Restarts of bursts in turbulence in a log-minimal channel

This paper investigates the restart mechanisms of wall-normal-velocity bursts in wall-bounded turbulence by combining a linearised Navier-Stokes framework with nonlinear forcing to identify minimal requirements for burst regeneration and demonstrate how nonlinearity facilitates this process by increasing linearly available energy through vortex breakup and merging.

The Gist

The Big Picture: The Turbulent "Heartbeat"

Imagine a river flowing past a rock. The water doesn't just flow smoothly; it churns, swirls, and occasionally sends a sudden, powerful splash of water upward toward the surface. In fluid dynamics, these sudden upward splashes are called "bursts."

For a long time, scientists thought these bursts were like a domino effect: long, stretched-out ribbons of fast water (called "streaks") would become unstable, break, and cause the burst. However, recent evidence suggests that even if you remove those ribbons, the bursts still happen.

This paper asks a different question: How does a burst start again after it has finished?

Think of a burst like a heartbeat. It beats (amplifies), then it rests (decays). For the heart to beat again, it needs to "recharge." This paper investigates exactly how that recharge happens in the chaotic world of turbulence, using a simplified model of a narrow channel.

The Two Main Questions

The authors tackle this problem in two ways, moving from a "what if" scenario to "what actually happens."

1. The "What If" Scenario: The Minimal Requirements

First, they asked: What is the absolute minimum amount of "push" needed to make a dying burst come back to life?

They used a simplified mathematical model (a linear system) where they could control the forces perfectly. They found that to restart a burst, you don't just need to push the water up; you need to do two specific things to the swirling eddies (vortices) inside the flow:

• The Breakup: Imagine a single, large, forward-leaning stack of cards. To restart the cycle, you have to break that stack apart into several smaller, separate layers.
• The Catch-Up: Once broken into layers, the "cards" (vortices) need to move. The layers further away from the wall move faster than the layers closer to the wall.
  • First, a backward-spinning layer catches up to a forward-spinning layer (a "counter-rotating catch-up").
  • Then, a forward-spinning layer catches up to another forward-spinning layer (a "co-rotating catch-up").

The Analogy: Think of a relay race where the runners are tired and slowing down. To get them running fast again, you don't just yell "Go!" You have to break their formation (breakup) so they can run in staggered lines. Then, the faster runners in the back catch up to the slower ones in the front, merging their energy to create a new surge of speed.

They also introduced a concept called Linearly Available Energy (LAE). Think of this as a "battery charge" for the burst.

• As a burst happens, it uses up its battery (LAE goes down).
• Eventually, the battery is too low to do anything.
• To restart, something must recharge the battery to a high enough level. The paper shows that the "push" needed to restart a burst is essentially a massive recharge of this specific energy.

2. The "Real World" Scenario: What Turbulence Actually Does

Next, they looked at real, messy turbulence data to see if nature follows these rules.

They found that in real turbulence, the "nonlinear" parts of the flow (the messy, chaotic interactions between different parts of the fluid) act exactly like the "push" they simulated in the first part.

• The Recharge: The chaotic interactions take a decaying burst and pump energy back into it, raising its "battery charge" (LAE) to a level where it can explode into a new burst.
• The Pattern: When they looked at the flow patterns during this recharge, they saw the exact same "breakup" and "merging" of vortex layers that their simple model predicted.

The Three Stages of a Restart

The paper breaks down the restart process into three conceptual phases, which they observed in both their models and real data:

1. Breakup: A single, coherent vortex structure gets shattered into multiple layers by external forces (or nonlinearity).
2. Counter-rotating Catch-up: The broken layers start moving. A layer spinning one way catches up to a layer spinning the opposite way. This creates a chaotic, disconnected phase.
3. Co-rotating Catch-up & Merge: Finally, layers spinning in the same direction catch up and merge. This reconnection creates a strong, backward-leaning structure that is ready to burst again.

The Key Takeaway

The paper challenges the old idea that long streaks of water are the main cause of bursts. Instead, it suggests that bursts have their own self-sustaining cycle.

The cycle works like this:

1. A burst happens and uses up its energy.
2. The burst decays and breaks apart into layers.
3. The chaotic nature of the fluid (nonlinearity) acts as a "recharger," breaking up the old structures and merging new ones to build up enough energy (LAE) to trigger the next burst.

In short, the paper proves that to restart a turbulent burst, the fluid must break apart and then reconnect in a very specific way, effectively recharging its internal energy battery to keep the turbulence alive.

Technical Summary

Technical Summary: Restarts of Bursts in Turbulence in a Log-Minimal Channel

Problem Statement
Wall-bounded turbulence is characterized by intermittent, intense events known as "bursts," which are responsible for the transfer of mass and momentum across wall distances. While the development of these bursts has been successfully linked to the linear Orr process (a forward-tilting of disturbances by mean shear), the mechanism for restarting a burst cycle—effectively a backward-tilting process required to sustain turbulence—remains less understood. Recent evidence challenges the classical view that streaks and bursts are mutually induced in a single self-sustaining process, suggesting instead that bursts can be sustained independently of streaks. This paper addresses the specific problem of how bursts restart in a log-minimal channel, aiming to identify the minimal requirements for burst restarting without relying on a priori knowledge of specific flow structures like streaks.

Methodology
The authors employ a dual approach combining a forced linearised Navier-Stokes system and analysis of direct numerical simulation (DNS) data from a log-minimal channel at $Re_\tau = 950$.

1. Forced Linearised System: The nonlinear terms of the Navier-Stokes equations are treated as external forcing terms acting on a linearised system. This allows the authors to isolate the "minimal requirements" for burst restarting through mathematical optimisation problems.

   • Optimisation Problems: The study formulates four key optimisation problems:
     • Problem 1: Maximising transient growth to define a "baseline" burst development.
     • Problem 2: Determining the minimal impulsive force required to induce a "weak-sense" re-burst (an increase in wall-normal velocity intensity) from a decaying state.
     • Problem 3: Determining the minimal continuous force required to drive a decaying state to a target state near the beginning of the amplification stage ("strong-sense" re-burst).
     • Problem 4: Identifying forces that sustain periodic bursts indefinitely.
   • Linearly Available Energy (LAE): A new metric, $E_v$, is defined as the integral of the squared wall-normal velocity intensity over the future evolution of a state in the absence of forcing. This quantity parameterises the "burstiness" or potential of a state to restart.

2. Real Turbulence Analysis: The authors analyse DNS data to extract "contributory features" of burst restarts. They distinguish between linear and nonlinear contributions to state evolution and use a "Dominant Joint Mode" (DJM) extraction technique to visualise flow patterns conditioned on specific states in the intensity-inclination phase space.

Key Contributions and Results

• Three Conceptual Periods of Restart: The optimisation solutions for burst restarting reveal three distinct phases governed by spanwise vorticity structures:

  1. Breakup: External forces (or nonlinear terms) break up forward-inclined vortices into multiple layers at different wall distances.
  2. Counter-rotating Catch-up: During a "latent period," vortices in different layers advect at different speeds, causing counter-rotating vortices to catch up to one another.
  3. Co-rotating Catch-up: Subsequently, co-rotating vortices merge, forming a backward-inclined structure ready to burst again.
     The "weak-sense" model shows that breakup alone can trigger a re-burst, but the "strong-sense" model indicates that the merging of co-rotating vortices is essential for a full recovery of the burst cycle.

• Role of Nonlinearity: In real turbulence, the nonlinear terms do not manifest as a simple backward-tilting mechanism on average. Instead, their essential role is to increase the Linearly Available Energy (LAE) of a decaying state to a level sufficient for the onset of the next burst. The nonlinear terms effectively "recharge" the LAE, compensating for its consumption by the linear Orr process.

• Flow Patterns: Analysis of real turbulence confirms the mechanisms identified in the linearised models. Flow patterns during the restarting stage exhibit both the breakup of vortices and the merging of catching-up vortices. These effects are facilitated by nonlinearity, suggesting that the structures observed in the linearised models are manifestations of real flow structures causing burst restarts.

• LAE as a Diagnostic: The paper demonstrates that the inclination angle ($\Psi_v$), while useful for describing burst development, is inadequate for characterising restarts. Two states with similar intensity and inclination can have vastly different futures. The LAE ($E_v$) provides a more robust characterisation, showing that the burst cycle in the $I_v$-$\sqrt{E_v}$ subspace is a counter-clockwise loop where nonlinearity drives the system back up the energy axis during the restart stage.

Significance
The paper claims to advance the understanding of wall-bounded turbulence by decoupling the burst restart mechanism from the streak-based self-sustaining process. By treating the problem through a structure-neutral, linearised framework with forcing, the authors identify that the "minimal requirements" for sustaining turbulence are the breakup of vortices and the subsequent merging of catching-up vortices, driven by nonlinear effects that replenish the Linearly Available Energy. This suggests that the self-sustaining process can be viewed as two separate sub-processes: the bursting cycle itself (which does not depend on streaks) and the regeneration of streaks (which are by-products of bursts). The findings support the hypothesis that bursts can be sustained even if long streaks are artificially damped, provided the nonlinear mechanisms for vortex breakup and merging remain active.