Waviness and self-sustained turbulence in plane Couette-Poiseuille flow
M. Etchevest, P. Dmitruk, S. Karmakar, B. Semin, R. Godoy-Diana, J.E. Wesfreid · 2025
Read Waviness and self-sustained turbulence in plane Couette-Poiseuille flow by M. Etchevest, P. Dmitruk, S. Karmakar, B. Semin, R. Godoy-Diana, J.E. Wesfreid (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-03885.txt.
Introduction
Manuel Etchevest and co-authors simulate a shear flow on the edge of turbulence to test a classic theory of how turbulence sustains itself, in a 2025 paper. In Waleffe's model, rolls stir up streaks, the streaks grow wavy, and the waviness regenerates the rolls, closing the loop. The team ran direct numerical simulations of plane Couette-Poiseuille flow across Reynolds numbers from five hundred to nine hundred and forty, with varied starting disturbance sizes.
Streaks, rolls, and waviness all grew at first, on a timescale matching linear theory for small disturbances but shorter for large ones. At higher Reynolds numbers with strong starts, the flow settled into lasting turbulence; otherwise it calmed back down. The key quantitative result: streak waviness is a quadratic function of the roll strength, once the rolls are strong enough. That relation is exactly the closing link the self-sustaining model needs, the authors argue.
Catalog record
| CREATOR | M. Etchevest, P. Dmitruk, S. Karmakar, B. Semin, R. Godoy-Diana, J.E. Wesfreid |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-03885
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Waviness and self-sustained turbulence in plane Couette-Poiseuille flow M. Etchevest, P. Dmitruk, S. Karmakar, B. Semin, R. Godoy-Diana, J.E. Wesfreid Abstract Direct numerical simulations of a Couette-Poiseuille flow were performed near the transition to turbulence to investigate the nonlinear relationship between streak waviness and rolls. This relationship is a key step in Waleffe’s model for a self-sustaining process (SSP). Simulations were conducted for Reynolds numbers ranging from 500 to 940, and a range of initial perturbation amplitudes were used. In these simulations, the streaks, rolls, and streak waviness initially increase. The optimal time for this growth is close to the linear transient growth period for small perturbations, but much shorter for large, highly nonlinear perturbations. For higher Reynolds numbers and large initial perturbations, the velocity field reaches a turbulent steady state, whereas it relaxes to a laminar state in other cases. The main result is that the waviness of the streaks is a quadratic function of the rolls, provided that its value is sufficiently large. I Introduction Shear flows are ubiquitous in both natural and industrial settings, ranging from atmospheric and oceanic currents to engineering systems such as pipelines and aerodynamics around transport vehicles. Understanding the transition to turbulence and its dynamics in these flows is crucial to optimizing energy efficiency, improving control strategies, and predicting flow behavior under various conditions. From a fundamental point of view, the study of turbulence, and in particular the knowledge of the mechanisms of transition, remains with many open problems to be solved. Streamwise elongated zones of high and low velocity ( streaks ) and streamwise vortices ( rolls ), are recognized as fundamental components in the transition to turbulence in wall-bounded shear flows Elder 1960 ; Klebanoff et al. 1962 ; Kline et al. 1967 . These structures are critical for understanding both the onset of turbulence and its maintenance at high Reynolds numbers. Above a critical Reynolds number, the turbulent regime can be achieved and maintained through a mechanism called the Self-Sustaining Process (SSP) Jiménez and Moin 1991 ; Hamilton et al. 1995 ; Waleffe 1997 . This mechanism consists of a cycle that involves a dynamic interplay between the streaks and rolls, forming a closed loop of energy transfer, via linear and non-linear interactions. The formation of elongated streaks results from the lift-up effect, where streamwise vortices advect velocities of different magnitudes from the walls, creating a striped streak pattern Landahl 1975 . These streaks can then develop wavy instabilities due to shear caused by the velocity difference between adjacent fast and slow streaks Hamilton et al. 1995 ; Waleffe 1997 ; Farrell and Ioannou 2012 . Eventually, the instability can result in a breakdown of the streaks, transferring energy back to the rolls through non-linear interactions. Experimentally and numerically, the breakdown of streaks has been analyzed in boundary layer and channel flows Mans et al. 2005 ; Mans et al. 2007 ; Brandt et al. 2004 ; Cossu and Hwang 2017 ; Duriez et al. 2009 and also in the context of centrifugally instabilities Swearingen and Blackwelder 1987 ; Martinand et al. 2014 ; Dessup et al. 2018 . A canonical example of wall-bounded shear flows is Plane Couette Flow (PCF), where a channel flow has either the upper or lower wall moving at a fixed velocity. PCF is linearly stable, so its critical Reynolds number ( Re_{c} ) is infinite Davey 1973 . Furthermore, its simplicity and well-defined boundary conditions make it an ideal model for investigating fundamental mechanisms such as the SSP. Numerous experimental observations in various shear flows, such as plane Couette flow, Poiseuille channel flow (PPF), and pipe flow, show turbulence onset at Reynolds numbers significantly below the linear threshold. This phenomenon, known as subcritical transition to turbulence Manneville 2016 , refers to the onset of turbulence from finite amplitude perturbations in a linearly stable flow. Key elements in this process include transient growth, responsible for the non-modal amplification of perturbations Schmid and Henningson 2012 , and exact coherent structures (ECS), such as edge states that act as boundaries in phase space between the laminar and turbulent regimes Duguet et al. 2009 . In addition, pattern formation was studied for these flows showing the coexistence of turbulent and laminar regions simultaneously in the form of patches. Recent developments on this subject can be found in Benavides and Barkley 2025 ; Ciola et al. 2025 ; Gomé et al. 2023 Figure 1: (a,b) Schematic representation of a typical Couette-Poiseuille laminar flow considered. (c) Representative 3D snapshots of the streamwise velocity field U_{x} (blue is low speed, red is high speed) for three cases, a decaying laminar flow at Re=625 , a transitioning flow at Re=806 , and a fully turbulent case at Re=940 , from top to bottom, respectively. Rendering using the software VAPOR Li et al. 2019 ; sgpearse et al. 2023 . Combining PPF and PCF results in Plane Couette-Poiseuille Flow (CPF)—see Fig. 1 —, which has also been shown to be linearly stable under certain conditions, such as when the upper plate velocity U_{\operatorname{belt}} exceeds 70\% of the centerline velocity Potter 1966 . The turbulent state of this type of flow has been observed experimentally Klotz et al. 2017 ; Liu et al. 2021 ; Liu et al. 2024 and numerically Kim and Lee 2018 ; Kashyap et al. 2020 ; Cheng et al. 2023 , exhibiting a rich interplay of laminar, transitional, and turbulent regimes. The goal of this work is to examine quantitatively the nonlinear interaction between streaks and rolls near the transition to turbulence using Direct Numerical Simulations (DNS) of Couette-Poiseuille flow at moderate Reynolds numbers. The streaks formation and further break down is represented in Fig. 2 (a), which shows snapshots at times t0=0 , t1=4 , t2=7 , t3=11 , t4=35 and t5=100 of the streamwise velocity component in a numerical simulation obtained for Re=806 (Simulation T16 see Table 1 below in Section II ) in the plane parallel to the plates (x-y plane). Panel (b) shows the same streak formation on a plane normal to the base flow (y-z plane), adding the observation of the streamwise rolls. Here the heatmap corresponds to the streamwise velocity component, and the cross-stream flow fields are superimposed. Fig. 3 shows the evolution of the streaks energy content (the streamwise energy E_{x} ), highlighting the times corresponding to the t0-t5 snapshots in Fig. 2 . The streaks energy is defined as E_{x}=\frac{1}{L_{x}L_{y}L_{z}}\int_{0}^{L_{x}}\int_{0}^{L_{y}}\int_{0}^{L_{z}}\left(U_{x}(x,y,z)-\langle U\rangle_{x,z}(y)\right)^{2}dxdydz, (1) where \langle U\rangle_{x,z}(y) is the flow averaged over x and z , which depends only on the wall-normal coordinate y and time t . The first two snapshots in Fig. 2 (a) ( t0 and t1 ) show streak generation from the noisy initial condition with imposed artificial rolls, which are observed in panel (b). The streaks become wavy at t2 . At time t1 in panel (b), advection of high-velocity fluid from the top wall to the bottom and low-velocity fluid from the bottom wall to the top creates streaks with a mushroom-like structure, similar to temperature fluctuations in Rayleigh-Bénard convection. This process corresponds to the increasing streak energy that reaches a maximum in Fig. 3 . The instability then develops, generating wall-normal vorticity and vortices visible in