Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles

arXiv:2604.20028 · physics.flu-dyn · Submitted 2026-04-21 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: We're looking at the paper: "Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles".

Mira: Here's the gist of the paper. Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles is investigated using direct numerical simulations to reveal a mean circulation driven by size-dependent wall-normal transport.

Kai: Study Motivation The study presents a first investigation of a fully developed wall-bounded flow of resolved cohesive particles.

Mira: Due to attractive inter-particle forces, cohesive particles suspended in turbulence undergo a complex process of aggregation, breakup, and restructuring.

Kai: The authors employ a population balance equation (PBE) framework to analyze aggregate dynamics.

Title and authors: Kai: So the title "Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles" tells us immediately that the focus is on three intertwined dynamics: how they stick together, how they separate, and how their size influences where they end up in the flow.

Mira: And it’s not just about the forces; it’s specifically about wall-bounded flows. This means we are looking at conditions near a surface, where the physics of turbulence and shear are changing rapidly across the gap between the wall and the bulk fluid.

Lev: I wonder what that means for real hardware. If you were trying to build an error correction system, knowing that aggregate size dictates their movement near a wall could affect how you model particle interactions in a dense medium.

Kai: It’s about understanding the underlying mechanism of flocculation in turbulent flows, which is hard because experimental observations are often limited to single events or simplified setups. This paper tackles the complexity by looking at resolved particles in these challenging environments.

Mira: The authors are using a population balance equation to analyze aggregate dynamics, and they're looking at how this affects the overall flow structure through circulation rather than just static measurements of particle concentration.

Lev: That’s interesting because when we run simulations on real hardware, we often have to make huge assumptions about how many particles are interacting simultaneously; if the size matters so much, that assumption changes things a lot.

The paper's summary: Kai: So they found regions where there is net aggregate production and depletion both in physical space and in size space, which points to a wall-normal imbalance between aggregation and breakup happening during equilibrium.

Mira: That imbalance is then closed by the wall-normal advection of these aggregates, which creates this concept of statistical circulation across the joint space of size and wall-normal coordinate.

Lev: So instead of just seeing particles clump up or break apart uniformly, they are moving in a kind of statistical loop where larger particles are driven to certain areas and smaller ones get cycled elsewhere.

Kai: Yes, specifically, the paper suggests that larger aggregates are preferentially produced in the channel center and then migrate toward the wall where they eventually break down.

Mira: And smaller aggregates, they get transported away from that wall, they grow through aggregation somewhere else, and then they reenter that cycle near the wall to break again. It’s a kind of self-sustaining statistical cycle.

Lev: From an engineering standpoint, if you were designing a system relying on stable particle distribution, this cyclical transport means you can't just look at one region; you have to model the entire flow field and size distribution simultaneously.

The paper's improvements: Kai: The paper suggests a few improvements in how we think about this, starting with making the predictive modeling better by incorporating that size-dependent wall-normal transport directly into our models.

Mira: That’s important because it moves us past parameterizations that are often oversimplified or agnostic to the specific local flow conditions, which is what they point out from prior experimental work.

Lev: I like the idea of quantifying the breakup fraction across different modes—collisional, lubricated, and hydrodynamic—as a function of cohesive strength Co. That gives us a much finer control knob for understanding when aggregation or breakup dominates.

Kai: They also map aggregate orientation and axial straining against the wall-normal position to show how aggregates get progressively more extended as they approach the wall, which directly promotes that breakup we just talked about.

Mira: And finally, they developed a statistical circulation map in both size and physical space by plotting the wall-normal advective flux. This visualization helps you spot exactly where aggregation and breakup are locally imbalanced, beyond just seeing the bulk balance of aggregation and breakup over the whole domain.

Lev: That map would be incredibly useful for designing experimental setups because it tells you precisely which regions to focus on when trying to measure these events in a real flow environment.

Conclusion: Kai: So, to wrap up, the main point of this work on "Aggregation Breakup and Size-Dependent Transport in a Turbulent Channel Flow with Cohesive Particles" is that equilibrium isn't just achieved by simple local balance but through spatially separated regions where net production and depletion occur.

Mira: The paper concludes that you need to include the size-dependent wall-normal advective transport to fully close the population balance equation, meaning aggregation and breakup must balance globally, but they don't have to do so locally in the wall-normal direction.

Lev: That suggests that when we run simulations on real hardware, we can’t just check if aggregation equals breakup at every single point; we have to account for the transport of particles across those spatial boundaries.

Kai: It shows that this paper provides a conceptual picture of statistical circulation in the joint space of size and wall-normal coordinate, where larger aggregates move toward the wall to break, and smaller ones are cycled back out.

Mira: This research really deepens our understanding of how these cohesive materials behave under complex, inhomogeneous turbulence near solid boundaries. It points us toward a more complete way to model particle dynamics in these flows.

Lev: Yeah, it gives us a better framework for predicting the actual distribution of particles you’d see on the floor of a channel flow experiment, which is where I spend most of my time thinking about real system constraints.

Kai: That’s what we were looking at with this paper—how size and transport drive the overall structure. It's a lot to digest, but it opens up new ways to think about these systems.

Alexandre D. Leonelli, Lukas Widmer, Eckart Meiburg

University of California Santa Barbara · ETH Zurich

physics.flu-dyn

Submitted: 2026-04-21

Updated: 2026-10-12

Comments: 29 pages, 13 figures, accepted to the Journal of Fluid Mechanics

DOI: 10.1017/jfm.2026.12202

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 90/100

The gist: Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles is investigated using direct numerical simulations to reveal a mean circulation driven by…

Key concepts

Population Balance Equation (PBE)
A mathematical tool used to track how the size of particle aggregates changes over time. It accounts for all events that create new aggregates (aggregation) or destroy existing ones (breakup), allowing researchers to model the evolution of aggregate populations in a fluid.
Cohesive Number ($Co$)
This number measures the strength of the attractive forces between particles relative to their inertia. A higher $Co$ indicates stronger cohesive forces, which dictates whether aggregates are more likely to stick together (aggregate) or break apart (breakup) when interacting with the wall.
Statistical Circulation
This refers to a conceptual flow pattern in the space defined by particle size and wall-normal position. It describes a cycle where small particles are produced near the wall, grow in the center, move back to the wall to break up, and then re-enter this cycle.
Size-Dependent Transport
This is the idea that particles of different sizes are transported differently by fluid motion. In this study, it means that aggregates of different sizes are carried toward or away from the wall at different rates, which is crucial for understanding why aggregation and breakup do not balance uniformly across the flow.

Terminology

Summary

Aggregation, breakup, and size-dependent transport in a turbulent channel flow with cohesive particles is investigated using direct numerical simulations to reveal a mean circulation driven by size-dependent wall-normal transport. p. 0

Study Motivation

The study presents a first investigation of a fully developed wall-bounded flow of resolved cohesive particles. Due to attractive inter-particle forces, cohesive particles suspended in turbulence undergo a complex process of aggregation, breakup, and restructuring. p. 0 The authors employ a population balance equation (PBE) framework to analyze aggregate dynamics. p. 0

Numerical Methodology

The fluid phase is governed by the continuity and Navier Stokes equations, solved using the immersed boundary method (IBM) for particle-fluid coupling. The cohesive force is modeled using a formulation from Vowinckel et al. p. 2 (2019), which provides an efficient approximation of short-range attractive forces predicted by DLVO theory. p. 2 Particle resolution is maintained at a grid spacing where the characteristic length scale at the wall, the viscous unit δν = ν/uτ, is sufficiently resolved. p. 4 The cohesive number Co quantifies the ratio of maximum cohesive force to a characteristic particle inertial force. p. 4

Population Balance Analysis

The evolution of the aggregate size concentration is described by a population balance equation (PBE) which includes source and sink terms for discrete aggregation and breakup events. The PBE is plane-averaged in the wall-normal direction, allowing for spatially inhomogeneous net aggregation and breakup during equilibrium. p. 7 The bulk PBE is balanced solely by aggregation and breakup over the entire domain at every n. p. 7

Key Findings on Circulation

Regions of net aggregate production and depletion are identified in both physical and size-space, revealing a wall-normal imbalance between aggregation and breakup. This imbalance is closed by the wall-normal advection of aggregates, giving rise to a conceptual picture of statistical circulation in the joint space of size and wall-normal coordinate. p. 1 Specifically, larger aggregates are preferentially produced in the channel center and migrate toward the wall where they break, while smaller aggregates are transported away from the wall, grow, and reenter the cycle. p. 0

Mechanistic Insights

The analysis of breakup fraction Φb shows that for Co = 1.5, categories separate nearly beyond the 95% confidence interval, indicating that only at the strongest cohesive strength, aggregates that interact directly with the wall by either the normal lubrication model or direct contact are increasingly likely to break rather than aggregate. Furthermore, near-wall aggregates align preferentially with the streamwise direction and undergo increasingly rapid extension along their long axis on approach to the wall, where breakup is most prominent. p. 1

Statistical Organization

The analysis of aggregate orientation reveals that aggregates maintain rotation rates similar to the fluid and become progressively more extended on approach to the wall, which promotes breakup. The resulting dynamics organize into a closed statistical cycle where small aggregates are produced near the wall, advected toward the channel center where they grow through aggregation, transported back toward the wall to break down, and re-enter the cycle. p. 0 This cyclical structure is observed at all Co in figure 8(a-d). p. 9

Conclusion

Equilibrium is instead maintained through spatially separated regions of net aggregate production and depletion and closure of the PBE is recovered with the inclusion of size-dependent wall-normal advective transport. The final result is that aggregation and breakup must balance globally, but they need not balance locally in the wall-normal direction. p. 0

Acknowledgements

A.D.L. and E.M. thank Dr. Ronald Chan and Prof. Filippo Coletti for many insightful discussions. p. 0 A slip velocity formula is provided to estimate the particle Reynolds number. p. 4 The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 The maximum Rep,i over all cases occurs for Co = 1. p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 45 and is estimated as maxi Rep,i = 49. p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 4 p. 458. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The correlation time, τint = Z τ0 ρ(τ)dτ quantifies the rate of relaxation of ρ(τ). p. 9 The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 p. 9 The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58. The PDF of Rep accumulates Rep,i for each case averaged over all particles and Teq. The number of independent samples is estimated by dividing the total sampling period for each case T by 2τint. The maximum Rep,i over all cases occurs for Co = 1.5 and is estimated as maxi Rep,i = 49.58 <ref:2604. p. 0

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved predictive modeling of aggregate dynamics in wall-bounded flows by incorporating size-dependent wall-normal transport, which reveals a "mean circulation: larger aggregates are preferentially produced in the channel center and migrate toward the wall where they break, while smaller aggregates are transported away from the wall, grow, and reenter the cycle."

  2. Enhanced ability to distinguish between near-wall breakup mechanisms by quantifying breakup fractions across collisional, lubricated, and hydrodynamic modes as a function of cohesive strength Co.

  3. Development of a statistical circulation map in size and physical space by plotting the wall-normal advective flux (Q) to visualize regions where aggregation and breakup are locally imbalanced.

  4. Predictive capability for aggregate orientation and axial straining by mapping these properties against wall-normal position, allowing for understanding how aggregates become progressively more extended on approach to the wall, promoting breakup.

Related papers