25 October 2019 in Journal of Fluid Mechanics
We report direct numerical simulation results that clearly elucidate the mechanism that leads to curvature dependence of drag enhancement (DE) in viscoelastic turbulent Taylor–Couette flow. Change in the angular momentum transport and its inherent link to transitions in vortical flow structures have been explored to depict the influence of the curvature of the flow geometry on DE. Specifically, it has been demonstrated that a transition in vortical structures with increasing radius ratio leads to weakening and elimination of the small-scale Görtler vortices and development and better organization (occupying the entire gap) of large-scale Taylor vortices as also evinced by the patterns of angular momentum current. The commensurate change in DE and its underlying mechanism are examined by contributions of convective flux and polymeric stress to the angular momentum current. The present finding paves the way for capturing highly localized elastic turbulence structures in direct numerical simulation by increasing geometry curvature in traditional turbulent curvilinear flows.
8 May 2018 in Journal of Fluid Mechanics
Drag reduction (DR) in plane Couette flow (PCF) induced by the addition of flexible polymers has been studied via direct numerical simulation (DNS). The similarities and differences in the drag reduction features of PCF and plane Poiseuille flow (PPF) have been examined in detail, particularly in regard to the polymer-induced modification of large-scale structures (LSSs) in the near-wall turbulence. Specifically, it has been demonstrated that in the near-wall region, drag-reduced PCF has features similar to those of drag-reduced PPF; however, in the core region, intriguing differences are found between these two drag-reduced shear flows. Chief among these differences is the significant polymer stretch that arises from the enhanced exchanges between elastic potential energy and turbulent kinetic energy and the commensurate observation of peak values of the conformation tensor components $\unicode[STIX]{x1D60A}_{yy}$ and $\unicode[STIX]{x1D60A}_{zz}$ in this region. This finding is in stark contrast to that of drag-reduced PPF where the polymer stretch and the exchanges between elastic potential energy and turbulent kinetic energy in the core region are insignificant; to this end, in drag-reduced PPF, peak values of the conformation tensor components appear in the near-wall region. Therefore, this study paves the way for understanding the underlying flow physics in drag-reduced PCF, particularly in the context of elastic theory. Moreover, the longitudinal large-scale streaks at the channel centre of drag-reduced PCF are greatly strengthened due to the increased production/dissipation ratio; the LSS imprint effects on the near-wall flow of drag-reduced PCF monotonically increase as the Weissenberg number is enhanced.
1 December 2015 in Physics of Fluids
Direct numerical simulations have been performed to study the Taylor-Couette (TC) flow between two rotating, coaxial cylinders in the presence of a radial temperature gradient. Specifically, the influence of the buoyant force and the outer cylinder rotation on the turbulent TC flow system with the radius ratio η = 0.912 was examined. For the co-rotating TC flows with Rei (inner cylinder) =1000 and Reo (outer cylinder) =100, a transition pathway to highly turbulent flows is realized by increasing σ, a parameter signifying the ratio of buoyant to inertial force. This nonlinear flow transition involves four intriguing states that emerge in sequence as chaotic wavy vortex flow for σ = 0, wavy interpenetrating spiral flows for σ = 0.02 and 0.05, intermittent turbulent spirals for σ = 0.1 and 0.2, and turbulent spirals for σ = 0.4. Overall, the fluid motion changes from a centrifugally driven flow regime characterized by large-scale wavy Taylor vortices (TVs) to a buoyancy-dominated flow regime characterized by small-scale turbulent vortices. Commensurate changes in turbulence statistics and heat transfer are seen as a result of the weakening of large-scale TV circulations and enhancement of turbulent motions. Additionally, the influence of variation of the outer cylinder rotation, −500 < Reo < 500 in presence of buoyancy (σ = 0.1) with Rei = 1000, has been considered. Specifically, it is demonstrated that this variation strongly influences the azimuthal and axial mean flows with a weaker influence on the fluctuating fluid motions. Of special interest, here are the turbulent dynamics near the outer wall where a marked decrease of turbulence intensity and a sign inversion of the Reynolds stress Rrz are observed for the strongly counter-rotating regimes (Reo = − 300 and −500). To this end, it has been shown that the underlying flow physics for this drastic modification are associated with the modification of the correlation between the radial and axial fluctuating motions. In turn, the intriguing effects of this modification on the mean axial flow, turbulent statistics, force balance, and dynamic processes such as turbulence production and dissipation are discussed.
29 October 2026 in Journal of Fluid Mechanics
Wall-bounded turbulent flows are chaotic and multiscale, rendering fast prediction at high Reynolds numbers computationally prohibitive in applications such as wind farms. Classical data assimilation is based on repeated solutions of the governing equations and, thus, inherits this cost. Generative models learn the probability distribution of flow states, enabling scalable probabilistic reconstruction. Our generative framework couples a variational autoencoder with a diffusion transformer to generate four-dimensional spatiotemporal samples. Bayesian conditioning enables data assimilation without retraining and allows statistical constraints to be imposed through sampling. The framework is applied to a subdomain of turbulent plane Couette flow, where the corresponding direct numerical simulations in this generation region requires upper O left parenthesis 10 Superscript 6 Baseline right parenthesis $O(10^6)$ spatial degrees of freedom. Using upper O left parenthesis 10 right parenthesis $O(10)$ latent spatial degrees of freedom, the model achieves a compression ratio of upper O left parenthesis 10 Superscript 5 Baseline right parenthesis $O(10^5)$ , which is one to two orders of magnitude above prior reports. It reproduces single-point statistics up to fourth order and the energy spectra, as well as the intermittency and phase-sensitive structure captured by velocity-increment and upper Q $Q$ – upper R $R$ statistics. Two assimilation scenarios demonstrate that, when observations are statistically consistent with the prior, conditional diffusion models with the proposed sampling strategy preserve complex turbulent statistics in the posterior. However, enforcing these constraints while preserving physical fidelity and sample diversity introduces an inherent trade-off. Excessive conditioning can distort the learned prior, paralleling limitations of classical ensemble-based data assimilation, where this can likewise degrade the prior covariance. These results highlight both the promise of diffusion models as probabilistic surrogates and the challenges of conditioning them.