FAST INFERENCE OF CRACK TIP POSITION AND STRESS INTENSITY FACTORS FROM DISPLACEMENT DATA

Fracture prognosis and characterization efforts require knowledge of crack tip position and the configurational driving force acting on the crack. Here, we present an efficient numerical approach to determine these characteristics under a consistent theoretical framework from displacement data. The novel approach utilizes the separable characteristics of the asymptotic linear elastic fracture mechanics model to expedite the search for crack tip position and is particularly useful for noisy displacement data.

The importance of accurately locating crack tip position is assessed when quantifying the crack driving force from observed displacements. The proposed separability approach for quickly inferring crack tip position is introduced, setting the stage for subsequent assessment of the utility of the separability approach. Comparing to the widely-used displacement correlation approach, we examine performance in cases involving bad starting guesses, noise, and non-conformance with the asymptotic linear elastic fracture mechanics model, e.g. inelastic material behavior and finite geometries. We envision our proposed separability method and the associated code that has been made freely available to be of use to those doing experiments (involving digital image correlation) and simulations where the crack tip position is not explicitly defined, e.g. finite elements with damage models and atomistic simulations of crack growth.
EXTENDED ABSTRACT