The field of fluid mechanics is rapidly advancing, driven by unprecedented volumes of data from experiments, field measurements, and large-scale simulations at multiple spatiotemporal scales. Machine learning presents us with a wealth of techniques to extract information from data that can be translated into knowledge about the underlying fluid mechanics. Moreover, machine learning algorithms can augment domain knowledge and automate tasks related to flow control and optimization. This article presents an overview of past history, current developments, and emerging opportunities of machine learning for fluid mechanics. We outline fundamental machine learning methodologies and discuss their uses for understanding, modeling, optimizing, and controlling fluid flows. The strengths and limitations of these methods are addressed from the perspective of scientific inquiry that links data with modeling, experiments, and simulations. Machine learning provides a powerful information processing framework that can augment, and possibly even transform, current lines of fluid mechanics research and industrial applications.

of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.

The impact of fluidic actuation on the wake and drag of a 3D blunt body is experimentally investigated. The wake is forced by jets pulsed tangentially to the main flow with variable frequency and velocity. Depending on the forcing conditions, two flow regimes can be identified. First, for a broadband range of frequencies comprising the natural wake instabilities, the convection of the jet structures enhances wake entrainment, shortening the recirculating flow length with an augmentation of the bluff body drag. Further increase of the actuation frequency induces a wake fluidic boat-tailing by shear-layer deviation. It additionally lowers turbulent intensity and entrainment of high momentum fluid in the shear layer, leading to an overall reduction of the wake fluctuating kinetic energy. The association of both mechanisms is responsible for a raise of base pressure and decrease of the model's drag. The physical features of such regimes are discussed on the basis of drag, pressure and velocity measurements at several upstream conditions and control parameters. By adding curved surfaces at the jet outlets to take advantage of the so-called Coanda effect, periodic actuation can be further reinforced leading to drag reductions of about 20 % in unsteady regime. In general, the unsteady Coanda blowing not only intensifies the base pressure recovery but also preserves the effect of unsteady high frequency forcing on the turbulent field. The present results encourage the development of fluidic control in road vehicles' aerodynamics as well as provide a complement to our current understanding of bluff body drag and its manipulation.Comment: 47 pages, 35 figures, extended versio

Turbulent fluid has often been conceptualized as a transient thermodynamic phase. Here, a finite-time thermodynamics (FTT) formalism is proposed to compute mean flow and fluctuation levels of unsteady incompressible flows. The proposed formalism builds upon the Galerkin model framework, which simplifies a continuum 3D fluid motion into a finite-dimensional phase-space dynamics and, subsequently, into a thermodynamics energy problem. The Galerkin model consists of a velocity field expansion in terms of flow configuration dependent modes and of a dynamical system describing the temporal evolution of the mode coefficients. Each mode is treated as one thermodynamic degree of freedom, characterized by an energy level. The dynamical system approaches local thermal equilibrium (LTE) where each mode has the same energy if it is governed only by internal (triadic) mode interactions. However, in the generic case of unsteady flows, the full system approaches only partial LTE with unequal energy levels due to strongly mode-dependent external interactions. The first illustrated by a traveling wave governed by a 1D Burgers equation. It is then applied to two flow benchmarks: the relatively simple laminar vortex shedding, which is dominated by two eigenmodes, and the homogeneous shear turbulence, which has been modeled with 1459 modes.

We investigate linear-quadratic dynamical systems with energy-preserving quadratic terms. These systems arise for instance as Galerkin systems of incompressible flows. A criterion is presented to ensure long-term boundedness of the system dynamics. If the criterion is violated, a globally stable attractor cannot exist for an effective nonlinearity. Thus, the criterion can be considered a minimum requirement for control-oriented Galerkin models of viscous fluid flows. The criterion is exemplified, for example, for Galerkin systems of two-dimensional cylinder wake flow models in the transient and the post-transient regime, for the Lorenz system and for wall-bounded shear flows. There are numerous potential applications of the criterion, for instance, system reduction and control of strongly nonlinear dynamical systems.

No abstract

We investigate open-and closed-loop active control for aerodynamic drag reduction of a car model. Turbulent flow around a blunt-edged Ahmed body is examined at Re H ≈ 3 × 10 5 based on body height. The actuation is performed with pulsed jets at all trailing edges combined with a Coanda deflection surface. The flow is monitored with pressure sensors distributed at the rear side. We apply a model-free control strategy building on Dracopoulos & Kent (1997) and Gautier et al. (2015). The optimized control laws comprise periodic forcing, multi-frequency forcing and sensor-based feedback including also time-history information feedback and combination thereof. Key enabler is linear genetic programming as simple and efficient framework for multiple inputs (actuators) and multiple outputs (sensors). The proposed linear genetic programming control can select the best open-or closed-loop control in an unsupervised manner. Approximately 33% base pressure recovery associated with 22% drag reduction is achieved in all considered classes of control laws. Intriguingly, the feedback actuation emulates periodic high-frequency forcing by selecting one pressure sensor in the optimal control law. Our control strategy is, in principle, applicable to all multiple actuators and sensors experiments.

No abstract