time the nonsequential data assimilation method minimizing the cost function using treestructured Parasen estimator (DMC-TPE method) to assimilate full-field strain measurement data into PFF simulations. In this paper we validate the proposed data assimilation method through numerical experiments, where data assimilation is performed using the pseudo-measurement data obtained from PFF simulations conducted with parameters assumed as true values. The results demonstrate that the proposed method enables the simultaneous inverse estimation of multiple parameters included in the PFF model. The validation results revealed that, although the DMC-TPE depends on the initially estimated values, the parameters that significantly impact the simulation results were accurately estimated.