CTTC - Centre Tecnològic de la Transferència de Calor
http://hdl.handle.net/2117/3190
Thu, 26 May 2022 08:21:46 GMT2022-05-26T08:21:46ZA CFD-based surrogate model for predicting flow parameters in a ventilated room using sensor readings
http://hdl.handle.net/2117/367590
A CFD-based surrogate model for predicting flow parameters in a ventilated room using sensor readings
Morozova, Nina; Trias Miquel, Francesc Xavier; Capdevila Paramio, Roser; Schillaci, Eugenio; Oliva Llena, Asensio
In this work, we develop a computational fluid dynamics (CFD)-based surrogate model, which predicts flow parameters under different geometrical configurations and boundary conditions in a benchmark case of a mechanically ventilated room with mixed convection. The model inputs are the temperature and velocity values in different locations, which act as a surrogate of the sensor readings. The model’s output is a set of comfort-related flow parameters, such as the average Nusselt number on the hot wall, jet separation point, average kinetic energy, average enstrophy, and average temperature. We tested four different machine learning methods, among which we chose the gradient boosting regression due to its accurate performance. We also adapted the developed model for indoor environment control applications by determining the optimal combinations of sensor positions which minimize the prediction error. This model does not require the repetition of CFD simulations in order to be applied since the structure of the input data imitates sensor readings. Furthermore, the low computational cost of the model execution and good accuracy makes it an effective alternative to CFD for applications where rapid predictions of complex flow configurations are required, such as model predictive control.
Fri, 20 May 2022 12:43:00 GMThttp://hdl.handle.net/2117/3675902022-05-20T12:43:00ZMorozova, NinaTrias Miquel, Francesc XavierCapdevila Paramio, RoserSchillaci, EugenioOliva Llena, AsensioIn this work, we develop a computational fluid dynamics (CFD)-based surrogate model, which predicts flow parameters under different geometrical configurations and boundary conditions in a benchmark case of a mechanically ventilated room with mixed convection. The model inputs are the temperature and velocity values in different locations, which act as a surrogate of the sensor readings. The model’s output is a set of comfort-related flow parameters, such as the average Nusselt number on the hot wall, jet separation point, average kinetic energy, average enstrophy, and average temperature. We tested four different machine learning methods, among which we chose the gradient boosting regression due to its accurate performance. We also adapted the developed model for indoor environment control applications by determining the optimal combinations of sensor positions which minimize the prediction error. This model does not require the repetition of CFD simulations in order to be applied since the structure of the input data imitates sensor readings. Furthermore, the low computational cost of the model execution and good accuracy makes it an effective alternative to CFD for applications where rapid predictions of complex flow configurations are required, such as model predictive control.Coupled radiation and natural convection: Different approaches of the slw model for a non-gray gas mixture
http://hdl.handle.net/2117/367585
Coupled radiation and natural convection: Different approaches of the slw model for a non-gray gas mixture
Colomer Rey, Guillem; Consul Serracanta, Ricard; Oliva Llena, Asensio
The coupling between non-gray radiation heat transfer and convection–conduction heat transfer is studied. The spectral line weighted sum of gray gases model (SLW) is used to account for non-gray radiation properties. The aim of this work is to analyze the influence of the different approaches used when calculating the parameters of the SLW model. Such strategies include the use of optimized model coefficients to reduce the number of operations, and the interpolation of the distribution function instead of the use of mathematical correlations. Non-gray calculations are also compared to gray solutions using the Planck mean absorption coefficient, which can be also calculated with the SLW model. The radiative transfer equation (RTE) is solved by means of the discrete ordinates method (DOM). A natural convection driven cavity is chosen to couple radiation and conduction–convection energy transfer. Several cases, with a significant variation of the ratio between radiation to convection heat transfer, as well as the ratio between radiation to conduction heat transfer, are discussed.
Fri, 20 May 2022 11:42:38 GMThttp://hdl.handle.net/2117/3675852022-05-20T11:42:38ZColomer Rey, GuillemConsul Serracanta, RicardOliva Llena, AsensioThe coupling between non-gray radiation heat transfer and convection–conduction heat transfer is studied. The spectral line weighted sum of gray gases model (SLW) is used to account for non-gray radiation properties. The aim of this work is to analyze the influence of the different approaches used when calculating the parameters of the SLW model. Such strategies include the use of optimized model coefficients to reduce the number of operations, and the interpolation of the distribution function instead of the use of mathematical correlations. Non-gray calculations are also compared to gray solutions using the Planck mean absorption coefficient, which can be also calculated with the SLW model. The radiative transfer equation (RTE) is solved by means of the discrete ordinates method (DOM). A natural convection driven cavity is chosen to couple radiation and conduction–convection energy transfer. Several cases, with a significant variation of the ratio between radiation to convection heat transfer, as well as the ratio between radiation to conduction heat transfer, are discussed.DNS of mass transfer in turbulent bubbly flow in a vertical pipe
http://hdl.handle.net/2117/367458
DNS of mass transfer in turbulent bubbly flow in a vertical pipe
Balcázar Arciniega, Néstor; Rigola Serrano, Joaquim; Oliva Llena, Asensio
The present research focuses on the DNS of mass transfer in gravity-driven turbulent bubbly flow in a vertical pipe, at a high Reynolds number (Re approximates to 1000). The objective is to compute the mass transfer coefficient included in the Sherwood number (Sh), as a function of the Reynolds (Re) number, Damkoler (Da) number for first-order chemical reaction, Schmidt (Sc) number, bubble fraction (BF), and confinement ratio (CR). Indeed, Sh = Sh(Re, Sc, BF, Da, CR), where Re=Re(Eo, Mo, BF, CR), Mo is the Morton number, Eo is the Eotvos number, and physical properties ratios are set to 100. A novel unstructured multiple-marker conservative level-set method for mass transfer in bubbly flows is employed to perform present simulations, in order to circunvent the numerical coalescence of bubbles. Navier-Stokes equations, conservative level-set (CLS) and chemical species transport equations are solved by the finite-volume method on collocated unstructured meshes. Thermodynamic equilibrium relates the concentration of chemical species at the interface by the so-called Henry's law. The pressure-velocity coupling is performed by the fractional-step projection method, whereas surface tension is computed by the continuous surface force model extended to the multiple markers CLS method. Verifications and validations of the numerical methods have been reported in our previous works. Based on our last efforts performed to research mass transfer in gravity-driven bubble swarms on unconfined domains, this research is a further step to include the wall's effect through the confinement ratio (CR). Thus, this work unravels the impact of CR on Sh, at Re = O(1000), whereas the remaining parameters are kept constant.
Tue, 17 May 2022 17:24:06 GMThttp://hdl.handle.net/2117/3674582022-05-17T17:24:06ZBalcázar Arciniega, NéstorRigola Serrano, JoaquimOliva Llena, AsensioThe present research focuses on the DNS of mass transfer in gravity-driven turbulent bubbly flow in a vertical pipe, at a high Reynolds number (Re approximates to 1000). The objective is to compute the mass transfer coefficient included in the Sherwood number (Sh), as a function of the Reynolds (Re) number, Damkoler (Da) number for first-order chemical reaction, Schmidt (Sc) number, bubble fraction (BF), and confinement ratio (CR). Indeed, Sh = Sh(Re, Sc, BF, Da, CR), where Re=Re(Eo, Mo, BF, CR), Mo is the Morton number, Eo is the Eotvos number, and physical properties ratios are set to 100. A novel unstructured multiple-marker conservative level-set method for mass transfer in bubbly flows is employed to perform present simulations, in order to circunvent the numerical coalescence of bubbles. Navier-Stokes equations, conservative level-set (CLS) and chemical species transport equations are solved by the finite-volume method on collocated unstructured meshes. Thermodynamic equilibrium relates the concentration of chemical species at the interface by the so-called Henry's law. The pressure-velocity coupling is performed by the fractional-step projection method, whereas surface tension is computed by the continuous surface force model extended to the multiple markers CLS method. Verifications and validations of the numerical methods have been reported in our previous works. Based on our last efforts performed to research mass transfer in gravity-driven bubble swarms on unconfined domains, this research is a further step to include the wall's effect through the confinement ratio (CR). Thus, this work unravels the impact of CR on Sh, at Re = O(1000), whereas the remaining parameters are kept constant.Transient Thermal Management of Vapor Cycles to Simulate Aircraft Electrical Environment Control Systems
http://hdl.handle.net/2117/367457
Transient Thermal Management of Vapor Cycles to Simulate Aircraft Electrical Environment Control Systems
Ablanque Mejía, Nicolás; Torras Ortiz, Santiago; Oliet Casasayas, Carles; Rigola Serrano, Joaquim; Pérez Segarra, Carlos David
A Modelica library has been developed to simulate both transient and steady state of vapor cycle systems (VCS) integrated within Electrical Environment Control Systems (E-ECS) architectures applied in the thermal management of aircrafts. The numerical resolution of each one of different components: heat exchangers with special attention on two-phase due to their non-linearity; compressors based on parametric curves, valves, accumulators, and auxiliary elements, are detailed, looking for multi-level approach and object-oriented nature. Strategies for transient loop resolution system are also analyzed. A compromise between fast and robust resolution with a enough accuracy has been assured. Numerical results obtained are devoted not only to highlight the system and components robustness with low computational time, but also allow having a better understanding of system behavior on heat exchanger performance, refrigerant mass flow management, initialization cases, system control and transient conditions, becoming a virtual tool for real working conditions.
Tue, 17 May 2022 17:11:02 GMThttp://hdl.handle.net/2117/3674572022-05-17T17:11:02ZAblanque Mejía, NicolásTorras Ortiz, SantiagoOliet Casasayas, CarlesRigola Serrano, JoaquimPérez Segarra, Carlos DavidA Modelica library has been developed to simulate both transient and steady state of vapor cycle systems (VCS) integrated within Electrical Environment Control Systems (E-ECS) architectures applied in the thermal management of aircrafts. The numerical resolution of each one of different components: heat exchangers with special attention on two-phase due to their non-linearity; compressors based on parametric curves, valves, accumulators, and auxiliary elements, are detailed, looking for multi-level approach and object-oriented nature. Strategies for transient loop resolution system are also analyzed. A compromise between fast and robust resolution with a enough accuracy has been assured. Numerical results obtained are devoted not only to highlight the system and components robustness with low computational time, but also allow having a better understanding of system behavior on heat exchanger performance, refrigerant mass flow management, initialization cases, system control and transient conditions, becoming a virtual tool for real working conditions.Advanced techniques for grau area mitigation in DES simulations and their effectes on the subsonic round jet acoustic spectra
http://hdl.handle.net/2117/367010
Advanced techniques for grau area mitigation in DES simulations and their effectes on the subsonic round jet acoustic spectra
Duben, Alexey; Ruano Pérez, Jesús; Trias Miquel, Francesc Xavier; Rigola Serrano, Joaquim
The research dedicated to the investigation of different gray-area mitigation approaches towards accurate aerodynamics and aeroacoustics is presented. The recent modifications of hybrid RANS-LES DDES approach based on combinations of new adapting subgrid length scales (~ !, SLA and lsq) and LES models ( and S3QR) are considered. The object of investigation is an immersed subsonic turbulent jet. The simulations are carried out on a set of refining meshes using two different scale-resolving numerical algorithms realized in the compressible codes NOISEtte and OpenFOAM. The evaluation of different approaches is focused on the analysis of far field noise. The results show that all the considered techniques provide appropriate accuracy to predict the noise generated by the turbulent jet. The study clearly demonstrated the importance of both numerical scheme and subgrid turbulence model. The peculiarities of considered approaches are revealed and discussed.
Fri, 06 May 2022 12:38:05 GMThttp://hdl.handle.net/2117/3670102022-05-06T12:38:05ZDuben, AlexeyRuano Pérez, JesúsTrias Miquel, Francesc XavierRigola Serrano, JoaquimThe research dedicated to the investigation of different gray-area mitigation approaches towards accurate aerodynamics and aeroacoustics is presented. The recent modifications of hybrid RANS-LES DDES approach based on combinations of new adapting subgrid length scales (~ !, SLA and lsq) and LES models ( and S3QR) are considered. The object of investigation is an immersed subsonic turbulent jet. The simulations are carried out on a set of refining meshes using two different scale-resolving numerical algorithms realized in the compressible codes NOISEtte and OpenFOAM. The evaluation of different approaches is focused on the analysis of far field noise. The results show that all the considered techniques provide appropriate accuracy to predict the noise generated by the turbulent jet. The study clearly demonstrated the importance of both numerical scheme and subgrid turbulence model. The peculiarities of considered approaches are revealed and discussed.NUMA-Aware Strategies for the Heterogeneous Execution of SPMV on Modern Supercomputers
http://hdl.handle.net/2117/366919
NUMA-Aware Strategies for the Heterogeneous Execution of SPMV on Modern Supercomputers
Álvarez Farré, Xavier; Gorobets, Andrei; Trias Miquel, Francesc Xavier; Oliva Llena, Asensio
The sparse matrix-vector product is a widespread operation amongst the scientific computing community. It represents the dominant computational cost in many large-scale simulations relying on iterative methods, and its performance is sensitive to the sparse pattern, the storage format, and kernel implementation, and the target computing architecture. In this work, we are devoted to the efficient execution of the sparse matrix-vector product on (potentially hybrid) modern supercomputers with non-uniform memory access configurations. A hierarchical parallel implementation is proposed to minimize the number of processes participating in distributed-memory parallelization. As a result, a single process per computing node is enough to engage all its hardware and ensure efficient memory access on manycore platforms. The benefits of this approach have been demonstrated on up to 9,600 cores of MareNostrum 4 supercomputer, at Barcelona Supercomputing Center.
Fri, 06 May 2022 07:52:13 GMThttp://hdl.handle.net/2117/3669192022-05-06T07:52:13ZÁlvarez Farré, XavierGorobets, AndreiTrias Miquel, Francesc XavierOliva Llena, AsensioThe sparse matrix-vector product is a widespread operation amongst the scientific computing community. It represents the dominant computational cost in many large-scale simulations relying on iterative methods, and its performance is sensitive to the sparse pattern, the storage format, and kernel implementation, and the target computing architecture. In this work, we are devoted to the efficient execution of the sparse matrix-vector product on (potentially hybrid) modern supercomputers with non-uniform memory access configurations. A hierarchical parallel implementation is proposed to minimize the number of processes participating in distributed-memory parallelization. As a result, a single process per computing node is enough to engage all its hardware and ensure efficient memory access on manycore platforms. The benefits of this approach have been demonstrated on up to 9,600 cores of MareNostrum 4 supercomputer, at Barcelona Supercomputing Center.A General Method to Compute Numerical Dispersion Error
http://hdl.handle.net/2117/366918
A General Method to Compute Numerical Dispersion Error
Ruano Pérez, Jesús; Baez Vidal, Aleix; Rigola Serrano, Joaquim; Trias Miquel, Francesc Xavier
This article presents a new methodology to compute numerical dispersion error. The analysis here presented is not restricted to uniform structured meshes nor linear discrete operators as it does not rely on sinusoids to compute the associated error. When using uniform meshes, the results obtained with the present method collapse onto the obtained with the classic one via an easy change of basis. If non-uniform meshes are used, a new kind of results are obtained which shed some light onto the role stretching has on dispersion error.
Fri, 06 May 2022 07:27:41 GMThttp://hdl.handle.net/2117/3669182022-05-06T07:27:41ZRuano Pérez, JesúsBaez Vidal, AleixRigola Serrano, JoaquimTrias Miquel, Francesc XavierThis article presents a new methodology to compute numerical dispersion error. The analysis here presented is not restricted to uniform structured meshes nor linear discrete operators as it does not rely on sinusoids to compute the associated error. When using uniform meshes, the results obtained with the present method collapse onto the obtained with the classic one via an easy change of basis. If non-uniform meshes are used, a new kind of results are obtained which shed some light onto the role stretching has on dispersion error.On the implementation of flux limiters in algebraic frameworks
http://hdl.handle.net/2117/366917
On the implementation of flux limiters in algebraic frameworks
Valle Marchante, Nicolás; Álvarez Farré, Xavier; Gorobets, Andrei; Castro González, Jesús; Oliva Llena, Asensio; Trias Miquel, Francesc Xavier
The use of flux limiters is widespread within the scientific computing community to capture shock dis- continuities and are of paramount importance for the temporal integration of high-speed aerodynamics, multiphase flows and hyperbolic equations in general. Meanwhile, the breakthrough of new computing architectures and the hybridization of supercomputer systems pose a huge portability challenge, particularly for legacy codes, since the computing subroutines that form the algorithms, the so-called kernels, must be adapted to various complex parallel program- ming paradigms. From this perspective, the development of innovative implementations relying on a minimalist set of kernels simplifies the deployment of scientific computing software on state-of-the-art supercomputers, while it requires the reformulation of algorithms, such as the aforementioned flux lim- iters. Equipped with basic algebraic topology and graph theory underlying the classical mesh concept, a new flux limiter formulation is presented based on the adoption of algebraic data structures and kernels. As a result, traditional flux limiters are cast into a stream of only two types of computing kernels: sparse matrix-vector multiplication and generalized pointwise binary operators. The newly proposed formulation eases the deployment of such a numerical technique in massively parallel, potentially hybrid, computing systems and is demonstrated for a canonical advection problem.
Fri, 06 May 2022 06:54:44 GMThttp://hdl.handle.net/2117/3669172022-05-06T06:54:44ZValle Marchante, NicolásÁlvarez Farré, XavierGorobets, AndreiCastro González, JesúsOliva Llena, AsensioTrias Miquel, Francesc XavierThe use of flux limiters is widespread within the scientific computing community to capture shock dis- continuities and are of paramount importance for the temporal integration of high-speed aerodynamics, multiphase flows and hyperbolic equations in general. Meanwhile, the breakthrough of new computing architectures and the hybridization of supercomputer systems pose a huge portability challenge, particularly for legacy codes, since the computing subroutines that form the algorithms, the so-called kernels, must be adapted to various complex parallel program- ming paradigms. From this perspective, the development of innovative implementations relying on a minimalist set of kernels simplifies the deployment of scientific computing software on state-of-the-art supercomputers, while it requires the reformulation of algorithms, such as the aforementioned flux lim- iters. Equipped with basic algebraic topology and graph theory underlying the classical mesh concept, a new flux limiter formulation is presented based on the adoption of algebraic data structures and kernels. As a result, traditional flux limiters are cast into a stream of only two types of computing kernels: sparse matrix-vector multiplication and generalized pointwise binary operators. The newly proposed formulation eases the deployment of such a numerical technique in massively parallel, potentially hybrid, computing systems and is demonstrated for a canonical advection problem.A new general method to compute dispersion errors on Cartesian stretched meshes for both linear and non-linear operators
http://hdl.handle.net/2117/366824
A new general method to compute dispersion errors on Cartesian stretched meshes for both linear and non-linear operators
Ruano Pérez, Jesús; Baez Vidal, Aleix; Rigola Serrano, Joaquim; Trias Miquel, Francesc Xavier
The present article presents a new analysis for the dispersion error and the methodology to evaluate it numerically. Here we present the spectral properties of several convective schemes, including non-linear ones, on Cartesian stretched grids for linear advection problems. Results obtained with this method when applied to uniform structured meshes, converge to the results obtained with the classical method for all the studied schemes. Additionally, effects on the time step depending on which scheme is used are considered using the proposed method. The extracted conclusions taken into account both errors and computational cost allow to propose an optimal scheme according to the selected meshing strategy.
Wed, 04 May 2022 15:07:54 GMThttp://hdl.handle.net/2117/3668242022-05-04T15:07:54ZRuano Pérez, JesúsBaez Vidal, AleixRigola Serrano, JoaquimTrias Miquel, Francesc XavierThe present article presents a new analysis for the dispersion error and the methodology to evaluate it numerically. Here we present the spectral properties of several convective schemes, including non-linear ones, on Cartesian stretched grids for linear advection problems. Results obtained with this method when applied to uniform structured meshes, converge to the results obtained with the classical method for all the studied schemes. Additionally, effects on the time step depending on which scheme is used are considered using the proposed method. The extracted conclusions taken into account both errors and computational cost allow to propose an optimal scheme according to the selected meshing strategy.On the Interpolation Problem for the Poisson Equation on Collocated Meshes
http://hdl.handle.net/2117/366809
On the Interpolation Problem for the Poisson Equation on Collocated Meshes
Santos Serrano, Daniel; Muela Castro, Jordi; Valle Marchante, Nicolás; Trias Miquel, Francesc Xavier
The appearence of unphysical velocities in highly distorted meshes is a common problem in many simulations. In collocated meshes, this problem arises from the interpolation of the pressure gradient from faces to cells. Using an algebraic form for the classical incompressible Navier-Stokes equations, this problem is adressed. Starting from the work of F. X. Trias et. al. [FX.Trias et al. JCP 258: 246-267, 2014], a new approach for studying the Poisson equation obtained using the Fractional Step Method is found, such as a new interpolator is proposed in order to found a stable solution, which avoid the appearence of these unpleasant velocities. The stability provided by the interpolator is formally proved for cartesian meshes and its rotations, using fully-explicit time discretizations. The construction of the Poisson equation is supported on mimicking the symmetry properties of the differential operators and the Fractional Step Method. Then it is reinterpreted using a recursive application of the Fractional Step Method in order to study the system as an stationary iterative solver. Furthermore, a numerical analysis for unstructured mesh is also provided.
Wed, 04 May 2022 12:52:14 GMThttp://hdl.handle.net/2117/3668092022-05-04T12:52:14ZSantos Serrano, DanielMuela Castro, JordiValle Marchante, NicolásTrias Miquel, Francesc XavierThe appearence of unphysical velocities in highly distorted meshes is a common problem in many simulations. In collocated meshes, this problem arises from the interpolation of the pressure gradient from faces to cells. Using an algebraic form for the classical incompressible Navier-Stokes equations, this problem is adressed. Starting from the work of F. X. Trias et. al. [FX.Trias et al. JCP 258: 246-267, 2014], a new approach for studying the Poisson equation obtained using the Fractional Step Method is found, such as a new interpolator is proposed in order to found a stable solution, which avoid the appearence of these unpleasant velocities. The stability provided by the interpolator is formally proved for cartesian meshes and its rotations, using fully-explicit time discretizations. The construction of the Poisson equation is supported on mimicking the symmetry properties of the differential operators and the Fractional Step Method. Then it is reinterpreted using a recursive application of the Fractional Step Method in order to study the system as an stationary iterative solver. Furthermore, a numerical analysis for unstructured mesh is also provided.