JAM-Flow: Hydrodynamic Traffic Jams

• JAM-Flow is a phenomenon driven by long-range hydrodynamic interactions that induce traffic jams and intermittent flows in microfluidic networks below close-packing density.
• It utilizes a nonlinear current–density relation and dipolar interaction model to explain droplet avalanche dynamics and the emergence of a maximal current threshold.
• These insights are crucial for designing microfluidic devices and advancing the understanding of flow instabilities in biological and porous systems.

JAM-Flow describes the discovery and characterization of nonlinear, hydrodynamic interaction-induced traffic jams and intermittent flow phenomena in microfluidic obstacle networks transporting discrete particles, as investigated experimentally and theoretically by J.-C. Baret et al. ("Traffic jams and intermittent flows in microfluidic networks," (1005.5003)). Unlike "hardcore" crowding or close-packing-based jamming found in granular flows, jams and intermittencies here occur well below geometrical close-packing, and arise from long-range hydrodynamic couplings specific to flow-driven, low-dimensional systems.

1. Hydrodynamic Origin of Traffic Jams and Intermittency

The system consists of droplets (particles) advected by external flow through a microfluidic network comprising a lattice of 1D channels. As droplet injection (i.e., the particle current $j$) increases, two core phenomena are observed:

• Jamming: Droplets accumulate in the central lane at the network entrance, forming a jam even when the average density is much less than close-packing and no physical collisions occur. This jam forms at surprisingly low droplet densities.
• Intermittent Flow: If the droplet current exceeds a critical threshold ($j > j^*$), the jam periodically ejects clusters (avalanches) of droplets into lateral lanes, generating highly irregular, non-stationary flow and subsequent network invasion.

These effects are not due to direct particle contact, but rather are a consequence of long-range, dipolar hydrodynamic interactions: each droplet impedes local hydraulic conductance, causing nonlocal modifications in the velocity field that affect the motion of distant droplets. This nonlocality is essential—interactions propagate even in the absence of collisions.

2. Theoretical Model: Nonlinear Constitutive Law and Hydrodynamic Couplings

The droplet flow is formulated as follows:

• Local Conductance Reduction: The channel conductance at a droplet's position is reduced:

$$G(\mathbf{r}) = G\left[1 - b^2 \sum_{i=1}^N \delta(\mathbf{r} - \mathbf{r}_i)\right]$$

where $b^2$ parameterizes the droplet's impedance.

• Fluid Velocity: The local fluid velocity is given by

$$\mathbf{v}_F(\mathbf{r}) = -G(\mathbf{r}) \nabla P(\mathbf{r}),$$

subject to $$\nabla \cdot \left[ G(\mathbf{r}) \nabla P(\mathbf{r}) \right] = 0$$.

• Droplet Motion: With mobility $\mu$,

$$\dot{\mathbf{r}}_i = \mu\, \mathbf{v}_F(\mathbf{r}_i).$$

• Hydrodynamic "Dipole" Interactions: The velocity disturbance at position $x$ due to a droplet at $x_i$ is

$$v_{\text{dip}}(x; \dot{x}_i) = \dot{x}_i \frac{b^2 (1-\mu)}{2\pi\mu} \partial_x \left[ \frac{1}{|x-x_i|} \right],$$

leading to a strong, nonlocal, density-dependent feedback slowing approaching droplets.

• Current-Density Relation (\emph{nonlinear}):

$$j = \frac{\rho v_D^0}{1 + \frac{\pi}{6}(1-\mu) b^2 \rho^2}$$

with $j$ the droplet current, $\rho$ the linear density, and $v_D^0$ the single-droplet velocity.

The denominator's quadratic growth with density (due to hydrodynamic hindrance) causes $j(\rho)$ to reach a maximum ("maximal current") beyond which steady flow cannot be maintained.

3. Maximal Current and Conditions for Jam Instability

The constitutive law above predicts that $j(\rho)$ is non-monotonic (bell-shaped), with the maximum ("maximal current" $j^*$) at density

$$\rho^* = \frac{1}{\sqrt{a}}, \quad a = \frac{\pi}{6}(1-\mu) b^2$$

and maximal current

$j^* = \frac{v_D^0}{2\sqrt{a}}.$

• $j < j^*$: Steady, stationary flow along the central path.
• $j = j^*$: System approaches instability; dwell time and local density diverge.
• $j > j^*$: No stationary single-lane flow solution exists—network entrance jams, and droplets intermittently exit into other lanes. This yields correlated "avalanches" (bursts) of droplet ejection and non-Gaussian statistics for ejection events.

The regime transition is experimentally confirmed via direct measurement and time series of droplet number and ejection events.

4. Broader Context Within Driven Particle Systems

The existence of a maximal current and subsequent intermittency is a universal feature of many driven, low-dimensional systems, including the Totally Asymmetric Simple Exclusion Process (TASEP) and vehicular traffic flow models. However, in JAM-Flow (as described here), the nonlinearity is hydrodynamic in origin, existing well below the close-packing density and without hard-core exclusion—arising solely from the nonlocal modification of fluid flow by each particle.

Above the threshold, the dynamics display:

• Avalanche statistics: Bursts of motion/group ejections, with non-Gaussian distributions.
• Critical-like behavior: Self-organized criticality in the flow, reminiscent of phase transitions.

5. Implications and Applications

A. Microfluidic and Porous Media Systems

Understanding and quantifying the maximal current is directly relevant for:

• Lab-on-chip design: To avoid throughput instabilities and device clogging, engineers must ensure operation below this hydrodynamically-defined limit.
• Filtration and separation: Similar hydrodynamic crowding may constrain performance in filtration membranes, chemical reactors, or biotechnological devices.

B. Biological System Parallels

• Capillary blood flow and cell sorting in microarchitectured environments may be subject to analogous hydrodynamically-mediated jamming, even at low volume fractions.

C. Theoretical Advances

• Hydrodynamic Nonlocality: Demonstration that transport bottlenecks and avalanching in driven-particle flows can arise from long-range, non-contact interactions—expanding the universality of maximal current effects to a new class of physical systems.
• Experimental Confirmation: Provides the first experimental evidence of such hydrodynamic maximal-current phenomena, previously mainly theoretical.

---

Summary Table: Key Model Formulas

Quantity	Expression	Interpretation
Constitutive law	$$j = \dfrac{\rho v_D^0}{1 + \dfrac{\pi}{6}(1-\mu) b^2 \rho^2}$$	Nonlinear current–density relation
Maximal current	$$j^* = \dfrac{v_D^0}{2 \sqrt{\frac{\pi}{6}(1-\mu) b^2}}$$	Threshold for jam onset/intermittency
Dipole velocity	$v_{\text{dip}}(x; \dot{x}_i)$ as above	Hydrodynamic back-action from one droplet to another

---

The paper establishes that, in microfluidic channel networks, traffic jams and intermittent transport are emergent phenomena produced by long-range hydrodynamic interactions, with a precise, nonlinear current–density relation and a sharp maximal current. This generalizes the "maximal current" scenario beyond exclusion-based models and identifies explicit design constraints and operational limits for microfluidic, porous, and biological systems employing particle-laden flows.