ADVANCED CRACK TIP FIELD QUANTIFICATION USING DIGITAL IMAGE CORRELATION, MACHINE LEARNING, AND INTEGRAL EVALUATION [Keynote]

We use higher-order Williams coefficients from full-field displacement data obtained by digital image
correlation (DIC) to approximate complex crack tip fields with simpler expressions. The methodology is
based on invariant path integrals and machine-learned crack detection. We demonstrate the framework for
fatigue crack growth experiments of aluminium alloys and compare the results to matching finite element
simulations.
EXTENDED ABSTRACT