A multiphase cubic MARS method for fourth- and higher-order interface tracking of two or more materials with arbitrary topology and geometry

This paper proposes a multiphase cubic MARS method that achieves high-order accuracy in both time and space for tracking interfaces of multiple materials with arbitrary topology and geometry by utilizing graph-based representations, adaptive marker distribution, and robust junction handling.

The Gist

Imagine you are a chef trying to track the movement of different ingredients in a giant, swirling pot of soup. You have oil, water, spices, and chunks of vegetables, all mixing and swirling together. Your goal is to keep a perfect, high-definition map of where every single ingredient is at every moment, even as they stretch, twist, and squish into bizarre shapes.

This is the problem of Interface Tracking (IT) in fluid dynamics. Scientists need to know exactly where one material ends and another begins.

The paper you provided introduces a new, super-smart method called the Multiphase Cubic MARS Method to solve this problem. Here is the breakdown using simple analogies:

1. The Problem: The "Messy Kitchen"

Existing methods (like VOF or Level-Set) are like trying to draw the boundary between oil and water using a pixelated grid or a blurry photo.

• The Issue: When ingredients meet at a "junction" (like three streams of water meeting at a point, or a sharp corner), these old methods get confused. They tend to round off sharp corners or accidentally create tiny holes (vacuums) or overlaps where materials shouldn't be.
• The Result: Over time, the simulation loses accuracy. A sharp corner becomes a blob, and the physics of the simulation become wrong.

2. The Solution: The "Smart String" (MARS)

The authors propose a new way to think about the boundaries. Instead of pixels, imagine the boundary between materials is made of flexible, magical strings.

• The Graph (The Skeleton): First, they build a skeleton of the situation. They identify all the "junctions" (where strings meet) and "kinks" (sharp corners). They treat these as fixed points on a map. Think of this as the skeleton of the shape.
• The Splines (The Muscle): Between these fixed points, they stretch out cubic splines. In math terms, these are smooth, curvy lines that fit perfectly.
  • Periodic Splines: For a closed loop (like a bubble), the string connects back to itself smoothly.
  • Not-a-Knot Splines: For a line that stops at a junction, the string fits perfectly without forcing a weird bend.

3. The Secret Sauce: "ARMS" (Adding and Removing Markers)

As the soup swirls, the strings stretch and squish. If you just let them stretch, some parts get too long (losing detail) and others get too short (wasting energy).

The paper introduces a strategy called ARMS (Adding and Removing Markers). Imagine these strings have tiny beads on them (markers).

• Stretching: If a part of the string stretches out too much (like a rubber band), the system automatically adds more beads to keep the detail sharp.
• Squishing: If a part gets squished together too tight, the system removes extra beads to save space.
• Curvature Awareness: The system is smart enough to know that a sharp curve needs more beads than a straight line. It adapts the density of beads based on how "curvy" the shape is at that exact moment.

4. Why This is a Big Deal

The authors claim their method is 4th, 6th, or even 8th-order accurate.

• Analogy: If old methods are like a low-resolution JPEG image that gets blurry when you zoom in, this new method is like a 4K vector graphic that stays perfectly crisp no matter how much you zoom in or how much the shape twists.
• Handling Complexity: It can handle any number of materials (2, 10, or 100) and any shape (circles, stars, weird blobs) without the computer getting confused or the simulation crashing.
• No Overlaps: It guarantees that materials never overlap or leave gaps, which is a common headache for other methods.

5. The "Yin Space" (The Mathematical Foundation)

To make all this work, the authors use a mathematical concept called Yin Sets.

• Analogy: Think of the universe as a giant puzzle. The "Yin Space" is a rulebook that says, "Every piece of the puzzle must be a solid, connected chunk with a clear edge." This rulebook prevents the computer from creating impossible shapes (like a line with a dot floating in the middle of nowhere), ensuring the simulation stays physically realistic.

Summary

The paper presents a new, highly accurate way to track moving liquids and solids.

• Old Way: Pixelated, blurry, gets confused at corners, creates holes.
• New Way (Cubic MARS): Uses smooth mathematical curves (splines) guided by a smart skeleton. It automatically adds or removes detail markers based on how curvy the shape is.
• Result: It can simulate complex, multi-material flows (like a raccoon made of 22 different fluids, or a piggy bank made of 15) with incredible precision, keeping sharp corners sharp and junctions perfect.

It's essentially upgrading the simulation from a "pixel art" game to a "high-definition vector" masterpiece.

Technical Summary

Here is a detailed technical summary of the paper "A multiphase cubic MARS method for fourth- and higher-order interface tracking of two or more materials with arbitrary topology and geometry."

1. Problem Statement

The paper addresses the Multiphase Interface Tracking (IT) problem, which involves determining the regions occupied by $N_p$ distinct materials (phases) moving under a given velocity field. While significant progress has been made for two-phase flows, tracking three or more phases remains a major challenge due to:

• Complex Topology and Geometry: Interfaces can contain junctions (where 3+ phases meet, e.g., T, Y, or X junctions) and kinks (non-smooth points or $C^1$ discontinuities).
• Limitations of Existing Methods:
  • Level-Set Methods: Struggle with junctions because a single signed distance function cannot represent the complex topology of a junction (which is not homeomorphic to a 1D manifold). They often suffer from first-order accuracy at junctions.
  • Volume-of-Fluid (VOF) Methods: Reconstruction at junctions is difficult. Many schemes rely on "onion-skin" models or material ordering, which can lead to large errors, overlaps, vacuums, or topological changes (e.g., an X-junction erroneously becoming two T-junctions) due to numerical diffusion.
  • Moment-of-Fluid (MOF) Methods: Generally limited to second-order accuracy and struggle with non-smooth interfaces like Y-junctions.
• High-Order Accuracy: Existing explicit front-tracking methods often achieve only second-order accuracy. Achieving fourth-order or higher accuracy requires handling the pairing of smoothly connected curve segments at junctions, which is non-trivial.

2. Methodology: The Multiphase Cubic MARS Method

The authors propose a Multiphase Cubic MARS (Mapping and Adjusting Regular Semianalytic sets) method. The core philosophy is to treat topological and geometric problems using tools from topology and geometry rather than converting them solely into differential equations.

A. Mathematical Framework: Yin Space

• The materials are modeled as Yin sets (regular open semianalytic sets with bounded boundaries).
• Theorem 2.2 establishes that the boundary of any connected Yin set can be uniquely represented by a finite set of oriented Jordan curves.
• The interface is represented by a poset (partially ordered set) of these curves, capturing both the topology (inclusion relationships) and geometry.

B. Boundary Representation (Static)

To handle arbitrary topology and geometry without overlaps or vacuums:

1. Interface Graph ($G_\Gamma$): The interface is discretized into a planar graph where vertices are junctions (T, Y, X, etc.) and non-smooth points (kinks), and edges are curve segments.
2. Separation of Topology and Geometry:
   • Topology: Encoded in the graph structure and directed cycles representing the boundary of each phase.
   • Geometry: Approximated by cubic splines.
3. Spline Strategy:
   • Periodic Cubic Splines: Used for smooth closed curves (circuits in the graph).
   • Not-a-Knot Cubic Splines: Used for smooth curve segments (trails) that connect at junctions or kinks.
   • Algorithm 1 (Edge Partitioning): A critical algorithm partitions the graph edges into circuits (closed loops) and trails (open segments) based on a "smoothness indicator" (edge pairing). This ensures that a single spline is fitted across a smooth junction, preventing the overlaps and vacuums that occur when fitting independent splines to each phase.

C. Interface Evolution (Dynamic)

The method evolves the interface using a generalized ARMS (Adding and Removing Markers Strategy):

1. Flow Map: The interface markers are advected by the flow map $\phi$.
2. Marker Management (ARMS):
   • Adaptive Density: The distance between markers ($h_L$) is adjusted based on local curvature. High-curvature regions get denser markers to maintain uniform error distribution.
   • Regularity: The method enforces $(r, h)$-regularity, ensuring the ratio of maximum to minimum chordal lengths remains within a user-specified range.
   • Characteristic Markers: Crucially, junctions and kinks are treated as "characteristic points" that are never removed during the ARMS process. This preserves the topological structure of the interface graph throughout the simulation.
3. Reconstruction: After advection and marker adjustment, new cubic splines are fitted to the updated markers.

3. Key Contributions

1. Mathematical Model: A rigorous framework using Yin sets and interface graphs to represent an arbitrary number of materials with arbitrary topology (including complex junctions) and geometry.
2. Extension of MARS: Generalization of the MARS framework from two-phase flows to multiphase flows, specifically handling junctions and kinks.
3. High-Order Accuracy: The method achieves fourth-, sixth-, and eighth-order accuracy in both space and time. This is a significant improvement over the typical second-order accuracy of VOF/MOF methods and the first-order accuracy of some level-set methods at junctions.
4. Topological Preservation: By separating topology (graph/cycles) from geometry (splines) and protecting characteristic markers, the method guarantees that junctions and phase connectivity are preserved exactly, preventing the creation of overlaps or vacuums.
5. Curvature-Adaptive Strategy: A novel curvature-based ARMS strategy that dynamically adjusts marker density to resolve high-curvature arcs efficiently while maintaining optimal computational complexity ($O(1/h_L)$).

4. Results

The authors validated the method through extensive benchmark tests:

• Vortex Shear Tests: Simulations of circular disks divided into 3, 5, and 6 phases under reversing vortex flows.
  • Accuracy: Demonstrated convergence rates of 4, 6, and 8 depending on the choice of $h_L$ (spatial resolution) and time integrator order.
  • Comparison: The proposed method was orders of magnitude more accurate than state-of-the-art VOF and MOF methods (e.g., LVIRA, Power Diagram, standard MOF). For instance, in a 3-phase test, MARS errors were $\sim 10^{-9}$ while VOF methods were $\sim 10^{-2}$.
• Complex Geometries:
  • "Piggy" Test: 15 phases with complex boundaries.
  • "Raccoon" Test: 22 phases including a lemniscate (figure-8) shape.
  • Results: The method successfully tracked all phases, preserving complex topological structures (like the self-intersecting lemniscate) and handling high-curvature regions without generating spurious overlaps or vacuums.
• Efficiency: CPU time analysis confirmed linear complexity with respect to the number of markers, making the method computationally efficient and scalable.

5. Significance

This work represents a significant advancement in computational fluid dynamics for multiphase flows:

• Solves the Junction Problem: It provides a robust, high-order solution for tracking interfaces with complex junctions (T, Y, X), a long-standing challenge for VOF and Level-Set methods.
• High Fidelity: The ability to achieve 4th-to-8th order accuracy makes it suitable for simulations requiring extreme precision, such as studying contact line dynamics or droplet impacts.
• Versatility: The method is independent of material ordering and can handle arbitrary numbers of phases with complex, evolving geometries.
• Future Impact: The framework lays the groundwork for future extensions to handle topological changes (merging/splitting) and coupling with high-order flow solvers for incompressible multiphase simulations.

In summary, the Multiphase Cubic MARS method successfully bridges the gap between topological rigor and geometric accuracy, offering a superior alternative to existing interface tracking techniques for complex multiphase systems.