Stability prediction of vortex induced vibrations of multiple freely oscillating bodies

This paper investigates the stability of vortex-induced vibrations in multiple freely oscillating bodies by deriving a Linear Arbitrary Lagrangian Eulerian method and proposing a low-cost impedance-based criterion to accurately predict instability thresholds, which are validated through global stability analysis and extensive parametric studies on tandem and three-body cylinder systems.

The Gist

Imagine you are standing by a river, watching two buoys tied to springs bobbing up and down in the current. Sometimes, they just drift lazily. Other times, they start shaking violently, dancing in a chaotic rhythm. This phenomenon is called Vortex-Induced Vibration (VIV). It happens when the water flowing past an object creates swirling eddies (like tiny whirlpools) that push the object back and forth.

This paper is about predicting exactly when and why these buoys (or cylinders) will start shaking uncontrollably, especially when there are two or more of them close together.

Here is the breakdown of the research using simple analogies:

1. The Problem: The "Tandem" Dance

The researchers looked at a specific setup: two cylinders placed one behind the other in a flow (like two buoys in a line).

• The Challenge: When the first cylinder shakes, it messes up the water flow for the second one. The second cylinder then shakes, which messes up the flow for the first one again. It's a constant feedback loop.
• The Difficulty: Calculating this interaction is like trying to predict the exact path of two dancers who are constantly reacting to each other's moves while the music (the water flow) is changing. Doing this math for every possible speed, weight, and distance is incredibly slow and expensive for computers.

2. The Solution: The "Impedance" Crystal Ball

The authors developed a clever shortcut. Instead of simulating the whole chaotic dance every time, they created a method called Impedance-Based Stability Prediction.

• The Analogy: Imagine you want to know if a car engine will stall. Instead of driving the car on a test track for every possible hill and load, you put the engine on a test bench. You shake the engine at different frequencies and measure how much force it takes to keep it moving.
• How it works:
  1. The Test: They forced the cylinders to wiggle at specific frequencies (like shaking a test tube) and measured the "resistance" or impedance the water offered.
  2. The Crystal Ball: They turned these measurements into a mathematical "recipe" (a 2x2 matrix).
  3. The Prediction: Now, instead of running a massive simulation, they just plug in the weight of the cylinders, the stiffness of the springs, and the water speed into this recipe. If the math says "danger," the cylinders will shake. If it says "safe," they won't.

It's like having a weather forecast that tells you if a storm is coming without needing to simulate every single raindrop.

3. The Findings: What Makes Them Shake?

Using this new "crystal ball," they explored thousands of scenarios. Here are the main discoveries:

• The "Ghost" and the "Dancer": They found two main types of shaking:
  • Mode A (The Ghost): This is a "fluid" mode. The water wants to shake, but the heavy cylinders barely move. It's like a ghost haunting the water; the flow is unstable, but the objects stay relatively still.
  • Mode B (The Dancer): This is a "structural" mode. The cylinders themselves are the ones shaking violently, dragging the water along with them. This is the dangerous one for engineers.
• Heavy is Good (Usually): Making the cylinders heavier (increasing mass) generally stops them from shaking. It's like adding a heavy anchor to a boat; it's harder to push around.
  • Exception: For the "Ghost" mode (Mode A), making them heavier actually made the instability worse in some cases.
• The Spacing Secret: How far apart the cylinders are matters immensely.
  • Too Close: They get stuck in each other's wake, sometimes canceling out the shaking.
  • Just Right: They can amplify each other's shaking, creating a massive resonance.
  • Far Apart: They start acting like two separate individuals, ignoring each other.
• The "Trio" Test: They even tested this method on three cylinders arranged in a triangle. The "crystal ball" worked perfectly, proving this method can be used for complex groups of objects, not just pairs.

4. Why Does This Matter?

Why do we care about shaking buoys?

• Preventing Disasters: Engineers need to know when bridges, oil rigs, or underwater cables might shake themselves apart. This method helps them design safer structures without running millions of expensive computer simulations.
• Harvesting Energy: Conversely, some people want things to shake! There are devices that use this shaking motion to generate electricity from ocean currents. This research helps engineers tune these devices to shake as much as possible to generate the most power.

Summary

The paper is essentially a new, super-fast way to predict when objects in a fluid will start dancing dangerously. By treating the fluid's resistance like an electrical circuit (impedance), the researchers turned a complex, slow physics problem into a quick, easy calculation. It's a tool that helps us build safer bridges and more efficient energy harvesters.

Technical Summary

1. Problem Statement

The paper investigates the Vortex-Induced Vibrations (VIV) of multiple spring-mounted cylinders arranged in a tandem configuration (and briefly a three-body cluster) immersed in a Newtonian, incompressible fluid. The study focuses on the linear stability of the fluid-structure interaction (FSI) system where bodies are free to oscillate in the direction orthogonal to the flow (transverse motion, 1DOF).

Key challenges addressed include:

• The high computational cost of performing exhaustive parametric studies (varying Reynolds number $Re$, reduced velocity $U^*$, mass ratio $m^*$, damping $\gamma$, and spacing $L/D$) using traditional global Linear Stability Analysis (LSA).
• The complex interplay between fluid modes (wake instabilities) and structural modes (natural frequencies) in multi-body systems, leading to phenomena like mode coupling, decoupling, and branch exchanges.

2. Methodology

The authors propose a novel, low-cost computational framework combining Linear Arbitrary Lagrangian-Eulerian (L-ALE) formulations with an Impedance-Based Stability Criterion.

A. Linear Arbitrary Lagrangian-Eulerian (L-ALE) Formulation

• Discretization: The fluid domain is treated in a reference (Eulerian) frame, while the moving boundaries are handled via a diffeomorphism.
• Discrete-ALE Ansatz: Instead of solving the mesh deformation for every time step or eigenvalue iteration, the authors use a superposition principle. The mesh displacement field is expressed as a linear combination of $N$ elementary fields, where each field corresponds to a unit displacement of a single cylinder. This reduces the computational burden significantly.
• Linearization: The Navier-Stokes equations and the structural equations of motion are linearized around a steady base flow. The system is formulated as a generalized eigenvalue problem: $-i\omega \mathbf{B}\mathbf{q} = \mathbf{A}\mathbf{q}$, where $\mathbf{q}$ contains both fluid variables and structural displacements/velocities.

B. Impedance-Based Stability Criterion

To avoid solving the large-scale eigenvalue problem repeatedly for different structural parameters, the authors derive an impedance-based method:

1. Forced Problem: The cylinders are forced to oscillate harmonically at a prescribed real frequency $\omega$ with unit amplitude.
2. Impedance Definition: The unsteady lift forces are calculated. A mechanical impedance matrix $\mathbf{Z}$ is defined, relating the force on body $i$ to the velocity of body $j$ ($Z_{ij} = F_{ij} / (i\omega)$).
3. Generalized Criterion: The stability of the coupled system is determined by the roots of the determinant of a compact $2 \times 2$ (for two bodies) or $N \times N$ matrix $\mathbf{Z}_T$, which combines the structural parameters (mass, stiffness, damping) and the fluid impedance functions.
   • Instability occurs if the determinant has a root with a positive imaginary part (growth rate).
   • Efficiency: Once the impedance functions $Z_{ij}(\omega)$ are tabulated for a specific geometry ($Re, L/D$), predicting stability for any combination of structural parameters ($m^*, U^*, \gamma$) requires only solving a small linear system, making parametric sweeps instantaneous.

3. Key Contributions

• Derivation of L-ALE Method: A rigorous derivation of the linearized FSI problem using the ALE framework, specifically tailored for multiple oscillating bodies.
• Impedance-Based Prediction: The development of a low-computational-cost criterion that replaces large-scale eigenvalue solvers with the inspection of a small matrix determinant. This method is validated as being "exact" within the linear regime, equivalent to the classical $p-k$ method but using exact hydrodynamic coupling from Navier-Stokes rather than approximate aerodynamic models.
• Extension to Multi-Body Systems: Demonstration that the method scales efficiently to systems with $N > 2$ bodies (validated on a 3-cylinder "fluidic pinball" configuration).

4. Results and Discussion

The study performs an extensive parametric analysis on a tandem of two cylinders ($N=2$) and a three-cylinder cluster.

A. Validation

• The impedance-based predictions show perfect agreement with direct Linear Stability Analysis (LSA) results for various $Re$, $m^*$, and $L/D$.
• The method correctly identifies neutral stability thresholds and eigenfrequencies.

B. Physical Modes Identification

Two primary types of unstable modes are identified:

1. Mode A (Fluid-Dominated): Resembles the wake instability of fixed cylinders. The cylinder displacements are negligible. Its frequency matches the Strouhal frequency of the fixed tandem wake.
2. Mode B (Structure-Dominated): Arises from fluid-structure coupling. The cylinders exhibit significant transverse motion. Its frequency tracks the natural frequency of the spring-mass system at low $U^*$ and shifts toward the wake frequency at high $U^*$.
3. Mode C: Observed at larger spacings, exhibiting characteristics similar to Mode B but with different spatial structures.

C. Parametric Effects

• Mass Ratio ($m^*$):
  • Increasing mass generally stabilizes the structure-dominated modes (B and C) by reducing the range of unstable reduced velocities.
  • Conversely, increasing mass destabilizes the fluid-dominated Mode A, extending its unstable range to higher reduced velocities.
  • At high mass ($m^*=20$), complex topological transitions (branch exchanges) occur between modes A and B, identified as codimension-2 exceptional points.
• Damping ($\gamma$):
  • Increasing damping stabilizes the system, narrowing the unstable regions for all modes.
  • High damping leads to codimension-3 points (double Hopf with strong resonance) where modes coalesce.
• Spacing ($L/D$):
  • The stability map evolves with spacing, mirroring wake interference regimes observed in fixed cylinders (slender body, reattachment, intermittent, binary vortex street).
  • Decoupling at Large Spacing ($L/D > 4.5$): A significant dynamic shift occurs. Modes B and C become decoupled, meaning the motion of one cylinder is dominated by the flow of the other (e.g., Mode B behaves like the rear cylinder fixed, Mode C like the front cylinder fixed).
  • Mode A remains coupled: Even at large spacing, Mode A requires the simultaneous motion of both cylinders and cannot be described by single-cylinder fixed configurations.

D. Three-Body System

The method was successfully applied to a triangular arrangement of three cylinders. It correctly predicted five leading eigenmodes (two fluid-dominated, three structure-dominated) and their stability thresholds, confirming the scalability of the impedance approach.

5. Significance

• Computational Efficiency: The impedance-based method drastically reduces the cost of parametric studies. While traditional LSA requires solving large eigenvalue problems for every parameter set, the proposed method requires only a one-time calculation of forced problems to generate stability maps for infinite structural parameter combinations.
• Physical Insight: The study provides a clear classification of instability mechanisms in multi-body VIV, distinguishing between fluid-driven and structure-driven modes and explaining the transition from coupled to decoupled dynamics as spacing increases.
• Engineering Application: The framework is directly applicable to the design of energy harvesting devices (e.g., VIVACE concepts) where optimizing mass, damping, and spacing is crucial. It allows engineers to rapidly screen optimal parameters without expensive full-scale simulations.
• Future Outlook: The authors suggest extending this linear framework to nonlinear regimes (to capture limit cycles and energy harvesting efficiency) and exploring periodic configurations using Floquet-Bloch approaches.