Energies

Research

14 pages, 1254 KiB

Open AccessArticle

Transition Prediction in Incompressible Boundary Layer with Finite-Amplitude Streaks

by Juan Ángel Martín and Pedro Paredes

Energies 2021, 14(8), 2147; https://doi.org/10.3390/en14082147 - 12 Apr 2021

Cited by 4 | Viewed by 1856

Modulating the boundary layer velocity profile is a very promising strategy for achieving transition delay and reducing the friction of the plate. By perturbing the flow with counter-rotating vortices that undergo transient, non-modal growth, streamwise-aligned streaks are formed inside the boundary layer, which [...] Read more.

Modulating the boundary layer velocity profile is a very promising strategy for achieving transition delay and reducing the friction of the plate. By perturbing the flow with counter-rotating vortices that undergo transient, non-modal growth, streamwise-aligned streaks are formed inside the boundary layer, which have been proved (theoretical and experimentally) to be very robust flow structures. In this paper, we employ efficient numerical methods to perform a parametric stability investigation of the three-dimensional incompressible flat-plate boundary layer with finite-amplitude streaks. For this purpose, the Boundary Region Equations (BREs) are applied to solve the nonlinear downstream evolution of finite amplitude streaks. Regarding the stability analysis, the linear three-dimensional plane-marching Parabolized Stability Equations (PSEs) concept constitutes the best candidate for this task. Therefore, a thorough parametric study is presented, analyzing the instability characteristics with respect to critical conditions of the modified incompressible zero-pressure-gradient flat-plate boundary layer, by means of finite-amplitude linearly optimal and suboptimal disturbances or streaks. The parameter space is extended from low- to high- amplitude streaks, accurately documenting the transition delay for low-amplitude streaks and the amplitude threshold for streak shear layer instability or bypass transition, which drastically displaces the transition front upstream. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

34 pages, 7908 KiB

Open AccessArticle

Development and Validation of a Machine Learned Turbulence Model

by Shanti Bhushan, Greg W. Burgreen, Wesley Brewer and Ian D. Dettwiller

Energies 2021, 14(5), 1465; https://doi.org/10.3390/en14051465 - 8 Mar 2021

Cited by 9 | Viewed by 2396

Abstract

A stand-alone machine learned turbulence model is developed and applied for the solution of steady and unsteady boundary layer equations, and issues and constraints associated with the model are investigated. The results demonstrate that an accurately trained machine learned model can provide grid [...] Read more.

A stand-alone machine learned turbulence model is developed and applied for the solution of steady and unsteady boundary layer equations, and issues and constraints associated with the model are investigated. The results demonstrate that an accurately trained machine learned model can provide grid convergent, smooth solutions, work in extrapolation mode, and converge to a correct solution from ill-posed flow conditions. The accuracy of the machine learned response surface depends on the choice of flow variables, and training approach to minimize the overlap in the datasets. For the former, grouping flow variables into a problem relevant parameter for input features is desirable. For the latter, incorporation of physics-based constraints during training is helpful. Data clustering is also identified to be a useful tool as it avoids skewness of the model towards a dominant flow feature. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

15 pages, 710 KiB

Open AccessArticle

Tracking Turbulent Coherent Structures by Means of Neural Networks

by Jose J. Aguilar-Fuertes, Francisco Noguero-Rodríguez, José C. Jaen Ruiz, Luis M. García-RAffi and Sergio Hoyas

Energies 2021, 14(4), 984; https://doi.org/10.3390/en14040984 - 13 Feb 2021

Cited by 6 | Viewed by 1540

Abstract

The behaviours of individual flow structures have become a relevant matter of study in turbulent flows as the computational power to allow their study feasible has become available. Especially, high instantaneous Reynolds Stress events have been found to dominate the behaviour of the [...] Read more.

The behaviours of individual flow structures have become a relevant matter of study in turbulent flows as the computational power to allow their study feasible has become available. Especially, high instantaneous Reynolds Stress events have been found to dominate the behaviour of the logarithmic layer. In this work, we present a viability study where two machine learning solutions are proposed to reduce the computational cost of tracking such structures in large domains. The first one is a Multi-Layer Perceptron. The second one uses Long Short-Term Memory (LSTM). Both of the methods are developed with the objective of taking the the structures’ geometrical features as inputs from which to predict the structures’ geometrical features in future time steps. Some of the tested Multi-Layer Perceptron architectures proved to perform better and achieve higher accuracy than the LSTM architectures tested, providing lower errors on the predictions and achieving higher accuracy in relating the structures in the consecutive time steps. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

16 pages, 4873 KiB

Open AccessArticle

Prediction of Dead Oil Viscosity: Machine Learning vs. Classical Correlations

by Fahimeh Hadavimoghaddam, Mehdi Ostadhassan, Ehsan Heidaryan, Mohammad Ali Sadri, Inna Chapanova, Evgeny Popov, Alexey Cheremisin and Saeed Rafieepour

Energies 2021, 14(4), 930; https://doi.org/10.3390/en14040930 - 10 Feb 2021

Cited by 31 | Viewed by 4160

Abstract

Dead oil viscosity is a critical parameter to solve numerous reservoir engineering problems and one of the most unreliable properties to predict with classical black oil correlations. Determination of dead oil viscosity by experiments is expensive and time-consuming, which means developing an accurate [...] Read more.

Dead oil viscosity is a critical parameter to solve numerous reservoir engineering problems and one of the most unreliable properties to predict with classical black oil correlations. Determination of dead oil viscosity by experiments is expensive and time-consuming, which means developing an accurate and quick prediction model is required. This paper implements six machine learning models: random forest (RF), lightgbm, XGBoost, multilayer perceptron (MLP) neural network, stochastic real-valued (SRV) and SuperLearner to predict dead oil viscosity. More than 2000 pressure–volume–temperature (PVT) data were used for developing and testing these models. A huge range of viscosity data were used, from light intermediate to heavy oil. In this study, we give insight into the performance of different functional forms that have been used in the literature to formulate dead oil viscosity. The results show that the functional form

f (γ_{A P I}, T),

has the best performance, and additional correlating parameters might be unnecessary. Furthermore, SuperLearner outperformed other machine learning (ML) algorithms as well as common correlations that are based on the metric analysis. The SuperLearner model can potentially replace the empirical models for viscosity predictions on a wide range of viscosities (any oil type). Ultimately, the proposed model is capable of simulating the true physical trend of the dead oil viscosity with variations of oil API gravity, temperature and shear rate. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

11 pages, 9848 KiB

Open AccessArticle

Deep Reinforcement Learning Control of Cylinder Flow Using Rotary Oscillations at Low Reynolds Number

by Mikhail Tokarev, Egor Palkin and Rustam Mullyadzhanov

Energies 2020, 13(22), 5920; https://doi.org/10.3390/en13225920 - 13 Nov 2020

Cited by 26 | Viewed by 3467

Abstract

We apply deep reinforcement learning to active closed-loop control of a two-dimensional flow over a cylinder oscillating around its axis with a time-dependent angular velocity representing the only control parameter. Experimenting with the angular velocity, the neural network is able to devise a [...] Read more.

We apply deep reinforcement learning to active closed-loop control of a two-dimensional flow over a cylinder oscillating around its axis with a time-dependent angular velocity representing the only control parameter. Experimenting with the angular velocity, the neural network is able to devise a control strategy based on low frequency harmonic oscillations with some additional modulations to stabilize the Kármán vortex street at a low Reynolds number

R e = 100

. We examine the convergence issue for two reward functions showing that later epoch number does not always guarantee a better result. The performance of the controller provide the drag reduction of 14% or 16% depending on the employed reward function. The additional efforts are very low as the maximum amplitude of the angular velocity is equal to

8 %

of the incoming flow in the first case while the latter reward function returns an impressive

0.8 %

rotation amplitude which is comparable with the state-of-the-art adjoint optimization results. A detailed comparison with a flow controlled by harmonic oscillations with fixed amplitude and frequency is presented, highlighting the benefits of a feedback loop. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

18 pages, 9757 KiB

Open AccessArticle

Insights into the Aeroacoustic Noise Generation for Vertical Axis Turbines in Close Proximity

by Manuel Viqueira-Moreira and Esteban Ferrer

Energies 2020, 13(16), 4148; https://doi.org/10.3390/en13164148 - 11 Aug 2020

Cited by 5 | Viewed by 2350

Abstract

We present Large Eddy Simulations and aeroacoustic spectra for three configurations of increasing flow complexity: an isolated NACA0012 airfoil, an isolated rotating vertical axis wind turbine composed of three rotating airfoils and a farm of four vertical axis turbines (with identical characteristics as [...] Read more.

We present Large Eddy Simulations and aeroacoustic spectra for three configurations of increasing flow complexity: an isolated NACA0012 airfoil, an isolated rotating vertical axis wind turbine composed of three rotating airfoils and a farm of four vertical axis turbines (with identical characteristics as the isolated turbine), which are located in close proximity. The aeroacoustic signatures of the simulated airfoil and the isolated turbine are validated using published numerical and experimental data. We provide theoretical estimates to predict tonal frequencies, which are used to identify the main physical mechanisms responsible for the tonal signature and for each configuration and enable the categorisation of the main tonal aeroacoustic sources of vertical axis turbines operating in close proximity. Namely, we identify wake, vortex, blade passing and boundary layer phenomena and provide estimates for the associated tonal frequencies, which are validated with simulations. In the farm, we observe non-linear interactions and enhanced mixing that decreases tonal frequencies in favour of larger broadband amplitudes at low frequencies. Comparing the spectrum with that of the isolated turbine, only the blade passing frequency and the boundary layer tones can be clearly identified. Variations in acoustic amplitudes, tonal frequencies and sound directivities suggest that a linear combination of sources from isolated turbines is not enough to characterise the aeroacoustic footprint of vertical axiswind turbines located in close proximity, and that farms need to be considered and studied as different entities. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Graphical abstract

19 pages, 17876 KiB

Open AccessArticle

Design and Numerical Analysis of Flow Characteristics in a Scaled Volute and Vaned Nozzle of Radial Turbocharger Turbines

by Andrés Omar Tiseira Izaguirre, Roberto Navarro García, Lukas Benjamin Inhestern and Natalia Hervás Gómez

Energies 2020, 13(11), 2930; https://doi.org/10.3390/en13112930 - 7 Jun 2020

Cited by 6 | Viewed by 2722

Abstract

Over the past few decades, the aerodynamic improvements of turbocharger turbines contributed significantly to the overall efficiency augmentation and the advancements in downsizing of internal combustion engines. Due to the compact size of automotive turbochargers, the experimental measurement of the complex internal aerodynamics [...] Read more.

Over the past few decades, the aerodynamic improvements of turbocharger turbines contributed significantly to the overall efficiency augmentation and the advancements in downsizing of internal combustion engines. Due to the compact size of automotive turbochargers, the experimental measurement of the complex internal aerodynamics has been insufficiently studied. Hence, turbine designs mostly rely on the results of numerical simulations and the validation of zero-dimensional parameters as efficiency and reduced mass flow. To push the aerodynamic development even further, a precise validation of three-dimensional flow patterns predicted by applied computational fluid dynamics (CFD) methods is in need. This paper presents the design of an up-scaled volute-stator model, which allows optical experimental measurement techniques. In a preliminary step, numerical results indicate that the enlarged geometry will be representative of the flow patterns and characteristic non-dimensional numbers at defined flow sections of the real size turbine. Limitations due to rotor-stator interactions are highlighted. Measurement sections of interest for available measurement techniques are predefined. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

23 pages, 4094 KiB

Open AccessArticle

Dynamic Mode Decomposition Analysis of Spatially Agglomerated Flow Databases

by Binghua Li, Jesús Garicano-Mena, Yao Zheng and Eusebio Valero

Energies 2020, 13(9), 2134; https://doi.org/10.3390/en13092134 - 28 Apr 2020

Cited by 12 | Viewed by 2940

Abstract

Dynamic Mode Decomposition (DMD) techniques have risen as prominent feature identification methods in the field of fluid dynamics. Any of the multiple variables of the DMD method allows to identify meaningful features from either experimental or numerical flow data on a [...] Read more.

Dynamic Mode Decomposition (DMD) techniques have risen as prominent feature identification methods in the field of fluid dynamics. Any of the multiple variables of the DMD method allows to identify meaningful features from either experimental or numerical flow data on a data-driven manner. Performing a DMD analysis requires handling matrices

V \in R^{n_{p} \times N}

, where

n_{p}

and N are indicative of the spatial and temporal resolutions. The DMD analysis of a complex flow field requires long temporal sequences of well resolved data, and thus the memory footprint may become prohibitively large. In this contribution, the effect that principled spatial agglomeration (i.e., reduction in

n_{p}

via clustering) has on the results derived from the DMD analysis is investigated. We compare twelve different clustering algorithms on three testcases, encompassing different flow regimes: a synthetic flow field, a

R e_{D} = 60

flow around a cylinder cross section, and a

R e_{τ} \approx 200

turbulent channel flow. The performance of the clustering techniques is thoroughly assessed concerning both the accuracy of the results retrieved and the computational performance. From this assessment, we identify DBSCAN/HDBSCAN as the methods to be used if only relatively high agglomeration levels are affordable. On the contrary, Mini-batch K-means arises as the method of choice whenever high agglomeration

\tilde{n_{p}} / n_{p} ≪ 1

is possible. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

21 pages, 4804 KiB

Open AccessArticle

A Novel Algebraic Stress Model with Machine-Learning-Assisted Parameterization

by Chao Jiang, Junyi Mi, Shujin Laima and Hui Li

Energies 2020, 13(1), 258; https://doi.org/10.3390/en13010258 - 4 Jan 2020

Cited by 19 | Viewed by 3539

Abstract

Reynolds-stress closure modeling is critical to Reynolds-averaged Navier-Stokes (RANS) analysis, and it remains a challenging issue in reducing both structural and parametric inaccuracies. This study first proposes a novel algebraic stress model named as tensorial quadratic eddy-viscosity model (TQEVM), in which nonlinear terms [...] Read more.

Reynolds-stress closure modeling is critical to Reynolds-averaged Navier-Stokes (RANS) analysis, and it remains a challenging issue in reducing both structural and parametric inaccuracies. This study first proposes a novel algebraic stress model named as tensorial quadratic eddy-viscosity model (TQEVM), in which nonlinear terms improve previous model-form failure due to neglection of nonlocal effects. Then a data-driven regression model based on a fully-connected deep neural network is designed to determine the TQEVM coefficients. The well-trained data-driven model using high-fidelity direct numerical simulation (DNS) data successfully learned the underlying input-output relationships, further obtaining spatial-dependent optimal values of these coefficients. Finally, detailed validations are made in wall-bounded flows where nonlocal effects are expected to be significant. Comparative results indicate that TQEVM provides improvements both for the stress-strain misalignment and stress anisotropy, which are clear advantages over linear and quadratic eddy-viscosity models. TQEVM extends to the scope of resolution to the wall distance

y^{+} \approx 9

as well as provides a realizable solution. RANS simulations with TQEVM are also carried out and the obtained mean-flow quantities of interest agree well with DNS. This work, therefore, results in a high-fidelity representation of Reynolds stresses and contributes to further understanding of machine-learning-assisted turbulence modeling and regression analysis. Full article

(This article belongs to the Special Issue Machine-Learning Methods for Complex Flows)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Machine-Learning Methods for Complex Flows

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Published Papers (9 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI