Continuation and Bifurcation of Grain Boundaries in the Swift-Hohenberg Equation

TL;DR

This paper analyzes the existence, bifurcations, and energetics of grain boundaries in the Swift-Hohenberg equation, revealing symmetry-breaking phenomena and classifying various boundary types through analytical and numerical methods.

Contribution

It introduces a novel far-field-core decomposition technique for studying grain boundaries and provides new insights into their bifurcations and energy preferences.

Findings

Symmetric grain boundaries select marginally zig-zag stable stripes.

A parity-breaking pitchfork bifurcation occurs as stripe angle decreases.

Different grain boundary types are energetically favored depending on the angle.

Abstract

We study grain boundaries between striped phases in the prototypical Swift-Hohenberg equation. We propose an analytical and numerical far-field-core decomposition that allows us to study existence and bifurcations of grain boundaries analytically and numerically using continuation techniques. This decomposition overcomes problems with computing grain boundaries in a large doubly periodic box with phase conditions. Using the spatially conserved quantities of the time-independent Swift-Hohenberg equation, we show that symmetric grain boundaries must select the marginally zig-zag stable stripes. We find that as the angle between the stripes is decreased, the symmetric grain boundary undergoes a parity-breaking pitchfork bifurcation where dislocations at the grain boundary split into disclination pairs. A plethora of asymmetric grain boundaries (with different angles of the far-field stripes either side of the boundary) is found and investigated. The energy of the grain boundaries is then mapped out. We find that when the angle between the stripes is greater than a critical angle, the symmetric grain boundary is energetically preferred while when the angle is less than the critical angle, the grain boundaries where stripes on one side are parallel to the interface are energetically preferred. Finally, we propose a classification of grain boundaries that allows us to predict various non-standard asymmetric grain boundaries.