Acceleration Based Interface Coupling for Fluid Structure Interaction

18 concepts6 papers1 proposal

The Physics of the Added-Mass Instability

In fluid-structure interaction (FSI) simulations, engineers often use partitioned solvers. These solvers separate the fluid and solid calculations, solving each domain independently and alternating between them at each time step. While computationally convenient, this approach suffers from a severe numerical vulnerability known as the added-mass instability.

As noted by Vito Pasquariello et al. in A cut-cell finite volume – finite element coupling approach for fluid–structure interaction in compressible flow, standard loose or staggered coupling schemes become highly unstable when the densities of the fluid and the solid are comparable. To understand why, imagine a very light or even zero-mass rigid body submerged in a heavy fluid. Because the solid has little inertia of its own, its motion is dominated by the inertia of the surrounding fluid—the "added mass."

In a naive partitioned scheme, the fluid solver calculates the fluid pressure and passes the resulting forces to the solid solver. Because the solid is light, this force causes it to accelerate wildly. In the next sub-step, the solid solver passes this extreme velocity back to the fluid solver. The fluid solver must then react to this sudden boundary movement, generating an enormous, unphysical pressure spike to enforce incompressibility. This artificial feedback loop amplifies numerical errors, causing the simulation to experience divergent oscillations and crash (Jeffrey W. Banks et al. in A stable partitioned FSI algorithm for rigid bodies and incompressible flow in three dimensions).

The Cost of Traditional Fixes

To keep the simulation from blowing up, traditional partitioned methods are forced to use extremely small time steps or perform expensive sub-time-step iterations. The fluid and solid solvers must pass information back and forth 10 to 100 times per time step to achieve numerical stability (Jeffrey W. Banks et al. in A stable partitioned FSI algorithm for rigid bodies and incompressible flow. Part I: Model problem analysis).

Alternatively, some frameworks turn to implicit coupling. For example, Hsin-Yu Hsieh et al. in A target-fixed immersed-boundary formulation for rigid bodies interacting with fluid flow couple the fluid and rigid body dynamics implicitly, explicitly including added-mass effects to ensure stability at low solid-to-fluid mass ratios. However, resolving forces and motions simultaneously in this manner introduces significant computational overhead.

Embedding Motion into Pressure Boundary Conditions

The breakthrough insight that eliminates the need for these expensive sub-iterations is to change what information is shared at the fluid-solid boundary. Instead of merely matching velocities at the interface, the Added-Mass Partitioned (AMP) algorithm incorporates the linear and angular acceleration of the solid directly into the fluid's pressure boundary conditions (Jeffrey W. Banks et al. in A stable partitioned FSI algorithm for rigid bodies and incompressible flow. Part II: General formulation).

The fluid solver typically utilizes a fractional-step method. In this framework, fluid velocity is first advanced independently without considering the boundary's full impact. Then, a pressure update step enforces interface compatibility. The authors derive a generalized Robin interface condition that mathematically embeds the rigid body's equations of motion—including added-mass and added-damping tensors—directly into the fluid's pressure Poisson equation.

Why Acceleration Matching Restores Stability

By incorporating the solid's acceleration into the fluid's pressure boundary condition, the fluid solver effectively "foresees" how the solid will react to the fluid's inertia before the pressure is even updated.

Instead of a sequential lag where the fluid guesses a force and the solid overreacts, the generalized Robin condition forces the fluid pressure and the solid acceleration to balance each other simultaneously at the interface. This modification suppresses the numerical errors that drive the added-mass instability without requiring any iterative sub-steps.

As a result, the AMP scheme achieves second-order temporal accuracy while remaining completely stable, even when the solid body has negligible or zero mass relative to the displaced fluid (Jeffrey W. Banks et al. in A stable partitioned FSI algorithm for rigid bodies and incompressible flow. Part I: Model problem analysis). In these extreme zero-mass scenarios, the solid's motion is dictated entirely by the fluid's added mass and damping effects, yet the partitioned solver handles it seamlessly (Jeffrey W. Banks et al. in A stable partitioned FSI algorithm for rigid bodies and incompressible flow in three dimensions).

Application to Immersed Boundary Methods

The challenge of managing pressure-force coupling when fluid inertia impacts solid motion is not unique to body-conforming meshes; it is also a major bottleneck for fixed-grid methods. In two-way coupled Immersed Boundary Methods (IBMs), the fluid's added-mass effect can easily destabilize the fixed-grid interpolation.

To address this, Rachel Yovel et al. in A SIMPLE-based preconditioned solver for the direct-forcing immersed boundary method developed a preconditioned SIMPLE algorithm that leverages the discrete Laplacian operator. By designing a solver that efficiently manages the strong pressure-force coupling inherent to these interactions, the authors achieve robust, scalable simulations under significant added-mass effects. This demonstrates that whether through modified boundary conditions or specialized algebraic preconditioners, explicitly accounting for the simultaneous balance of fluid pressure and structural inertia is the universal key to stable fluid-structure coupling.

History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_added_mass_instability_c421ca50,
  title = {added-mass instability},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/added-mass-instability/rdr_c421ca5034f94c9195ece46370b86db8},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_added_mass_instability_c421ca50_v1,
  title = {added-mass instability},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/added-mass-instability/rdr_c421ca5034f94c9195ece46370b86db8/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake