Particle, kinetic and hydrodynamic models for sea ice floes. Part II: Rotating floes with nonlinear contact forces

This paper extends a multiscale modeling framework for sea-ice floes by generalizing the particle, kinetic, and hydrodynamic descriptions to include rotational dynamics and nonlinear contact forces, thereby deriving a comprehensive system of macroscopic equations that capture complex stress, transport, and dissipative mechanisms in sea-ice rheology.

The Gist

Imagine the Arctic Ocean not as a solid sheet of ice, but as a giant, chaotic dance floor filled with thousands of floating ice chunks, or "floes." These aren't just sliding around; they are bumping, spinning, grinding against each other, and being pushed by the wind and ocean currents.

This paper is the second part of a study (Part I covered the simpler case where the ice chunks didn't spin) that tries to build a mathematical "rulebook" to predict how this icy dance floor behaves. The authors, Deng, Ha, and Lee, created a three-layered system to model this, moving from the tiny details to the big picture.

Here is the breakdown of their work using simple analogies:

1. The Three Layers of the Model

Think of understanding a crowd of people at a concert. You can look at it in three ways:

• Layer 1: The Particle Model (The Individual Dancers)

  • What it is: This looks at every single ice chunk individually.
  • The New Twist: In this paper, they realized that ice chunks aren't just sliding blocks; they are spinning tops. When two chunks bump, they don't just bounce; they might spin faster, slow down, or change direction because of the friction of the bump.
  • The Physics: They used complex rules (like Hertzian contact mechanics) to describe how hard the chunks hit each other. It's like calculating exactly how much a billiard ball spins when it hits another one at an angle, but with ice and water drag involved.

• Layer 2: The Kinetic Model (The Crowd Density)

  • What it is: Instead of tracking 10,000 individual dancers, this layer asks: "If I look at a specific spot on the dance floor, what is the average speed and spin of the people there?"
  • The Analogy: Imagine a heat map. Instead of seeing Person A and Person B, you see a "cloud" of probability showing where people are likely to be and how fast they are moving on average. This layer bridges the gap between the messy individual collisions and the smooth flow of the whole crowd.

• Layer 3: The Hydrodynamic Model (The River Flow)

  • What it is: This is the "big picture" view. It treats the entire pack of ice like a thick, slow-moving fluid (like honey or a river).
  • The Result: This gives us equations that predict the overall movement of the ice sheet without needing to know where every single chunk is. It tells us how the ice mass moves, how much momentum it has, and how it spins as a whole.

2. The "Secret Sauce": Rotation and Friction

The biggest breakthrough in this paper is adding rotation and non-linear friction.

• The Spinning Top Effect: In the old models, ice chunks were like sliding pucks. In reality, when a chunk hits another, it often starts to spin. This paper shows that this spin matters. If the ocean current is swirling (has "vorticity"), it tries to make the ice chunks spin to match it. If the chunks are spinning fast, they collide differently than if they were just sliding.
• The Bumpy Road: The authors added rules for "non-linear contact." Imagine driving a car. If you hit a bump gently, it's a small jolt. If you hit it hard, the suspension compresses differently. Similarly, when ice chunks collide hard, the force isn't just a simple "bounce"; it involves complex compression and friction that eats up energy (dissipation).

3. What Did They Prove?

The authors didn't just write equations; they proved that their model makes physical sense:

• Energy Loss: They showed that because of these bumpy, spinning collisions, the system naturally loses energy over time (just like a spinning top eventually stops). The ice pack settles down.
• Alignment: If the ocean current is steady, the ice chunks will eventually stop spinning wildly and start moving in the same direction as the water, like a school of fish swimming in unison.
• Consistency: They ran computer simulations to prove that if you take the "Individual Dancer" model and average it out, you get the exact same result as the "River Flow" model. This confirms that their math connects the tiny scale to the huge scale perfectly.

4. Why Does This Matter?

Sea ice is crucial for our planet's climate. It reflects sunlight and insulates the ocean. However, as the ice melts, it breaks into smaller, more chaotic chunks (the "Marginal Ice Zone").

• The Problem: Old models treated the ice like a solid sheet or a simple fluid, which fails when the ice is broken into millions of spinning, colliding pieces.
• The Solution: This new framework allows scientists to predict how this fragmented ice will move and interact with storms and currents much more accurately.
• The Future: The authors hint at the next steps: adding temperature (melting and freezing) and breaking (ice chunks shattering into smaller pieces). This would make the model even more realistic, helping us predict climate change impacts with greater precision.

Summary

Think of this paper as upgrading a video game physics engine.

• Old Version: Ice chunks were sliding blocks that bounced off each other.
• New Version (This Paper): Ice chunks are spinning, tumbling, friction-heavy objects that collide, lose energy, and eventually align with the ocean currents.

The authors built a mathematical bridge that connects the chaotic dance of individual ice chunks to the smooth, predictable flow of the entire ice sheet, giving us a much clearer picture of how our polar regions are changing.

Technical Summary

1. Problem Statement

Sea ice dynamics, particularly in the Marginal Ice Zone (MIZ), are characterized by fragmented ice floes interacting through complex mechanical processes. While continuum models (e.g., viscous-plastic) work for compact ice packs, they fail to capture granular behaviors like collisions, ridging, and fracturing in fragmented regimes.

Previous work (Part I of this series) established a multiscale framework for non-rotating floes with linear contact forces. However, real sea ice floes possess rotational degrees of freedom, and their interactions involve nonlinear contact mechanics (Hertzian normal forces, velocity-dependent restitution, and tangential friction). These factors generate torques and complex energy dissipation mechanisms that the non-rotating model cannot capture.

The core challenge addressed in this paper is: How to rigorously extend the particle–kinetic–hydrodynamic hierarchy to include rotational dynamics and nonlinear contact forces to provide a more realistic, multiscale description of sea ice rheology.

2. Methodology

The authors employ a systematic multiscale modeling approach, deriving equations at three distinct scales:

A. Particle Scale (Discrete Element Model)

• Model: Floes are modeled as rigid cylindrical bodies with position $x_i$, linear velocity $v_i$, orientation $\theta_i$, angular velocity $\omega_i$, radius $r_i$, and thickness $h_i$.
• Governing Equations: Newton's equations of motion for translation and rotation.
  • Forces: Include ocean drag (quadratic in relative velocity) and contact forces.
  • Contact Mechanics:
    • Normal Force: Based on Hertz contact theory, incorporating nonlinear elastic deformation and velocity-dependent damping (restitution).
    • Tangential Force: Modeled via Coulomb friction laws, limiting tangential force relative to normal force.
    • Torques: Generated by tangential forces and ocean drag, coupling translational and rotational motions.
• Analysis: The authors derive balance laws for total linear momentum, total angular momentum (orbital + spin), and total energy. They prove that in the absence of external forcing, the system is dissipative due to collision-induced energy loss.

B. Kinetic Scale (Mean-Field Limit)

• Derivation: Using the BBGKY hierarchy (Bogoliubov–Born–Green–Kirkwood–Yvon) and the molecular chaos assumption (Stosszahlansatz), the authors derive a Vlasov-type kinetic equation for the one-particle distribution function $F(t, x, v, \theta, \omega, r, h)$.
• Phase Space: The phase space is extended to include angular velocity and orientation.
• Equation: The resulting equation describes the evolution of the probability density, accounting for:
  • Advection in physical and orientation space.
  • Nonlinear interaction terms representing the mean-field effect of collisions (stress and torque).
  • External forcing from ocean drag.
• Analysis: Macroscopic moments of the kinetic equation are analyzed to establish conservation of mass and balance laws for momentum and energy.

C. Hydrodynamic Scale (Continuum Limit)

• Closure: The kinetic equation is closed using a monokinetic ansatz, assuming that all particles at a given location share the same mean velocity $u$ and mean angular velocity $\bar{\omega}$.
• Resulting System: A set of macroscopic partial differential equations (PDEs) for:
  • Mass density ($\rho$).
  • Linear momentum ($\rho u$).
  • Angular momentum ($\hat{\omega}$).
  • Total energy.
• Stress Terms: The derivation explicitly identifies additional stress contributions (stress and couple-stress tensors) arising from the nonlinear contact interactions and rotational dynamics.

3. Key Contributions

1. Extension to Rotational Dynamics: The framework is generalized from non-rotating to rotating rigid bodies, introducing angular velocity and orientation as state variables.
2. Nonlinear Contact Mechanics: The model incorporates physically realistic Hertzian normal forces and Coulomb friction, moving beyond linear spring-dashpot models. This captures the nonlinear dependence of contact forces on deformation and relative velocity.
3. Rigorous Multiscale Derivation: The paper provides a mathematically rigorous derivation of the hierarchy:
   • Particle $\to$ Kinetic (via mean-field limit).
   • Kinetic $\to$ Hydrodynamic (via moment closure).
4. Energy and Momentum Analysis:
   • Proved that total linear and angular momentum are balanced by ocean drag, while internal contact forces conserve momentum (action-reaction).
   • Demonstrated that total energy is dissipative due to inelastic collisions and friction, establishing lower bounds and asymptotic convergence to equilibrium states.
5. Asymptotic Behavior: Proved that under constant ocean forcing, particle velocities converge to the ocean velocity, and angular velocities converge to zero (for irrotational flow) or the local ocean vorticity (for rotational flow).

4. Results

The theoretical findings were validated through numerical simulations:

• Example 1 (Constant Ocean Forcing):
  • Simulated 100 rotating floes in a periodic domain with constant ocean velocity.
  • Observations:
    • Translational velocities converged to the ocean velocity.
    • Angular velocities decayed to zero (due to zero ocean vorticity).
    • Total energy dissipated over time, with bursts of strain energy corresponding to collision events.
    • Momentum balance laws held true, with total momentum relaxing to the ocean-drift momentum.
• Example 2 (Consistency Test):
  • Compared the discrete particle model (10,000 floes) with the continuum hydrodynamic model (4.15) under a spatially varying rotational ocean flow.
  • Observations:
    • The coarse-grained particle statistics (density, velocity, angular velocity) matched the continuum solution closely.
    • This validated the particle–kinetic–hydrodynamic hierarchy, confirming that the macroscopic equations accurately reproduce the statistical behavior of the underlying particle system.

5. Significance

• Improved Rheology: This work provides a more physically consistent framework for modeling sea ice rheology in fragmented regimes, where granular effects and rotation are dominant.
• Bridging Scales: It offers a systematic pathway to couple discrete particle models (used for detailed local interactions) with continuum models (used for large-scale climate simulations), facilitating data assimilation and improved predictions.
• Physical Realism: By including nonlinear contact forces and torques, the model captures phenomena such as spin alignment, rotational clustering, and collision-induced jamming, which are critical for understanding ice pack mechanics but were previously neglected in analytical multiscale frameworks.
• Future Directions: The authors identify coupling with thermodynamics (melting/freezing) and handling variable particle numbers (fracture/aggregation) as the next critical steps for a fully comprehensive sea ice model.

In summary, this paper successfully extends the mathematical theory of sea ice dynamics to include rotation and nonlinear interactions, providing a robust theoretical foundation for next-generation sea ice modeling.