Using Principal Component Analysis to Distinguish Different Dynamic Phases in Superconducting Vortex Matter

Abstract

Vortices in type-II superconductors driven over random disorder are known to exhibit a remarkable variety of distinct nonequilibrium dynamical phases that arise due to the competition between vortex-vortex interactions, the quenched disorder, and the drive. These include pinned states, elastic flows, plastic or disordered flows, and dynamically reordered moving crystal or moving smectic states. The plastic flow phases can be particularly difficult to characterize since the flows are strongly disordered. Here we perform principal component analysis (PCA) on the positions and velocities of vortex matter moving over random disorder for different disorder strengths and drives. We find that PCA can distinguish the known dynamic phases as well as or better than previous measures based on transport signatures or topological defect densities. In addition, PCA recognizes distinct plastic flow regimes, a slowly changing channel flow and a moving amorphous fluid flow, that do not produce distinct signatures in the standard measurements. Our results suggest that this position and velocity based PCA approach could be used to characterize dynamic phases in a broader class of systems that exhibit depinning and nonequilibrium phase transitions.

Similar papers