Skip to main content

High Fidelity of turbulent flows around bluff bodies

RYC2018-025949-I

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal de Promoción del Talento y su Empleabilidad en I+D+i
Subprogram Subprograma Estatal de Incorporación
Call Ramón y Cajal
Call year 2018
State Plan 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre BARCELONA SUPERCOMPUTING CENTER CENTRO NACIONAL DE SUPERCOMPUTACION

Resultados relacionados

Resultados totales (Incluyendo duplicados): 7
Encontrada(s) 1 página(s)

Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Active flow control of a turbulent separation bubble through deep reinforcement learning

  • Font, Bernat
  • Alcántara Ávila, Francisco
  • Rabault, Jean
  • Vinuesa, Ricardo
  • Lehmkuhl Barba, Oriol
Journal of Physics: Conference Series 5th Madrid Turbulence Workshop 29/05/2023 - 30/06/2023 Madrid, Spain, The control efficacy of classical periodic forcing and deep reinforcement learning (DRL) is assessed for a turbulent separation bubble (TSB) at Reτ = 180 on the upstream region before separation occurs. The TSB can resemble a separation phenomenon naturally arising in wings, and a successful reduction of the TSB can have practical implications in the reduction of the aviation carbon footprint. We find that the classical zero-net-mas-flux (ZNMF) periodic control is able to reduce the TSB by 15.7%. On the other hand, the DRL-based control achieves 25.3% reduction and provides a smoother control strategy while also being ZNMF. To the best of our knowledge, the current test case is the highest Reynolds-number flow that has been successfully controlled using DRL to this date. In future work, these results will be scaled to well-resolved large-eddy simulation grids. Furthermore, we provide details of our open-source CFD–DRL framework suited for the next generation of exascale computing machines., This work was performer in part during the Fifth Madrid Summer Workshop, funded by the European Research Council (ERC) under the Caust grant ERC-AdG-101018287. OL has been partially funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under grant agreement no. 101093393, and by MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR (PCI2022-134996-2), project CEEC. OL has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I). RV and FAA acknowledge financial support from ERC grant no. ‘2021-CoG-101043998, DEEPCONTROL’. The authors thank Laia Juli´o from the BSC support services and Dr. Andrew Shao from the SmartSim team for the help and fruitful discussions about the CFD–DRL software framework. The authors also thank the UPM Turbulence Summer School and our hosts Dr. Miguel P´ erez Encinar and Dr. Adri´an Lozano-Dur´an for the organization of the workshop. Finally, we acknowledge the National Academic Infrastructure for Supercomputing in Sweden (NAISS) for the computational time provided at the large-GPU systems Alvis and Berzelius., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Atmospheric boundary layer over urban roughness: validation of large-eddy simulation

  • Teng, Ming
  • Duró Diaz, Josep Maria
  • Mestres Manzanares, Ernest
  • Muela Castro, Jordi
  • Lehmkuhl Barba, Oriol
  • Rodríguez Pérez, Ivette María
The study presents wall-modeled large-eddy simulations (LES) characterizing the flow features of a neutral atmospheric boundary layer over two urban-like roughness geometries: an array of three-dimensional square prisms and the “Michel-Stadt” geometry model. The former is an arrangement of idealized building blocks and incorporates a 7x7 array of wall-mounted prisms with identical spacing ratios in both transversal and longitudinal directions. The latter mimics a typical central European urban geometry, which presents spatial inhomogeneity in all directions. In both cases, the incident wind angle is 0°. The Reynolds numbers for each case are Reh = 5.0*10^6 and 8.0*10^6¿, respectively (¿ Reh = Uref*H/v with Uref and H denoting the reference velocity and building height, respectively, and v the kinematic viscosity). The LES employs a high-order, low-dissipation numerical scheme with a spatial resolution of 0.75¿m within the urban canopy. An online precursor simulation ensures realistic turbulent inflow conditions, improving the accuracy of the results. The simulations performed successfully capture mean-velocity profiles, wake regions, and rooftop acceleration, with excellent agreement in the streamwise velocity component. While turbulent kinetic energy is well predicted at most locations, minor discrepancies are observed near the ground, partially due to insufficient near-wall resolution and measurement constraints. The analysis of scatter plots and validation metrics (FAC2 and hit rate) shows that LES predictions outperform the standard criteria commonly used in urban flow simulations, while spectral analysis verifies that LES accurately resolves the turbulent energy cascade over approximately two frequency decades. The Kolmogorov -2/3 slope in the pre-multiplied spectra has been well reproduced below and above the urban canopy. These findings reinforce the importance of spectral analysis in LES validation and highlight the potential of high-order methods for LES of urban flows., This work has been partially financially supported by “Agencia de Gestiod’Ajuts Universitaris i de Recerca” under the call CLIMA 2023 (ref. 2023 CLIMA 00097) and the APPWIND project (ref. PLEC2021-007943). O. Lehmkuhl’s work is financed by a Ramon y Cajal postdoctoral contract by the Ministerio de Economía y Competitividad, Secretaría de Estado de Investigacion, Desarrollo e Innovacion, Spain (RYC2018-025949-I). The authors acknowledge the support of the Departament de Recerca i Universitats de la Generalitat de Catalunya through the research group Large-scale Computational Fluid Dynamics (ref. 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (ref. 2021 SGR 01051)., Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Deep reinforcement learning for active flow control in a turbulent separation bubble

  • Font, Bernat
  • Alcántara Ávila, Francisco
  • Rabault, Jean
  • Vinuesa, Ricardo
  • Lehmkuhl, Oriol
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., This work was performed in part during the Fifth Madrid Summer Workshop, funded by the European Research Council (ERC) under the Caust grant ERC-AdG-101018287. OL has been partially funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under grant agreement no. 101093393, and by MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR (PCI2022-134996-2), project CEEC. OL has been partially supported by a Ramon y Cajal postdoctoral contract (ref: RYC2018-025949-I). RV and FAA acknowledge financial support from ERC grant no. ‘2021-CoG-101043998, DEEPCONTROL’. The authors thank Laia Julió from the BSC support services and Dr. Andrew Shao from the SmartSim team for the help and fruitful discussions about the CFD–DRL software framework. The authors also thank the UPM Turbulence Summer School and our hosts Javier Jiménez, Miguel Pérez Encinar and Adrián Lozano-Durán for the organization of the workshop. Finally, we acknowledge the National Academic Infrastructure for Supercomputing in Sweden (NAISS) for the computational time provided at the large-GPU systems Alvis and Berzelius., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

On deep-learning-based closures for algebraic surrogate models of turbulent flows

  • Eiximeno Franch, Benet
  • Sanchís Agudo, Marcial
  • Miró Jané, Arnau
  • Rodríguez Pérez, Ivette María
  • Vinuesa Motilva, Ricardo
  • Lehmkuhl Barba, Oriol
A deep-learning-based closure model to address energy loss in low-dimensional surrogate models based on proper-orthogonal-decomposition (POD) modes is introduced. Using a transformer-encoder block with an easy-attention mechanism, the model predicts the spatial probability density function of fluctuations not captured by the truncated POD modes. The methodology is demonstrated on the wake of the Windsor body at yaw angles of d=[2.5°,5°,7.5°,10°,12.5°] , with d=7.5° as a test case, and in a realistic urban environment at wind directions of d=[-45°,-22.5°,0°,22.5°,45°] , with d=0° as a test case. Key coherent modes are identified by clustering them based on dominant frequency dynamics using Hotelling’s T2 on the spectral properties of temporal coefficients. These coherent modes account for nearly 60% and 75% of the total energy for the Windsor body and the urban environment, respectively. For each case, a common POD basis is created by concatenating coherent modes from training angles and orthonormalising the set without losing information. Transformers with different size on the attention layer, (64, 128 and 256), are trained to model the missing fluctuations in the Windsor body case. Larger attention sizes always improve predictions for the training set, but the transformer with an attention layer of size 256 slightly overshoots the fluctuation predictions in the Windsor body test set because they have lower intensity than in the training cases. A single transformer with an attention size of 256 is trained for the urban flow. In both cases, adding the predicted fluctuations close the energy gap between the reconstruction and the original flow field, improving predictions for energy, root-mean-square velocity fluctuations and instantaneous flow fields. For instance, in the Windsor body case, the deepest architecture reduces the mean energy error from 37% to 12% and decreases the Kullback–Leibler divergence of velocity distributions from DKL=0.2 to below DKL=0.026., The research leading to this work has been partially funded by the project TIFON with reference PLEC2023-010251/ AEI/10.13039/501100011033. B. Eiximeno’s work was funded by a contract from the Subprograma de Ayudas Predoctorales given by the Ministerio de Ciencia e Innovación (PRE2021-096927). O. Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I). The authors acknowledge the support of Departament de Recerca i Universitats de la Generalitat de Catalunya to the Research Group Large-scale Computational Fluid Dynamics (Code: 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (Code: 2021 SGR 01051). M. Sanchis-Agudo and R. Vinuesa would like to acknowledge the support from Marie Sklodowska-Curie Actions project MODELAIR, funded by the European Commission under the Horizon Europe program through grant agreement number 101072559., Peer Reviewed, 9.5 - Augmentar la investigació científica i millorar la capacitat tecnològica dels sectors industrials de tots els països, en particular els països en desenvolupament, entre d’altres maneres fomentant la innovació i augmentant substancialment, d’aquí al 2030, el nombre de persones que treballen en el camp de la investigació i el desenvolupament per cada milió d’habitants, així com la despesa en investigació i desenvolupament dels sectors públic i privat, 9 - Indústria, Innovació i Infraestructura, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Atmospheric boundary layer over urban roughness: Insights from mean-flow statistics

  • Teng, Ming
  • Duró Diaz, Josep Maria
  • Mestres Manzanares, Ernest
  • Muela Castro, Jordi
  • Lehmkuhl Barba, Oriol
  • Rodríguez Pérez, Ivette María
A wall-modeled large-eddy simulation database of an atmospheric flow over urban configurations (Teng et al. Phys. Fluids 37, 065129, 2025), obtained using a high-fidelity spectral-element method, is analyzed. The database comprises two urban geometries: (i) an in-line array of cubic prisms and (ii) a heterogeneous Michelstadt model, with plan area densities of 𝜆𝑝 = 0.25 and 0.31, and frontal area densities of 𝜆𝑓 = 0.25 and 0.24, respectively. In both cases, the incident wind angle is 0◦. The corresponding Reynolds numbers based on the mean building heights are 𝑅𝑒𝐻 = 5.0 × 106 and 6.0 × 106, respectively (𝐻 denotes the mean building height). The present study addresses the mean-flow behavior of the atmospheric boundary layer over these two urban configurations, with particular emphasis on the effects of urban configuration within the roughness sublayer. In addition to conventional mean statistics, turbulence anisotropy is characterized through the anisotropy invariant function and the Lumley triangle. The results reveal that heterogeneity in the Michelstadt configuration generates a more complex and spatially varying flow field, in contrast to the periodicity of the cube array. The thickness of the roughness sublayer in Michelstadt case is substantially higher than that in the array of cubic prisms. In both cases, the spatial distribution of the anisotropy invariant function exhibits an inverse relationship with the turbulent kinetic energy, attributed to the dominance of the streamwise Reynolds stress. These findings provide deeper insights into roughness-induced flow patterns and reference data for parameterizations within urban canopy., This work has been partially financially supported by ‘Agència de Gestió d’Ajuts Universitaris i de Recerca’ under the call CLIMA 2023 (ref. 2023 CLIMA 00097) and the APPWIND project (ref.: PLEC2021007943). O. Lehmkuhl’s work is financed by a Ramón y Cajal postdoctoral contract by the Ministerio de Economía y Competitividad, Secretaría de Estado de Investigación, Desarrollo e Innovación, Spain (RYC2018-025949-I). The authors ackowledge the Departament de Recerca i Universitats de la Generalitat de Catalunya for supporting the research groups Large-scale Computational Fluid Dynamics (ref.: 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (ref.: 2021 SGR 01051)., Peer Reviewed, 3 - Salut i Benestar, 9 - Indústria, Innovació i Infraestructura, 3.d - Reforçar la capacitat de tots els països, en particular els països en desenvolupament, en matèria d’alerta primerenca, reducció de riscos i gestió dels riscos per a la salut nacional i mundial, 9.5 - Augmentar la investigació científica i millorar la capacitat tecnològica dels sectors industrials de tots els països, en particular els països en desenvolupament, entre d’altres maneres fomentant la innovació i augmentant substancialment, d’aquí al 2030, el nombre de persones que treballen en el camp de la investigació i el desenvolupament per cada milió d’habitants, així com la despesa en investigació i desenvolupament dels sectors públic i privat, 11 - Ciutats i Comunitats Sostenibles, 11.3 - Per a 2030, augmentar la urbanització inclusiva i sostenible, així com la capacitat de planificar i gestionar de manera participativa, integrada i sostenible els assentaments humans a tots els països, 11.6 - Per a 2030, reduir l’impacte ambiental negatiu per capita de les ciutats, amb especial atenció a la qualitat de l’aire, així com a la gestió dels residus municipals i d’altre tipus, 13 - Acció per al Clima, 13.3 - Millorar l’educació, la conscienciació i la capacitat humana i institucional en relació amb la mitigació del canvi climàtic, l’adaptació a aquest, la reducció dels efectes i l’alerta primerenca, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Towards a machine learning model for explicit algebraic Reynolds stress modelling using multi-expression programming

  • Miró Jané, Arnau
  • Wallin, Stefan
  • Colombo, Alessandro
  • Temmerman, Lionel
  • Wunsch, Dirk
  • Lehmkuhl Barba, Oriol
A novel framework for a 5-term explicit algebraic Reynolds stress model has been proposed, which is suitable for machine learning procedures. Under this framework, high-fidelity datasets have been explored and post-processed. Through multi-expression programming, models have been obtained using these data. They have been thoroughly tested and present all the desirable features of a model. Implementation of these models in a numerical code has also been possible. Results have shown that they present a small upgrade over classical models for certain kinds of flows, thus showing that the physics have been captured well. It is proved that the present workflow can obtain valid expressions with similar to better performance than the baseline model., The research leading to this work has received funding within the HiFi-TURB project from the European Union’s Horizon2020 research and innovation programme under grant agreement No 814837. Oriol Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Towards the usage of geometry agnostic ß-variational-autoencoders for model order reduction of turbulent flows

  • Eiximeno Franch, Benet
  • Miró Jané, Arnau
  • Kumar, Vishal
  • Rodríguez Pérez, Ivette María
  • Lehmkuhl Barba, Oriol
The present work introduces a novel data compression framework for model order reduction of turbulent flows. The compression algorithm relies on the Geometry Agnostic Variatonal-autoencoders Integration (GAVI) technique. GAVI decouples the geometry and mesh information from the process of reducing a dataset via a CNN-based autoencoder via the QR factorization of the snapshot matrix. The framework is tested on a channel flow at Ret = 950. After reducing 9 times the amount of storage required, the root mean square of the velocity fluctuations and the budgets of the Reynolds stresses are recovered with excellent agreement when compared to the original DNS., The research leading to this work has been partially funded by the project TIFON with reference PLEC2023-010251/ AEI/10.13039/501100011033. B. Eiximeno’s work was funded by a contract from the Subprograma de Ayudas Predoctorales given by the Ministerio de Ciencia e Innovación (PRE2021- 096927). O. Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I). V. Kumar acknowledges his AI4S fellowship within the Generación D initiative by Red.es, Ministerio para la Transformación Digital y de la Función Pública, for talent attraction (C005/24-ED CV1), funded by NextGenerationEU through PRTR. The authors acknowledge the support of the Departament de Recerca i Universitats de la Generalitat de Catalunya through the research group Large-scale Computational Fluid Dynamics (ref.: 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (ref.: 2021 SGR 01051)., Peer Reviewed, 9 - Indústria, Innovació i Infraestructura, Postprint (published version)