Loss of localization at high crack speeds

Abstract

At the IUTAM Symposium on “High Velocity Deformation of Solids” in Tokyo 1977, I argued that localization of the process region at a crack edge to a size determined by an intrinsic length parameter of the material was lost at a very high crack speeds [1]. A few years later several experimental results appeared, giving indirect support for such a loss by showing lack of a unique relation between stress intensity factor or surface roughness and crack velocity [2],[3],[4],[5]. Some of the experiments also showed a remarkable preference for constant crack velocity in the high velocity region. These observations were later parallelled by numerical simulations by Johnson [6],[7]. The constant velocity reached did not seem to be a material property. Here it is argued that the preference for a constant velocity is in accordance with a perception of the process region as a continuum at very high crack velocities, a consequence of the lost significance of an intrinsic length parameter.