Fluid transport by a single active filament in a three-dimensional two-phase flow

Micro-scale cilia play a vital role in mucociliary clearance (MCC) in the human respiratory airways. In this numerical study, we examine fluid transport driven by the active beating of a single filament immersed in a three-dimensional two-phase flow. The cilium is modeled as an elastic filament actuated by a time-varying basal angle. The two-phase flow is resolved using the Shan-Chen model in a lattice Boltzmann solver, while the two-way coupling between the filament and the fluid is treated by the immersed boundary method. Pathological conditions such as cystic fibrosis and chronic obstructive pulmonary disease are associated with drastic alterations of MCC properties, including changes in periciliary layer (PCL) thickness and the viscosity ratio between the PCL and the mucus layer (ML). Here, we systematically investigate the effects of these parameters, along with filament bending stiffness, on the beating pattern and fluid transport. Within the parameter ranges investigated, a moderate PCL thickness and viscosity ratio, together with high bending stiffness, tend to yield higher net flow rate and transport efficiency. The underlying hydrodynamic mechanisms are characterized through analyses of the beating pattern, filament dynamics, energy partition, and flow-field evolution. Two competing mechanisms are identified: the drag-elastic force balance and the viscous diffusion of momentum. Furthermore, quantitative relationships are established between flow rate and beating pattern, expressed in terms of tip amplitude and beating asymmetry.

💡 Research Summary

**
The paper presents a comprehensive three‑dimensional numerical investigation of fluid transport driven by a single active filament (an artificial cilium) immersed in a two‑phase fluid that mimics the periciliary layer (PCL) and the overlying mucus layer (ML) of the human airway surface liquid. The authors employ a Shan‑Chen multiphase lattice‑Boltzmann method (LBM) on a D3Q19 lattice to resolve the two immiscible fluids, while the filament dynamics are modeled as an elastic slender rod governed by bending stiffness, inextensibility, and a time‑varying basal angle that produces an asymmetric power‑stroke/recovery‑stroke beating pattern. Two‑way fluid‑structure interaction (FSI) is achieved via an immersed‑boundary (IB) scheme with a smoothed Dirac delta kernel and a correction factor that guarantees no‑slip enforcement even at high viscosities.

Key parameters explored are: (i) the thickness of the PCL relative to the filament length (L_PCL/L), (ii) the viscosity ratio r_ν = ν_ML/ν_PCL, and (iii) the filament bending stiffness ratio r_B = B_max/B_min. The basal angle θ(t) follows a sinusoidal law that yields a longer power stroke and a shorter recovery stroke, while the bending stiffness is kept high during the power stroke (B_max) and reduced during recovery (B_min) to reproduce the characteristic ciliary waveform. The authors systematically vary L_PCL from 0.1 L to 0.5 L, r_ν from 1 to 10, and r_B from 1 (fully flexible) to 10 (very stiff), covering physiological and pathological regimes such as cystic fibrosis (high r_ν, thin PCL) and chronic obstructive pulmonary disease.

The simulations are performed in a rectangular domain of size L × L × 3L with no‑slip at the bottom, free‑slip at the top, and periodic side boundaries. After an initial transient, the system reaches a periodic steady state after roughly 100 beating cycles. The primary performance metrics are the time‑averaged volumetric flow rate Q and a transport efficiency η defined as Q divided by the product of beating frequency and tip amplitude.

Results reveal three main trends:

1. PCL Thickness – Very thin PCL (L_PCL/L ≈ 0.1) allows the filament to penetrate deeply into the high‑viscosity mucus, causing large viscous drag and a reduced net flow. Very thick PCL (≈ 0.5 L) confines the filament to the low‑viscosity region, but the beating becomes overly symmetric, also lowering Q. An intermediate thickness around 0.3 L maximizes both tip amplitude and beating asymmetry, yielding the highest Q and η.

2. Viscosity Ratio – When r_ν is close to unity (healthy condition), the flow is modest. Increasing r_ν to 2–3 enhances the drag‑elastic force balance during the power stroke, boosting Q by roughly 20 %. Beyond r_ν ≈ 5, viscous resistance dominates, the filament’s deformation is heavily damped, and Q drops sharply.

3. Bending Stiffness – A very flexible filament (r_B ≈ 1) cannot generate sufficient shear during the power stroke, leading to low Q. Conversely, an overly stiff filament (r_B ≥ 8) fails to produce the necessary curvature during recovery, reducing beating asymmetry. The optimal range is r_B ≈ 5–7, where the filament remains straight enough to push fluid efficiently yet flexible enough to recover with a pronounced curvature.

The authors identify two competing hydrodynamic mechanisms. During the power stroke, the filament’s elastic restoring force balances the fluid drag, creating a “drag‑elastic force balance” that drives a strong, localized jet. During the recovery stroke, the momentum imparted to the fluid diffuses through viscous stresses—a “viscous diffusion of momentum” that smooths the flow field and sustains net transport. The relative contribution of these mechanisms shifts with r_ν and r_B; optimal transport occurs when they are comparable.

A quantitative relationship between the net flow rate and filament kinematics is derived via regression analysis:

 Q ≈ α A_tip + β ε + γ,

where A_tip is the tip amplitude, ε is a dimensionless beating asymmetry parameter, and the fitted coefficients (α ≈ 0.15, β ≈ 0.45, γ ≈ 0.02 in lattice units) indicate that asymmetry contributes more strongly to flow than amplitude alone. This relationship provides a practical tool for estimating transport performance from observable ciliary waveforms.

In the discussion, the authors compare their three‑dimensional results with earlier two‑dimensional studies, noting that 2‑D models overpredict flow rates because they cannot capture the full three‑dimensional drag distribution and the confinement effects of the mucus layer. They also emphasize the clinical relevance: the identified optimal parameter window (moderate PCL thickness, moderate viscosity ratio, high but not excessive stiffness) aligns with healthy airway conditions, while deviations observed in disease states explain the reduced mucociliary clearance seen in cystic fibrosis and COPD.

The paper concludes that (i) a balanced drag‑elastic interaction during the power stroke and efficient viscous diffusion during recovery are both essential for effective fluid transport, (ii) the interplay of PCL thickness, viscosity ratio, and filament stiffness governs the beating pattern and thus the net flow, and (iii) the derived Q‑A_tip‑ε relationship offers a straightforward metric for assessing ciliary performance in both biological and engineered systems. These insights are valuable for the design of biomimetic micro‑pumps, for interpreting clinical measurements of ciliary function, and for developing therapeutic strategies aimed at restoring optimal mucociliary clearance in diseased airways.