Inertial effects on the interphase drag force and rheology of dilute suspensions of buoyant droplets at low Reynolds number

This study employs the reciprocal theorem to demonstrate that inertial effects in dilute suspensions of buoyant droplets at low Reynolds numbers introduce quadratic dependencies on relative velocity and velocity variance into the interphase drag force and effective stress of the continuous phase.

The Gist

Imagine a crowded room where everyone is trying to walk in the same direction, but some people are floating balloons (buoyant droplets) and others are the air (the continuous fluid). Usually, when we study how these balloons move through the air, we pretend the air is perfectly still and the balloons move very slowly, like a snail in a thick syrup. In that slow world, the rules are simple: the faster the balloon moves, the harder the air pushes back.

However, in the real world, things aren't always that slow or that simple. Sometimes the air has a little bit of "oomph" (inertia), and the balloons might be jiggling around a bit, not just moving in a straight line. This paper, written by Nicolas Fintzi and Jean-Lou Pierson, asks a specific question: What happens to the forces on these floating balloons when the air has a tiny bit of speed, and when the balloons are bouncing around with their own little bit of energy?

Here is the breakdown of their discovery, using everyday analogies:

1. The "Reciprocal Theorem" as a Magic Mirror

To solve this, the authors didn't just simulate every single drop of air and every balloon. That would be like trying to count every grain of sand on a beach to understand how the tide moves. Instead, they used a mathematical tool called the Reciprocal Theorem.

Think of this like a magic mirror. Instead of looking directly at the messy, complex reality of a balloon moving through slightly windy air, they looked at a "mirror image" of the problem where the rules are simpler (like a perfectly still room). By comparing the real problem to this simple mirror image, they could calculate the complex forces without doing all the heavy lifting. It's a shortcut that lets them see the hidden details of how the air pushes and pulls on the balloon.

2. The "Jitter" Matters (Velocity Variance)

In many old models, scientists assumed all the balloons were moving at the exact same speed. But in reality, some balloons might be drifting faster, some slower, and some might be wobbling up and down. This "jitter" or velocity variance is like a crowd of people walking; if everyone walks at the exact same pace, it's orderly. But if some are sprinting and others are strolling, the crowd creates a different kind of pressure.

The authors found that this "jitter" creates extra forces.

• The Drag Force: The air doesn't just push back based on the average speed of the balloons. It also pushes back based on how much the balloons are jittering around that average.
• The Stress (The "Squeeze"): When you look at the whole group of balloons, their jittering creates an extra "squeeze" or pressure on the air around them. It's like a crowd of people shuffling nervously; even if they aren't running, their fidgeting creates a sense of pressure in the room.

3. The "Speed Squared" Effect

One of the most important findings is how these forces behave when the balloons move faster.

• In the very slow, syrup-like world, the force is directly proportional to speed (double the speed, double the push).
• In this new, slightly faster world, the force starts to depend on the square of the speed.

Imagine pushing a shopping cart. If you push it gently, it's easy. If you push it twice as hard, it's not just twice as hard; the air resistance and the way the wheels interact with the floor make it feel much harder. The authors showed that for these floating droplets, the "push back" from the air grows much faster than the speed itself, and it also depends heavily on how much the droplets are jittering.

4. Why This Changes the "Recipe" for Fluids

The paper concludes that if you want to describe how a mixture of air and floating droplets behaves (like in a bubble column or a flotation tank), you can't just use the old, simple recipes.

• The Old Recipe: "Add viscosity (thickness) based on how many droplets there are."
• The New Recipe: "Add viscosity, but also add a term that depends on how fast the droplets are moving relative to the air, and another term that depends on how much they are jittering."

This means the mixture acts less like a simple thick liquid (like honey) and more like a smart material that changes its behavior depending on how fast and how chaotically the droplets are moving.

Summary

In short, Fintzi and Pierson used a clever mathematical mirror to show that when floating droplets move through a fluid with a little bit of speed:

1. Inertia matters: The "oomph" of the fluid changes the drag force.
2. Jitter matters: The random speed differences between droplets create extra forces and pressures.
3. Non-linear behavior: The forces don't just grow with speed; they grow with the square of the speed and the square of the jitter.

This helps engineers understand that to predict how these mixtures flow (like in industrial separation tanks), they need to account for the "fidgeting" of the droplets, not just their average speed.

Technical Summary

Technical Summary: Inertial Effects on Interphase Drag and Rheology of Dilute Buoyant Droplet Suspensions

Problem Formulation
This work addresses the modeling of dispersed two-phase flows composed of buoyant droplets in the regime of low but finite Reynolds numbers ($Re = aU\rho_f/\mu_f$) and dilute volume fractions ($\phi \ll 1$). While extensive literature exists for Stokes flow (viscous-dominated) and potential flow (inviscid) limits, there is a scarcity of closure models for higher-order hydrodynamic force moments in the low-but-finite $Re$ regime for buoyant droplets. Specifically, the study seeks to quantify how fluid inertia modifies the interphase drag force and the first and second moments of the hydrodynamic force. These moments are critical for closing the averaged momentum equations in two-fluid models, as they contribute to the effective stress of the continuous phase and the momentum exchange between phases. The authors focus on the specific configuration of translating spherical droplets where the relative velocity induces inertia, while surface tension is sufficient to maintain spherical shape ($Ca \ll 1$).

Methodology
The authors employ a rigorous statistical averaging procedure combined with the reciprocal theorem to derive closure relations.

1. Averaged Equations: The study begins by defining ensemble-averaged mass and momentum equations for the dispersed and continuous phases. The closure problem involves expressing ensemble-averaged terms (such as interphase force $F$ and effective stresses $\Sigma^{\text{eff}}$) in terms of macroscopic variables (number density, volume fraction, mean velocities).
2. Conditional Averaging: To resolve the closure problem, the authors utilize conditionally averaged fields. They consider a "test droplet" with a specific center-of-mass velocity $w$ immersed in a background flow. The governing equations for the disturbance fields around this droplet are derived by conditionally averaging the Navier-Stokes equations.
3. General Reciprocal Theorem: A central methodological contribution is the derivation of a general reciprocal theorem applicable to droplets. This theorem relates the force moments acting on a droplet in an arbitrary flow to known auxiliary solutions (Stokes flow solutions and singularity solutions like point sources and point forces). Unlike previous works that focused solely on force or stresslet, this formulation systematically computes moments of arbitrary order.
4. Perturbation Analysis: The authors apply this theorem to a droplet in a uniform steady flow. They compute the $O(Re)$ inertial corrections to the drag force, the first moment (stresslet), and the second moment of the force. This involves matching inner (Stokes) and outer (Oseen) solutions to handle the non-uniformity of the disturbance field at finite $Re$.
5. Ensemble Averaging: Finally, the derived force moments are ensemble-averaged over the distribution of droplet velocities. This step explicitly introduces terms dependent on the velocity variance of the dispersed phase ($\langle \delta_p u'_\alpha u'_\alpha \rangle$) into the macroscopic equations.

Key Results
The study yields explicit analytical expressions for the inertial corrections to the hydrodynamic forces and moments:

• Drag Force: The $O(Re)$ correction to the drag force scales as $O(\rho_f \phi U^2)$. The result recovers the classic Taylor and Acrivos (1964) solution for the drag on a translating droplet but is derived within the context of the reciprocal theorem and ensemble averaging.
• First Moment (Stresslet): The inertial correction to the first moment of the force scales as $O(\rho_f \phi U^2)$. Notably, the authors find that the droplet density ratio ($\zeta = \rho_p/\rho_f$) does not appear in the leading-order inertial corrections for the first and second moments, provided the droplets remain spherical. This contrasts with the drag force where density enters through deformation in other contexts.
• Second Moment: The $O(Re)$ correction to the second moment of the force scales as $O(a \rho_f \phi U^2)$. The authors derive the complete tensor expression, including its trace and traceless parts.
• Velocity Variance Contributions: A critical finding is that the ensemble averaging of the force moments over the droplet velocity distribution introduces additional contributions proportional to the velocity variance of the dispersed phase. Specifically, the averaged drag force and the effective stress of the continuous phase depend quadratically on the relative velocity and linearly on the velocity variance.
• Effective Stress: The resulting effective stress tensor for the continuous phase includes terms proportional to $\rho_f \phi U_r U_r$ and $\rho_f \phi \langle u' u' \rangle$. These terms indicate that the suspension exhibits non-Newtonian behavior, with the effective stress depending on the square of the relative velocity and spatial gradients of the volume fraction and relative velocity.

Significance and Claims
The authors position this work as a necessary step toward more accurate two-fluid modeling of buoyant emulsions and bubbly flows, particularly in industrial processes like flotation and liquid-liquid extraction where buoyancy and turbulence coexist.

• Novelty in Framework: The paper claims to be the first to investigate buoyant droplets at low but finite $Re$ within the framework of averaged two-phase flow equations, explicitly including rheological contributions from the first and second force moments.
• Methodological Rigor: By deriving a general reciprocal relation, the authors provide a systematic tool for computing force moments of arbitrary order, extending previous methodologies used for solid particles or potential flow.
• Closure Implications: The study highlights that standard drag laws derived for isolated particles must be modified when applied in averaged frameworks. The averaging process over the particle velocity distribution naturally generates higher-order moments (velocity variance terms) that are often neglected or modeled empirically. The authors show that these terms are physically grounded consequences of the averaging procedure.
• Rheological Behavior: The work demonstrates that dilute suspensions of buoyant droplets at low $Re$ behave as second-gradient fluids (or Reiner-Rivlin fluids in certain limits), where the effective stress depends on velocity gradients and variances. The authors caution, however, that while the inertial contribution of the first moment is significant, the inertial contribution of the second moment may be negligible in practical models due to scale separation assumptions ($a/L \ll 1$).
• Limitations: The authors modestly note that the system of equations is not fully closed, as no explicit model is proposed for the velocity-variance terms themselves, though they reference existing kinetic theory and turbulence modeling approaches for potential closures. Additionally, the results assume spherical droplets and clean interfaces, excluding Marangoni stresses and significant droplet deformation.