On the Underlying Drag-Reduction Mechanisms of Flow-Control Strategies in a Transitional Channel Flow: Temporal Approach

We have demonstrated the effectiveness of reinforcement learning (RL) in bluff body flow control problems both in experiments and simulations by automatically discovering active control strategies for drag reduction in turbulent flow. Specifically, we aimed to maximize the power gain efficiency by properly selecting the rotational speed of two small cylinders, located parallel to and downstream of the main cylinder. By properly defining rewards and designing noise reduction techniques, and after an automatic sequence of tens of towing experiments, the RL agent was shown to discover a control strategy that is comparable to the optimal strategy found through lengthy systematically planned control experiments. Subsequently, these results were verified by simulations that enabled us to gain insight into the physical mechanisms of the drag reduction process. While RL has been used effectively previously in idealized computer flow simulation studies, this study demonstrates its effectiveness in experimental fluid mechanics and verifies it by simulations, potentially paving the way for efficient exploration of additional active flow control strategies in other complex fluid mechanics applications.

Download Full-text

Deep Reinforcement Learning for Active Flow Control around a Circular Cylinder Using Unsteady-mode Plasma Actuators

10.22541/au.160912628.89631259/v1 ◽

2020 ◽

Author(s):

Mohamed Elhawary

Keyword(s):

Fluid Dynamics ◽

Reinforcement Learning ◽

Drag Reduction ◽

Flow Control ◽

Control Strategies ◽

Active Flow Control ◽

Plasma Actuators ◽

Burst Frequency ◽

Computational Fluid Dynamics Cfd ◽

Active Flow

Deep reinforcement learning (DRL) algorithms are rapidly making inroads into fluid mechanics, following the remarkable achievements of these techniques in a wide range of science and engineering applications. In this paper, a deep reinforcement learning (DRL) agent has been employed to train an artificial neural network (ANN) using computational fluid dynamics (CFD) data to perform active flow control (AFC) around a 2-D circular cylinder. Flow control strategies are investigated at a diameter-based Reynolds number Re_D = 100 using advantage actor-critic (A2C) algorithm by means of two symmetric plasma actuators located on the surface of the cylinder near the separation point. The DRL agent interacts with the computational fluid dynamics (CFD) environment through manipulating the non-dimensional burst frequency (f+) of the two plasma actuators, and the time-averaged surface pressure is used as a feedback observation to the deep neural networks (DNNs). The results show that a regular actuation using a constant non-dimensional burst frequency gives a maximum drag reduction of 21.8 %, while the DRL agent is able to learn a control strategy that achieves a drag reduction of 22.6%. By analyzing the flow-field, it is shown that the drag reduction is accompanied with a strong flow reattachment and a significant reduction in the mean velocity magnitude and velocity fluctuations at the wake region. These outcomes prove the great capabilities of the deep reinforcement learning (DRL) paradigm in performing active flow control (AFC), and pave the way toward developing robust flow control strategies for real-life applications.

Download Full-text

Pumping or drag reduction?

Journal of Fluid Mechanics ◽

10.1017/s0022112009007629 ◽

2009 ◽

Vol 635 ◽

pp. 171-187 ◽

Cited By ~ 50

Author(s):

JÉRÔME HŒPFFNER ◽

KOJI FUKAGATA

Keyword(s):

Pressure Gradient ◽

Drag Reduction ◽

Flow Control ◽

Channel Flow ◽

Travelling Waves ◽

Weakly Nonlinear ◽

Flow Response ◽

Flow Motion ◽

Strongly Connected ◽

Wall Deformation

Two types of wall actuation in channel flow are considered: travelling waves of wall deformation (peristalsis) and travelling waves of blowing and suction. The flow response and its mechanisms are analysed using nonlinear and weakly nonlinear computations. We show that both actuations induce a flux in the channel in the absence of an imposed pressure gradient and can thus be characterized as pumping. In the context of flow control, pumping and drag reduction are strongly connected, and we seek to define them properly based on these two actuation examples. Movies showing the flow motion for the two types of actuation are available with the online version of this paper (journals.cambridge.org/FLM).

Download Full-text

Drag-Reduction Mechanisms of Suction-and-Oscillatory-Blowing Flow Control

AIAA Journal ◽

10.2514/1.j052903 ◽

2014 ◽

Vol 52 (11) ◽

pp. 2491-2505 ◽

Cited By ~ 14

Author(s):

David Schatzman ◽

Jacob Wilson ◽

Eran Arad ◽

Avraham Seifert ◽

Tom Shtendel

Keyword(s):

Drag Reduction ◽

Flow Control ◽

Reduction Mechanisms

Download Full-text

Drag Reduction for Trucks by Active Flow Control of the Wake Behind the Trailer

Proceedings of the Sixth International Symposium On Turbulence, Heat and Mass Transfer ◽

10.1615/ichmt.2009.turbulheatmasstransf.1380 ◽

2009 ◽

Cited By ~ 5

Author(s):

M. El-Alti ◽

P. Kjellgren ◽

Lars Davidson

Keyword(s):

Drag Reduction ◽

Flow Control ◽

Active Flow Control ◽

Active Flow

Download Full-text

DRAG REDUCTION IN TURBULENT CHANNEL FLOW WITH TWO-TIERED SUPERHYDROPHOBIC WALLS

10.1615/tfec2017.bev.018117 ◽

2017 ◽

Author(s):

Richard Perkins ◽

Julie Crockett ◽

Daniel Maynes

Keyword(s):

Drag Reduction ◽

Channel Flow ◽

Turbulent Channel Flow

Download Full-text

Noise Reduction Effect of Superhydrophobic Surfaces with Streamwise Strip of Channel Flow

Applied Sciences ◽

10.3390/app11093869 ◽

2021 ◽

Vol 11 (9) ◽

pp. 3869

Author(s):

Chen Niu ◽

Yongwei Liu ◽

Dejiang Shang ◽

Chao Zhang

Keyword(s):

Reynolds Number ◽

Drag Reduction ◽

Noise Reduction ◽

Channel Flow ◽

Superhydrophobic Surface ◽

Sound Field ◽

Flow Noise ◽

Slip Boundary ◽

Eddy Simulation ◽

Reduction Effect

Superhydrophobic surface is a promising technology, but the effect of superhydrophobic surface on flow noise is still unclear. Therefore, we used alternating free-slip and no-slip boundary conditions to study the flow noise of superhydrophobic channel flows with streamwise strips. The numerical calculations of the flow and the sound field have been carried out by the methods of large eddy simulation (LES) and Lighthill analogy, respectively. Under a constant pressure gradient (CPG) condition, the average Reynolds number and the friction Reynolds number are approximately set to 4200 and 180, respectively. The influence on noise of different gas fractions (GF) and strip number in a spanwise period on channel flow have been studied. Our results show that the superhydrophobic surface has noise reduction effect in some cases. Under CPG conditions, the increase in GF increases the bulk velocity and weakens the noise reduction effect. Otherwise, the increase in strip number enhances the lateral energy exchange of the superhydrophobic surface, and results in more transverse vortices and attenuates the noise reduction effect. In our results, the best noise reduction effect is obtained as 10.7 dB under the scenario of the strip number is 4 and GF is 0.5. The best drag reduction effect is 32%, and the result is obtained under the scenario of GF is 0.8 and strip number is 1. In summary, the choice of GF and the number of strips is comprehensively considered to guarantee the performance of drag reduction and noise reduction in this work.

Download Full-text