Thursday, October 3, 2019 - 3:45pm
How can we predict when and where a material will fail? For disordered solids, including many food and cosmetic products, screens and cases for smartphones, and even mud and gravel perched on a hillside, this fundamental question remains a challenge. I will discuss how our group and others are extending standard techniques from solid state and statistical physics to address this question. Specifically, we use theory and computation to quantify phonon-like (and non-phonon-like) low frequency vibrational modes in solids, and find that quasi-localized low-frequency modes are surprisingly good predictors of future deformation and failure. In order to really understand how “good” a predictor is, we are also engaged in a large collaborative effort that compares many different recently identified methods for predicting failure, and I will highlight some preliminary results from that effort. I will also discuss a related project that compares the nonlinear response of hard granular materials to granular materials with a softer interaction potential. We find a beautiful finite-size scaling collapse that, quite surprisingly, suggests the hard sphere limit is singular, which means that well-understood hard spheres may not be a good starting point for
predicting the behavior of real materials.
September 23, 2019 - 4:43pm