HOC simulations of miscible viscous fingering of a finite slice: A new insight

This study employs high-order compact finite difference simulations to demonstrate that while the onset of viscous fingering in miscible slices is independent of boundary conditions, long-time solute mixing and instability growth are significantly enhanced by permeable boundaries, offering new insights for chromatography separation processes.

The Gist

The Big Picture: A Race Through a Sponge

Imagine you have a giant, thick sponge (this represents porous media, like soil or rock underground). You are trying to push a specific blob of liquid through this sponge.

• The Blob: A finite slice of liquid (like a drop of dye or a specific chemical).
• The Pusher: A different liquid pushing it from behind.
• The Problem: The two liquids mix easily (they are miscible), but they have different "thicknesses" (viscosities). One is runny like water; the other is thick like honey.

When a runny liquid pushes a thick liquid, or vice versa, the interface between them doesn't stay smooth. It gets wobbly and starts growing little fingers that reach out and invade the other side. This is called Viscous Fingering.

This paper asks a simple but crucial question: Does the shape of the container (the boundaries) change how these "fingers" grow?

The Three Scenarios (The Boundaries)

The researchers simulated this process on a computer to see what happens under three different "rules" for the sides of the sponge:

1. The "Infinite Loop" (Periodic): Imagine the sponge is wrapped around a cylinder. If a finger hits the right edge, it instantly reappears on the left edge. It's like a video game world where you walk off the screen and come back on the other side.
2. The "Solid Wall" (Impermeable): The sides of the sponge are sealed tight with solid walls. Nothing can get in or out. The liquid can't flow sideways, and the dye can't leak out.
3. The "Sieve" (Permeable): The sides are like a sieve or a screen. The liquid can flow through the sides (entering or leaving), but the dye particles themselves cannot leak out. It's like a crowd of people (the fluid) moving through a room with open doors, but the specific people wearing red shirts (the dye) are told they can't leave the room, even though the crowd can.

The Key Discovery: The "Magic" of the Sieve

The researchers found that at the very beginning, all three scenarios look the same. The fingers start growing at the same speed and in the same pattern. It's like three runners starting a race; for the first few seconds, they are all side-by-side.

However, as time goes on, the "Sieve" (Permeable) scenario becomes the wild card.

• In the "Solid Wall" and "Infinite Loop" cases: The amount of liquid and dye stays exactly the same. The fingers grow, mix, and eventually slow down as they get diluted.
• In the "Sieve" case: Something surprising happens. Because the liquid can flow in and out of the sides, the system actually gains mass (more liquid enters the system than leaves).

The Analogy:
Think of the "fingers" as a group of runners trying to break through a barrier.

• In the Solid Wall scenario, the runners are tired and running out of energy. They hit the wall and stop.
• In the Sieve scenario, it's like the runners are being handed fresh energy drinks (extra mass) from the sides as they run. This extra energy keeps them strong. They don't just stop at the barrier; they smash right through it, creating much longer, more chaotic fingers and mixing the liquids much more thoroughly.

Why Does This Matter?

The paper concludes that while the start of the instability doesn't care about the walls, the long-term behavior depends entirely on them.

1. Mixing: The "Sieve" boundaries cause the most mixing. The liquids blend together much faster and more completely because the extra mass keeps the instability alive.
2. Real World Applications: This isn't just about math; it's about real life.
   • Oil Recovery: If you are trying to push oil out of a rock formation, knowing if the rock has "leaky" sides helps you predict how much oil you can get out.
   • Pollution: If a toxic spill happens in the ground, "leaky" boundaries might mean the pollution spreads much further and mixes more dangerously than we thought.
   • Chromatography: This is a lab technique used to separate chemicals (like in drug testing). If the boundaries aren't controlled, the separation might fail because the chemicals mix too much.

The Bottom Line

The researchers used a super-accurate computer method (like a high-definition camera) to watch this process. They found that boundaries matter.

If you assume the sides are sealed (which is common in simple models), you might underestimate how much mixing will happen in the real world if those sides are actually permeable. The "leaky" boundaries act like an engine, feeding the instability and causing the fluids to mix in a much more dramatic and complex way than previously thought.

Technical Summary

1. Problem Statement

The study investigates the dynamics of miscible viscous fingering (VF) occurring when a finite-width slice of fluid (a "sample") is displaced by another fluid within a homogeneous, isotropic porous medium.

• Physical Context: Unlike traditional semi-infinite displacement problems, this study focuses on a finite sample embedded in a displacing fluid. The viscosity of the system depends on the solute concentration, creating a viscosity contrast ($\mu(c) = e^{Rc}$) that drives hydrodynamic instability.
• Key Variable: The log-mobility ratio, $R = \ln(\mu_2/\mu_1)$, determines the stability:
  • $R > 0$: A more viscous sample is displaced by a less viscous fluid (instability at the rear interface).
  • $R < 0$: A less viscous sample is displaced by a more viscous fluid (instability at the frontal interface).
• Research Gap: Previous studies often relied on periodic boundary conditions (Type-I) due to the limitations of spectral methods. This restricts the ability to model realistic physical scenarios where lateral boundaries may be impermeable (no-flow) or permeable (allowing flow but no solute diffusion). The paper aims to systematically analyze how transverse boundary conditions affect the onset, growth, and long-term mixing of viscous fingers.

2. Methodology

The authors employ a high-fidelity numerical approach to solve the governing equations.

• Governing Equations:
  • Flow: Governed by incompressible Darcy's law coupled with a solute transport equation (advection-diffusion).
  • Formulation: The system is transformed into a stream-function ($\psi$) and vorticity/concentration ($c$) formulation to handle the two-way coupling between velocity and concentration fields.
• Numerical Scheme:
  • Spatial Discretization: A fourth-order accurate Higher-Order Compact (HOC) finite difference method is used. This provides high accuracy on uniform grids and allows for the use of non-periodic boundary conditions.
  • Temporal Integration: The Crank-Nicolson technique is used for time integration, ensuring second-order temporal accuracy.
  • Solver: The resulting coupled nonlinear algebraic systems are solved iteratively using the BiConjugate Gradient Stabilized (BiCGStab) method with an Incomplete LU (ILU) preconditioner.
• Boundary Conditions Investigated:
  1. Type-I (Periodic): Standard periodic conditions in the transverse direction.
  2. Type-II (Impermeable): Zero normal velocity ($\psi=0$) and zero solute flux ($\nabla c \cdot \mathbf{n} = 0$) at lateral boundaries.
  3. Type-III (Permeable): Allows non-zero normal velocity (flow through boundaries) but maintains zero diffusive solute flux ($\nabla c \cdot \mathbf{n} = 0$).
• Validation: The code was validated against existing literature (Mishra et al., 2008) for periodic conditions and grid independence tests confirmed that a $1025 \times 129$ grid is sufficient for convergence.

3. Key Contributions

• Novel Boundary Analysis: This is the first systematic study comparing periodic, impermeable, and permeable transverse boundaries for finite slice miscible displacement.
• Discovery of Mass Enhancement: The study identifies that permeable boundaries (Type-III) allow for a net increase in solute mass within the domain over time, a phenomenon absent in periodic or impermeable cases.
• Mechanism of Instability Amplification: The authors demonstrate that the mass accumulation in Type-III boundaries sustains higher concentration gradients, which in turn maintains a higher viscosity contrast, leading to stronger and more persistent fingering instabilities.
• Non-Monotonic Interfacial Dynamics: The paper reveals that permeable boundaries lead to non-monotonic evolution of interfacial length (peaks and valleys), contrasting with the monotonic decay seen in other boundary types.

4. Key Results and Findings

A. Onset and Early-Time Behavior

• The onset time of viscous fingering and the initial pattern formation are independent of the boundary condition type.
• The number of fingers formed is determined by the initial perturbation and physical parameters, not the lateral boundaries.

B. Long-Time Behavior and Breakthrough

• Breakthrough Time ($t_{bk}$): The time when fingers from the unstable interface interact with the stable interface.
  • For $R > 0$ (viscous sample), fingers grow upstream and hit the stable front.
  • For $R < 0$ (less viscous sample), fingers grow downstream and hit the stable rear.
• Post-Breakthrough Dynamics:
  • Type-I & Type-II: The stable interface acts as a barrier. Fingers cannot penetrate it; instead, they reorient and move upstream (for $R>0$) or downstream (for $R<0$) along the barrier. Mixing is limited by diffusion and reorientation.
  • Type-III (Permeable): The stable interface fails to act as a barrier. Fingers penetrate through the stable interface. This is attributed to the mass enhancement caused by the permeable boundaries, which keeps the concentration gradient high, sustaining the instability.

C. Quantitative Metrics

• Mass Balance:
  • Type-I and Type-II: Total solute mass is conserved.
  • Type-III: Solute mass increases over time due to advective flux across the lateral boundaries.
• Mixing Length:
  • Type-III boundaries result in the largest mixing lengths for both forward and backward directions. The enhanced instability allows the sample to spread significantly further than in periodic or impermeable cases.
• Interfacial Length ($I(t)$):
  • Type-I/II: $I(t)$ decays monotonically after breakthrough due to finger merging and dilution.
  • Type-III: $I(t)$ exhibits non-monotonic behavior (oscillations of peaks and valleys). Peaks correspond to transient interface growth driven by localized mass accumulation, while valleys correspond to finger coalescence.
• Skewness:
  • Less viscous samples ($R<0$) show right-skewed distributions (faster downstream spread).
  • More viscous samples ($R>0$) show left-skewed distributions.
  • Periodic boundaries (Type-I) tend to reduce skewness compared to the other types.

5. Significance and Implications

• Chromatography: The findings are critical for chromatographic separation, where finite samples are displaced through porous media. The study suggests that boundary permeability can significantly alter band broadening and separation efficiency.
• Environmental Remediation: In groundwater contamination scenarios, the assumption of impermeable boundaries (common in models) may underestimate the spread of contaminants if the actual geological formation allows for lateral flow (permeable boundaries).
• Methodological Advancement: The successful application of HOC methods to these complex, non-periodic boundary problems provides a robust framework for future studies of hydrodynamic instabilities in realistic geometries.

Conclusion: The study concludes that while boundary conditions do not affect the initiation of viscous fingering, they fundamentally alter the long-term evolution, mixing efficiency, and mass transport of the system. Specifically, permeable boundaries create a feedback loop of mass accumulation and gradient maintenance that leads to significantly stronger instabilities and broader spreading compared to traditional periodic or impermeable models.