Stable fluid-rigid body interaction algorithm using the direct-forcing immersed boundary method (DF-IBM)

This paper presents an extended direct-forcing immersed boundary method (DF-IBM) that couples Navier-Stokes equations with Newton-Euler equations through an implicit, computationally efficient algorithm to stably simulate free-moving fluid-rigid body interactions across various density ratios.

The Gist

Imagine you are trying to simulate a game of air hockey, but with a twist: the puck isn't just sliding on a table; it’s floating in a thick, swirling pool of honey, and the puck itself is alive and reacting to every tiny ripple in the honey.

This scientific paper describes a new, smarter way for computers to calculate exactly how a solid object (like a puck or a leaf) moves when it is tossed into a moving fluid (like water or air).

Here is the breakdown of how they did it, using everyday analogies.

1. The Problem: The "Ghost in the Machine" (Internal Mass Effect)

When you push a toy boat through water, you aren't just moving the boat; you are also pushing a "ghost" of water that is trapped inside the shape of the boat. In physics, this is called the Internal Mass Effect (IME).

If a computer program forgets to account for this "ghost water," the simulation breaks. It’s like trying to predict how a heavy pendulum swings, but forgetting that the pendulum is actually a hollow balloon filled with water. The math gets "jittery," the object might suddenly fly off the screen, or the simulation might simply crash. This is especially hard when the object and the fluid have almost the same weight (like a piece of wood floating in water).

2. The Solution: The "Smart Negotiator" (Implicit Coupling)

Most computer simulations use a method called "Partitioned Coupling." Imagine two people—one representing the Water and one representing the Solid—trying to coordinate a dance.

• The Old Way (Weak Coupling): The Water person says, "I'm moving this way!" and the Solid person says, "Okay, I'll move there!" They only talk once per step. If the water moves too fast, the solid person trips, and the dance becomes a chaotic mess.
• The New Way (Strong/Implicit Coupling): This paper introduces a "Smart Negotiator." Instead of just shouting instructions once, the Water and the Solid enter a rapid-fire conversation (iterations) at every single micro-second. They keep adjusting their positions—"Wait, if I move there, you'll push me here..."—until they both agree on a stable position. This ensures the "dance" remains smooth, even if the object is very light or the water is very turbulent.

3. The "Speed Bump" (Fixed Relaxation)

Even with a negotiator, sometimes the conversation gets too intense, and the two sides start overreacting to each other's movements (this is called numerical instability).

To fix this, the researchers added a "Fixed Relaxation" technique. Think of this as a speed bump or a buffer. Instead of the solid object instantly snapping to the new position the water suggests, it moves only partway there. This prevents the system from "oversteering" and keeps the simulation stable and calm.

4. Why does this matter? (The "Real World" Impact)

Why spend all this time on math? Because this "Smart Negotiator" approach allows scientists to simulate much more complex and realistic scenarios without the computer exploding:

• Medical Science: Simulating how a tiny heart valve or a blood cell moves through a vein (where the "solid" and "fluid" are almost the same density).
• Nature: Watching how a leaf tumbles in a stream or how a seed falls through the air.
• Engineering: Designing better shapes for underwater drones or aircraft wings that can react to wind and waves naturally.

In short: The researchers built a more stable, "conversational" mathematical bridge between fluids and solids, allowing computers to simulate the messy, swirling reality of nature with much higher accuracy and less crashing.

Technical Summary

Technical Summary: Stable Fluid-Rigid Body Interaction Algorithm using the Direct-Forcing Immersed Boundary Method (DF-IBM)

1. Problem Statement

The paper addresses the challenges inherent in Fluid-Structure Interaction (FSI) problems where a rigid body undergoes free (flow-induced) motion rather than prescribed motion. Specifically, the authors focus on the numerical instabilities that arise when using the Immersed Boundary Method (IBM) in partitioned (segregated) coupling frameworks.

The primary technical hurdle is the Internal Mass Effect (IME). In incompressible flows, the inertia of the fluid enclosed within the rigid body's boundary significantly impacts translational and rotational dynamics. When the density of the solid ($\rho_s$) is close to the density of the fluid ($\rho_f$)—a condition known as neutral buoyancy—standard explicit coupling methods often fail, leading to spurious numerical energy production and divergence. Furthermore, the "rigid body approximation" (assuming internal fluid follows the rigid body's motion) introduces additional instabilities in rotational dynamics.

2. Methodology

The authors propose an enhanced Direct-Forcing Immersed Boundary Method (DF-IBM) coupled with the Newton-Euler equations of motion. The methodology is characterized by several key technical innovations:

• Implicit Partitioned Coupling: Unlike weakly coupled (explicit) methods, the authors developed an implicit algorithm that iteratively enforces the kinematic and dynamic interface conditions. This is achieved through a fixed-point strategy within a partitioned framework, allowing the use of existing fluid solvers (like OpenFOAM).
• Optimized Predictor-Corrector Strategy: The algorithm leverages a modified PISO (Pressure Implicit with Splitting of Operators) approach. To improve efficiency, the authors omit the costly momentum predictor and the standard pressure Poisson equation (PPE) corrector loops, instead using a modified PPE (mPPE) that inherently accounts for the boundary force.
• Mitigation of IME Instabilities:
  • Explicit IME Treatment: To handle the density ratio limitation, the authors utilize a first-order forward Euler scheme for the IME terms.
  • Fixed Relaxation Technique: To ensure stability during the implicit iterations—especially for neutrally buoyant bodies and complex rotational motions—a fixed relaxation parameter ($\alpha_r$) is applied to the linear and angular velocities. This damps high-velocity fluctuations and prevents divergence.
• Mapping Operators: The method uses a direct-forcing approach where velocity interpolation ($I$) and force spreading ($S$) operators map data between the Eulerian fluid grid and the Lagrangian immersed boundary.

3. Key Contributions

• Robustness across Density Ratios: The algorithm remains stable for a wide range of solid-to-fluid density ratios, including extremely challenging cases where $\rho_s/\rho_f \approx 1$ (neutrally buoyant) and even as low as $0.3$.
• Handling Complex Dynamics: The method successfully integrates both translational and rotational Newton-Euler equations, specifically addressing the instabilities caused by the rigid body approximation in rotational motion.
• Computational Efficiency: By optimizing the PISO loops and leveraging the fact that the Lagrangian domain is much smaller than the Eulerian domain, the method provides a high-fidelity solution without the massive computational overhead of body-conforming (ALE) mesh updates.

4. Results and Validation

The algorithm was validated against several benchmark cases:

• 2D Disk Sedimentation: Confirmed accurate terminal velocity and kinetic energy, demonstrating that the inclusion of IME is vital for matching experimental/literature data.
• Shear Flow-Induced Rotation: Validated the rotational dynamics of neutrally buoyant disks. The relaxation technique successfully eliminated the oscillations typically seen in explicit IME treatments.
• NACA0012 Airfoil Rotation: Demonstrated the ability to capture complex, flow-induced oscillations in an airfoil. While a slight phase shift was noted (a known characteristic of the explicit IME treatment for high rotational Reynolds numbers), the oscillation peaks were highly accurate.
• 2D Ellipse Sedimentation: Successfully captured coupled translational and rotational trajectories, showing strong agreement with ALE-based finite element results.
• Extreme Density Ratios: Successfully simulated both "free fall" (dense) and "free rise" (light) scenarios, matching literature values for drag coefficients, lift coefficients, and Strouhal numbers.

5. Significance

This work provides a highly stable and efficient numerical framework for simulating complex FSI problems. By overcoming the "added mass" instability through implicit coupling and relaxation, it opens the door for more accurate simulations in fields where fluid and solid densities are similar, such as biomechanics (e.g., blood flow interacting with vessel walls) and marine biology (e.g., plankton or organism movement in water). The use of a fixed Cartesian grid via IBM also makes it significantly more practical for complex, moving geometries compared to traditional mesh-deformation methods.