Search

In this work, we aim to shed light on the following research question: can we find a subgrid-scale (SGS) heat flux model with good physical and numerical properties, such that we can obtain satisfactory predictions for buoyancy-driven turbulence at high Rayleigh numbers? This is motivated by our previous findings showing the limitations of existing SGS heat flux models for LES. On one hand, the most popular models rely on the eddy-diffusivity assumption despite their well-known lack of accuracy in a priori studies. On the other hand, the gradient model, which is the leading term of the Taylor series of the SGS flux, is much more accurate a priori but cannot be used as a standalone model since it produces a finite-time blow-up. In this context, we firstly aim to reconcile accuracy and stability for the gradient model. To do so, it is expressed as a linear combination of regularized (smoother) forms of the convective operator. The new alternative form can indeed be viewed as an approximate deconvolution of the exact SGS flux. Moreover, it facilitates the mathematical analysis of the gradient model, neatly identifying those terms that may cause numerical instabilities, leading to a new unconditionally stable non-linear model that can indeed be viewed as a stabilized version of the gradient model. In this way, we expect to combine the good a priori accuracy of the gradient model with the stability required in practical numerical simulations., F.X.T. and A.O. are supported by SIMEX project (PID2022-142174OB-I00) of Ministerio de Ciencia e Innovación and the RETOtwin project (PDC2021-120970-I00) of Ministerio de Econom´ ıa y Competitividad, Spain. Calculations were carried out on MareNostrum 4 supercomputer at BSC. The authors thankfully acknowledge these institutions., Peer Reviewed, Postprint (published version)

©2024 American Physical Society., Wereveal and analyze an efficient magnetic dynamo action due to precession-driven hydrodynamic turbulence in the local model of a precessional flow, focusing on the kinematic stage of this dynamo. The growth rate of the magnetic field monotonically increases with the Poincaré number Po, characterizing precession strength, and the magnetic Prandtl number Pm, equal to the ratio of viscosity to resistivity, for the considered ranges of these parameters. The critical Poc for the dynamo onset decreases with increasing Pm. To understand the scale-byscale evolution (growth) of the precession dynamo and its driving processes, we perform spectral analysis by calculating the spectra of magnetic energy and of different terms in the induction equation in Fourier space. To this end, we decompose the velocity field of precession-driven turbulence into two-dimensional (2D) vortical and three-dimensional (3D) inertial wave modes. It is shown that the dynamo operates across a broad range of scales and exhibits a remarkable transition from a primarily vortex-driven regime at lower Po to a more complex regime at higher Po where it is driven jointly by vortices, inertial waves, and the shear of the background precessional f low. Vortices and shear drive the dynamo mostly at large scales comparable to the flow system size, and at intermediate scales, while at smaller scales it is mainly driven by inertial waves. This study can be important not only for understanding the magnetic dynamo action in precession-driven flows, but also in a general context of f lows where vortices emerge and govern the flow dynamics and evolution., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 787544), Shota Rustaveli National Science Foundation of Georgia (SRNSFG) [Grant No. FR-23-1277] and EPSRC Grant No. EP/R014604/1. A.J.B. was funded by STFC Grants No. ST/S000275/1 and No. ST/W000873/1., Peer Reviewed, Postprint (author's final draft)

A comprehensive analysis of the unsteady flow dynamics past the 30P30N three-element high lift wing is performed by means of large eddy simulations at different angles of attack (AoA=5°, 9°, and 23°) and at a Reynolds number of Rec=750.000 (based on the nested chord). Results are compared with experimental and numerical investigations, showing a quantitatively good agreement and, thus, proving the reliability and accuracy of the present simulations. Within the slat and main coves, large recirculation bubbles are bounded by shear layers, where the onset of turbulence is triggered by Kelvin–Helmholtz instabilities. In the energy spectrum of the velocity fluctuations, the footprint of these instabilities is detected as a broadband peak; its frequency being moved toward lower values as the angle of attack increases. Kelvin–Helmholtz vortices roll-up and break down into small scales that eventually impinge into the slat and main coves lower surfaces. The slat impingement shows to be more prominent, and hence, larger velocity and pressure fluctuations are observed. The impingement strength diminishes with the angle of attack in both coves, while higher fluctuations are originated on the slat and main respective suction sides, leading to larger boundary layers. This is associated with the displacement of the stagnation point with the angle of attack. Another salient feature observed is the laminar-to-turbulent flow transition in the main and flap leading edges although the average location of this transition seems to not be affected by the angle of attack. Tollmien–Schlichting instabilities precede this transition, with the disturbances amplified by the inviscid mode at low angles of attack, while at AoA=23°, the local Reynolds number on the main suction side is incremented and the viscous mode becomes important. The analysis shows that the turbulent wake formed at the trailing edge of all elements dominates the dynamics downstream. This is especially true at the higher an, This research has received financial support from the Ministerio de Ciencia e Innovación of Spain (Nos. PID2020-116937RB-C21 and PID2020-116937RB-C22). Furthermore, simulations were conducted with the assistance of the Red Española de Supercomputación (RES) and the EuroHPC JU, who granted us computational resources at the HPC facilities of MareNostrum IV, at Barcelona Supercomputing Center (IM-2022-3-0005), and Vega, at Institute of Information Science (REG- 2022R01-030), respectively. Finally, R. Montalà would like to express gratitude to the Departament de Recerca i Universitats de la Generalitat de Catalunya for awarding the FI-SDUR grant (No. 2022 FISDU 00066), which supports his doctoral studies., Peer Reviewed, Postprint (published version)

Microduct flows are known for their inherent laminar regimes resulting from the characteristic small dimensions and low velocities. In this regard, direct numerical simulations are employed to investigate an innovative approach that harnesses the unique thermophysical properties of high-pressure transcritical fluids to achieve significantly higher rates of mixing and heat transfer in microduct geometries. The strategy is based on the sizeable changes in properties that supercritical fluids, at pressures and temperatures exceeding their critical value, undergo across the pseudo-boiling region. To this end, four different cases are considered, and systematically analyzed, in which the bulk pressure and temperature difference between walls are varied. The results obtained indicate that laminar flow prevails at low-pressure conditions, while flow regimes with turbulent characteristics can be achieved when operating at high-pressure conditions with a transversal temperature difference. The transition to the turbulence-like regime is assessed by quantifying variations in velocity and temperature profiles, accompanied by the observation of secondary flow motions. As a result, substantial increases in the Nusselt number of roughly 20×, indicative of enhanced heat transfer, are obtained at the hot wall in comparison to cases with same temperature differences at low pressure., This work is funded by the European Union (ERC, SCRAMBLE, 101040379). Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them., Peer Reviewed, Postprint (published version)

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)

Over the last years, the reduction of CO2 pollution has been of a major concern in the automotive industry, since it is one of the biggest contributors to the overall emissions of this gas. Besides, the European Union is taking action to reduce its emissions. On the other hand, exposure to high levels of noise is considered harmful to health, of which the automotive industry is also a major contributor, especially in large cities. Both problems can be addressed by performing scale-resolving simulations of the flow over a passenger car, whose results can be used to reduce the aerodynamically generated drag and noise. Nevertheless, in recent years this type of simulations have not been feasible due to their high computational requirements. Nowadays, with the emergence of supercomputing architectures together with the use of high-order methods, allowing the algorithm to be parallelized, the computational resources of this models have been optimized. The high-order methods stand out since they can obtain solutions with higher accuracy and with less computational cost than linear methods. However, the lack of having a robust method to generate 3D curvilinear meshes has made difficult their implementation into real life cases. Therefore, the present work aims to establish a suitable quality criteria, which allows to obtain high-order 3D meshes in a robust way, by choosing the adequate methodologies already present in the literature. The main drawbacks of the present approach are also analyzed. The aerodynamic analysis is performed with the DrivAer geometry, which represents a generic but realistic car configuration which was specifically created to have a wide range of representative results to validate CFD algorithms when dealing with complex geometries. The results show good agreement with the literature, which opens the way for further research to get the most out of high order methods, and reduce the computational demands of scale-resolving simulations.

Transcritical turbulent flows are governed by the compressible Navier–Stokes equations along with a real-gas equation of state. Their computation is strongly susceptible to numerical instabilities and requires kinetic-energy- and pressure-equilibrium-preserving schemes to yield stable and non-dissipative scale-resolving simulations. Building upon a recently developed kinetic-energy- and pressure-equilibrium-preserving discretization framework based on transporting a pressure equation, the objectives of this paper are to (i) derive a filtered set of equations suitable for large-eddy simulation, and (ii) characterize the properties of the resulting subfilter-scale terms by performing a priori analyses of transcritical wall-bounded turbulence direct numerical simulation data. The filtering operation leads to three unconventional subfilter-scale terms that emerge from the pressure equation and require dedicated modeling. The subfilter-scale stress tensor is dissected in terms of magnitude, shape and orientation based on an eigendecomposition analysis, and compared with existing subfilter-scale models. A priori analyses confirm that models of eddy-viscosity type are favorable for this framework, although the tensor shape is not fully captured. Closure expressions are finally proposed and tested for the novel subfilter terms, showing acceptable performances., Peer Reviewed, Postprint (published version)

This work presents the investigation of a solution to accelerate computations of high-pressure transcritical fluids at low-Mach number regimes based on the artificial compressibility method. The highly adaptable nature of this approach is particularly well suited for conducting complex, high-fidelity simulations on GPU-based computer architectures. The initial exploratory results presented provide speedups of 10× without significantly compromising the simulation accuracy in terms of flow physics., This work is funded by the European Union (ERC, SCRAMBLE, 101040379). Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. Support is also acknowledged from the Beatriz Galindo program (BGP18/00026, Spain), the SRG program (2021-SGR-01045, Catalonia), and the computer resources at FinisTerrae III provided by CESGA (RES-IM-2023-1-0005)., Peer Reviewed, Postprint (published version)

In this work, we aim to shed light to the following research question: is the complexity of numerically solving Poisson’s equation increasing or decreasing for very large DNS and LES simulations of incompressible flows? Physical and numerical arguments are combined to derive power law scalings at very high Reynolds numbers., This work is supported by the Ministerio de Economía y Competitividad, Spain, RETOtwin project (PDC2021-120970-I00). A.A-B is supported by the predoctoral grants DIN2018-010061 and 2019-DI-90, by MCIN/AEI/10.13039/501100011033 and the AGAUR. Cal- culations are being performed on the MareNostrum 4 supercomputer at the BSC. The authors thankfully acknowledge these institutions., Postprint (published version)

Preserving the operators’ symmetries at the discrete level is crucial to enable reliable DNS and LES simulations of turbulent flows. Moreover, real-world applications demand robust and stable numerical methods suitable for complex geometries. In this regard, this work describes TFA, our novel in-house code, which relies on a symmetry-preserving discretisation for unstructured collocated grids that, apart from being virtually free of artificial dissipation, is shown to be unconditionally stable. To ensure cross-platform portability, the implementation of such a discretisation relies on a minimal set of algebraic kernels. Doing this poses challenges that need to be addressed, like the low arithmetic intensity of the sparse matrix-vector product, the reformulation of boundary conditions and flux limiters, or the efficient computation of eigenbounds to determine the time-step. With the aim of analysing the advantages and disadvantages of this “algebraic” approach, a comparison with OpenFOAM, probably the most widespread open-source CFD code, will be made. Finally, a relevant case from the CSP industry will be presented in order to assess the feasibility of overnight industrial simulations., A.A.B., G.C., J.A.H., F.X.T., C.D.P.S. and A.O.were financially supported by the competitive R+D project RETOtwin (PDC2021-120970-I00), given by MCIN/AEI/10.13039/501100011033 and European Union Next Generation EU. A.A.B. has also been supported by the predoctoral grants DIN 2018-010061 and 2019-DI-90, given by MCIN/AEI/10.13039/501100011033 and the Catalan Agency for Management of University and Research Grants (AGAUR), respectively. J.A.H.has also been supported by the predoctoral grant 2022 FI_B1 00204, given by AGAUR. Calculations were carried out on Snellius supercomputer at SURF. The authors thankfully acknowledge these institutions., Postprint (published version)

Preserving the operators symmetries at discrete level is the key aspect to enable reliable DNS and LES simulations of turbulent flows. On the other hand, realworld applications demand robust and stable numerical methods suitable for complex geometries. In this regard, this work presents a symmetry-preserving discretization for unstructured collocated grids that, apart from being virtually free of artificial dissipation, it is shown to be unconditionally stable. Special attention is given to the (de)construction of the discrete operators into fundamental geometrical entities. On this basis, a new inexpensive method to compute strict eigenbounds for the convective and diffusive operators, which are needed to determine the time-step à la CFL, is presented. Apart from providing better estimations than previous methods, it is relies on a sparsematrix vector product where only vectors change on time. Hence, both implementation in existing codes and cross-platform portability are straightforward. Altogether leads to a simple and robust approach for DNS and LES simulations of complex turbulent flows., F.X.T., J.A.H,D. Sand A.O. are supported by the Ministerio de Economía y Competitividad, Spain, RETO twin project (PDC2021-120970-I00). Calculations were carried out on MareNostrum4 supercomputer at BSC.We thankfully acknowledge these institutions., Postprint (published version)

In this work, we aim to shed light to the following research question: is the complexity of numerically solving Poisson’s equation increasing or decreasing for very large DNS and LES simulations of incompressible flows? Physical and numerical arguments are combined to derive power-law scalings at very high Reynolds numbers. Preliminary results of forced homogeneous isotropic turbulence seems to confirm these scalings. Theoretical convergence analysis for both Jacobi and multigrid solvers defines a two-dimensional phase space divided into two regions depending whether the number of solver iterations tend to decrease or increase with the Reynolds number. Preliminary results seem to confirm that we are indeed in the second region., This work was financially supported by two competitive R+D projects: RETOtwin (PDC2021- 120970-I00), given by MCIN/AEI/ 10.13039/501100011033 and European Union Next GenerationEU, and FusionCAT (001-P-001722), given by Generalitat de Catalunya RIS3CATFEDER. A` del Alsalti-Baldellou has also been supported by the predoctoral grants DIN2018- 010061 and 2019-DI-90, given by MCIN/AEI/10.13039/501100011033 and the Catalan Agency for Management of University and Research Grants (AGAUR), respectively., Postprint (author's final draft)

An Inertia Particle Separator (IPS) is a device situated in the intake of the aircraft to remove particles from entering the engine; consisting of two types of outlets: core and scavenge. This work presents the numerical simulation of the Reynolds-Average Navier-Stokes (RANS) and the Large Eddy Simulation (LES) approach for an IPS. First, the fluid comparison of RANS and LES is carried out for three cases. Results show that the profile velocities of the continuous phase for RANS and LES are in good agreement. In the following, the separation efficiency is analyzed through different Reynolds numbers (Re) of 52K, 83K and 110K and coreto-scavenge mass flow rate ratios (ß = 0.10,0.12,0.18) for two types of particles with diameters of d p = 10µm and d p = 120µm. The elastic particle-wall collision is considered here. The particles with the smallest Stokes number (St <<1) show the smallest separation efficiency because they follow the flow pattern. Their separation efficiency decreases in lower flow split values. Also, decreasing the Re decreases the separation efficiency for the particle diameter of 10 µm. The particles with 120 µm have a separation efficiency almost near one, which does not change with decreasing the flow split. The discrepancies in separation efficiency between RANS and LES show up in the small particles where these particles are affected by the flow unsteadiness. In contrast, RANS can predict the separation efficiency correctly for the particles with large inertia and Stokes numbers (St >> 1)., This work has been developed within the EU H2020 Clean Sky 2 research project “A New proTection devIce for FOD - ANTIFOD” (grant agreement Nº 821352). Linda Bahramian acknowledges the financial support from the Secretariat of Universities and Research of the Generalitat de Catalunya and the European Social Fund, FI AGAUR Grant (2019 FI B 01205) and the UPC-Santander Grant. Carles Oliet, as a Serra Húnter associate professor, acknowledges the Catalan Government for the support through this Programme., Postprint (published version)

Postprint (published version)

This master's thesis explores the implementation of convective schemes in incompressible flow modeling, emphasizing their impact on energy conservation, momentum, and angular momentum. Specifically the skew-symmetric term and the EMAC term recently proposed. The study begins with a literature review on incompressible fluid simulation and introduces the variational multiscale method, applied specifically to stabilize formulations in incompressible fluid flow. Then the study of conservation properties for the convective terms and finally some numerical tests.

El presente Trabajo de Final de Máster se enfoca principalmente en la implementación de un sistema de acoplamiento multiescalar utilizando la metodología Pseudo-DNS, con el objetivo de reducir la complejidad de los modelos de simulación de dinámica de fluidos. El sistema involucra la confección y simulación de modelos de Elementos Volumétricos Representativos (RVE), que permiten la obtención de los efectos o patrones de las turbulencias de pequeña escala que no pueden resolver las mallas gruesas. A través de esta información, en el proyecto se ha confeccionado una nueva ley constitutiva modifica la viscosidad en cada caso para que se puedan insertar de forma adecuada estos efectos capturados previamente. A partir de aquí, se han verificado los resultados obtenidos a través de la creación de un modelo de una tubería, el cual se poseen los resultados de referencia. Por último, se ha utilizado la nueva ley constitutiva para la simulación del paso de un flujo de agua por una de las fundamentaciones que sustentan el Puente de la Bahía de Hangzhou para poder comprender la aplicabilidad real que tiene el sistema confeccionado dentro del campo de la ingeniería civil., This Master's thesis focuses mainly on the implementation of a multiscale coupling system using the Pseudo-DNS methodology, with the aim of reducing the complexity of fluid dynamics simulation models. The system involves the creation and simulation of Representative Volume Element (RVE) models, which allow the effects or patterns of small-scale turbulence that cannot be resolved by coarse meshes to be obtained. Using this information, the project has created a new constitutive law that modifies the viscosity in each case so that these previously captured effects can be inserted appropriately. From here, the results obtained have been verified by making a model of a pipe, which has the reference results. Finally, the new constitutive law has been used to simulate the passage of a water flow through one of the foundations supporting the Hangzhou Bay Bridge in order to understand the real applicability of the system created within the field of civil engineering., Aquest Treball de Final de Màster s'enfoca principalment a la implementació d'un sistema d'acoblament multiescalar utilitzant la metodologia Pseudo-DNS, amb l'objectiu de reduir la complexitat dels models de simulació de dinàmica de fluids. El sistema involucra la confecció i simulació de models d'Elements Volumètrics Representatius (RVE), que permeten l'obtenció dels efectes o els patrons de les turbulències de petita escala que no poden resoldre les malles més grosses. A través d'aquesta informació, al projecte ha confeccionat una nova llei constitutiva modifica la viscositat en cada cas perquè es puguin inserir de forma adequada aquests efectes capturats prèviament. A partir d'aquí, s'han verificat els resultats obtinguts a través de la creació d'un model d'una canonada, el qual es posseeixen els resultats de referència. Finalment, s'ha utilitzat la nova llei constitutiva per a la simulació del pas d'un flux d'aigua per una de les fonamentacions que sustenten el pont de la badia de Hangzhou per poder comprendre l'aplicabilitat real que té el sistema confeccionat dins del camp de l'enginyeria civil.

En muchos escenarios físicos complejos, como la resolución de determinados problemas de dinámica de fluidos en contextos astrofísicos, necesitamos recurrir inevitablemente a métodos numéricos, con las limitaciones que ello comporta. A grandes escalas tenemos que el coste computacional de resolver estos problemas es muy elevado, por lo que resulta necesario buscar algún método para reducirlo, tratando de perder la menor cantidad de información posible acerca de los fenómenos físicos que están teniendo lugar. En el presente trabajo, expondremos un estudio que hemos realizado sobre un modelo sub-grid, tratando de tener en cuenta la contribución de las escalas no resueltas en las ecuaciones de evolución del sistema, aplicado a un flujo ideal incompresible con ecuación de estado genérica.

The large-scale laminar/turbulent spiral patterns that appear in the linearly unstable regime of counter-rotating Taylor–Couette flow are investigated from a statistical perspective by means of direct numerical simulation. Unlike the vast majority of previous numerical studies, we analyse the flow in periodic parallelogram-annular domains, following a coordinate change that aligns one of the parallelogram sides with the spiral pattern. The domain size, shape and spatial resolution have been varied and the results compared with those in a sufficiently large computational orthogonal domain with natural axial and azimuthal periodicity. We find that a minimal parallelogram of the right tilt significantly reduces the computational cost without notably compromising the statistical properties of the supercritical turbulent spiral. Its mean structure, obtained from extremely long time integrations in a co-rotating reference frame using the method of slices, bears remarkable similarity with the turbulent stripes observed in plane Couette flow, the centrifugal instability playing only a secondary route. This article is part of the theme issue ‘Taylor–Couette and related flows on the centennial of Taylor’s seminal Philosophical transactions paper (Part 2)., KD’s research was supported by Australian Research Council Discovery Early Career Researcher Award (DE170100171). BW, RA, FM and AM research was supported by the Spanish Ministerio de Economía y Competitividad (grant numbers FIS2016-77849-R and FIS2017-85794-P) and Ministerio de Ciencia e Innovación (grant number PID2020-114043GB-I00), and the Generalitat de Catalunya (grant 2017-SGR-785). BW’s research was also supported by the Chinese Scholarship Council (grant CSC No. 201806440152). FM is a Serra-Húnter Fellow., Peer Reviewed, Preprint