Side-wall wetting and linear stability of falling films

This study employs a biglobal stability framework to demonstrate that side-wall wetting exerts a dual influence on falling film stability, acting as a relative destabilizing mechanism in confined channels by weakening boundary-layer stabilization, while simultaneously providing significant long-wave stabilization in weakly-confined channels through enhanced capillary anchoring.

The Gist

Imagine a thin sheet of water sliding down a tilted windowpane. In the world of physics, this is called a "falling film." Usually, if the window is very wide, the water flows smoothly until it gets fast enough to start rippling and breaking apart. Scientists have known for a long time how to predict when this happens on a wide, open surface.

But what happens if you put that same water in a narrow gutter or a channel with walls on the sides? And what if the water likes to "stick" to those walls a little bit (a phenomenon called wetting)?

This paper, written by Mohamed and Sesterhenn, explores exactly that. They built a sophisticated mathematical model to see how the side walls and the water's tendency to climb up them (like a tiny mountain range of water at the edges) change the rules of stability.

Here is the story of their findings, broken down into simple concepts:

The Two Main Characters: The Walls and the Water's "Stickiness"

1. The Walls (Confinement): When water flows in a narrow channel, the walls act like a brake. The water right next to the wall slows down due to friction, creating a "cushion" of slow-moving fluid. This cushion usually helps stabilize the flow, stopping ripples from growing too fast.
2. The Stickiness (Wetting): Water doesn't just hit the wall and stop; it often curves up the side, forming a little hill or "meniscus." Because the water is thicker at the edges, gravity pulls it faster there, creating a speed bump right next to the wall.

The authors discovered that these two characters play a very different game depending on how wide the channel is.

Scenario A: The Narrow Gutter (Confined Channels)

The Setup: Imagine a relatively narrow channel where the walls are close enough that their "braking effect" (the slow-moving cushion) is strong.

The Surprise: In this narrow setting, the water's "stickiness" actually makes things worse.

• The Analogy: Think of the wall's braking effect as a team of people holding a rope to stop a runaway cart. The "stickiness" of the water is like a gust of wind that pushes the cart faster right next to the people holding the rope.
• What Happens: The water climbing up the side (wetting) creates a speed bump (velocity overshoot) that thins out the braking cushion. This weakens the walls' ability to stop the ripples. So, in a narrow channel, wetting acts like a villain, making the flow unstable sooner than it would be otherwise.

Scenario B: The Wide River (Weakly-Confined Channels)

The Setup: Now, imagine a very wide channel where the walls are so far away that their braking effect is barely noticeable in the middle. The flow behaves mostly like it's on an open, infinite surface.

The Surprise: Here, the water's "stickiness" becomes a hero.

• The Analogy: Imagine the water at the edges is like a tight rubber band anchoring the whole sheet of water. Even though the walls are far away, the "stickiness" pulls the edges down tight.
• What Happens: This anchoring effect makes it much harder for long, slow ripples to start. It's as if the water is being "tensioned" or tightened by the walls. This pushes the point of instability to much higher speeds. In this wide setting, wetting acts as a stabilizer, keeping the flow smooth for longer.

The "Phase Diagram": Finding the Switch

The authors created a map (a phase diagram) to show where the switch happens.

• If the channel is narrow, wetting is a troublemaker (destabilizing).
• If the channel is wide, wetting is a protector (stabilizing).
• There is a smooth transition zone in between where the behavior shifts from one to the other.

Did They Check the Real World?

Yes. The authors compared their mathematical predictions with real-world experiments done by other scientists using glycerol-water mixtures.

• The Result: Their model matched the real-world data very well. When the experiments showed that wetter surfaces made the flow more stable in wide channels, the math predicted the exact same thing.

The "Secret Sauce": What the Water Looks Like Inside

To understand why this happens, they looked at the invisible swirls and movements inside the water (eigenmodes).

• In the Narrow Channel: The wetting creates little whirlpools right near the walls. These swirls mess up the smooth braking effect, making the flow chaotic.
• In the Wide Channel: The water at the edges acts like a strong anchor. The ripples try to wiggle, but the "anchored" edges hold them back, preventing the instability from growing.

Summary

In short, this paper tells us that context is everything.

• In a narrow channel, the water sticking to the walls destabilizes the flow by weakening the natural friction that usually keeps it calm.
• In a wide channel, that same sticking effect stabilizes the flow by acting like a tight anchor against the walls.

The authors successfully built a mathematical tool that explains this complex dance between the shape of the channel, the water's speed, and how much the water likes to hug the walls.

Technical Summary

Technical Summary: Side-wall wetting and linear stability of falling films

Problem Statement
Gravity-driven liquid films flowing down inclined surfaces are inherently unstable, a phenomenon classically described by the interplay of inertia, viscosity, and gravity. While the stability of unbounded films is well-understood, the behavior of films confined in the spanwise direction introduces complexities related to geometric confinement and side-wall interactions. Previous theoretical work (Mohamed et al. 2023) established that spanwise confinement stabilizes moderate-wavenumber perturbations via viscous boundary layers, identifying distinct hydrodynamic (H-mode) and wall-confined (W-mode) stability regimes. However, these models neglected side-wall wetting effects. Experimental studies (Vlachogiannis et al. 2010; Georgantaki et al. 2011; Pollak et al. 2011) have indicated that wetting alters the local interface shape, creating a meniscus and a streamwise velocity overshoot near the walls, which can shift instability thresholds. A comprehensive theoretical framework linking these wetting effects to the stability of confined films has been lacking. This work addresses that gap by investigating how side-wall wetting influences the linear stability of falling films in both confined and weakly-confined regimes.

Methodology
The authors develop a biglobal linear stability framework based on the Navier–Stokes equations, incorporating inertia, viscosity, gravity, capillarity, and geometric confinement.

• Base State: The steady-state solution accounts for the curved meniscus and the resulting spanwise variation in film thickness and velocity. The contact line singularity at the side walls is resolved using a Navier slip condition, where the tangential velocity is proportional to the local viscous stress. The physical domain, defined by the curved free surface $\bar{h}(z)$, is mapped to a rectangular computational domain using Chebyshev spectral collocation methods.
• Linear Stability Analysis: The flow is decomposed into a steady base state and infinitesimal perturbations. A temporal biglobal stability analysis is performed, where perturbations are assumed to vary as $\exp(ikx - i\omega t)$. The contact angle is prescribed as a static value in the base state, while the perturbation of the contact angle is set to zero (free-edge condition) at the leading order.
• Parameters: The study varies the confinement ratio ($W/H$), the contact angle ($\theta$), and the ratio of channel width to capillary length ($W/l_c$). Two limiting regimes are examined: "confined" channels ($W/H = 20$) where viscous boundary layers are dynamically significant, and "weakly-confined" channels ($W/H = 100$) where confinement effects are negligible compared to the unconfined 1D limit.

Key Results
The study reveals that the influence of side-wall wetting on stability is regime-dependent and often counter-intuitive:

1. Confined Channels ($W/H = 20$):

   • In the absence of wetting, spanwise confinement provides significant stabilization at moderate wavenumbers due to viscous boundary layers.
   • Introducing wetting (decreasing contact angle) weakens this confinement-induced stabilization. As the contact angle decreases, the neutral stability curves shift toward the unconfined 1D limit.
   • Mechanism: Wetting induces a capillary elevation (meniscus) and a corresponding streamwise velocity overshoot near the side walls. This overshoot reduces the thickness of the stabilizing viscous boundary layer and introduces near-wall vortical structures that reduce effective viscous damping. Thus, wetting acts as a relative destabilizing mechanism in confined channels.

2. Weakly-Confined Channels ($W/H = 100$):

   • In the absence of wetting, the stability behavior mirrors the classical 1D unconfined case, with no significant confinement-induced stabilization.
   • Introducing wetting produces a net stabilization of long-wave perturbations ($k \to 0$), significantly increasing the critical Reynolds number.
   • Mechanism: In wide channels, the velocity overshoot is localized and negligible for stability. Instead, the dominant effect is the "tensioning" of the interface through strong anchoring of perturbations at the side walls. This stabilizing effect strengthens as the contact angle decreases, representing a competition between destabilizing inertia and stabilizing wetting-induced capillary forces.

3. Transition and Scaling:

   • A phase diagram of the critical Reynolds number shift versus $W/H$ and wavenumber $k$ illustrates a smooth transition between the relative destabilization (moderate $k$ in confined channels) and long-wave stabilization (low $k$ in weakly-confined channels).
   • The long-wave stabilization threshold is shown to depend on the Kapitza number ($Ka$) and the ratio $W/l_c$. Theoretical predictions for the critical Reynolds number as a function of $Ka$ show consistent trends and magnitudes with experimental data from Georgantaki et al. (2011).

4. Eigenmode Structures:

   • In the destabilizing confined case, perturbation eigenmodes exhibit intense structures within the meniscus region, linked to the velocity overshoot.
   • In the stabilizing weakly-confined case, interface perturbations are strongly anchored at the walls, while velocity perturbations remain concentrated in the channel core.

Significance and Claims
The paper claims to provide the first theoretical and numerical framework that explicitly resolves the interaction between side-wall wetting and spanwise confinement in falling film stability. The primary contribution is the identification of a dual role for wetting: it acts as a relative destabilizer in geometrically confined channels by undermining the viscous boundary layer stabilization, yet acts as a stabilizer in weakly-confined (wide) channels by suppressing long-wave instabilities through interface anchoring.

The authors emphasize that their quantitative predictions align with existing experimental observations, particularly the shift in instability thresholds observed in wide channels with strong wetting. The work clarifies that the "destabilizing" velocity overshoot observed in experiments is secondary to the dominant stabilizing capillary effects in wide channels, whereas in narrow channels, the overshoot becomes the primary mechanism reducing stability. The study concludes that a complete understanding of falling film stability requires accounting for the competition between geometric confinement, inertia, and wetting-induced capillary forces.