Inference of phase field fracture models

Abstract

The phase field approach to modeling fracture uses a diffuse damage field to represent a crack. This addresses the singularities that arise at the crack tip in computations with sharp interface models, mollifying some of the difficulties associated with the mathematical and numerical treatment of fracture. The introduction of the diffuse field helps with crack propagation dynamics, enabling phase-field approaches to model all phases of damage from crack initiation to propagation, branching, and merging. Specific formulations, beginning with brittle fracture, have also been shown to converge to classical solutions. Extensions to cover the range of material failure, including ductile and cohesive fracture, leads to an array of possible models. There exists a large body of studies of these models and their consequences for crack evolution. However, there have not been systematic studies into how optimal models may be chosen. Here we take a first step in this direction by developing formal methods for identification of the best parsimonious model of phase field fracture given full-field data on the damage and deformation fields. We consider some of the main models that have been used to model damage, its degradation of elastic response, and its propagation. Our approach builds upon Variational System Identification (VSI), a weak form variant of the Sparse Identification of Nonlinear Dynamics (SINDy). In this first communication we focus on synthetically generated data but we also consider central issues associated with the use of experimental full-field data, such as data sparsity and noise.