Turbulent Pipe Flow of Thixotropic Fluids

Through direct numerical simulations and a stochastic Lagrangian model, this study demonstrates that turbulent pipe flow of thixotropic fluids can be accurately described by an effective purely viscous analogue across all thixotropic kinetic regimes, thereby revealing the fundamental feedback mechanisms between microstructure, rheology, and turbulence.

The Gist

Imagine you are stirring a pot of thick soup. If you stir it slowly, it feels thick and gloopy. If you stir it fast, it suddenly becomes runny and easy to mix. This is a property called thixotropy: the fluid's thickness changes over time depending on how much it has been "worked" or sheared.

Now, imagine that soup is flowing through a giant, high-speed pipe, churning around in a chaotic, turbulent mess. This is the world of thixotropic turbulence. The scientists in this paper wanted to understand exactly how this chaotic mixing works when the fluid is constantly changing its own thickness.

Here is the story of their discovery, broken down into simple concepts:

1. The Problem: A Fluid with a Memory

Most fluids (like water) are simple. If you push them, they move. If you stop pushing, they stop. But thixotropic fluids (like ketchup, paint, or certain biological slurries) have a "memory."

• The Microstructure: Think of the fluid as being made of tiny, fragile Lego structures floating inside.
• The Breakdown: When the fluid flows fast (high shear), the turbulence smashes these Lego structures apart, making the fluid thinner.
• The Rebuild: When the fluid sits still or flows slowly, the structures slowly rebuild themselves, making the fluid thick again.

The big question was: In a wild, swirling pipe flow, how does the fluid know whether to be thick or thin? Does it react instantly to the speed it's moving right now, or does it remember how fast it was moving a second ago?

2. The Experiment: The Digital Pipe

The researchers built a super-accurate computer simulation of a pipe. They didn't use real soup; they used a mathematical model of a "thixotropic fluid" and ran it through a digital pipe at high speeds. They tested three different "speeds of memory":

• Fast Memory: The fluid reacts instantly. If it gets hit by turbulence, it breaks down immediately. If it stops, it rebuilds immediately.
• Slow Memory: The fluid is stubborn. It takes a long time to break down or rebuild, regardless of what the turbulence is doing right now.
• Medium Memory: The fluid reacts at a pace that matches the swirling of the turbulence. This is the tricky, complex middle ground.

3. The Discovery: The "Time Travel" Insight

The team realized that to understand the fluid, they couldn't just look at a snapshot of the pipe (like a photo). They had to follow individual tiny particles as they traveled through the pipe, like a time traveler watching a drop of water on a rollercoaster.

They found that the fluid's thickness at any given moment depends on the history of the ride that specific drop of water just took.

• If a drop of water just passed through a violent, fast-spinning whirlpool, its internal structures are smashed, and it is thin.
• If it just drifted through a calm zone, it has had time to rebuild, and it is thick.

4. The Big Surprise: The "Simple" Answer

The most exciting part of the paper is what they found when they tried to predict the flow. They expected the "Medium Memory" case to be a chaotic nightmare that required incredibly complex math to solve.

Instead, they discovered a magic shortcut.

They found that even though the fluid is changing its thickness in real-time, the overall behavior of the turbulent pipe flow acts exactly as if the fluid were not changing at all.

• The Analogy: Imagine a crowd of people running through a hallway. Some people are wearing heavy coats (thick fluid), and some are in t-shirts (thin fluid). The coats change based on how fast the person is running.
  • The researchers found that you don't need to track every single coat changing. You can just pretend that everyone in a specific part of the hallway is wearing a "standard average coat" for that spot.
  • If you use this "average coat" idea, your prediction of how the crowd moves is almost perfectly accurate (within 2.4% error).

5. The Three Rules They Found

The paper总结出 three simple rules based on the "speed of memory" (which they call the thixoviscous number, $\Lambda$):

1. Super Fast Memory ($\Lambda \gg 1$): The fluid reacts so instantly that it behaves like a standard "shear-thinning" fluid (like ketchup). It gets thinner the faster you push it, and that's it.
2. Super Slow Memory ($\Lambda \ll 1$): The fluid is so slow to react that it doesn't notice the turbulence at all. It behaves like a standard, boring, thick fluid (like honey) that never changes.
3. Medium Memory ($\Lambda \approx 1$): This is the sweet spot. The fluid reacts at the same speed as the turbulence. Surprisingly, the researchers found that you can still treat this complex, changing fluid as a simple, unchanging fluid—if you just calculate the "average thickness" based on where the fluid is in the pipe.

The Bottom Line

The paper claims that turbulent flow of these complex, time-changing fluids is actually much simpler than we thought.

Even though the fluid is constantly breaking down and rebuilding its internal structure, the chaotic swirling of the pipe averages everything out. You can predict how the fluid will flow by pretending it is a simple, static fluid with a "smart" thickness that changes depending on how far it is from the pipe wall.

This is a huge deal because it means engineers might not need super-complex, slow computers to design pipes for these fluids. They can use simpler, faster models that treat the fluid as if it were "frozen" in time, and they will still get the right answer.

In short: The fluid has a memory, but the turbulence is so good at mixing things up that, in the end, the fluid acts like it has no memory at all. It just behaves like a simple, thick liquid that knows exactly how to flow.

Technical Summary

Technical Summary: Turbulent Pipe Flow of Thixotropic Fluids

Problem Statement
Complex materials with internal microstructures, such as suspensions and emulsions, exhibit time-dependent rheology characterized by viscoelasticity and thixotropy. In many large-scale industrial applications, such as turbulent pipe flow, the elastic response timescale is significantly shorter than the thixotropic timescale, allowing these flows to be approximated as purely thixotropic. However, the fundamental dynamics of thixotropic turbulence remain poorly understood, particularly the feedback mechanisms between microstructural state, rheology, and turbulence structure. While thixotropic turbulence involves a complex interplay between microstructural evolution, macroscopic rheology, and turbulence-driven shear, there is a lack of understanding regarding how these interactions evolve across different thixotropic kinetic timescales relative to the eddy turnover time.

Methodology
To address these gaps, the authors conducted Direct Numerical Simulations (DNS) of fully developed turbulent pipe flow for a model thixotropic (Moore) fluid. The study utilized a high-resolution spectral element code (Semtex) extended to handle thixotropic fluids. The simulations covered a range of thixoviscous numbers ($\Lambda$), which characterizes the ratio of the thixotropic kinetic timescale to the advective timescale, spanning from slow kinetics ($\Lambda \ll 1$) to fast kinetics ($\Lambda \gg 1$).

The governing equations included the incompressible Navier-Stokes equations coupled with an advection-diffusion-reaction equation for the structural parameter $\lambda$, which quantifies the state of the microstructure. The study analyzed results in both the Eulerian frame (focusing on mean fields and turbulence statistics) and the Lagrangian frame (tracking fluid particle histories). A stochastic model was developed to describe the effective viscosity in the intermediate kinetic regime ($\Lambda \sim 1$) by modeling the Lagrangian shear history as a non-stationary random walk governed by a Fokker-Planck equation for radial particle position.

Key Contributions and Results

1. Limiting Behaviors: The analysis confirms that in the limits of very fast ($\Lambda \to \infty$) and very slow ($\Lambda \to 0$) kinetics, the time-dependent thixotropic flows behave as time-independent, purely viscous (generalized Newtonian) analogues.

   • For fast kinetics, the structural parameter instantly equilibrates with the local shear rate, resulting in shear-thinning behavior similar to a Cross fluid.
   • For slow kinetics, the structural parameter evolves so slowly that it samples the ensemble of shear rates along pathlines, effectively behaving as a Newtonian fluid with a viscosity determined by the mean shear rate.

2. Intermediate Kinetics and Effective Viscosity: For intermediate kinetics ($\Lambda \sim 1$), the rheology is governed by a path integral of the thixotropic fading memory kernel over the distribution of Lagrangian shear history. The authors developed a stochastic model to approximate this history.

   • This model treats the effective viscosity as a radially dependent, time-independent quantity derived from the conditionally averaged structural parameter $\langle \lambda | r \rangle$.
   • DNS computations using this effective viscosity closure showed excellent agreement (within 2.4% error for Reynolds stresses and mean velocity) with the fully thixotropic model for $\Lambda = 1$.

3. Universality of the Analogue: The study demonstrates that the turbulent flow of time-dependent thixotropic fluids can be accurately captured by a time-independent (generalized Newtonian) effective viscosity across the entire spectrum of $\Lambda \in [0, \infty)$. This holds true even for nonlinear rheology models, provided the effective viscosity is defined via the conditional average of the structural parameter.

Significance
The paper claims that these findings uncover the feedback mechanisms between microstructure, rheology, and turbulence, offering fundamental insights into the structure of thixotropic turbulence. The primary significance lies in the discovery that, despite the complex coupling and time-dependence of the microstructure, thixotropic turbulent flows can be simplified to time-independent purely viscous analogues. This simplification is attributed to the fact that thixotropic flows are fundamentally viscous (stress-strain relationship) rather than viscoelastic, and the ergodic nature of turbulent flow allows the time-dependent viscosity to be represented by a spatially varying effective viscosity. This result provides a valuable framework for predicting and controlling the flow of thixotropic fluids in engineering applications without the computational cost of solving full time-dependent thixotropic kinetic equations.