Skip to main content

NEXT GENERATION OF INDUSTRIAL AERODYNAMIC SIMULATION CODE

PCI2021-121962

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal para Afrontar las Prioridades de Nuestro Entorno
Subprogram Subprograma Estatal de Internacionalización
Call Proyectos Programación Conjunta Internacional
Call year 2021
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): 8
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Toward the usage of deep learning surrogate models in ground vehicle aerodynamics

  • Eiximeno Franch, Benet
  • Miró Jané, Arnau|||0000-0002-2772-6050
  • Rodríguez Pérez, Ivette María|||0000-0002-3749-277X
  • Lehmkuhl Barba, Oriol|||0000-0002-2670-1871
This study introduces a deep learning surrogate model designed to predict the evolution of the mean pressure coefficient on the back face of a Windsor body across a range of yaw angles from 2.5◦ to 10◦ . Utilizing a variational autoencoder (VAE), the model effectively compresses snapshots of back pressure taken at yaw angles of 2.5◦, 5◦, and 10◦ into two latent vectors. These snapshots are derived from wall-modeled large eddy simulations (WMLESs) conducted at a Reynolds number of ReL = 2.9 × 106. The frequencies that dominate the latent vectors correspond closely with those observed in both the drag’s temporal evolution and the dynamic mode decomposition. The projection of the mean pressure coefficient to the latent space yields an increasing linear evolution of the two latent variables with the yaw angle. The mean pressure coefficient distribution at a yaw angle of 7.5◦ is predicted with a mean error of e = 3.13% when compared to the WMLESs results after obtaining the values of the latent space with linear interpolation., The research leading to this work has been partially funded by the European Project NextSim, which has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and cofounded by the Spanish Agencia Estatal de Investigacion (AEI) under grant agreement PCI2021-121962. This work was partially financially supported by the Ministerio de Economía, Industria, y Competitividad, as well as the Secretaría de Estado de Investigación, Desarrollo, e Innovación in Spain (refs PID2020-116937RB-C21 and PID2020-116937RB-C22). 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). Oriol Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: 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 (Code: 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (Code: 2021 SGR 01051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

An algebraic global linelet preconditioner for incompressible flow solvers

  • Olazábal, Ramiro de|||0000-0002-1936-0992
  • Borrell Pol, Ricard
  • Lehmkuhl Barba, Oriol|||0000-0002-2670-1871
Given the wide range of scales present in fluid dynamics, highly anisotropic meshes are often employed to resolve boundary layer flow at high Reynolds numbers. However, such meshes result in distorted elements, which can deteriorate the convergence of the preconditioned conjugate gradient solver (PCG) used to solve the discrete pressure equations. To address this issue, the Linelet Preconditioner is commonly used to accelerate the convergence rate of the PCG by building line segments (linelets) along the direction of the strongest couplings and applying a specific operation to each linelet. Under the current state of the art, linelets are constrained to operate independently within each domain partition. In these scenarios, PCG convergence was observed to deteriorate when the number of domain partitions is increased. This work proposes a communication step to be integrated into the preconditioning step, enabling different linelets to exchange information. The proposed method is observed to improve the convergence rate of linear systems with high stiffness due to highly stretched elements. Also, an algorithm is developed to generate the preconditioning matrix by purely algebraic considerations, in contrast to the usual approach of building linelets based on geometrical considerations. Furthermore, since the current approach is agnostic to the domain partition, it does not impose any constraint on the domain decomposition. Hence, the proposed method is a robust preconditioner with a numerical performance almost independent of the quality of the domain decomposition, providing a way for a more balanced load distribution., This work has been partially financially supported by the Ministerio de Ciencia e Innovacion, Spain; Contract/grant numbers PRE2018-086548 (Project: “Towards high-fidelity methodologies for innovative aircraft design”). This project has received funding from the European High-Performance Computing Joint Undertaking Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigación (AEI) under grant agreement PCI2021-121962. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Spain, France, Germany., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Self-induced large-scale motions in a three-dimensional diffuser

  • Miró Jané, Arnau|||0000-0002-2772-6050
  • Eiximeno Franch, Benet
  • Rodríguez Pérez, Ivette María|||0000-0002-3749-277X
  • Lehmkuhl Barba, Oriol|||0000-0002-2670-1871
A direct numerical simulation of a three-dimensional diffuser at Reynolds number Re = 10,000 (based on inlet bulk velocity) has been performed using a low-dissipation finite element code. The geometry chosen for this work is the Stanford diffuser, introduced by Cherry et al. (Int. J. Heat Fluid Flow 29:803–811, 2008). Results have been exhaustively compared with the published data with a quite good agreement. Additionally, further turbulent statistics have been provided such as the Reynolds stresses or the turbulent kinetic energy. A proper orthogonal decomposition and a dynamic mode decomposition analyses of the main flow variables have been performed to identify the main characteristics of the large-scale motions. A combined, self-induced movement of the large-scales has been found to originate in the top-right expansion corner with two clear features. A low-frequency diagonal cross-stream travelling wave first reported by Malm et al. (J. Fluid Mech. 699:320–351, 2012), has been clearly identified in the spatial modes of the stream-wise velocity components and the pressure, associated with the narrow band frequency of St¿[0.083,0.01] . This movement is caused by the geometrical expansion of the diffuser in the cross-stream direction. A second low-frequency trait has been identified associated with the persisting secondary flows and acting as a back and forth global accelerating-decelerating motion located on the straight area of the diffuser, with associated frequencies of St<0.005 . The smallest frequency observed in this work has been St=0.0013 . This low-frequency observed in the Stanford diffuser points out the need for longer simulations in order to obtain further turbulent statistics., The research leading to this work has been partially funded by the European Project NextSim which has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigacion (AEI) under grant agreement PCI2021-121962. Benet Eiximeno also acknowledges the financial support by the Ministerio de Economía y Competitividad, Secretaría de Estado de Investigación, Desarrollo e Innovación, Spain (Refs: PID2020-116937RB-C21 and PID2020-116937RB-C22). Oriol Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I). He also acknowledges the support of the European Project HiFi-TURB which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 814837. We also acknowledge the Barcelona Supercomputing Center for awarding us access to the MareNostrum IV machine based in Barcelona, Spain. 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)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

SOD2D: A GPU-enabled spectral finite elements method for compressible scale-resolving simulations

  • Silva, Lucas Gasparino Ferreira da
  • Spiga, Filippo
  • Lehmkuhl Barba, Oriol|||0000-0002-2670-1871
As new supercomputer architectures become more heavily focused on using hardware accelerators, in particular general-purpose graphical processors, it is therefore relevant that algorithms for computational fluid dynamics, especially those targeting scale-resolving simulations, be designed in such a way as to make efficient use of such hardware. In this paper, we propose one such hardware-accelerated Continuous Galerkin Finite Elements model, aimed at handling simulations of turbulent compressible flows over complex geometries. As this model is intended for use in Large-Eddy and Direct Numerical simulations, it is necessary that the resulting scheme introduces only small amounts of artificial (numerical) diffusion for stabilization purposes. We achieve this through a combination of Spectral Finite Elements, operator splittings on the convective term, and use of the Entropy Viscosity stabilization model adapted to spectral elements scheme. The paper will present the resulting algorithm, how it is made to work efficiently on the accelerators, and results obtained so far that demonstrate its high-accuracy capabilities.As new supercomputer architectures become more heavily focused on using hardware accelerators, in particular general-purpose graphical processors, it is therefore relevant that algorithms for computational fluid dynamics, especially those targeting scale-resolving simulations, be designed in such a way as to make efficient use of such hardware. In this paper, we propose one such hardware-accelerated Continuous Galerkin Finite Elements model, aimed at handling simulations of turbulent compressible flows over complex geometries. As this model is intended for use in Large-Eddy and Direct Numerical simulations, it is necessary that the resulting scheme introduces only small amounts of artificial (numerical) diffusion for stabilization purposes. We achieve this through a combination of Spectral Finite Elements, operator splittings on the convective term, and use of the Entropy Viscosity stabilization model adapted to spectral elements scheme. The paper will present the resulting algorithm, how it is made to work efficiently on the accelerators, and results obtained so far that demonstrate its high-accuracy capabilities., This project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigación (AEI) under grant agreement PCI2021-121962. O. Lehmkuhl 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). Lucas Gasparino has received financial support from ’la Caixa’ Foundation (ID 100010434). The fellowship grant code is LCF/BQ/DI18/11660051., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

An algebraic global linelet preconditioner for incompressible flow solvers

  • Olazábal, Ramiro de
  • Borrell Pol, Ricard
  • Lehmkuhl Barba, Oriol
Given the wide range of scales present in fluid dynamics, highly anisotropic meshes are often employed to resolve boundary layer flow at high Reynolds numbers. However, such meshes result in distorted elements, which can deteriorate the convergence of the preconditioned conjugate gradient solver (PCG) used to solve the discrete pressure equations. To address this issue, the Linelet Preconditioner is commonly used to accelerate the convergence rate of the PCG by building line segments (linelets) along the direction of the strongest couplings and applying a specific operation to each linelet. Under the current state of the art, linelets are constrained to operate independently within each domain partition. In these scenarios, PCG convergence was observed to deteriorate when the number of domain partitions is increased. This work proposes a communication step to be integrated into the preconditioning step, enabling different linelets to exchange information. The proposed method is observed to improve the convergence rate of linear systems with high stiffness due to highly stretched elements. Also, an algorithm is developed to generate the preconditioning matrix by purely algebraic considerations, in contrast to the usual approach of building linelets based on geometrical considerations. Furthermore, since the current approach is agnostic to the domain partition, it does not impose any constraint on the domain decomposition. Hence, the proposed method is a robust preconditioner with a numerical performance almost independent of the quality of the domain decomposition, providing a way for a more balanced load distribution., This work has been partially financially supported by the Ministerio de Ciencia e Innovacion, Spain; Contract/grant numbers PRE2018-086548 (Project: “Towards high-fidelity methodologies for innovative aircraft design”). This project has received funding from the European High-Performance Computing Joint Undertaking Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigación (AEI) under grant agreement PCI2021-121962. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Spain, France, Germany., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Toward the usage of deep learning surrogate models in ground vehicle aerodynamics

  • Eiximeno Franch, Benet
  • Miró Jané, Arnau
  • Rodríguez Pérez, Ivette María
  • Lehmkuhl Barba, Oriol
This study introduces a deep learning surrogate model designed to predict the evolution of the mean pressure coefficient on the back face of a Windsor body across a range of yaw angles from 2.5◦ to 10◦ . Utilizing a variational autoencoder (VAE), the model effectively compresses snapshots of back pressure taken at yaw angles of 2.5◦, 5◦, and 10◦ into two latent vectors. These snapshots are derived from wall-modeled large eddy simulations (WMLESs) conducted at a Reynolds number of ReL = 2.9 × 106. The frequencies that dominate the latent vectors correspond closely with those observed in both the drag’s temporal evolution and the dynamic mode decomposition. The projection of the mean pressure coefficient to the latent space yields an increasing linear evolution of the two latent variables with the yaw angle. The mean pressure coefficient distribution at a yaw angle of 7.5◦ is predicted with a mean error of e = 3.13% when compared to the WMLESs results after obtaining the values of the latent space with linear interpolation., The research leading to this work has been partially funded by the European Project NextSim, which has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and cofounded by the Spanish Agencia Estatal de Investigacion (AEI) under grant agreement PCI2021-121962. This work was partially financially supported by the Ministerio de Economía, Industria, y Competitividad, as well as the Secretaría de Estado de Investigación, Desarrollo, e Innovación in Spain (refs PID2020-116937RB-C21 and PID2020-116937RB-C22). 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). Oriol Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: 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 (Code: 2021 SGR 00902) and the Turbulence and Aerodynamics Research Group (Code: 2021 SGR 01051)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

SOD2D: A GPU-enabled spectral finite elements method for compressible scale-resolving simulations

  • Silva, Lucas Gasparino Ferreira da
  • Spiga, Filippo
  • Lehmkuhl Barba, Oriol
As new supercomputer architectures become more heavily focused on using hardware accelerators, in particular general-purpose graphical processors, it is therefore relevant that algorithms for computational fluid dynamics, especially those targeting scale-resolving simulations, be designed in such a way as to make efficient use of such hardware. In this paper, we propose one such hardware-accelerated Continuous Galerkin Finite Elements model, aimed at handling simulations of turbulent compressible flows over complex geometries. As this model is intended for use in Large-Eddy and Direct Numerical simulations, it is necessary that the resulting scheme introduces only small amounts of artificial (numerical) diffusion for stabilization purposes. We achieve this through a combination of Spectral Finite Elements, operator splittings on the convective term, and use of the Entropy Viscosity stabilization model adapted to spectral elements scheme. The paper will present the resulting algorithm, how it is made to work efficiently on the accelerators, and results obtained so far that demonstrate its high-accuracy capabilities.As new supercomputer architectures become more heavily focused on using hardware accelerators, in particular general-purpose graphical processors, it is therefore relevant that algorithms for computational fluid dynamics, especially those targeting scale-resolving simulations, be designed in such a way as to make efficient use of such hardware. In this paper, we propose one such hardware-accelerated Continuous Galerkin Finite Elements model, aimed at handling simulations of turbulent compressible flows over complex geometries. As this model is intended for use in Large-Eddy and Direct Numerical simulations, it is necessary that the resulting scheme introduces only small amounts of artificial (numerical) diffusion for stabilization purposes. We achieve this through a combination of Spectral Finite Elements, operator splittings on the convective term, and use of the Entropy Viscosity stabilization model adapted to spectral elements scheme. The paper will present the resulting algorithm, how it is made to work efficiently on the accelerators, and results obtained so far that demonstrate its high-accuracy capabilities., This project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigación (AEI) under grant agreement PCI2021-121962. O. Lehmkuhl 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). Lucas Gasparino has received financial support from ’la Caixa’ Foundation (ID 100010434). The fellowship grant code is LCF/BQ/DI18/11660051., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Self-induced large-scale motions in a three-dimensional diffuser

  • Miró Jané, Arnau
  • Eiximeno Franch, Benet
  • Rodríguez Pérez, Ivette María
  • Lehmkuhl Barba, Oriol
A direct numerical simulation of a three-dimensional diffuser at Reynolds number Re = 10,000 (based on inlet bulk velocity) has been performed using a low-dissipation finite element code. The geometry chosen for this work is the Stanford diffuser, introduced by Cherry et al. (Int. J. Heat Fluid Flow 29:803–811, 2008). Results have been exhaustively compared with the published data with a quite good agreement. Additionally, further turbulent statistics have been provided such as the Reynolds stresses or the turbulent kinetic energy. A proper orthogonal decomposition and a dynamic mode decomposition analyses of the main flow variables have been performed to identify the main characteristics of the large-scale motions. A combined, self-induced movement of the large-scales has been found to originate in the top-right expansion corner with two clear features. A low-frequency diagonal cross-stream travelling wave first reported by Malm et al. (J. Fluid Mech. 699:320–351, 2012), has been clearly identified in the spatial modes of the stream-wise velocity components and the pressure, associated with the narrow band frequency of St¿[0.083,0.01] . This movement is caused by the geometrical expansion of the diffuser in the cross-stream direction. A second low-frequency trait has been identified associated with the persisting secondary flows and acting as a back and forth global accelerating-decelerating motion located on the straight area of the diffuser, with associated frequencies of St<0.005 . The smallest frequency observed in this work has been St=0.0013 . This low-frequency observed in the Stanford diffuser points out the need for longer simulations in order to obtain further turbulent statistics., The research leading to this work has been partially funded by the European Project NextSim which has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 956104 and co-founded by the Spanish Agencia Estatal de Investigacion (AEI) under grant agreement PCI2021-121962. Benet Eiximeno also acknowledges the financial support by the Ministerio de Economía y Competitividad, Secretaría de Estado de Investigación, Desarrollo e Innovación, Spain (Refs: PID2020-116937RB-C21 and PID2020-116937RB-C22). Oriol Lehmkuhl has been partially supported by a Ramon y Cajal postdoctoral contract (Ref: RYC2018-025949-I). He also acknowledges the support of the European Project HiFi-TURB which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 814837. We also acknowledge the Barcelona Supercomputing Center for awarding us access to the MareNostrum IV machine based in Barcelona, Spain. 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)., Peer Reviewed, Postprint (published version)