Wave Particle Turbulent Simulation of Spatially Developing Round Jets Using a Non Equilibrium Transport Model with a Mixing Length Characteristic Time Closure

This paper demonstrates that the Wave-Particle Turbulent Simulation (WPTS) framework, when coupled with a mixing-length-based characteristic-time closure, accurately predicts the similarity behavior and characteristic features of spatially developing round jets across different Reynolds numbers.

The Gist

Imagine you are trying to predict how a crowd of people moves through a busy train station. There are two ways to look at this: you can look at the "big picture" (the overall flow of the crowd moving toward the platforms), or you can look at the "individuals" (the single person who suddenly trips, changes direction, or sprints to catch a train).

This scientific paper describes a new way to simulate turbulence—the chaotic, swirling motion seen in fluids like air or water—by combining these two perspectives into one smart system.

Here is the breakdown of how it works, using everyday analogies.

1. The Problem: The "Blurry Vision" Dilemma

When scientists simulate things like jet engines or ocean currents, they face a massive math problem. To see every tiny swirl (turbulence) perfectly, you need a supercomputer the size of a building. If you use a regular computer, you have to "blur" the small details to save time.

The problem with "blurring" is that you lose the chaos. It’s like watching a video of a storm where the lightning is just a soft, steady glow instead of sharp, jagged bolts. You get the general shape of the storm, but you miss the energy that actually makes it dangerous.

2. The Solution: The "Wave and the Particle" (WPTS)

The researchers use a method called WPTS (Wave-Particle Turbulent Simulation). Think of this like a high-tech video game engine that uses two different "rendering" modes at the same time:

• The Wave (The Background): This is the "big picture." It treats the fluid like a smooth, continuous wave or a calm river. It’s great for seeing the general direction the water is flowing, but it’s too "smooth" to show chaos.
• The Particle (The Chaos): When the simulation detects a "rough" or "turbulent" area, it suddenly spawns "digital particles." These are like individual, energetic actors dropped into the scene. These particles don't just follow the smooth flow; they zip around, carry energy, and "bump" into the background flow, creating the realistic swirling motion of a real jet of air.

3. The "Mixing Length" Secret Sauce

The "secret ingredient" this paper introduces is a new way to decide how long these particles should stay active.

They used an old idea from 1925 called the "Mixing Length Hypothesis." Imagine you are in a crowded room. If you are in a small, quiet corner, you stay put. But if you are in the middle of a mosh pit, you are going to be tossed around and travel a certain distance before you settle down.

The researchers created a mathematical rule (a "closure model") that tells the simulation: "In this part of the jet, the 'mosh pit' is huge, so let the particles travel far and stay wild. In this other part, the flow is calm, so turn the particles off and just use the smooth 'wave' mode."

4. The Test: The Round Jet

To prove it works, they tested it on a "Round Jet" (imagine a high-pressure fire hose spraying into the air).

They tested it at two speeds: a "slow" speed (Reynolds number 5,000) and a "fast, chaotic" speed (Reynolds number 20,000). Even at the high, chaotic speed—which is usually incredibly hard for computers to handle—their "Wave + Particle" method accurately predicted how the jet spreads out and how the wind speed drops as you move away from the center.

Summary: Why does this matter?

Instead of choosing between a "smooth but boring" simulation or a "detailed but impossibly expensive" one, these researchers have built a "smart" simulation.

It’s like a camera that stays in "low-resolution mode" to save battery while you're looking at a still landscape, but instantly switches to "ultra-high-speed slow motion" the moment a bird flies through the frame. This makes it much faster and cheaper to design better airplanes, more efficient engines, and safer industrial tools.

Technical Summary

Technical Summary: Wave–Particle Turbulent Simulation of Spatially Developing Round Jets

1. Problem Statement

The paper addresses the challenge of accurately simulating turbulent flows, particularly spatially developing round jets, using coarse computational grids. Traditional methods face a trade-off: Direct Numerical Simulation (DNS) provides high fidelity but is computationally prohibitive at high Reynolds numbers ($Re$); Large Eddy Simulation (LES) requires significant resolution to capture subgrid-scale motions; and Reynolds-Averaged Navier–Stokes (RANS) methods often rely on equilibrium assumptions (eddy-viscosity models) that fail to capture non-local, non-equilibrium transport effects. The goal is to develop a multiscale modeling approach that can predict the self-similar behavior of jets across different Reynolds numbers using a mesh resolution comparable to RANS.

2. Methodology

The authors utilize the Wave–Particle Turbulent Simulation (WPTS) framework, a kinetic-theory-based approach derived from the Unified Gas-Kinetic Scheme (UGKS).

• Wave–Particle Decomposition: The flow is represented by two coupled components:
  • Wave (Eulerian) Component: Represents the large-scale, grid-resolved hydrodynamic structures using the Navier–Stokes equations.
  • Particle (Lagrangian) Component: Stochastic particles represent unresolved small-scale turbulent motions (unresolved Turbulent Kinetic Energy, TKE).
• Non-Equilibrium Transport: Unlike eddy-viscosity models that modify local diffusion, WPTS allows particles to travel a finite distance (free transport) before interacting with the background flow. This enables the simulation to capture non-local effects, where flow information is transported across multiple grid cells.
• Mixing-Length Characteristic-Time Closure ($\tau_t$): The core innovation is a new turbulence characteristic-time model tailored for jets. It is derived from Prandtl’s mixing-length hypothesis:
  • The turbulent viscosity is modeled as $\nu_t = (C_{mll} l)^2 |S|$, where $S$ is the strain rate tensor.
  • The mixing length $l$ is defined as $l = \sqrt{D x + b_{ml} l}$, which increases with the downstream distance $x$.
  • This $\tau_t$ governs the particle sampling and the transport distance of the Lagrangian particles, allowing the model to adaptively transition between laminar and turbulent regimes.

3. Key Contributions

• Development of a Jet-Specific Closure: The authors successfully integrated a mixing-length-based characteristic-time model into the WPTS framework, specifically designed to capture the spatial evolution of jets.
• Unified Framework: The method provides a seamless transition between laminar and turbulent flows; in laminar regions, particles vanish, and the method reverts to the standard gas-kinetic scheme (GKS).
• Reynolds-Number Scaling Strategy: The study establishes empirical scaling laws for the model: the minimum grid size scales with $Re^{-0.2}$ and the model coefficient $C_{mll}^2$ scales with $Re^{-0.4}$.

4. Results

The model was validated against DNS and experimental data for two cases: $Re = 5,000$ and $Re = 20,000$.

• $Re = 5,000$ Case:
  • The centerline velocity decay constant ($B_u$) was predicted as 5.55, showing excellent agreement with DNS ($B_u = 5.50$) and experimental data (ranging from 5.80 to 6.06).
  • The simulation accurately reproduced the self-similar profiles of mean velocity and Reynolds stresses (r.m.s. of $U'U'$, $U'_rU'_r$, and $U'U'_r$) across different downstream stations ($x = 25D, 30D, 35D$).
• $Re = 20,000$ Case:
  • Despite the significantly higher Reynolds number and the use of a coarse mesh, the WPTS accurately captured the jet's self-similarity.
  • The decay constant $B_u$ was 6.18, deviating by only ~2% from experimental values ($B_u = 6.06$).
  • The profiles for mean velocity and Reynolds stresses collapsed onto single curves, confirming the robustness of the model at higher $Re$.

5. Significance

This work demonstrates that the WPTS framework, when coupled with a physically motivated mixing-length closure, serves as a powerful and efficient tool for turbulence modeling. Its ability to capture non-local, non-equilibrium transport on coarse grids makes it a highly promising alternative to traditional RANS and LES for engineering applications where high-fidelity simulation of complex, spatially developing flows is required at a fraction of the computational cost.