🔢 mathematics

Stream function -- pressure virtual element methods for the Stokes--Darcy interface problem

This paper presents a novel Virtual Element Method that utilizes a stream function formulation for the Stokes domain and a pressure formulation for the Darcy domain to efficiently solve the coupled Stokes–Darcy interface problem on general polygonal meshes while naturally handling complex interface geometries.

Original authors: Franco Dassi, Rekha Khot, David Mora, Andres E. Rubiano, Ricardo Ruiz-Baier

Published 2026-07-03

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Franco Dassi, Rekha Khot, David Mora, Andres E. Rubiano, Ricardo Ruiz-Baier

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where fluids behave in two very different ways depending on where they are flowing. In one area, like a wide river or a pipe, the fluid moves freely and smoothly (this is the Stokes part). In another area, like a sponge or a sponge-like tissue, the fluid has to squeeze through tiny, crowded holes, moving much more slowly and reluctantly (this is the Darcy part).

The big challenge for scientists is figuring out exactly how these two worlds talk to each other at the boundary where they meet. How much water leaks from the river into the sponge? How does the pressure change?

This paper introduces a new, clever mathematical tool called a Virtual Element Method (VEM) to solve this puzzle. Here is how it works, broken down into simple concepts:

1. The "Stream Function" Trick: Turning a 2-Step Dance into a Solo

Usually, when modeling the free-flowing river, mathematicians have to track two things at once: how fast the water is moving (velocity) and how hard it is pushing (pressure). It's like trying to choreograph a dance where two partners must stay perfectly in sync; if one step is wrong, the whole math breaks down.

The authors decided to use a "Stream Function." Think of this as a topographic map of the water's path. Instead of tracking the speed and pressure separately, they just track the "contour lines" of the flow.

The Benefit: Because the water is incompressible (it doesn't get squished), following these contour lines automatically guarantees that the water isn't disappearing or appearing out of nowhere. It's like drawing a closed loop; you know the water stays inside. This simplifies the math significantly and removes the need for that tricky "dance partner" synchronization.

2. The "Virtual Element" Advantage: Building with Irregular Bricks

Traditional math tools often require the computer to break the problem into perfect squares or triangles (like a grid). If the boundary between the river and the sponge is jagged or curved, you have to force the grid to fit, which is like trying to pave a winding garden path with square bricks—it looks messy and requires a lot of cutting and fitting.

The Virtual Element Method is like having a box of Lego bricks of any shape.

You can use hexagons, pentagons, or weird, star-shaped pieces.
If the interface between the river and the sponge is jagged, the math tool just uses a polygon that fits that shape perfectly.
Why it matters: You don't have to redraw the map or "remesh" the area just because the boundary is weird. The tool handles the irregular shapes naturally.

3. The "Handshake" at the Border

The most important part of this paper is how they made the two different worlds (the free-flowing river and the porous sponge) shake hands. They enforced three rules at the boundary:

Mass Conservation: The amount of water leaving the river must equal the amount entering the sponge.
Stress Balance: The push of the water against the sponge must match the resistance of the sponge.
The Slip Condition: The water doesn't stop dead at the edge of the sponge; it "slips" a little bit along the surface before going in. This is known as the Beavers–Joseph–Saffman condition.

The authors created a new mathematical "glue" that holds these two different types of equations together without breaking the system.

4. Real-World Tests (The "Proof")

To prove their new tool works, they ran several simulations:

The "Dead-End Filter": They simulated a quarter-circle filter (like a coffee filter) where fluid flows from a free space into a porous material. The math worked perfectly, even with messy, irregular grids.
The "Bio-Artificial Organ": They modeled a simplified version of a bio-artificial pancreas. Imagine a tiny blood vessel running through a sponge-like scaffold. The blood flows through the vessel, and nutrients/oxygen filter out into the sponge to feed cells.
- They tested different "sponge densities" (permeability).
- Result: When the sponge was very open (high permeability), the blood flowed easily out of the vessel and into the scaffold. When the sponge was tight, the blood stayed in the vessel. The model accurately predicted how the flow changed based on the sponge's properties.

Summary

In short, this paper presents a new, flexible way to simulate fluids moving between open spaces and spongy materials. By using a "stream function" to simplify the free-flow math and "virtual elements" to handle messy shapes, they created a robust tool that works well on complex, irregular geometries. It's a new set of mathematical "Lego bricks" that makes modeling complex fluid problems easier and more accurate.

Technical Summary: Stream function – pressure virtual element methods for the Stokes–Darcy interface problem

Problem Statement
The paper addresses the numerical simulation of the coupled Stokes–Darcy interface problem, a fundamental challenge in fluid mechanics governing the interaction between free-flow regions (Stokes) and porous media (Darcy). The coupling is achieved through continuity of normal velocities (mass conservation), balance of normal stresses, and the Beavers–Joseph–Saffman tangential slip condition. While various Finite Element (FE), discontinuous Galerkin (dG), and Hybrid High–Order (HHO) methods exist, often relying on velocity–pressure formulations that require inf–sup stable pairs, this work focuses on a specific reformulation in two-dimensional simply connected domains. Here, the incompressibility constraint is satisfied identically by introducing a scalar stream function, thereby eliminating the pressure variable from the Stokes equations and reducing the system to a fourth-order biharmonic problem coupled with a second-order Darcy pressure equation.

Methodology
The authors propose a novel Virtual Element Method (VEM) discretization for this stream function–pressure formulation. The methodology leverages VEM's capacity to handle general polygonal meshes, which is particularly advantageous for complex interface geometries and non-matching meshes.

Formulation: The continuous problem is cast in a primal-primal form. The Stokes domain utilizes a stream function $\chi \in H^2$ , while the Darcy domain uses a pressure potential $\phi \in H^1$ . The weak formulation results in a perturbed saddle-point problem.
Discretization:
- Stokes Domain: A $C^1$ -conforming lowest-order VEM space is employed for the stream function. This involves quadratic polynomials ( $P_2$ ) and requires $H^2$ -conformity, achieved through enhanced virtual spaces with degrees of freedom (DoFs) defined by nodal values and nodal gradients.
- Darcy Domain: A $C^0$ -conforming lowest-order VEM space is used for the pressure, based on linear polynomials ( $P_1$ ).
- Coupling: The interface conditions are enforced naturally within the weak formulation. The method accommodates non-matching meshes at the interface without remeshing.
Implementation: The scheme is implemented using the open-source object-oriented library vem++. A specific feature, the vem::vemMesh2dDofFilter class, is utilized to manage DoFs across different subdomains, allowing for the simultaneous definition of different VEM spaces on a single mesh and filtering them based on element markers.
Stabilization: The discrete bilinear forms include stabilization terms ( $S_S^K$ and $S_D^K$ ) to ensure coercivity on the kernel of the projection operators, following standard VEM practices (e.g., diagonal recipe and DOFI-DOFI).

Key Contributions
The paper outlines five primary contributions:

Continuous Analysis: Development of a continuous analysis for the Stokes–Darcy interface problem in stream function–pressure formulation, which inherently avoids the need for inf–sup stability conditions.
Discrete Scheme: Design of a $C^1$ – $C^0$ lowest-order VEM discretization. The authors highlight the computational efficiency of this scheme, noting it involves only $3N_S + N_D + 4N_\Sigma$ degrees of freedom (where $N_S$ , $N_D$ , and $N_\Sigma$ are the counts of interior vertices in the Stokes domain, Darcy domain, and on the interface, respectively).
Error Analysis: Derivation of a priori error estimates for the proposed discretization, proving optimal convergence rates under standard regularity assumptions.
Software Implementation: Integration of new functionality into the vem++ library to facilitate the straightforward implementation of interface problems via mesh filtering.
Numerical Validation: Execution of numerical experiments to confirm theoretical findings.

Results
The numerical experiments validate the theoretical convergence rates and the method's robustness:

Manufactured Solutions: Two experiments using manufactured solutions (one with zero interface values, one with non-zero) on various mesh types (quadrilateral, non-convex, Voronoi, and perturbed Voronoi) demonstrate an asymptotic linear decay in error ( $O(h)$ ), consistent with the theoretical predictions for the lowest-order scheme.
Dead-End Filtration: A simulation of a quarter-annulus dead-end filter successfully recovers results from previous studies, demonstrating the method's ability to handle curved interfaces and mixed boundary conditions.
Bioartificial Organs: A simulation of blood flow in a bioartificial organ (encapsulation-based) with varying permeability constants illustrates the method's capability to model complex physiological flows and the influence of permeability on flux across the interface.

Significance and Claims
The paper positions its work as bridging a gap in the literature regarding the extension of stream function formulations to coupled Stokes–Darcy problems using Virtual Elements. The authors claim that the proposed structure offers a "completely inf–sup-free strategy," simplifying the design of discrete spaces compared to traditional velocity-pressure approaches. The ability to handle irregular interface geometries and non-matching meshes without remeshing is highlighted as a key advantage of the VEM framework.

The authors remain modest regarding future challenges, explicitly noting that extending this stream-function formulation to multiply connected domains (where the stream function is not uniquely determined by standard boundary conditions) and scaling the framework to three dimensions (requiring vector potentials and high-order $C^1$ spaces for vector-valued functions) remain open challenges. Additionally, the integration of stabilization-free techniques and the derivation of robust a posteriori error estimators are identified as directions for future research.

1. The "Stream Function" Trick: Turning a 2-Step Dance into a Solo

2. The "Virtual Element" Advantage: Building with Irregular Bricks

3. The "Handshake" at the Border

4. Real-World Tests (The "Proof")

Summary

More like this