[Paper Review] The structure of fluctuations in stochastic homogenization

Mitia Duerinckx, Antoine Gloria+1 more|arXiv (Cornell University)|Feb 4, 2016

Advanced Mathematical Modeling in Engineering48 references22 citations

TL;DR

This paper establishes a new theory of fluctuations in stochastic homogenization by introducing the homogenization commutator as the dominant driver of fluctuations in the field, flux, and corrector for elliptic systems with random coefficients. It proves that fluctuations scale with the commutator in a strong probabilistic norm, and that the rescaled commutator converges to Gaussian white noise, enabling optimal error estimates in the representative volume element method.

ABSTRACT

Four quantities are fundamental in homogenization of elliptic systems in divergence form and in its applications: the field and the flux of the solution operator (applied to a general deterministic right-hand side), and the field and the flux of the corrector. Homogenization is the study of the large-scale properties of these objects. In case of random coefficients, these quantities fluctuate and their fluctuations are a priori unrelated. Depending on the law of the coefficient field, and in particular on the decay of its correlations on large scales, these fluctuations may display different scalings and different limiting laws (if any). In this contribution, we identify another crucial intrinsic quantity, motivated by H-convergence, which we refer to as the \\emph{homogenization commutator} and is related to variational quantities first considered by Armstrong and Smart. In the simplified setting of the random conductance model, we show what we believe to be a general principle, namely that the homogenization commutator drives at leading order the fluctuations of each of the four other quantities in a strong norm in probability, which is expressed in form of a suitable two-scale expansion and reveals the \\emph{pathwise structure} of fluctuations in stochastic homogenization. In addition, we show that the (rescaled) homogenization commutator converges in law to a Gaussian white noise, and we analyze to which precision the covariance tensor that characterizes the latter can be extracted from the representative volume element method. This collection of results constitutes a new theory of fluctuations in stochastic homogenization that holds in any dimension and yields optimal rates. Extensions to the (non-symmetric) continuum setting are also discussed, the details of which are postponed to forthcoming works.

Motivation & Objective

• To identify a universal mechanism governing fluctuations in stochastic homogenization of elliptic systems with random coefficients.
• To establish that the homogenization commutator drives the leading-order fluctuations of the solution field, flux, and corrector in a strong probabilistic norm.
• To prove the rescaled homogenization commutator converges in law to a Gaussian white noise and characterize its covariance structure.
• To analyze the precision with which the fluctuation tensor can be approximated via the representative volume element method.
• To extend the theory to the non-symmetric continuum setting, with full details reserved for future work.

Proposed method

• Introduces the homogenization commutator as a variational quantity related to H-convergence and first considered by Armstrong and Smart, derived from the difference between local and global energy densities.
• Uses a two-scale expansion to express the solution gradient as a sum of the homogenized gradient and oscillatory corrector terms, with the commutator quantifying the leading-order fluctuation.
• Applies spectral gap estimates and the Helffer-Sjöstrand representation formula to control moments and covariances of the commutator and its impact on the fluctuation tensor.
• Employs periodized corrector equations and discrete calculus to derive moment bounds and error estimates in the representative volume element method.
• Uses the L2-regularity theory for elliptic equations and the spectral gap inequality to control the variance of the effective conductivity tensor.
• Establishes convergence rates via a combination of stochastic estimates and deterministic two-scale asymptotics, achieving optimal scaling in any dimension.

Experimental results

Research questions

• RQ1 What intrinsic quantity governs the leading-order fluctuations in stochastic homogenization of elliptic systems?
• RQ2 How does the homogenization commutator relate to the fluctuations of the solution field, flux, and corrector in a strong probabilistic norm?
• RQ3 Does the rescaled homogenization commutator converge in law to a Gaussian white noise, and what is the structure of its covariance tensor?
• RQ4 To what extent can the fluctuation tensor be approximated using the representative volume element method, and what is the associated error rate?
• RQ5 How does the theory extend to non-symmetric, continuum elliptic systems?

Key findings

• The homogenization commutator drives the leading-order fluctuations of the solution field, flux, and corrector in the L2 norm, establishing a pathwise two-scale expansion structure.
• The rescaled homogenization commutator converges in law to a Gaussian white noise, with the limiting covariance tensor characterized by a specific integral formula involving the corrector and its dual.
• The fluctuation tensor of the effective conductivity can be approximated via the representative volume element method with an error rate of order $ L^{-d/2} \log^{d/2} L $.
• The variance of the effective conductivity tensor decays at the optimal rate $ N^{-1} $ when averaging over $ N $ independent realizations of the representative volume element.
• The theory achieves optimal convergence rates in any dimension $ d \geq 2 $, with explicit bounds derived from spectral gap and moment estimates.
• The method applies to the random conductance model with i.i.d. coefficients and extends to the non-symmetric continuum setting, though full details are deferred to future work.