Fracture is largely a microstructure-based phenomenon, but for experimentalists, computational mechanicians, and fleet managers operating at the macroscale, this fact might seem inexplicable, inapplicable, or even inconsequential. The latter response is what the three chapters of this dissertation address. Together, they attempt to dispel the notion that microstructural effects do not translate in any useful way to the structural scale. They also present models which are verified and validated herein to ease this disconnect. These three chapters are individual papers submitted to refereed journals for publication. The paper in the first chapter appears in Engineering Fracture Mechanics (DOI: http://dx.doi.org/10.1016/j.engfracmech.2014.03.010). It generalizes the Park-PaulinoRoesler potential-based cohesive zone model to three-dimensions, a means to model fracture even under a high degree of mode-mixity at both the macro- and micro-scales. The generalization is validated against several material tests at the macroscale: T-Peel, MMB, ECT, and BDWT. Its ability to model intergranular fracture at the microscale is also explored. The paper in the second chapter fills a void in the Digital Twin community- it presents for the first time a straight-forward use case which both clarifies and motivates this new paradigm in fleet management. Specifically, ductile fracture is modeled in a non-standardized specimen which fails along one of two likely crack paths. This crack path ambiguity, the result of grain-size deviations in specimen geometry, underpins the importance of considering as-manufactured component geometry in the design, assessment, and certification of structural systems, a cornerstone of Digital Twin. It also highlights the limitations of a continuum plasticity damage model in resolving accurately this ambiguity particularly close to the bifurcation, on the order of a few grain sizes, and motivates the need to consider crack nucleation at the microscale. The paper in the third chapter demonstrates Digital Twin at the microscale. It details the implementation, verification, and validation of a microstructure-based, Digital Twin framework which accounts for the predominant microcrack nucleation mechanism in the nickel-based superalloy LSHR. Also included is an extensive grain boundary analysis, an investigation that would otherwise be impossible to conduct to any appreciable fidelity without the as-processed, Digital Twin microstructural model.