Postdoc Estelle Berthier has been working to understand how disordered materials fail, using structures created with a variety of connectivities via a laser-cutter. In collaboration with Silke Henkes (U. Bristol) and Jen Schwarz (Syracuse) and former postdoc Jonathan Kollmer (U. Duisburg-Essen), we have observed that a rigid-cluster analysis reveals the origins of a brittle-ductile transition. In the image shown above, low-connectivity materials (left) have few, small rigid clusters and fail in a ductile matter, while high-connectivity materials (right) have large rigid clusters and fail in a brittle manner.
We have also made progress, in collaboration with Mason Porter (UCLA), on understanding how to forecast the locations of failure, which are statistically more-likely to be located on edges which score high on a particular network measure called edge betweenness centrality. This quantity measures how commonly each edge is used to connect shortest-path distances across the network. This work has been highlighted in several articles and press releases:
- Tracey Peake, NC State Press Release
- Paolo Moretti and Michael Zaiser, “Network analysis predicts failure of materials and structures” PNAS
- Abigail Klopper, “Network failure forecast” Nature Physics