| Abstract Scope |
Metal deformation depends on structures and mechanisms that span a vast range of spatiotemporal resolutions. Modeling these mechanisms can require atomistic, mesoscale, or continuum mechanical frameworks. In multiscale approaches, the unification of multiple such models often relies on a shared set of physics-based parameters (hierarchical), intermediate idealized boundary conditions (concurrent or hierarchical approaches), and/or a homogenization of statistically varied local configurations into effective properties. These scale transitions are as essential to multiscale modeling as the single-scale constituent models. Often, they escape scrutiny in uncertainty quantification (UQ).
Case studies examine the nature of scale-transition uncertainty, in the top-down and bottom-up calibration of a physics-based crystal plasticity flow rule for bcc Fe, reduced-order modeling of crystal plasticity, and in inverse methods in indentation testing of Al 7075. To address questions raised by these cases, new perspectives are offered on how to make multiscale validation more robust using information-rich observations. |