A Hybrid Experimental-Computational Approach for the Analysis of Dynamic Fracture

Abstract

The fracture behavior of materials resulting from rapidly applied loads remains poorly understood, and requires consideration of the inertial forces and rapidly changing crack front. At the same time, with the advent of improved high-speed cameras and full-field optical techniques such as digital image correlation (DIC), the spatial and temporal resolution of measurable field data has increased. As such, this thesis presents a novel hybrid experimental-computational technique that extracts dynamic fracture criterion from DIC data using improved, iterative elastodynamic expressions relating stress intensity factors (SIFs) to displacement fields around the crack front. Two variations of the solution are explored: an over-deterministic least squares regression is used when the crack tip location is known, such as at initiation of fracture, and an iterative Newton-Raphson optimization is used when the crack tip location is unknown. Dynamic fracture experiments are conducted on single-edge-notch specimens, impacted on the opposite face by a striker in mode I (or crack opening), to determine the efficacy of the method. Three materials with varying microstructures are studied: brittle polymer polymethyl methacrylate (PMMA), nanolayered metal-ceramic MAX phase Ti3SiC2, and human femoral cortical bone, a natural hierarchical ceramic matrix composite. PMMA is a historically well-studied material in classical dynamic fracture mechanics and is chosen as a model material, and when combined with data from literature is used to conduct validation and uncertainty quantification arising from key experimental and numerical parameters. Material rate dependency and the effect of transient wave interactions are investigated in the dynamic fracture behavior of MAX phase Ti3SiC2, and found to hold little influence due to the energy absorbing mechanisms of kink banding and delamination unique to the layered microstructure. Mode-mixity arising from an orthotropic microstructure is investigated in human femoral cortical bone, and lead to significant mode II SIFs even in the absence of any applied mode II loading, though mode I fracture remains dominant. These collective results demonstrate the wide variety of material microstructures with varying deformation mechanisms to which the hybrid experimental-computational dynamic fracture analysis is well-suited to handle.%%%%Ph.D., Mechanical Engineering and Mechanics – Drexel University, 2017