[Paper Review] Generation of fine transition layers and their dynamics for the stochastic Allen--Cahn equation

Matthieu Alfaro, Dimitra C. Antonopoulou+2 more|arXiv (Cornell University)|Dec 10, 2018

Advanced Mathematical Modeling in Engineering39 references4 citations

TL;DR

This paper studies the stochastic Allen-Cahn equation with a small parameter ε and a time-dependent noise term ξ^ε(t) of order O(ε^−γ) with 0 < γ < 1/3, showing that steep transition layers of thickness O(ε) form within time O(ε²|ln ε|), and that their motion converges to a stochastic curvature flow V = (n−1)κ + cẆ_t in the sharp interface limit ε → 0. The solution profile near the interface remains robustly close to a traveling wave despite the noise, confirming the stability of the interface structure under stochastic perturbations.

ABSTRACT

We study an $\ep$-dependent stochastic Allen--Cahn equation with a mild random noise on a bounded domain in $\mathbb{R}^n$, $n\geq 2$. Here $\ep$ is a small positive parameter that represents formally the thickness of the solution interface, while the mild noise $ξ^\ep(t)$ is a smooth random function of $t$ of order $\mathcal O(\ep^{-γ})$ with $0

Motivation & Objective

• To analyze the formation and dynamics of sharp transition layers in the stochastic Allen-Cahn equation with a singular time-dependent noise term.
• To establish the sharp interface limit as ε → 0 for ε-independent initial data, extending prior results that required ε-dependent initial data.
• To prove that the solution profile near the interface remains close to a traveling wave despite the presence of noise, demonstrating robustness of the interfacial structure.
• To derive a stochastic motion law for the interface that matches curvature flow with additive white noise, improving on earlier results with weaker assumptions.

Proposed method

• Use of a mild noise ξ^ε(t) that converges to space-time white noise as ε → 0, with amplitude O(ε^−γ), 0 < γ < 1/3.
• Application of rescaling techniques around the interface to derive a limiting equation in the moving frame, leading to a traveling wave solution U₀(z^(n) − z*).
• Establishment of uniform estimates on the solution profile via energy methods and comparison principles to control deviations from the traveling wave.
• Use of compactness arguments and subsequential convergence to extract a limit solution w(z, τ) satisfying w_τ = Δ_z w + f(w), which is then identified as a shifted traveling wave.
• Proof of interface thickness O(ε) via contradiction and asymptotic analysis, relying on the strict positivity of U₀′.
• Derivation of the stochastic motion law by analyzing the normal velocity of the interface and showing convergence to V = (n−1)κ + cẆ_t in the limit ε → 0.

Experimental results

Research questions

• RQ1 Does a steep transition layer of thickness O(ε) form in the stochastic Allen-Cahn equation within a time of order O(ε²|ln ε|), even with a singular time-dependent noise?
• RQ2 Can the motion of the interface be described by a stochastic curvature flow V = (n−1)κ + cẆ_t in the sharp interface limit ε → 0, for general ε-independent initial data?
• RQ3 Is the solution profile near the interface robustly close to a traveling wave profile despite the presence of noise, provided the noise is temporally correlated and spatially uniform?
• RQ4 How does the thickness of the interfacial layer behave over time, and can it be bounded uniformly in ε as ε → 0?
• RQ5 What is the precise regularity and geometric structure of the interface Γ_t^ε, and can it be expressed as a smooth graph over a reference manifold γ_t^ε?

Key findings

• The interface thickness is rigorously shown to be O(ε), which is the optimal estimate and improves upon earlier O(ε^α) bounds with α < 1.
• The generation of interface occurs within a time of order O(ε²|ln ε|), confirming the rapid formation of sharp layers even under stochastic perturbations.
• The motion of the interface converges to a stochastic curvature flow V = (n−1)κ + cẆ_t as ε → 0, where Ẇ_t is space-time white noise.
• The solution profile near the interface remains uniformly close to a (squeezed) traveling wave profile, indicating robustness against the noise as long as it is spatially uniform.
• The interface Γ_t^ε is shown to be a smooth hypersurface and can be expressed as a graph over a reference manifold γ_t^ε, with uniform control on the normal derivative.
• The results extend and improve upon earlier works by Funaki (1999) and Weber (2010), particularly by removing restrictions on initial data and proving profile robustness.