Corrector estimates and homogenization errors of unsteady flow ruled by Darcy's law

Li Wang, Qiang Xu, Zhifei ZhangarXiv 2022 · DOI · arXiv

Abstract

Focusing on Darcy's law incorporating memory effects, this paper studies non-stationary Stokes equations on perforated domains. We establish a sharp homogenization error for both velocity and pressure in terms of the energy norm. The main challenge lies in gauging the boundary layers induced by the incompressibility condition. To address this, we construct boundary-layer correctors using Bogovskii's operator. Also, the present work provides detailed regularity estimates for these correctors, where a significant difficulty arises from the incompatibility between initial and boundary values. The methodologies developed herein hold great potential for tackling the same issue in other evolutionary models beyond a homogenization setting.

Introduction

When fluid flows through a porous medium, such as water through a sponge or oil through sandstone, the underlying geometry is incredibly complex—a "perforated domain" filled with microscopic obstacles. Mathematically, this is described at the microscopic level using the Stokes equations. However, because simulating every tiny pore is computationally prohibitive, researchers use homogenization to derive a simpler macroscopic model: Darcy’s law.

For unsteady (time-dependent) flows, the macroscopic limit is "Darcy’s law with memory," where the current flow depends on the history of the pressure gradient. While qualitative convergence for these models has been established, obtaining sharp quantitative error estimates remained a major challenge. Specifically, researchers sought to determine how fast the microscopic solution approaches the macroscopic one as the pore size ($\epsilon$) shrinks to zero.

The Challenge of Boundary Layers

The primary hurdle in obtaining sharp estimates is the "boundary layer"—the chaotic fluid behavior near the edges of the domain. Standard mathematical techniques used to handle these layers often violate the incompressibility (divergence-free) condition of fluids. This makes it notoriously difficult to obtain sharp error estimates for both velocity and pressure simultaneously.

Methodology: Bogovskii’s Operator and Refined Expansions

To address these challenges, the authors employ a multi-step analytical approach centered on a refined asymptotic expansion. The core of their method involves constructing boundary-layer correctors that respect the physical constraints of the fluid.

Breakdown of the asymptotic expansion for the velocity field, highlighting the zero-order, first-order, and boundary layer terms used to define the remainder $w_\epsilon$.

Key components of the methodology include:

  1. Asymptotic Expansion: A proposed expansion for velocity that includes an oscillating part (the main flow), a boundary-layer part (to fix edge effects), and a small remainder.
  2. Bogovskii’s Operator: This is the critical innovation. To ensure the boundary-layer correctors remain incompressible, the authors use Bogovskii’s operator to "fix" boundary mismatches without violating the divergence-free condition.
  3. Radial Cut-off Functions: These are designed to respect the geometry of the boundary, allowing for superior control over energy norms compared to standard cut-offs.
  4. Handling Incompatibility: The authors developed regularity estimates for cases where the initial fluid state does not perfectly match boundary conditions at $t=0$. By utilizing time-weighted estimates ($t^\alpha$), they successfully managed the resulting singularities.

Key Insights

  • The Memory Effect: In unsteady flow, permeability is not a constant but a tensor that evolves over time. The paper confirms that the "memory" of the fluid is a fundamental component of the macroscopic limit.
  • Incompressibility as a Bottleneck: Previous suboptimal estimates were largely due to the difficulty of maintaining a divergence-free velocity field while zeroing out boundary errors. The structured redistribution of fluid divergence via Bogovskii’s operator resolves this.
  • Start-up Singularities: The mismatch between initial and boundary data creates problems primarily near $t=0$. Using specific semigroup decay estimates and interior Caccioppoli-type inequalities, the authors demonstrated that this initial "shock" does not degrade the overall $O(\sqrt{\epsilon})$ convergence rate.

Results and Convergence Rates

The study establishes sharp error estimates for both velocity and pressure at a rate of $O(\sqrt{\epsilon})$ in the $L^2$ (energy) norm. This matches the optimal rates known for simpler steady-state versions of the problem.

The microscopic pressure $p_\epsilon$ was shown to converge to the macroscopic pressure $p_0$ at the same $O(\sqrt{\epsilon})$ rate, representing a significant improvement over previous qualitative results. Furthermore, the derived asymptotic formula for velocity provides a precise framework for understanding how the microscopic flow $u_\epsilon$ relates to the macroscopic pressure gradient and boundary correctors.

Conclusion and Future Work

While the current proof is restricted to 2D and 3D spaces due to the properties of the curl operator, the methodology offers a robust framework for handling incompatible initial/boundary data. Future work may explore the optimality of the time-weight exponents used or extend these techniques to other evolutionary PDEs, such as the non-stationary Navier-Stokes equations in complex, perforated geometries.

Summary updated

Abstract

Focusing on Darcy's law incorporating memory effects, this paper studies non-stationary Stokes equations on perforated domains. We establish a sharp homogenization error for both velocity and pressure in terms of the energy norm. The main challenge lies in gauging the boundary layers induced by the incompressibility condition. To address this, we construct boundary-layer correctors using Bogovskii's operator. Also, the present work provides detailed regularity estimates for these correctors, where a significant difficulty arises from the incompatibility between initial and boundary values. The methodologies developed herein hold great potential for tackling the same issue in other evolutionary models beyond a homogenization setting.
BibTeX
@misc{wang2022corrector,
  author = {Li Wang and Qiang Xu and Zhifei Zhang},
  title = {{Corrector estimates and homogenization errors of unsteady flow ruled by Darcy's law}},
  year = {2022},
  eprint = {2202.04826},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2202.04826},
}
Figure