Coupled Vlasov and non-Newtonian fluid dynamics: existence and large-time behavior

Technical Summary: Coupled Vlasov and Non-Newtonian Fluid Dynamics

Problem Formulation
This paper investigates the mathematical analysis of a coupled kinetic-fluid system describing the interaction between a dilute particle system and an incompressible non-Newtonian fluid on a periodic domain $\mathbb{T}^3$. The system is governed by the Vlasov equation for the particle distribution function $f(t, x, v)$ and a modified incompressible Navier–Stokes equation for the fluid velocity $u(t, x)$. The fluid stress tensor $\tau(Du)$ is assumed to follow a power-law relation, $\tau(Du) = \mu|Du|^{p-2}Du$, where $p > 1$ is the power-law exponent. This model captures generalized Newtonian fluids, including shear-thinning ($p < 2$) and shear-thickening ($p > 2$) behaviors. The coupling occurs through a drag force term in the fluid momentum equation and a corresponding force in the particle transport equation.

The primary analytical challenges addressed are:

1. Global Existence: Establishing the existence of global-in-time weak solutions for the coupled system, particularly in the regime where the fluid stress is purely nonlinear (lacking a linear Newtonian component) and the exponent $p$ is in the lower range of the shear-thinning regime.
2. Large-Time Behavior: Quantifying the decay rates of the system toward velocity alignment (monokinetic state) and identifying how the nonlinearity of the fluid viscosity influences these rates.

Methodology
The authors employ a multi-step approximation and compactness strategy to prove global existence, followed by energy-based arguments for large-time asymptotics.

• Approximation Scheme: To construct weak solutions, the authors introduce a two-level regularization. First, they mollify the nonlinear terms (specifically the drag force and convective terms) and truncate the initial data in velocity space using a parameter $\varepsilon > 0$. Second, they apply a Galerkin projection onto a finite-dimensional subspace of divergence-free vector fields.
• Fixed Point Argument: For the regularized system, the authors decouple the kinetic and fluid equations by freezing the macroscopic quantities (density and momentum). They define a solution map on a Banach space and utilize Schauder's fixed point theorem to establish the existence of solutions to the regularized system on a local time interval.
• Compactness and Limits: The global existence is achieved by passing to the limit in two stages:
  1. Galerkin Limit ($n \to \infty$): Using uniform energy estimates and velocity averaging lemmas, the authors establish strong compactness for the macroscopic quantities (density and momentum) and the fluid velocity. This allows them to identify the limit of the nonlinear drag term and the stress tensor.
  2. Mollification Limit ($\varepsilon \to 0$): By removing the mollification, they recover a weak solution to the original system. A critical aspect of this step is handling the identification of the nonlinear stress tensor $\tau(Du)$ without the aid of linear dissipation. The authors adapt the $L^\infty$-truncation method (capacity method) combined with monotonicity arguments to prove the almost everywhere convergence of the velocity gradients.
• Large-Time Analysis: For the asymptotic behavior, the authors define a modulated energy functional $E(t)$ measuring the deviation of the system from a monokinetic state. They derive a differential identity linking the time derivative of $E(t)$ to a total dissipation functional $D(t)$, which includes fluid viscosity and drag dissipation. By establishing an inequality of the form $E(t) \lesssim D(t)^\gamma$, they derive differential inequalities that yield explicit decay rates.

Key Contributions and Results

1. Global Existence of Weak Solutions:
   The paper proves the existence of global-in-time weak solutions for the coupled Vlasov–non-Newtonian system for all power-law exponents $p > 8/5$. This result extends previous theories which often required $p \ge 11/5$ or relied on the presence of a linear viscous term to ensure coercivity. The authors explicitly address the difficulty of controlling the spatial average of the fluid velocity in the periodic setting, where total momentum is conserved but individual averages may evolve. They achieve this by decomposing the fluid velocity into a zero-mean component and a spatial average, allowing for refined a priori estimates.

2. Large-Time Decay Estimates:
   Under an additional assumption that the macroscopic particle density remains uniformly bounded in time, the authors establish quantitative decay estimates for the modulated energy $E(t)$. The decay rate is shown to depend critically on the exponent $p$:

   • Algebraic Decay: For $p > 2$ (shear-thickening regime), the modulated energy decays algebraically.
   • Exponential Decay: For $6/5 \le p \le 2$ (shear-thinning and Newtonian-like regimes), the system exhibits exponential decay.
     This dichotomy highlights that in the purely nonlinear regime without linear dissipation, the large-time dynamics are governed by the interplay between nonlinear fluid diffusion and kinetic drag. The exponential decay for $p \le 2$ is attributed to the stronger dissipative effects of the fluid in this range, whereas the algebraic decay for $p > 2$ reflects weaker diffusion.

3. Convergence to Alignment:
   As a consequence of the decay estimates, the paper demonstrates that the average velocities of the fluid and particles converge to a common limit $v_\infty$ as $t \to \infty$. Furthermore, the particle distribution function converges in the bounded-Lipschitz distance to a monokinetic profile $\rho_f(t, x) \otimes \delta_{v_\infty}(v)$.

Significance and Claims
The authors position this work as the first global-in-time existence theory for weak solutions to this specific coupled kinetic-fluid system in the periodic domain with a purely nonlinear $p$-power law viscosity, specifically covering the range $8/5 < p < 11/5$. They note that previous results typically relied on a linear (Newtonian) component in the stress tensor to guarantee $L^2$ coercivity and facilitate compactness arguments. By removing this linear part, the paper addresses a more challenging analytical setting where standard coercivity estimates involving $\|\nabla u\|_{L^p}$ cannot be directly applied due to the conservation of total momentum.

Regarding large-time behavior, the paper claims to fill a gap in the literature where existing quantitative studies of decay rates fundamentally relied on linear viscous terms to derive exponential convergence. This work provides the first quantitative study of large-time dynamics in the purely nonlinear regime, revealing that exponential decay is not generally expected when $p > 2$, and that algebraic decay is the natural outcome of the nonlinear diffusion-drag interplay. The results are confined to the periodic domain $\mathbb{T}^3$, distinguishing the analysis from works on bounded domains with Dirichlet conditions or the whole space.