🔢 mathematics

Second order explicit splitting scheme for fluid-poroelastic structure interaction problems

This paper presents and rigorously analyzes a fully discrete, second-order, explicit splitting scheme for fluid-poroelastic structure interaction problems that achieves unconditional stability under a parabolic CFL condition and optimal-order convergence by combining BDF2 time stepping with second-order Adams-Bashforth extrapolation within a Robin reformulation framework.

Original authors: Yifan Wang, Jeonghun Lee, Suncica Canic

Published 2026-06-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Yifan Wang, Jeonghun Lee, Suncica Canic

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 busy highway where two very different types of traffic are trying to move in sync. On one side, you have fluid (like water or blood) flowing smoothly. On the other side, you have a porous sponge (like a soft, squishy rock or biological tissue) that can soak up liquid and deform.

The problem this paper solves is how to calculate the movement of both the fluid and the sponge at the same time, especially where they touch. This is called Fluid-Poroelastic Structure Interaction.

Here is the breakdown of what the authors did, using simple analogies:

1. The Problem: The "Tug-of-War" at the Interface

When the fluid pushes against the sponge, the sponge moves. When the sponge moves, it pushes back on the fluid. They are locked in a tight dance.

The Old Way (Monolithic): Imagine trying to solve this dance by having one giant brain calculate the position of every single water molecule and every single sponge fiber simultaneously. It works, but it's incredibly slow and requires a supercomputer for every single step.
The New Way (Partitioned): The authors wanted to split the brain into two smaller brains. One brain handles the fluid, and the other handles the sponge. They solve their own problems separately and then talk to each other. This is much faster and allows them to work in parallel (like two people typing on different computers at the same time).

2. The Trap: The "Explicit" Shortcut

Usually, when two separate systems talk to each other, they wait for the other to finish before taking the next step. This is safe but slow.
The authors wanted to be even faster. They used an "Explicit" method.

The Analogy: Imagine two dancers. Instead of waiting for the partner to finish a move before starting the next, they try to predict where the partner will be based on what they did a moment ago.
The Risk: If your prediction is wrong, the dancers might trip and fall (the math becomes unstable and crashes). The paper's main challenge was proving that their specific prediction method is safe and won't cause a crash.

3. The Solution: The "Second-Order" Prediction

The authors developed a specific recipe for this prediction:

BDF2 (The Memory): Instead of just looking at the last step to guess the next one, they look at the last two steps. It's like a dancer remembering their last two moves to predict the rhythm of the next one. This makes the guess much more accurate.
AB2 (The Extrapolation): They use a mathematical trick (Adams-Bashforth) to project the interface data (the "handshake" between fluid and sponge) forward in time.
Robin Reformulation: They changed the rules of the handshake slightly (using something called a "Robin" condition). Think of this as adding a spring between the dancers. It allows them to pull and push in a way that keeps the math stable even when they are guessing the next move.

4. The Proof: "Stability" and "Accuracy"

The authors didn't just guess this would work; they did the heavy math to prove it.

Stability: They proved that as long as the time steps are small enough (a rule they call a "CFL condition," like a speed limit), the simulation will never explode or go crazy. The "energy" of the system stays under control.
Accuracy: They proved that if you make the time steps smaller, the error drops by the square of that size.
- Analogy: If you cut your time step in half, your error doesn't just get half as big; it gets four times smaller. This is called "Second-Order Accuracy." It means the method gets very precise very quickly.

5. The Experiments: "Manufactured Solutions" and "Blood Flow"

To test their theory, they ran two types of simulations:

The "Fake" World: They created a problem where they already knew the exact answer (a "manufactured solution"). They ran their algorithm and compared the result to the known answer.
- Result: The error matched their predictions perfectly. The method was indeed second-order accurate in time and optimal in space.
The "Moving" World: They applied the method to a more complex scenario: blood flowing through an artery where the walls are porous and moving. Even though their strict math proof was for a fixed room, they showed the method works robustly even when the room itself is changing shape (using a technique called ALE).
- Result: The simulation produced smooth, realistic pressure waves and deformations, similar to what is seen in other high-quality studies.

Summary

This paper presents a fast, parallel, and mathematically proven way to simulate how fluids and spongy materials interact.

Why it matters: It allows scientists to run complex simulations much faster than before because the fluid and the sponge can be solved on different computers at the same time.
The Catch: You have to follow a specific "speed limit" (time step size) to keep the math stable, but if you do, the results are highly accurate.

The authors conclude that this method is a strong candidate for solving even more complex, real-world problems in the future, provided the math is extended to handle moving boundaries.

Technical Summary: Second Order Explicit Splitting Scheme for Fluid-Poroelastic Structure Interaction Problems

Problem Statement
The paper addresses the numerical simulation of fluid–poroelastic structure interaction (FPSI) problems, specifically the time-dependent Stokes–Biot system on fixed domains. These problems arise in contexts such as biological tissue perfusion and flow through deformable porous materials. A central challenge in this field is the efficient and stable treatment of the interface coupling conditions, which enforce continuity of normal flux and balance of normal and tangential stresses. While monolithic methods are robust, they require solving large coupled systems at every time step. Partitioned methods are computationally attractive as they allow the use of specialized solvers for the fluid (Stokes) and poroelastic (Biot) subproblems independently. However, explicit or loosely coupled partitioned schemes often struggle with stability, particularly when the interface coupling is treated explicitly.

Methodology
The authors propose a fully discrete, second-order explicit splitting scheme for the Stokes–Biot problem. The methodology relies on the following components:

Robin Reformulation: The interface conditions are rewritten in a Robin form using penalty parameters $\gamma > 0$ (tangential) and $L > 0$ (normal). This reformulation decouples the subproblems while maintaining a consistent coupling mechanism suitable for partitioned algorithms.
Time Discretization: The scheme employs the second-order Backward Differentiation Formula (BDF2) for the time evolution of the subdomain equations.
Explicit Extrapolation: To achieve a fully explicit, parallelizable scheme, the interface data required for the Robin boundary conditions at time $t_{n+1}$ are approximated using second-order Adams–Bashforth (AB2) extrapolation based on data from previous time steps ( $t_n$ and $t_{n-1}$ ).
Algorithm Structure: At each time step, the Stokes and Biot subproblems are solved independently and in parallel. The kinematic relation between structure displacement and velocity is handled via a BDF2 relation within the Biot subproblem.

Key Contributions
The primary contributions of the work are theoretical and analytical:

Discrete Stability Analysis: The authors provide a rigorous stability proof for the proposed scheme. By utilizing BDF2 energy identities and a sharp decomposition of the extrapolated interface terms, they establish a closed stability bound. This bound holds under a parabolic Courant–Friedrichs–Lewy (CFL) condition linking the time step $\Delta t$ and the mesh size $h$ (specifically $\Delta t \leq c^* h^2$ ).
A Priori Error Estimates: The paper derives an a priori error estimate using a projection-based framework. This involves:
- A Fortin projection for the fluid velocity to handle the divergence-free constraint.
- Ritz-type projections for the poroelastic displacement and pore pressure.
- A careful analysis of consistency defects arising from the BDF2 time discretization, AB2 interface extrapolation, and the projected kinematic relation.
Convergence Rates: The analysis demonstrates that the total errors in fluid velocity, structure velocity, pore pressure, and elastic displacement are bounded by $C(h^k + \Delta t^2)$ in bulk energy norms for $1 \leq k \leq 3$ . This confirms second-order temporal accuracy and optimal-order spatial convergence.

Numerical Results
The theoretical findings are validated through two numerical experiments:

Fixed-Domain Benchmark (Manufactured Solution): A test case with known exact solutions was used to verify convergence rates. The results confirmed the predicted second-order convergence in time for all variables (fluid velocity/pressure, structure displacement/velocity, and pore pressure). Spatial convergence tests on refined meshes showed optimal-order rates consistent with the chosen finite element spaces (Taylor–Hood $P_2/P_1$ for Stokes and $P_2$ elements for Biot variables).
Moving-Domain Example (Navier–Stokes–Biot): To demonstrate the applicability of the partitioned strategy beyond the fixed-domain Stokes–Biot setting, the authors applied the scheme to a 2D blood flow problem in a moving domain governed by the Navier–Stokes equations coupled with the Biot system. Using an Arbitrary Lagrangian–Eulerian (ALE) formulation, the simulation successfully captured pressure wave propagation and structural deformation, showing results consistent with previous literature.

Significance and Claims
The paper claims to provide a mathematically justified, second-order accurate, and parallelizable explicit partitioned method for fluid–poroelastic interaction. The significance lies in overcoming the stability challenges typically associated with explicit coupling in FPSI problems. By proving stability under a parabolic CFL condition and deriving rigorous error bounds, the work establishes that explicit splitting can be both efficient and accurate without sacrificing mathematical rigor.

The authors note that while the stability and error analysis are strictly for the fixed-domain Stokes–Biot system, the numerical example suggests the strategy is robust enough to be extended to more general moving-domain flow problems, such as those involving the Navier–Stokes equations. They conclude that the scheme offers a practical compromise between modularity, parallel efficiency, and provable accuracy, though they modestly acknowledge that further work is needed to investigate relaxing the CFL restriction or extending the framework to geometrically nonlinear moving-interface problems.