Thermodynamic Approach to Momentum Transport in Dense Fluids

This paper introduces a new thermodynamic framework for extending Chapman-Enskog theory to dense fluids by utilizing an exchange function linked to thermodynamic properties, proposing an alternative to Modified Enskog Theory that incorporates potential interaction energy and demonstrates high accuracy in predicting shear viscosity for Lennard-Jones and Weeks-Chandler-Anderson fluids across a wide range of densities and temperatures.

The Gist

Imagine you are trying to predict how thick and sticky a liquid is (its "viscosity") just by knowing how hot it is and how crowded the molecules are. For simple, hard balls bouncing around, scientists have had a good recipe for this for a long time. But real fluids are messy: their molecules aren't perfect hard balls; they are soft, they attract each other from a distance, and sometimes they even vibrate like little dumbbells.

This paper presents a new, smarter recipe for predicting how thick these messy fluids will get, without needing to guess or fit a million numbers to make it work.

The Old Way: The "Hard Ball" Problem

Think of the old method (Chapman-Enskog theory) like trying to describe a crowd of people by pretending they are all rigid, unyielding steel balls.

• The Problem: Real molecules are like people in a crowded room. They are soft, they hug (attract), and they push away (repel) before they actually touch.
• The Old Fix: Scientists tried to pretend these soft, hugging people were just "effective" steel balls with a slightly different size. But this only works when the room is empty. As the room gets crowded (high density), the "steel ball" idea breaks down because it ignores the hugging and the softness.

The New Approach: The "Thermodynamic Exchange"

The authors propose a new framework. Instead of trying to force real molecules into a "steel ball" box, they look at the energy exchange happening in the fluid.

Imagine a busy dance floor.

• The Old View: You only count how many times dancers bump into each other (collisions).
• The New View: You also count how much energy is stored in the music and the mood of the room (potential energy).

The authors introduce a concept called an "exchange function." Think of this as a scorecard that tracks how much momentum (the "push") is being swapped between molecules.

• They realized that for simple hard balls, this scorecard is easy to calculate.
• For complex fluids, they found a way to calculate this scorecard using the fluid's thermodynamic properties (like pressure and temperature) and the potential energy of the molecules.

Essentially, they replaced the guesswork of "what size ball should we pretend this is?" with a direct calculation of "how much energy is involved in the interaction?"

What They Tested

To see if their new recipe worked, they simulated three different types of "fluids" on a computer:

1. The "Soft Repellers" (WCA Fluid): Molecules that only push each other away but don't stick together. Like people who only want personal space.
2. The "Full Interaction" (Lennard-Jones Fluid): Molecules that push away when close but pull together when a bit further away. Like magnets that also have a repulsive force.
3. The "Dumbbell" (Diatomic Molecules): Molecules made of two atoms connected by a spring. These are tricky because they can wiggle and vibrate, meaning collisions aren't perfectly bouncy (elastic).

The Results: How Well Did It Work?

The authors compared their new predictions against the computer simulations (which act as the "ground truth").

• For the Simple and "Full Interaction" fluids: The new method was incredibly accurate.

  • At low and medium crowds (densities), the prediction was off by only 2% to 4%.
  • Even in very crowded conditions, the error rarely exceeded 8%.
  • Analogy: It's like predicting traffic flow in a city with 95% accuracy without needing to know the color of every car.

• For the "Dumbbell" (Diatomic) fluids: The method struggled a bit more, with errors between 15% and 30%.

  • Why? The new recipe assumed the collisions were perfectly bouncy. But because these molecules vibrate (like a spring), they absorb some energy during a crash, making the "bounciness" different.
  • The Fix: The authors showed that if they added a simple "tuning knob" (a single number) to account for this wiggling, the accuracy jumped back up to 1.5% to 5%.

The Bottom Line

This paper doesn't claim to cure diseases or build new engines. It claims to have found a better mathematical way to describe how fluids flow.

They proved that you don't need to pretend complex fluids are made of hard balls to predict their behavior. Instead, by looking at the energy involved in how molecules interact, you can get a very accurate prediction of how thick the fluid will be. It's a more honest way of looking at the physics, one that respects the "softness" and "stickiness" of the real world.

Technical Summary

Technical Summary: Thermodynamic Approach to Momentum Transport in Dense Fluids

Problem Statement
The prediction of transport properties, specifically shear viscosity, in dense fluids remains a significant challenge in non-equilibrium statistical mechanics. While the Chapman-Enskog theory provides a framework for calculating transport coefficients based on the Boltzmann transport equation, its application to real fluids typically relies on the "hard-sphere" model. Existing extensions, such as Modified Enskog Theory (MET) and effective hard-sphere diameter approaches, often fail at densities approaching the critical point. These methods struggle because they rely on heuristic effective parameters that do not adequately capture the complex density dependence and the impact of long-range or soft interactions on the momentum exchange rate. Consequently, quantitative predictions for real fluids often require ad-hoc empirical fitting.

Methodology
The authors propose a new general framework that extends Chapman-Enskog theory by replacing effective hard-sphere parameters with a thermodynamic exchange function derived from the system's state variables.

1. Theoretical Framework:

   • The authors reformulate the Enskog expression for shear viscosity ($\eta$) by introducing an exchange function, $\hat{g}$, which combines the excluded volume fraction and the radial distribution function at contact.
   • They derive a general relation: $\hat{g} = Z^{-1} - X_{res}$, where $Z$ is the compressibility factor and $X_{res}$ is a residual function that vanishes for hard-sphere fluids.
   • The core of the proposed method is the selection of $X_{res}$. The authors demonstrate that existing theories like MET and Dipolar Hard-Sphere (DHS) kinetic theory fit into this framework with specific choices for $X_{res}$.
   • Proposed Relation: The authors propose setting $X_{res} = \beta u_{ex}$, where $\beta = 1/k_B T$ and $u_{ex}$ is the excess internal energy per particle. This choice explicitly incorporates the potential interaction energy associated with inter-particle forces, rather than relying on effective diameters.

2. Simulation and Validation:

   • To test the validity of the proposed relation ($X_{res} = \beta u_{ex}$), the authors performed Molecular Dynamics (MD) simulations using the non-equilibrium method developed by Müller-Plathe to calculate shear viscosity.
   • Fluid Models: Three distinct systems were simulated across a wide range of reduced densities ($0.05 \leq \rho^* \leq 0.8$) and temperatures ($1.5 \leq T^* \leq 4.0$):
     1. Cut and Shifted Mie Fluids: Including the Weeks-Chandler-Anderson (WCA) fluid and variations with different repulsive/attractive exponents. These models isolate repulsive soft-core interactions.
     2. Full Mie Fluids: Including the Lennard-Jones (LJ) fluid, which introduces long-range attractive interactions.
     3. Diatomic Lennard-Jones Molecules: A system with internal degrees of freedom (rotational and vibrational) and inelastic collisions due to bond stretching.
   • Parameter Calculation: The low-density limit viscosity ($\eta_0$) and the exchange function $\hat{g}$ were calculated using thermodynamic relations derived from simulation data (pressure and excess internal energy) or semi-empirical expressions where available.

Key Results

• Atomic Mie and LJ Fluids: For fluids characterized by single-atom Mie-type interactions (both cut/shifted and full potentials), the proposed expression ($X_{res} = \beta u_{ex}$) yields high accuracy.
  • At low and intermediate densities ($\rho^* \leq 0.3$), the mean relative prediction error is between 2% and 4%.
  • Across the entire density range, the maximum mean relative errors are 4.4% for the WCA fluid and 8.1% for the full LJ fluid.
  • The method outperforms MET, particularly at elevated densities where MET tends to under-predict viscosity.
• Diatomic Molecular Fluids: For diatomic LJ molecules, the unmodified approach ($X_{res} = \beta u_{ex}$) performs poorly, with prediction errors diverging immediately and stabilizing between 20% and 30%.
  • The authors attribute this failure to the inelastic nature of molecular collisions (energy transfer to internal degrees of freedom), which the pure thermodynamic relation does not account for.
  • By introducing a heuristic fit parameter $C$ such that $X_{res} = C\beta u_{ex}$, the accuracy is restored. The fitted values of $C$ are consistently less than 1 (ranging from ~0.87 to 0.91), reflecting the reduced effective interaction strength due to inelasticity. With this modification, the mean relative error drops to 1.5%–5%.

Significance and Claims
The paper claims to provide a general framework for calculating transport coefficients that moves beyond the limitations of effective hard-sphere descriptions. The primary significance lies in demonstrating that the shear viscosity of dense fluids can be accurately predicted using a thermodynamic relation involving the excess internal energy ($\beta u_{ex}$), without resorting to empirical fitting for simple atomic fluids.

The authors assert that their approach successfully captures the physics of momentum exchange in systems with soft repulsive cores and long-range attractive forces, achieving accuracy comparable to or better than existing extensions like MET. However, the paper remains modest regarding molecular fluids with internal degrees of freedom; it acknowledges that the standard thermodynamic relation fails for inelastic systems and that a heuristic correction is currently required. The work suggests that the limitations of Chapman-Enskog theory are less severe than commonly believed, provided the exchange function is chosen carefully to reflect the underlying thermodynamic state and interaction energy.