Laura Michel | Searcharxiv

arXiv2026

Laddering of a knitted fabric: a topology-induced failure

Laddering is the propagation of a topological defect in an everyday-life material: weft knitted fabrics, following a broken thread or a dropped stitch. What is a minor frustration when damaging a pair of tights is a more serious issue for industrial-scale production, but might inspire new solutions to limit and mitigate damage to architected materials. In this work, laddering is investigated in a pre-stressed model knit through experiments and Discrete Element Rod simulations. The control parameter is the initial tension applied on the fabric. A force threshold due to the stitch's natural curvature is evidenced. It controls both the propagation onset and arrest, as tension is relaxed by the thread length freed by ladder growth, and enables damage prediction at moderate tension. Furthermore, we uncovered that the laddering velocity is of the order of the velocity of bending waves and exhibits an unexpected linear scaling with the fabric tension, that arises from a complex combination of elastic and friction forces. Finally, we discuss the implications of our results from the perspective of damage control and mitigation.

cond-mat.soft↗

arXiv2022

Dynamically writing coupled memories using a reinforcement learning agent, meeting physical bounds

Traditional memory writing operations proceed one bit at a time, where e.g. an individual magnetic domain is force-flipped by a localized external field. One way to increase material storage capacity would be to write several bits at a time in the bulk of the material. However, the manipulation of bits is commonly done through quasi-static operations. While simple to model, this method is known to reduce memory capacity. In this paper, we demonstrate how a reinforcement learning agent can exploit the dynamical response of a simple multi-bit mechanical system to restore its memory to full capacity. To do so, we introduce a model framework consisting of a chain of bi-stable springs, which is manipulated on one end by the external action of the agent. We show that the agent manages to learn how to reach all available states for three springs, even though some states are not reachable through adiabatic manipulation, and that both the training speed and convergence within physical parameter space are improved using transfer learning techniques. Interestingly, the agent also points to an optimal design of the system in terms of writing time. In fact, it appears to learn how to take advantage of the underlying physics: the control time exhibits a non-monotonic dependence on the internal dissipation, reaching a minimum at a cross-over shown to verify a mechanically motivated scaling relation.

cond-mat.soft↗