The failure of materials and interfaces is mediated by cracks, nearly singular dissipative structures that propagate at velocities approaching the speed of sound. Crack initiation and subsequent propagation -- the dynamic process of fracture -- couples a wide range of time and length scales. Crack dynamics challenge our understanding of the fundamental physics processes that take place in the extreme conditions within the nearly singular region where material failure occurs. Here, we first briefly review the classic approach to dynamic fracture, "Linear Elastic Fracture Mechanics" (LEFM), and discuss its successes and limitations. We show how, on the one hand, recent experiments performed on straight cracks propagating in soft brittle materials have quantitatively confirmed the predictions of this theory to an unprecedented degree. On the other hand, these experiments show how LEFM breaks down as the singular region at the tip of a crack is approached. This breakdown naturally leads to a new theoretical framework coined "Weakly Nonlinear Fracture Mechanics", where weak elastic nonlinearities are incorporated. The stronger singularity predicted by this theory gives rise to a new and intrinsic length scale, $\ell_{nl}$. These predictions are verified in detail through direct measurements. We then theoretically and experimentally review how the emergence of $\ell_{nl}$ is linked to a new equation for crack motion, which predicts the existence of a high-speed oscillatory crack instability whose wave-length is determined by $\ell_{nl}$. We conclude by delineating outstanding challenges in the field.
Bridging perturbation and variational approaches in brittle fracture
Navier, Ecole des Ponts, Univ Gustave Eiffel, CNRS（纳维尔实验室，巴黎国立桥梁学院，古斯塔夫·埃菲尔大学，法国国家科学研究中心）；Computational Mechanics of Building Materials, Institute for Building Materials, ETH Zürich（建筑材料研究所，苏黎世联邦理工学院）；Department of Microsystems Engineering (IMTEK), University of Freiburg（弗莱堡大学微系统工程系）
