A new direct numerical simulation of a Poiseuille channel flow has been conducted for a friction Reynolds number of $10000$, using the pseudospectral code LISO. The mean streamwise velocity presents a long logarithmic layer, extending from 400 to 2500 wall units, longer than it was thought. The maximum of the intensity of the streamwise velocity increases with the Reynolds number, as expected. Also, the elusive second maximum of this intensity has not appeared yet. In case it exists, its location will be around $y^{+}\approx 120$, for a friction Reynolds number extrapolated to approximately 13 500. The small differences in the near-wall gradient of this intensity for several Reynolds numbers are related to the scaling failure of the dissipation, confirming this hypothesis. The scaling of the turbulent budgets in the center of the channel is almost perfect above 1000 wall units. Finally, the peak of the pressure intensity grows with the Reynolds number and does not scale in wall units. If the pressure at the wall is modeled as an inverse quadratic power of ${\text{Re}}_{\tau }$, then $p_{\infty }^{\prime +}\approx 4.7$ at the limit of infinite Reynolds number.

Rayleigh–Bénard convection (RBC) is a recurrent phenomenon in a number of industrial and geoscience flows and a well-studied system from a fundamental fluid-mechanics viewpoint. In the present work, we conduct numerical simulations to apply deep reinforcement learning (DRL) for controlling two-dimensional RBC using sensor-based feedback control. We show that effective RBC control can be obtained by leveraging invariant multi-agent reinforcement learning (MARL), which takes advantage of the locality and translational invariance inherent to RBC flows inside wide channels. MARL applied to RBC allows for an increase in the number of control segments without encountering the curse of dimensionality that would result from a naive increase in the DRL action-size dimension. This is made possible by the MARL ability for re-using the knowledge generated in different parts of the RBC domain. MARL is able to discover an advanced control strategy that destabilizes the spontaneous RBC double-cell pattern, changes the topology of RBC by coalescing adjacent convection cells, and actively controls the resulting coalesced cell to bring it to a new stable configuration. This modified flow configuration results in reduced convective heat transfer, which is beneficial in a number of industrial processes. We additionally draw comparisons with a conventional single-agent reinforcement learning (SARL) setup and report that in the same number of episodes, SARL is not able to learn an effective policy to control the cells. Thus, our work both shows the potential of MARL for controlling large RBC systems and demonstrates the possibility for DRL to discover strategies that move the RBC configuration between different topological configurations, yielding desirable heat-transfer characteristics.

The increase in emissions associated with aviation requires deeper research into novel sensing and flow-control strategies to obtain improved aerodynamic performances. In this context, data-driven methods are suitable for exploring new approaches to control the flow and develop more efficient strategies. Deep artificial neural networks (ANNs) used together with reinforcement learning, i.e., deep reinforcement learning (DRL), are receiving more attention due to their capabilities of controlling complex problems in multiple areas. In particular, these techniques have been recently used to solve problems related to flow control. In this work, an ANN trained through a DRL agent, coupled with the numerical solver Alya, is used to perform active flow control. The Tensorforce library was used to apply DRL to the simulated flow. Two-dimensional simulations of the flow around a cylinder were conducted and an active control based on two jets located on the walls of the cylinder was considered. By gathering information from the flow surrounding the cylinder, the ANN agent is able to learn through proximal policy optimization (PPO) effective control strategies for the jets, leading to a significant drag reduction. Furthermore, the agent needs to account for the coupled effects of the friction- and pressure-drag components, as well as the interaction between the two boundary layers on both sides of the cylinder and the wake. In the present work, a Reynolds number range beyond those previously considered was studied and compared with results obtained using classical flow-control methods. Significantly different forms of nature in the control strategies were identified by the DRL as the Reynolds number $Re$ increased. On the one hand, for $Re\le 1000$, the classical control strategy based on an opposition control relative to the wake oscillation was obtained. On the other hand, for $Re=2000$, the new strategy consisted of energization of the boundary layers and the separation area, which modulated the flow separation and reduced the drag in a fashion similar to that of the drag crisis, through a high-frequency actuation. A cross-application of agents was performed for a flow at $Re=2000$, obtaining similar results in terms of the drag reduction with the agents trained at $Re=1000$ and 2000. The fact that two different strategies yielded the same performance made us question whether this Reynolds number regime ($Re=2000$) belongs to a transition towards a nature-different flow, which would only admits a high-frequency actuation strategy to obtain the drag reduction. At the same time, this finding allows for the application of ANNs trained at lower Reynolds numbers, but are comparable in nature, saving computational resources.

The calculation of turbulence statistics is considered the key unsolved problem of fluid mechanics, i.e., precisely the computation of arbitrary statistical velocity moments from first principles alone. Using symmetry theory, we derive turbulent scaling laws for moments of arbitrary order in two regions of a turbulent channel flow. Besides the classical scaling symmetries of space and time, the key symmetries for the present work reflect the two well-known characteristics of turbulent flows: non-Gaussianity and intermittency. To validate the new scaling laws we made a new simulation at an unprecedented friction Reynolds number of 10 000, large enough to test the new scaling laws. Two key results appear as an application of symmetry theory, which allowed us to generate symmetry invariant solutions for arbitrary orders of moments for the underlying infinite set of moment equations. First, we show that in the sense of the generalization of the deficit law all moments of the streamwise velocity in the channel center follow a power-law scaling, with exponents depending on the first and second moments alone. Second, we show that the logarithmic law of the mean streamwise velocity in wall-bounded flows is indeed a valid solution of the moment equations, and further, all higher moments in this region follow a power law, where the scaling exponent of the second moment determines all higher moments. With this we give a first complete mathematical framework for all moments in the log region, which was first discovered about 100 years ago.

The control efficacy of deep reinforcement learning (DRL) compared with classical periodic forcing is numerically assessed for a turbulent separation bubble (TSB). We show that a control strategy learned on a coarse grid works on a fine grid as long as the coarse grid captures main flow features. This allows to significantly reduce the computational cost of DRL training in a turbulent-flow environment. On the fine grid, the periodic control is able to reduce the TSB area by 6.8%, while the DRL-based control achieves 9.0% reduction. Furthermore, the DRL agent provides a smoother control strategy while conserving momentum instantaneously. The physical analysis of the DRL control strategy reveals the production of large-scale counter-rotating vortices by adjacent actuator pairs. It is shown that the DRL agent acts on a wide range of frequencies to sustain these vortices in time. Last, we also introduce our computational fluid dynamics and DRL open-source framework suited for the next generation of exascale computing machines.

A direct numerical simulation of turbulent heat transfer in a channel flow has been conducted for a Reynolds number of is increased. Finally, the turbulent budgets are presented. A relation between the increment of the inner peak of the temperature intensities and the scaling failure of the dissipation and viscous diffusion terms is provided. The statistics of all simulations can be downloaded from the web page of our group: http://personales.upv.es/serhocal/.