Introduction
Simulating the interaction between fluids and solid structures (Fluid-Structure Interaction, or FSI) is a cornerstone of modern engineering, essential for designing everything from heart valves to aircraft wings. However, researchers have long struggled with the "added-mass" instability. This numerical phenomenon occurs when a solid body is light relative to the fluid it displaces; the fluid acts as an additional inertial weight that standard partitioned algorithms fail to account for, leading to catastrophic numerical oscillations.
Traditional partitioned schemes, which solve fluid and solid equations separately for modularity, typically require many expensive sub-iterations per time step to remain stable in these "light-body" scenarios. This paper introduces the Added-Mass Partitioned for Rigid Bodies (AMP-RB) algorithm, a breakthrough that achieves unconditional stability without the need for sub-iterations.
The AMP-RB Algorithm
The core innovation of the AMP-RB scheme is the use of a generalized Robin interface condition. Unlike standard methods that simply pass velocity and stress back and forth between solvers, AMP derives a boundary condition for the fluid pressure that explicitly incorporates the linear and angular acceleration of the rigid body. This effectively "bakes" the added-mass effect directly into the fluid solver's pressure step.
To handle viscous instabilities—often referred to as "added-damping"—the authors introduce added-damping tensors. These are derived through a linearization of force and torque integrals, allowing the algorithm to anticipate how viscous shear will affect the body's motion. A key discovery in the methodology is the introduction of a scaling parameter, $\beta_d$. The authors found that setting $\beta_d \approx 1$ provides a robust "sweet spot" for stability across a wide range of physical parameters. For extremely light bodies, an optional Velocity Correction (VC) step further ensures stability against viscous drag.
Stability Analysis and Model Problems
To prove the algorithm's robustness, the authors performed rigorous mathematical analysis using Fourier and GKS stability theory. They focused on simplified model problems that capture the essential physics of the FSI challenge, such as 1D piston motion and 2D disks in annular geometries.
The analysis demonstrates that while traditional schemes are only stable if the body's mass exceeds the fluid's added mass, the AMP scheme remains stable even when the body mass is zero. This theoretical foundation was verified using several benchmarks, including newly derived exact solutions for translating disks in annular Stokes flow.
Key Results
- Unconditional Stability: AMP-RB is shown to be stable for any mass ratio, effectively solving the added-mass and added-damping instability problems.
- Second-Order Accuracy: Numerical tests confirm that the scheme maintains second-order accuracy in both space and time, with errors dropping quadratically as the grid is refined.
- Computational Efficiency: By eliminating the need for sub-iterations, the AMP scheme is significantly faster than traditional methods, which might otherwise require 10–100 iterations per time step to maintain stability for light bodies.
Conclusion and Future Work
This work establishes a stable and efficient framework for FSI simulations involving rigid bodies. By addressing the fundamental mathematical causes of instability, the AMP-RB algorithm allows for high-fidelity simulations of light or even massless structures in incompressible flow. While this first part focuses on rigid bodies and simplified geometries, the authors plan to extend these principles to general, complex geometries and deforming elastic structures in future research.
Summary updated