Turbophoresis of inertial particles in inhomogeneous turbulence produced by oscillating grids

This paper experimentally demonstrates through oscillating grid turbulence that inertial particles accumulate in regions of lower turbulence intensity due to turbophoresis, a phenomenon where particle migration is driven by the gradient of turbulence intensity.

The Gist

Imagine you are at a crowded dance floor where the music is loud and chaotic. The crowd represents turbulence (swirling, unpredictable air), and the dancers represent particles (tiny solid bits floating in the air).

This paper is about a specific phenomenon called turbophoresis. In simple terms, it explains why heavy dancers (inertial particles) tend to get pushed away from the wildest, most energetic parts of the dance floor and end up gathering in the calmer, quieter corners.

Here is a breakdown of the paper's story using everyday analogies:

1. The Setup: The Chaotic Dance Floor

The researchers built a giant, clear box filled with air. To create the "dance floor" (turbulence), they used special oscillating grids (like giant, rapidly shaking combs) that moved back and forth in the air.

• One Grid: Created a flow that was very strong near the comb and got weaker as you moved away.
• Two Grids: Created a more symmetrical flow, like two combs shaking from opposite sides.

They wanted to see how different types of "dancers" would move in this chaotic air.

2. The Two Types of Dancers

The researchers used two types of particles to see how they behaved:

• The "Ghost" Dancers (Non-inertial particles): These were tiny smoke particles (0.7 microns). They are so light that the wind carries them everywhere instantly. They follow the air perfectly, like a leaf caught in a breeze. They spread out evenly.
• The "Heavy" Dancers (Inertial particles): These were slightly larger glass beads (10 microns). They have weight and "stubbornness" (inertia). When the air swirls, these particles can't turn instantly. They keep going straight for a split second before the air pulls them around.

3. The Phenomenon: The "Centrifugal" Push

The paper explains that because the "Heavy" dancers have inertia, they react differently to the swirling air than the "Ghost" dancers.

The Analogy: Imagine you are on a spinning merry-go-round. If you try to run toward the center, your body wants to keep going in a straight line (inertia), so you feel like you are being pushed outward.

• In the experiment, the air near the shaking grids is a wild, high-energy swirl (high turbulence).
• The air further away is calmer (low turbulence).
• The "Heavy" particles, trying to follow the wild swirls, actually get flung out of the high-energy zones because they can't turn fast enough. They drift toward the calm zones where the turbulence is weaker.

This movement toward the calm zones is called turbophoresis.

4. The Experiment: How They Measured It

To prove this wasn't just the wind blowing the particles to a specific spot, the researchers did a clever trick:

1. They measured where the "Ghost" dancers went. Since they follow the wind perfectly, they showed the "natural" path of the air.
2. They measured where the "Heavy" dancers went.
3. The Comparison: They divided the "Heavy" dancer map by the "Ghost" dancer map.

The Result:
Where the "Ghost" dancers were spread out evenly, the "Heavy" dancers were missing. But in the areas where the air was calmer (lower turbulence intensity), the "Heavy" dancers had piled up.

It's as if the wild music near the shaking grids pushed the heavy dancers away, leaving them to gather in the quiet corners of the room.

5. The Conclusion

The paper confirms that inertia + uneven turbulence = clustering in calm spots.

• What they found: The heavy particles didn't just randomly scatter; they actively accumulated in the regions where the turbulence was the weakest.
• Why it matters (according to the paper): This is a fundamental physics rule. It explains how solid particles (like dust or droplets) naturally sort themselves out in a chaotic fluid without needing any external force to push them there. The "push" comes from the particles' own inability to keep up with the rapid changes in the swirling air.

In a nutshell: If you throw heavy marbles into a stormy ocean, they won't stay in the biggest waves. They will drift toward the calmer patches of water because their weight makes them slip out of the chaotic swirls. That is turbophoresis.

Technical Summary

Technical Summary: Turbophoresis of Inertial Particles in Inhomogeneous Turbulence Produced by Oscillating Grids

Problem Statement
The turbulent transport of inertial particles remains a subject of significant theoretical and experimental inquiry, particularly regarding the formation of large-scale, inhomogeneous concentrations of particles in scales much larger than the integral turbulence scale. While two primary mechanisms—turbulent thermal diffusion (in temperature-stratified turbulence) and turbophoresis (in small-scale inhomogeneous turbulence)—are known to drive such formations, isolating and experimentally verifying turbophoresis in non-stratified flows presents challenges. Turbophoresis is a phenomenon where particle inertia, combined with turbulence inhomogeneity, causes particles to drift toward regions of minimum turbulent intensity. Although theoretically predicted and observed in numerical simulations (DNS), experimental confirmation in controlled laboratory settings, specifically distinguishing this effect from boundary accumulation or gravitational settling, is required.

Methodology
The authors conducted experiments in a rectangular transparent chamber ($26 \times 53 \times 26$ cm) using airflow generated by one and two oscillating grids. The grids, arranged in a square array, oscillated at $10.5$ Hz with an amplitude of $6$ cm, creating an isothermal, inhomogeneous turbulence with a grid Reynolds number of approximately $2520$.

The study employed Particle Image Velocimetry (PIV) to measure fluid velocity fields and spatial distributions of inertial particles.

• Fluid Characterization: The PIV system utilized incense smoke particles ($0.7 \mu$m diameter) as tracers to map mean velocity, velocity shear, root-mean-square (r.m.s.) turbulent velocities, anisotropy, integral length scales, and Reynolds numbers across the flow domain.
• Particle Tracking: Inertial particles were introduced using hollow borosilicate glass spheres ($10 \mu$m diameter, $\rho_p \approx 1.4$ g/cm$^3$). Their spatial distribution was determined via Mie light scattering intensity measured by the CCD camera.
• Isolation of Turbophoresis: To isolate the turbophoretic effect from other transport mechanisms, the number density of inertial particles ($n_{in}$) at every point was normalized by the number density of non-inertial particles ($n_{nin}$) obtained in separate experiments under identical flow conditions. This ratio ($n_{in}/n_{nin}$) effectively removes the influence of the mean flow and turbulent diffusion, highlighting the specific contribution of turbophoresis.
• Theoretical Framework: The analysis relied on the steady-state solution of the particle number density equation, which posits that the ratio of inertial to non-inertial particle densities follows a power-law dependence on the normalized turbulence intensity: $n_{in}/n_{nin} \propto (\langle u^2 \rangle / \langle u^2 \rangle_{max})^{-a^* \kappa_{turboph} / \tau_0}$.

Key Results

1. Flow Characteristics: The experiments successfully generated inhomogeneous turbulence where turbulent velocity fluctuations were stronger near the oscillating grids and decayed with distance. The turbulence produced by two grids exhibited greater symmetry and lower anisotropy compared to the single-grid configuration. Integral turbulence scales were observed to scale linearly with the distance from the grid, consistent with previous grid-stirred turbulence studies.
2. Particle Accumulation: The spatial distributions of the normalized particle number density ($n_{in}/n_{nin}$) revealed a clear tendency for inertial particles to accumulate in regions of lower turbulence intensity.
3. Correlation with Theory: The experimental data showed that the ratio of inertial to non-inertial particle densities was less than unity in most of the domain, except in regions corresponding to the minimum of the normalized turbulence intensity ($\langle u^2 \rangle / \langle u^2 \rangle_{max}$). In these low-intensity regions, the ratio exceeded unity, indicating large-scale clustering.
4. Mechanism Verification: The observed accumulation patterns aligned with the theoretical dependence derived from Eq. (16). The study confirmed that the accumulation is a bulk effect driven by the gradient of turbulent intensity rather than boundary effects. Furthermore, the authors demonstrated that gravitational settling (terminal fall velocity $\approx 0.33$ cm/s) was negligible compared to the turbulent velocities and did not dominate the observed horizontal transport patterns.

Significance and Claims
The paper claims to provide experimental evidence supporting the theoretical predictions of turbophoresis in inhomogeneous, non-stratified turbulence. By normalizing inertial particle distributions against non-inertial baselines, the authors successfully isolated the turbophoretic velocity component. The results confirm that inertial particles are accumulated within large-scale concentrations located in regions of minimum turbulence intensity, a phenomenon driven by the competition between the effective turbophoretic pumping velocity (proportional to the Stokes time and the gradient of turbulence intensity) and turbulent diffusion.

The study concludes that turbophoresis is a distinct mechanism from turbulent thermal diffusion, originating from the averaging of the particle velocity field in inhomogeneous turbulence rather than temperature stratification. The findings validate the theoretical model where the formation of large-scale particle concentrations is governed by the gradient of turbulent intensity, offering a controlled experimental verification of a phenomenon previously largely explored through direct numerical simulations.