The Signficance of Hyperelasticity for Dynamic Fracture

Presented by:Professor Markus J. BuehlerMaterials Science, California Institute of TechnologyAbstract:One of the most fundamental questions in materials science and engineering is the understanding of behavior, deformation and failure of materials under extreme conditions and under small geometric confinement. We employ large-scale atomistic modeling using massively parallelized computing resources to investigate materials behavior from a fundamental perspective. Together with theoretical concepts of continuum mechanics theories of deformation, we apply atomistic simulations to solve timely research problems the area of deformation and fracture of different types of materials. The main focus of this talk is on deformation and failure of brittle materials, and in particular on the role of hyperelasticity, the elasticity of large strains. We demonstrate by large-scale molecular dynamics (MD) simulations that linear elastic theories are incapable of capturing all phenomena, and that hyperelasticity is crucial to form a complete picture of dynamic fracture. For example, we show that cracks can move faster than the speed of sound (Buehler et al., Nature, 2003). This phenomenon is in clear contradiction to classical theories. This phenomenon of supersonic cracking, first predicted by our modeling, was recently verified by experiment. We explain this finding by our newly discovered characteristic energy length scale c that describes the region of energy transport near a crack tip. The characteristic energy length scale c is a new theoretical concept that helps to explain outstanding questions in dynamic fracture. We further demonstrate that hyperelasticity governs the instability dynamics of cracks, and report a new theory that incorporates the characteristic energy length scale c in explaining experimental and computational results of the instability dynamics of cracks.Our results allow, for the first time, linkage of classical linear elastic theories of crack tip instabilities as proposed by Eshelby and Yoffe with purely hyperelastic concepts within a new theoretical framework of the generalized instability model and explain long-standing questions of the onset dynamics of crack instabilities. More recently, we have extended our models to include the effect of complex chemistry during crack motion. Based on hybrid models encompassing regions of reactive potentials (ReaxFF) embedded in non-reactive regions, we model crack propagation in silicon, nickel and aluminum, including oxidative processes due to an environment of O2 molecules in the vicinity of the crack surfaces. Our models demonstrate the competing mechanisms of formation of oxide layers and crack growth. These calculations are carried out within our newly developed Computational Materials Design Facility (CMDF), a new multi-paradigm multi-scale simulation framework enabling seamless integration of quantum mechanical scales with macroscopic theories. Refreshments will be served at 2:30p.m.

Location: Vivian Hall of Engineering (VHE) - 217

Audiences: Everyone Is Invited

Contact: Petra Pearce