Wall-bounded turbulence needs not be long
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "Wall-bounded turbulence needs not be long".
Mira: Here's the gist of the paper. Wall-bounded turbulence needs not be long experiments on the regeneration of long streaks in flows in which they had originally been damped show that their initial growth is due to the interaction of the mean shear with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped.
Kai: Core Findings on Streak Regeneration The research investigates how streamwise velocity streaks regenerate after being inhibited by damping, revealing that their initial growth is driven by the interaction between the mean shear and long cross-flow velocities termed ‘rollers’ which persist even when streaks are damped.
Mira: Furthermore, simulations demonstrate that turbulence persists in cases where only long rollers are damped while long streaks remain, and these flows recover when the damping is removed.
Kai: Simulations involving both streaks and rollers longer than a certain length survive and regenerate, providing an interpretation based on their energy balance.
Title and authors: Kai: So, this paper is called "Wall-bounded turbulence needs not be long," and the authors are Carlos Martínez-López and Javier Jiménez from the School of Aeronautics at U. Politecnica Madrid. They're tackling a really fundamental idea about how turbulence works near a wall.
Mira: It sounds like they're challenging the traditional view that you need infinitely long structures to explain wall turbulence, which is a big thing because most classic models rely on those very long features > Wall-bounded turbulence needs not be long.
Lev: From an error correction standpoint, if we can model this better, it means our minimal self-sustaining units might not need to be infinitely large for the simulation to capture the correct physics.
Kai: Exactly, and what's really interesting here is that they are looking at experiments where they actually damped some of these long streaks and seeing what happens when they let them grow back.
Mira: They seem to be suggesting that you don't need those infinitely long structures in the first place for the growth process to start > Wall-bounded turbulence needs not be long.
Lev: If this holds up, it shifts how we think about the required fidelity for modeling these flows, because we might be able to use smaller basis sets without losing accuracy.
The paper's summary: Kai: So, the core finding here is that the initial growth of streaks isn't just about them spontaneously appearing; it’s due to an interaction between the mean shear and these long cross-flow velocities called 'rollers,' which stay around even when the streaks are damped > Wall-bounded turbulence needs not be long.
Mira: That's a key mechanism they found, and they also show that turbulence can actually persist even if you only dampen those rollers while leaving the long streaks in place > Wall-bounded turbulence needs not be long.
Lev: The fact that the flow recovers when damping is removed in that scenario suggests the dynamics aren't as fragile as we might assume, which would be good for designing robust error correction protocols > Wall-bounded turbulence needs not be long.
Kai: And they even tested this by damping both streaks and rollers longer than lambda plus x equal to six hundred and those structures survived and regenerated in terms of their energy balance > Wall-bounded turbulence needs not be long <ref:2607.08878#pg1,Wall-bounded turbulence needs not be long>.
Mira: That result is really telling because it contrasts with the classical minimal channels that assume infinitely long structures > Wall-bounded turbulence needs not be long.
Lev: So, they're showing that the physics is governed by something more localized than just those massive scale structures we usually think about > Wall-bounded turbulence needs not be long.
The paper's improvements: Kai: One of the main things they point out is that the new flows they create don't contain features longer than the damping wavelength, which supports a model where wall turbulence only depends on processes with an aspect ratio of order unity > Wall-bounded turbulence needs not be long.
Mira: That really means we can simplify our models by focusing on things that are roughly one-dimensional in their cross-section rather than assuming infinite geometry > Wall-bounded turbulence needs not be long.
Lev: From an engineering view, if the aspect ratio is order unity, the computational cost for simulating these flows drops significantly because you don't have to resolve those enormous structures > Wall-bounded turbulence needs not be long.
Kai: They also looked at how streaks actually get elongated, and they found that the new fluctuations are advected by the wall-normal velocity, but those advection velocities are higher than previously seen values > Wall-bounded turbulence needs not be long.
Mira: Then they propose an alternative idea where the recovery time scale is determined locally by the shear at each y position, which might explain why it looks like a constant advection velocity > Wall-bounded turbulence needs not be long.
Lev: If we can model that local determination of the recovery time scale, it means we can predict how quickly a structure will grow without needing to track everything across the entire domain > Wall-bounded turbulence needs not be long.
Conclusion: Kai: So, to wrap up this paper "Wall-bounded turbulence needs not be long," the main implication is that wall turbulence doesn't require input from very long scales at all, as the 'engine' itself has an aspect ratio of order unity > Wall-bounded turbulence needs not be long.
Mira: They conclude that the streaks are mostly recreated by the effect of residual rollers—those transverse velocity components—when you remove the damping > Wall-bounded turbulence needs not be long.
Lev: For real hardware, this means we can expect a more robust recovery mechanism when we truncate our models and reintroduce those residual scales > Wall-bounded turbulence needs not be long.
Kai: It’s different from the classical minimal channels that assume infinitely long structures, which is a big departure > Wall-bounded turbulence needs not be long.
Mira: They also note that both nonlinear transport and the linear effect of rollers contribute to streak reformation in those new flows > Wall-bounded turbulence needs not be long.
Lev: So, for future work, it seems like they need to build models that capture how these short structures organize themselves into groups at the long end of the spectrum > Wall-bounded turbulence needs not be long.
School of Aeronautics, U. PolitǬnica Madrid
physics.flu-dyn
Submitted: 2026-07-09
Updated: 2026-11-16
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
Importance score: 65/100
The gist: Wall-bounded turbulence needs not be long experiments on the regeneration of long streaks in flows in which they had originally been damped show that their initial growth is due to the interaction of…
Key concepts
- Streamwise Velocity Streaks
- These are regions in the flow where the velocity is aligned with the main direction of flow. The study focuses on how these streaks grow back after being artificially inhibited (damped) to understand their regeneration mechanism.
- Rollers
- 'Rollers' are long cross-flow velocities that remain present in the flow even when streaks are damped. They interact with the mean shear to drive the initial growth and regeneration of velocity streaks.
- Geometric Aspect Ratio
- This refers to the ratio of a structure's length to its width. The paper concludes that wall turbulence only requires structures where this aspect ratio is of order unity, meaning it doesn't depend on infinitely long structures.
Terminology
Summary
Wall-bounded turbulence needs not be long experiments on the regeneration of long streaks in flows in which they had originally been damped show that their initial growth is due to the interaction of the mean shear with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped. p. 0 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. Polit´ecnica Madrid, 28040 Madrid Spain
(Received 13 July 2026)
Experiments on the regeneration of long streaks in flows in which they had originally
been damped show that their initial growth is due to the interaction of the mean shear
with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped.Where the paper says it
Core Findings on Streak Regeneration
The research investigates how streamwise velocity streaks regenerate after being inhibited by damping, revealing that their initial growth is driven by the interaction between the mean shear and long cross-flow velocities termed ‘rollers’ which persist even when streaks are damped. Furthermore, simulations demonstrate that turbulence persists in cases where only long rollers are damped while long streaks remain, and these flows recover when the damping is removed. p. 0 Polit´ecnica Madrid, 28040 Madrid Spain
(Received 13 July 2026)
Experiments on the regeneration of long streaks in flows in which they had originally
been damped show that their initial growth is due to the interaction of the mean shear
with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped. More surprisingly, turbulence also persists in simulations in which only the long rollers
are damped while long streaks remain, and these flows are also able to recover when the
damping is removed. Finally, simulations are presented in which both streaks and rollers
longer than λ
+
x ≈ 600 are damped. They survive and regenerate, and an interpretation
in terms of their energy balance is provided. They survive and regenerate, and an interpretation
in terms of their energy balance is provided. In contraposition to the classical minimal
channels, which include infinitely long structures, these new flows do not contain features
longer than the damping wavelength, and support a model in which wall turbulence only
depends on processes for which the geometric aspect ratio is of order unity.Where the paper says it
Where the paper says it
Where the paper says it
Mechanisms of Elongation and Growth
The paper examines the first stages of how streaks become much longer than the bursts that presumably create them by starting with initial conditions where long streaks are inhibited and observing their evolution after inhibition is removed. The analysis shows that the new fluctuations are advected by the wall-normal velocity, but the implied advection velocities are higher than previously observed values. p. 3 It initially overshoots, as in figure 1(a), moves away from the wall and eventually relaxes
towards equilibrium. This outwards drift could be interpreted as that the new fluctuations
are advected by the wall-normal velocity, but the implied advection velocities, 1. 8uτ at t = 0), are higher than previously observed values, which are
closer to uτ (Flores & Jim´enez 2010; Lozano-Dur´an & Jim´enez 2014b). An alternative
explanation is that the time scale of the recovery is determined at each y by the local
shear, which also leads to an apparently constant advection velocity,
trec ∼ S
−1
(y) ≈ κy/uτ ⇒ dy/dt ≈ y/trec ≈ uτ /κ ≈ 2. Figure 2(a) is
the evolution of the streamwise velocity spectrum during the initial stages of the regeneration. The streaks grow fastest near the wall, where they have almost reached their
equilibrium length after 0.Where the paper says it
Where the paper says it
Where the paper says it
Role of Different Flow Features
The study explores the dependence of streak growth on local shear by releasing only some distances from the wall to observe independent elongation of layers. The outer-released case does not grow for the wall distances that remain unreleased. p. 5 Figure
5(c) displays the maximum elongation velocities of the partially released cases, compared
to the fully released one from figure 3(c), and shows that the similarity is quantitative. The outer-released case in figure 5(a) does not grow for the wall distances that remain Damping the rollers
The discussion in the previous section suggests that long rollers, rather than long
streaks, are the key ingredient of wall turbulence because, unless inhibited by damping,
they deform the mean profile to also build long streaks. However, although the experiments discussed in §3 show that streaks are not necessary if rollers are present, they
do not prove that the rollers are required. Turbulence survives, implying that the only velocity fluctuations
necessary for wall turbulence are short. Recovery experiments of these fully truncated
flows show that long rollers are maintained by nonlinear turbulent transport from shorter
units, while the streaks reform both through nonlinear transport and by the linear effect
of the rollers on the mean shear.Where the paper says it
Where the paper says it
Where the paper says it
Implications for Turbulence Models
The findings suggest that wall-bounded turbulence requires essentially no input from long scales, as the turbulence ‘engine’ has an aspect ratio of order unity. The paper concludes that the streaks are predominantly recreated by the effect of residual long ‘rollers’ of the transverse velocities when damping is removed. p. 8 Conclusions
This paper continues the damping experiments in Jim´enez (2022), which showed that
the long streaks of the streamwise velocity in wall-bounded flows are essentially passive
products of shorter structures with λ
+
x ≲ 600. We have first performed recovery experiments in which the damping is removed, and shown that the streaks are predominantly
recreated by the effect of residual long ‘rollers’ of the transverse velocities. Turbulence survives, implying that the only velocity fluctuations
necessary for wall turbulence are short. Recovery experiments of these fully truncated
flows show that long rollers are maintained by nonlinear turbulent transport from shorter
units, while the streaks reform both through nonlinear transport and by the linear effect
of the rollers on the mean shear. Recovery experiments of these fully truncated
flows show that long rollers are maintained by nonlinear turbulent transport from shorter
units, while the streaks reform both through nonlinear transport and by the linear effect
of the rollers on the mean shear. Note that the new truncated flows are different from the
minimal units in Jim´enez & Moin (1991), which retain infinitely long structures.Where the paper says it
Where the paper says it
Where the paper says it
Damping and Robustness of Structures
When damping both long components, the flow is shown to recover well because truncation is more robust than only damping the roller. The streaks grow simultaneously with the action of the roller on the mean shear initially predominant, and later by nonlinear transfer. p. 7 The roller and the streak grow simultaneously, with
the action of the roller on the mean shear initially predominant (figure 7. Note that
only the long roller harmonics can initially participate in this essentially linear process,
and that only later does the nonlinear transfer become relevant. Adapting their analysis to our case, the picture is of shorter
structures organising themselves into groups that populate the long end of the spectrum. The above truncation experiments suggest that these groups do not feed back into the
shorter wavelengths, which thus constitute an essentially autonomous turbulence ‘engine’. Adapting their analysis to our case, the picture is of shorter
structures organising themselves into groups that populate the long end of the spectrum. The above truncation experiments suggest that these groups do not feed back into the
shorter wavelengths, which thus constitute an essentially autonomous turbulence ‘engine’. It was
hypothesised by Lozano-Dur´an & Jim´enez (2014) that these features mimic the effect of
long but finite structures in natural flow, but even the infinite component is absent in
the new short units. This work was supported by the European Research Council under the Caust grant
ERC-AdG-101018287.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
REFERENCES
Adrian, R. J. 2007 Hairpin vortex organization in wall turbulence. p. 9 10 Carlos Mart´ınez–L´opez and Javier Jim´enez
REFERENCES
Adrian, R. 2007 Hairpin vortex organization in wall turbulence. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Adrian, R. J., Meinhart, C. p. 9 19, 041301. , Meinhart, C. , Meinhart, C. & Tomkins, C. & Tomkins, C. 2000 Vortex organization in the outer region
of the turbulent boundary layer. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. Under consideration for publication in J. Fluid Mech. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
del Alamo, J. C., Jim ´ enez, J. p. 0 Fluid Mech. 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. Fluid Mech. 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. Fluid Mech. 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. , Jim ´ enez, J. , Zandonade, P. , Jim ´ enez, J. , Zandonade, P. , Jim ´ enez, J. , Zandonade, P. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. 2014), but
Couette flow has much longer streaks (Pirozzoli et al. 2014; Lee & Moser 2018), and the
existence of flows with similar wall-normal velocity structures (Sillero et al. & Moser, R. ´ 2004 Scaling of the energy
spectra of turbulent channels. Fluid Mech. 500, 135–144.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
del Alamo, J. C., Jim ´ enez, J., Zandonade, P. & Moser, R. D.’ 2006 Self-similar vortex clusters in the turbulent logarithmic region in J. p. 9 & Moser, R. ´ 2006 Self-similar vortex
clusters in the turbulent logarithmic region. Fluid Mech. 561, 329–358.Where the paper says it
Where the paper says it
del Alamo, J. C., Jim ´ enez, J., Zandonade, P. & Moser, R. D.’ 2006 Linear energy amplification in turbulent channels in J. p. 9 & Jim ´ enez, J. ´ 2006 Linear energy amplification in turbulent channels. Fluid Mech. 559, 205–213.Where the paper says it
Where the paper says it
Baltzer, J. R., Adrian, R. p. 1 (2012) showed that the down- and
up-draft components of bursts are statistically found in pairs associated with quasistreamwise vortices located along the edges of streamwise-velocity streaks that organise
them longitudinally. Adrian et al. , Adrian, R. 2013 Structural organization of large and very large
scales in turbulent pipe flow simulation. Fluid Mech. 720, 236–279.Where the paper says it
Where the paper says it
Where the paper says it
Butler, K. M. & Farrell, B. p. 1 An obvious possibility is that streaks are wakes left in the mean shear by differentially
advected bursts (del Alamo ´ et al. Optimal initial conditions for the transient
linearised growth of perturbations of the mean velocity profile of a turbulent channel
result in infinitely long streaks and shorter bursts (Butler & Farrell 1993; del Alamo & ´
Jim´enez 2006; Jim´enez 2013), although this is mostly because longer features decay more
slowly, and Vaughan et al. & Farrell, B. 1993 Optimal perturbations and streak spacing in wallbounded shear flow. 1993 Optimal perturbations and streak spacing in wallbounded shear flow. Fluids A 5, 774–777.Where the paper says it
Where the paper says it
Where the paper says it
Cole, J. D. 1968 Perturbation methods in applied mathematics in Blaisdell <ref:2607. p. 9 Fluids A 5, 774–777. 1968 Perturbation methods in applied mathematics.Where the paper says it
Flores, O. & Jimenez, J. p. 9 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1968 Perturbation methods in applied mathematics. & Jimenez, J. & Jimenez, J. ´ 2010 Hierarchy of minimal flow units in the logarithmic layer. ´ 2010 Hierarchy of minimal flow units in the logarithmic layer. Fluids 22, 071704.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Hamilton, J. M., Kim, J. & Waleffe, F. p. 1 They showed that both streaks and vortices are required, and
numerous minimal kinematic and dynamical models have resulted in a widely accepted
minimal regeneration cycle (Jim´enez 1994; Hamilton et al. 1995; Waleffe 1997). & Waleffe, F. 1995 Regeneration mechanisms of near-wall turbulence structures. Fluid Mech. 287, 317–348.Where the paper says it
Where the paper says it
Where the paper says it
Hoyas, S. & Jimenez, J.’ 2006 Scaling of the velocity fluctuations in turbulent channels up to Reτ = 2003 in Phys. p. 9 & Jimenez, J. ´ 2006 Scaling of the velocity fluctuations in turbulent channels up to
Reτ = 2003. ´ 2006 Scaling of the velocity fluctuations in turbulent channels up to
Reτ = 2003. Fluids 18, 011702.Where the paper says it
Where the paper says it
Hoyas, S. & Jimenez, J.’ 2008 Reynolds number effects on the Reynolds-stress budgets in turbulent channels in Phys. p. 9 & Jimenez, J. ´ 2008 Reynolds number effects on the Reynolds-stress budgets in
turbulent channels. ´ 2008 Reynolds number effects on the Reynolds-stress budgets in
turbulent channels. Fluids 20, 101511.Where the paper says it
Where the paper says it
Hutchins, Nicholas & Marusic, Ivan 2007 Large-scale influences in near-wall turbulence in Phil. Trans. Roy. Soc. A 365, 647–664 <ref:2607. p. 9 Hutchins, Nicholas & Marusic, Ivan 2007 Large-scale influences in near-wall turbulence. A 365, 647–664.Where the paper says it
Jimenez, J.’ 1994 On the structure and control of near wall turbulence in Phys. p. 9 Jimenez, J. ´ 1994 On the structure and control of near wall turbulence. ´ 1994 On the structure and control of near wall turbulence. Fluids 6, 944–953.Where the paper says it
Where the paper says it
Jimenez, J.’ 2013 How linear is wall-bounded turbulence? p. 9 Jimenez, J. ´ 2013 How linear is wall-bounded turbulence? ´ 2013 How linear is wall-bounded turbulence? Fluids 25, 110814.Where the paper says it
Where the paper says it
Jimenez, J.’ 2018 Coherent structures in wall-bounded turbulence in J. p. 9 Jimenez, J. ´ 2018 Coherent structures in wall-bounded turbulence. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Jimenez, J. & Moin, P.’ 1991 The minimal flow unit in near-wall turbulence in J. p. 9 Jimenez, J. ´ 1991 The minimal flow unit in near-wall turbulence. Fluid Mech. 225, 213–240.Where the paper says it
Where the paper says it
Jimenez, J. & Pinelli, A. p. 1 2 Carlos Mart´ınez–L´opez and Javier Jim´enez
two short bursting vortices and a streak that is infinitely long when compared to the
computational box. Jim´enez & Pinelli (1999) modified different terms of the evolution
equations of such a unit. & Pinelli, A. ´ 1999 The autonomous cycle of near-wall turbulence. Fluid Mech. 389, 335–359.Where the paper says it
Where the paper says it
Where the paper says it
Kim, H. T., Kline, S. p. 0 In both cases, the
ratio between the inner and outer scales is proportional to some power of the Reynolds
number that provides a potentially infinite expansion parameter. Less often discussed is the elongated geometry of the streamwise velocity streaks (Kline
et al. Wall-bounded turbulence provides the additional example of the momentum transfer across a hierarchy of attached eddies (Townsend 1961). In both cases, the
ratio between the inner and outer scales is proportional to some power of the Reynolds
number that provides a potentially infinite expansion parameter. & Reynolds, W. 1971 The production of turbulence near a
smooth wall in a turbulent boundary layer. Fluid Mech. 50, 133–160.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Kim, J., Moin, P. & Moser, R. D. 1987 Turbulence statistics in fully developed channel flow at low Reynolds number in J. p. 9 & Moser, R. 1987 Turbulence statistics in fully developed channel flow
at low Reynolds number. Fluid Mech. 177, 133–166.Where the paper says it
Where the paper says it
Kline, S J, Reynolds, W C, Schraub, F A & Runstadler, P. W 1967 The structure of turbulent boundary layers in J. p. 9 Kline, S J, Reynolds, W C, Schraub, F A & Runstadler, P. W 1967 The structure of
turbulent boundary layers. Fluid Mech. 30, 741–773.Where the paper says it
Where the paper says it
Kolmogorov, A. N. 1941 The local structure of turbulence in incompressible viscous fluids at very large Reynolds numbers in Dokl. p. 9 Kolmogorov, A. 1941 The local structure of turbulence in incompressible viscous fluids at
very large Reynolds numbers. 1941 The local structure of turbulence in incompressible viscous fluids at
very large Reynolds numbers. SSSR 30, 301–305, reprinted in Proc. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Lee, M. & Moser, R. D. 2018 Extreme-scale motions in turbulent plane Couette flows in J. p. 9 & Moser, R. 2018 Extreme-scale motions in turbulent plane Couette flows. Fluid Mech. 842, 128–145.Where the paper says it
Where the paper says it
Lozano-Duran, A., Flores, O. p. 1 1995; Waleffe 1997). This was
extended to the logarithmic layer by Flores & Jim´enez (2010), but the length of these
minimal units is never much greater than their spanwise width. & Jim ´ enez, J. ´ 2012 The three-dimensional structure of momentum transfer in turbulent channels. Fluid Mech. 694, 100–130.Where the paper says it
Where the paper says it
Where the paper says it
Lozano-Duran, A. & Jimenez, J.’ 2014 Effect of the computational domain on direct simulations of turbulent channels up to Reτ = 4200 in Phys. p. 9 & Jim ´ enez, J. ´ 2014 Effect of the computational domain on direct simulations of turbulent channels up to Reτ = 4200. ´ 2006 Scaling of the velocity fluctuations in turbulent channels up to
Reτ = 2003. Fluids 18, 011702.Where the paper says it
Where the paper says it
Lozano-Duran, A. & Jimenez, J.’ 2014b Time-resolved evolution of coherent structures in turbulent channels: characterization of eddies and cascade in J. p. 9 & Jim ´ enez, J. ´ 2014b Time-resolved evolution of coherent structures in
turbulent channels: characterization of eddies and cascade. Fluid Mech. 759, 432–471.Where the paper says it
Where the paper says it
Mizuno, Y. 2016 Spectra of energy transport in turbulent channel flows for moderate Reynolds numbers in J. p. 9 759, 432–471. 2016 Spectra of energy transport in turbulent channel flows for moderate Reynolds
numbers. Fluid Mech. 805, 171–187.Where the paper says it
Where the paper says it
Pirozzoli, S., Bernardini, M. & Orlandi, P. 2014 Turbulence statistics in Couette flow at high Reynolds number in J. p. 9 & Moser, R. 1987 Turbulence statistics in fully developed channel flow
at low Reynolds number. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Schoppa, W. & Hussain, F. p. 0 The small-scale limit of this relation has been studied the most; meandering instabilities
with wavelengths of the order of the width of the streaks create bursts with aspect ratios
of order unity (Kim et al. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. Jimenez, J. ´ 1994 On the structure and control of near wall turbulence. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Sillero, J. A., Jimenez, J. p. 9 1968 Perturbation methods in applied mathematics. & Jimenez, J. 1971), but the two-point correlation of the latter is much shorter than that of the former (Jim´enez 2018), and the reason
for the discrepancy is not well understood. Admittedly, this is a less natural asymptotic
problem than the two examples above, because the ratio of the two correlation lengths
in boundary layers and channels is at most of the order of 10–20 (Sillero et al. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Toh, S. & Itano, T. p. 1 A possibility is that the causal scale interactions take place across different wall distances, because a burst in the outer part of the boundary layer is large when compared to
the width of the near-wall streaks, and the footprints of these outer structures reach the
wall (Townsend 1961; Hoyas & Jim´enez 2006; Hutchins & Marusic 2007). Conversely, Toh
& Itano (2005) proposed that the outer streaks are created by a ‘bunching’ instability of
near-wall ones. A possibility is that the causal scale interactions take place across different wall distances, because a burst in the outer part of the boundary layer is large when compared to
the width of the near-wall streaks, and the footprints of these outer structures reach the
wall (Townsend 1961; Hoyas & Jim´enez 2006; Hutchins & Marusic 2007). Conversely, Toh
& Itano (2005) proposed that the outer streaks are created by a ‘bunching’ instability of
near-wall ones. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Townsend, A. A. 1961 Equilibrium layers and wall turbulence in J. p. 0 Turbulence is no stranger to this segregation:
energy is generated at large scales and dissipates at the smallest ones, and one of the
central objects in turbulence research is the cascade connecting the two regimes (Kolmogorov 1941). Wall-bounded turbulence provides the additional example of the momentum transfer across a hierarchy of attached eddies (Townsend 1961). 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Vaughan, T. L., Farrell, B. p. 1 An obvious possibility is that streaks are wakes left in the mean shear by differentially
advected bursts (del Alamo ´ et al. Optimal initial conditions for the transient
linearised growth of perturbations of the mean velocity profile of a turbulent channel
result in infinitely long streaks and shorter bursts (Butler & Farrell 1993; del Alamo & ´
Jim´enez 2006; Jim´enez 2013), although this is mostly because longer features decay more
slowly, and Vaughan et al. Optimal initial conditions for the transient
linearised growth of perturbations of the mean velocity profile of a turbulent channel
result in infinitely long streaks and shorter bursts (Butler & Farrell 1993; del Alamo & ´
Jim´enez 2006; Jim´enez 2013), although this is mostly because longer features decay more
slowly, and Vaughan et al. (2015), among others, have presented plausible self-sustaining
models of wall turbulence based on time-dependent infinitely long streaks. 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
Where the paper says it
Waleffe, F. 1997 On a self-sustaining process in shear flows in Phys. p. 0 08878v1 [physics. flu-dyn] 9 Jul 2026
self-sustaining unit of the buffer layer of plane Poiseuille turbulence contains one or 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Where the paper says it
--- Page 1 ---
Under consideration for publication in J. Fluid Mech. 1 Wall-bounded turbulence needs not be long By Carlos Mart´ınez–L´opez and Javier Jim´enez School of Aeronautics, U. p. 0 Fluid Mech. 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. 1
Wall-bounded turbulence needs not be long
By Carlos Mart´ınez–L´opez and Javier Jim´enez
School of Aeronautics, U. Polit´ecnica Madrid, 28040 Madrid Spain
(Received 13 July 2026)
Experiments on the regeneration of long streaks in flows in which they had originally
been damped show that their initial growth is due to the interaction of the mean shear
with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped.Where the paper says it
Where the paper says it
Experiments on the regeneration of long streaks in flows in which they had originally been damped show that their initial growth is due to the interaction of the mean shear with long cross-flow velocities (‘rollers’) that remain even when the streaks are damped <ref:2607.08878#pg2>
More surprisingly, turbulence also persists in simulations in which only the long rollers are damped while long streaks remain, and these flows are also able to recover when the damping is removed <ref:2607.08878#pg2>
Finally, simulations are presented in which both streaks and rollers longer than λ+x ≈ 600 are damped. They survive and regenerate, and an interpretation in terms of their energy balance is provided <ref:2607. p. 0 Finally, simulations are presented in which both streaks and rollers
longer than λ
+
x ≈ 600 are damped. They survive and regenerate, and an interpretation
in terms of their energy balance is provided.Where the paper says it
In contraposition to the classical minimal channels, which include infinitely long structures, these new flows do not contain features longer than the damping wavelength, and support a model in which wall turbulence only depends on processes for which the geometric aspect ratio is of order unity <ref:2607.08878#pg2>
--- Page 2 ---
Carlos Mart´ınez–L´opez and Javier Jim´enez 2 Carlos Mart´ınez–L´opez and Javier Jim´enez The small-scale limit of this relation has been studied the most; meandering instabilities with wavelengths of the order the width of the streaks create bursts with aspect ratios of order unity <ref:2607.08878#pg2>
A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal self-sustaining unit of the buffer layer of plane Poiseuille turbulence contains one or two short bursting vortices and a streak that is infinitely long when compared to the computational box <ref:2607.08878#pg2>
Jim´enez & Pinelli (1999) modified different terms of the evolution equations of such a unit. They showed that both streaks and vortices are required, and numerous minimal kinematic and dynamical models have resulted in a widely accepted minimal regeneration cycle <ref:2607. p. 1 Jim´enez & Pinelli (1999) modified different terms of the evolution
equations of such a unit. They showed that both streaks and vortices are required, and
numerous minimal kinematic and dynamical models have resulted in a widely accepted
minimal regeneration cycle (Jim´enez 1994; Hamilton et al.Where the paper says it
This was extended to the logarithmic layer by Flores & Jim´enez (2010), but the length of these minimal units is never much greater than their spanwise width <ref:2607.08878#pg2>
An obvious possibility is that streaks are wakes left in the mean shear by differentially advected bursts (del Alamo ´ et al. 2004) <ref:2607. p. 0 1971; Jim´enez & Moin 1991; Schoppa & Hussain 2002), but
the route from short bursts to long streaks has been explored less. A useful model was introduced by Jim´enez & Moin (1991), who showed that a minimal
arXiv:2607.Where the paper says it
Optimal initial conditions for the transient linearised growth of perturbations of the mean velocity profile of a turbulent channel result in infinitely long streaks and shorter bursts <ref:2607.08878#pg2>
However, this optimal amplification also requires an infinitely long time (del Alamo & Jim´enez 2006), ´ and del Alamo ´ et al. (2006) noted that the lifetime of individual bursts is too short for their wakes to explain the observed streak length <ref:2607. p. 1 However, this
optimal amplification also requires an infinitely long time (del Alamo & Jim´enez 2006), ´
and del Alamo ´ et al. (2006) noted that the lifetime of individual bursts is too short for
their wakes to explain the observed streak length.Where the paper says it
The problem of creating long streaks becomes that of creating long packets of bursts <ref:2607.08878#pg2>
In fact, there is ample evidence that long streaks are associated with several bursts <ref:2607.08878#pg2>
Del Alamo ´ et al. (2006) and Lozano-Dur´an et al. p. 1 Del Alamo ´ et al. (2006) and Lozano-Dur´an et al. (2006) and Lozano-Dur´an et al. (2012) showed that the down- and
up-draft components of bursts are statistically found in pairs associated with quasistreamwise vortices located along the edges of streamwise-velocity streaks that organise
them longitudinally.Where the paper says it
Where the paper says it
Adrian et al. 2000 and Adrian (2007) propose a plausible model in terms of ‘hairpin’ packets, and demonstrate the formation of short packets in the absence of background fluctuations <ref:2607. p. 1 Adrian et al. (2000) and Adrian (2007) propose a plausible model
in terms of ‘hairpin’ packets, and demonstrate the formation of short packets in the
absence of background fluctuations.Where the paper says it
Composite packets in pipes are described by Baltzer et al. 2013, organised by large-scale weak streamwise rolls <ref:2607. p. 1 Composite packets in pipes are described by Baltzer
et al. (2013), organised by large-scale weak streamwise rolls.Where the paper says it
--- Page 3 ---
Case goes to Ubh/ν Reτ (Λx × Λz) ∆x + ∆z + ∆y + max λf /h λ+ f(x, y, z) grid points F550 SS550:10/15 10060 550 (8 × 4)πh 8.9 5.5 6.7 ∞ ∞ 1536 × 257 × 1, five or more words copied word for word from that page>
Table of the parameters of the reference and damped simulations used as initial conditions for the recovery experiments are presented in Table 1 <ref:2607.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Enhanced turbulence modeling for wall-bounded flows by incorporating short-scale structures as essential components rather than assuming infinitely long ones, as suggested by
the only velocity fluctuations necessary for wall turbulence are short.
This allows AI systems to accurately predict flow behavior in regimes where classical models based on long streaks fail. -
Improved simulation capabilities for recovering turbulent states after damping by using a two-stage recovery process: first allowing the flow to settle into a short-streak equilibrium, and then removing the damping, which is shown to be
predominantly recreated by the effect of residual long rollers of the transverse velocities.
This provides a more physically realistic mechanism for streak regeneration. -
Development of anisotropic turbulence engines in AI models capable of understanding how
short structures organising themselves into groups that populate the long end of the spectrum
sustain turbulence, as seen in truncation experiments wherethese groups do not feed back into the shorter wavelengths, which thus constitute an essentially autonomous turbulence engine.
This enables better prediction of flow stability and energy transfer mechanisms. -
Ability to distinguish between different damping strategies by analyzing how flow components respond to structural inhibition; for instance, the paper suggests that
keeping one of the two velocity components of the Reynolds stress while damping the other one is less stable than damping both of them.
This allows for more nuanced control over turbulent features in computational simulations. -
Refinement of streak elongation prediction by modeling its transient phase, which involves a
fast initial transient followed by a slower evolution,
and understanding thatrapid streak elongation only lasts until the initial length increases by a factor of approximately three.
This provides tighter constraints on the timescales for structural growth in turbulent boundary layers.
Abstract
Experiments on the regeneration of long streaks in flows in which they had originally been damped show that their initial growth is due to the interaction of the mean shear with weak long cross-flow velocities (`rollers') that remain even when the streaks are damped. More surprisingly, turbulence also persists in simulations in which only the long rollers are damped while long streaks remain, and these flows are also able to recover when the damping is removed. Finally, simulations are presented in which both streaks and rollers longer than λ x+ about 600 are damped. They survive and regenerate, and an interpretation in terms of their energy balance is provided. In contraposition to the classical minimal channels, which include infinitely long structures, these new flows do not contain any feature longer than the damping wavelength, and support a model in which wall turbulence only depends on processes for which the geometric λ x-λ z aspect ratio is of order unity.
Related papers
- Data-driven impeller model for efficient large eddy simulations of metastable von K'arm'an flows
- Numerical study of nitrogen oxides chemistry during plasma assisted combustion in a sequential combustor
- Emergent Transfer of a Physics Foundation Model from Simulation to Laboratory Turbulence
- Early onset of secondary shear instability in Kelvin-Helmholtz braids at high Reynolds number
- Rotational covariance, continuous spectrum and non-modal growth of three-dimensional planar weak viscous shock layers
- Deep Learning-based Algebraic Reynolds Stress Closures for RANS Simulations of Turbulent Flows