Searcharxiv⌕ Search

arXiv subjects

Oriol Lehmkuhl

Publications and source records attributed to Oriol Lehmkuhl.

At least 19 recordsLinked to original sources

Impact of Wind Direction on Flow and Turbulent Statistics Over a Realistic Urban Area: A Large-Eddy Simulation Study

Effects of wind direction in realistic urban canopies remain difficult to characterize systematically because local flow patterns, building-height variability, and turbulent statistics within a realistic urban canopy may respond differently to changes in the approaching wind. To investigate these multi-scale directional impacts, we conducted high-resolution large-eddy simulations of the atmospheric boundary layer over the Zona Universitaria Pedralbes district in Barcelona. The computational meshes employ fourth-order spectral elements, yielding pedestrian-level resolutions below 1 m and resulting in approximately 5.0e8 degrees of freedom. Sixteen wind directions uniformly distributed over 360deg are simulated to capture the full directional response of the urban fabric.

physics.flu-dyn↗

Leveraging unstructured grids for direct numerical simulations of wall turbulence

Towards computational cost saving for direct numerical simulations (DNSs) of wall turbulence, we formulate an unstructured grid-generation framework, termed $η$-grid, where the wall-normal ($y$) and spanwise ($z$) grid sizes are proportional to the local Kolmogorov scale $η$. The framework consists of an inner layer, with a thickness $\sim 50$ viscous units, with viscous-scaled grid sizes similar to a conventional DNS grid: $0.3 \lesssim Δy^+ \lesssim 4, Δz^+ \simeq 5$ over a smooth wall, and $\ell^+/30 \lesssim Δy^+, Δz^+ \lesssim 4$ over uneven surfaces, where $\ell^+$ is the smallest surface wavelength. Above the inner layer, $Δy^+ \simeq Δz^+ \simeq 2η^+$. We test $η$-grid with finite volume and spectral element solvers, and conduct DNSs of turbulent channel flows and boundary layers over smooth wall and various streamwise-aligned riblets, up to friction Reynolds number $δ^+_0 = 1000$. We assess the accuracy of $η$-grid against the conventional Cartesian grids, through comparison with the reference DNS and experimental data. Results from $η$-grid and the Cartesian grids differ by less than $1\%$, in terms of turbulence statistics up to second-order, and the energy spectra. For turbulent channel flows with $10^3 \lesssim δ^+_0 \lesssim 10^4$, the number of grid points with $η$-grid ($N_η$) scales $\propto {δ^+_0}^{2.47}$ over a smooth wall, and $\propto {δ^+_0}^{2.0-2.47}$ over riblets, whereas the number of grid points with a Cartesian grid and hyperbolic-tangent $y$-grid ($N_\mathrm{Tanh}$) scales $\propto {δ^+_0}^{3.0}$. By $δ^+_0 = 6000$, $N_η/N_\mathrm{Tanh} \simeq 0.1$ over a smooth wall, and $N_η/N_\mathrm{Tanh} \simeq 0.04$ over typical drag-reducing riblets, with viscous-scaled spacing $15$.

physics.flu-dyn↗

Can Explicit Subgrid Models Enhance Implicit LES Simulations? A Very High-Order Solver Perspective

High-order discontinuous Galerkin (DG) methods offer excellent accuracy for turbulent-flow simulations and are increasingly attractive on GPU-oriented architectures, where high polynomial orders can improve arithmetic intensity. However, very high-order under-resolved simulations remain sensitive to the balance between numerical and modeled dissipation. We investigate how explicit Vreman subgrid-scale (SGS) modeling interacts with dissipation from split-form stabilization and Riemann solvers in a DGSEM framework. Using the three-dimensional Taylor-Green vortex at Re=1600 and in the inviscid limit, we assess kinetic-energy dissipation, spectral accuracy, and stability across well-resolved, under-resolved viscous, and strongly under-resolved regimes, comparing lower- and very high-order configurations at similar degrees of freedom. The usefulness of explicit SGS modeling depends strongly on resolution, polynomial order, and the numerical dissipation already present. In the well-resolved cases considered, Vreman modeling does not improve accuracy because its active wavenumber range overlaps with the scheme's inherent dissipation. At similar degrees of freedom, lower-order simulations introduce stronger damping near the smallest resolved scales, whereas very high-order simulations preserve more spectral content but are more susceptible to high-wavenumber energy accumulation when dissipation is insufficient. Under stronger under-resolution, a weak SGS contribution can control this accumulation, while excessive SGS dissipation degrades intermediate scales. These results identify regimes in which explicit SGS modeling is beneficial, neutral, or detrimental, and provide practical guidance for selecting dissipation mechanisms in very high-order DG turbulence simulations suited to modern GPU architectures.

physics.flu-dyn↗

Optimization-Embedded Active Multi-Fidelity Surrogate Learning for Multi-Condition Airfoil Shape Optimization

Active multi-fidelity surrogate modeling is developed for multi-condition airfoil shape optimization to reduce high-fidelity CFD cost while retaining RANS-consistent aerodynamic metrics. The framework couples a low-fidelity-informed Gaussian process regression transfer model with uncertainty-triggered sampling and a synchronized elitism rule embedded in a hybrid genetic algorithm. Low-fidelity XFOIL evaluations provide inexpensive features, while sparse RANS simulations are adaptively allocated when predictive uncertainty exceeds a threshold; elite candidates are mandatorily validated at high fidelity, and the population is re-evaluated to prevent evolutionary selection based on outdated fitness values produced by earlier surrogate states. The method is demonstrated for a two-point problem at $Re=6\times10^6$ with cruise at $α=2^\circ$ (maximize $E=L/D$) and take-off at $α=10^\circ$ (maximize $C_L$) using a 12-parameter CST representation. Independent multi-fidelity surrogates per flight condition enable decoupled refinement. The optimized design improves cruise efficiency by 41.05% and take-off lift by 20.75% relative to the best first-generation individual. Over the full campaign, RANS evaluations were required for only 14.78% and 9.5% of the condition-specific candidate evaluations at cruise and take-off, respectively. These percentages quantify the reduction in high-fidelity usage relative to the fixed automated RANS workflow adopted as the high-fidelity reference in this study.

physics.flu-dyn↗

High-lift Wing Separation Control via Bayesian Optimization and Deep Reinforcement Learning

This study investigates active flow control (AFC) of a 30P30N high-lift wing at a Reynolds number Re$_c$ = 450,000 and angle of attack $α$ = 23$^\circ$ using wallresolved large-eddy simulations (LES). Two optimization strategies are explored: open-loop Bayesian optimization (BO) and closed-loop deep reinforcement learning (DRL), both targeting the mitigation of stall and the improvement of aerodynamic efficiency via synthetic jets on the slat, main, and flap elements. The uncontrolled configuration was validated against literature data, confirming the reliability of the LES setup. The BO framework successfully identified steady jet velocities that increased efficiency by +10.9% through a -9.7% drag reduction while maintaining lift. In contrast, the DRL agent, despite leveraging instantaneous flow information from distributed sensors, achieved only minor improvements in lift and drag, with negligible efficiency gain. Training analysis indicated that the penalty-dominated reward constrained exploration. These results highlight the need for carefully designed rewards and computational acceleration strategies in DRL-based flow control at high Reynolds numbers.

physics.flu-dyn↗

Non-equilibrium effects in turbulent boundary layers over riblets: DNS of step changes in surface texture

We computationally study the response of zero-pressure-gradient (ZPG) turbulent boundary layers (TBLs) to streamwise step changes from a smooth wall to riblets (SM_RI), and vice versa (RI_SM). To quantify the departure from equilibrium due to the step changes, we conduct reference calculations of ZPG TBLs over an entirely smooth wall, and an entirely riblet-covered surface. To save the computational cost, we generate an optimal grid for an unstructured spectral-element code, consistent with the size of turbulent scales across the TBL. By the step change, the momentum thickness Reynolds number reaches $Re_{θ_0} \simeq 680$ (friction Reynolds number $Re_{τ_0} \simeq 283$), and by the domain outlet downstream of the step change, $Re_θ \simeq 1000$ ($Re_τ \simeq 400$). The TBL departure from equilibrium due to the step change, and its subsequent relaxation, recall previous studies on step changes in surface roughness. Downstream of the step change, growth of the internal equilibrium layer thickness $δ_\text{IEL}$, hence recovery to equilibrium, follows two stages. Stage I corresponds to the recovery up to the buffer region ($y^+ \simeq 10$), which is slower during the RI_SM step change than the SM_RI counterpart. For the RI_SM cases during Stage I, $δ_\text{IEL} \propto (x/k)^{0.6}$, and this stage is completed by $x \simeq 100k \simeq (5δ_0 - 20δ_0)$ downstream of the step change, where $k$ is the riblet height. Stage II recovery i.e.\ recovery of the outer region, is quite slow. Therefore, for drag-increasing riblets with $k^+ \ge 25$, $δ_\text{IEL}$ does not reach the boundary layer thickness, even up to $50δ_0$ downstream of the step change, owing to the advected frozen wake from upstream. As a result, skin-friction coefficient reaches to more than $90\%$ of its equilibrium counterpart, but does not reach its $100\%$.

physics.app-ph↗

Reduced-order modeling of large-scale turbulence using Koopman $β$-variational autoencoders

Reduced-order models (ROMs) are very popular for surrogate modeling of full-order computational fluid dynamics (CFD) simulations, allowing for real-time approximation of complex flow phenomena. However, their application to CFD models including large eddy simulation (LES) and direct numerical simulaton (DNS) is limited due to the highly chaotic and multi-scale nature of resolved turbulent flow. Due to the large amounts of noise present in small-scale turbulent structures, error accumulation becomes a major issue, making long-term prediction of unsteady flow infeasible. While linear subspace methods like dynamic mode decomposition (DMD) can be used to pre-process turbulent flow data to remove small-scale structures, this often requires a very large number of modes and a non-trivial mode selection process. In this work, a ROM framework using Koopman $β$-variational autoencoders ($β$-VAEs) is introduced for reduced-order modeling of large-scale turbulence. The Koopman operator captures the variation of a non-linear dynamical system through a linear representation of state observables. By constraining the latent space of a $β$-VAE to grow linearly using a Koopman-inspired loss function, small-scale turbulent structures are filtered out in reconstructions of input data and latent variables are denoised in an unsupervised manner so that they can be sufficiently modeled over time. Combined with a long short-term memory (LSTM) ensemble for time series prediction of latent variables, the model is tested on LES flow past a Windsor body at multiple yaw angles, showing that the Koopman $β$-VAE can effectively denoise latent variables and remove small-scale structures from reconstructions while acting globally over multiple cases.

physics.flu-dyn↗

Application of riblets to separating turbulent boundary layers

We conduct direct numerical simulations of separating turbulent boundary layers (TBLs) over triangular riblets with tip angles $90^o$ (T9) and $60^o$ (T6). Our setup follows the separating TBL study of Wu et al.\ ({\it J. Fluid Mech.}, vol.\ 883, 2020, p.\ A45). An equilibrium zero pressure-gradient (ZPG) TBL is generated at a reference location, followed by imposition of a Gaussian suction profile to create a separation bubble. The ZPG TBLs over the riblets and the benchmark smooth case have matched momentum thickness Reynolds number $Re_{θ_0} = 583$ (friction Reynolds number 224). We employ a well-validated spectral-element solver, and leverage its unstructured-grid nature to generate an optimal grid, based on the size of turbulent scales across the TBL. At the reference location, the T9 and T6 riblets respectively increase and reduce drag, with viscous-scaled spacings $52$ and $13$. We discover that for both riblet cases, the mean separation location occurs at a distance of $140θ_0$ downstream of the reference location, $18\%$ shorter than the mean separation distance for the smooth case ($170θ_0$). This outcome is related to the progressive enhancement of the Kelvin-Helmholtz (KH) rollers over the riblets, owing to the continuous rise in the adverse pressure-gradient. The KH rollers penetrate into the turbulent separation bubble, with significantly larger size and coherence compared to their counterparts upstream of the mean separation location.

physics.app-ph↗

Impact of Wind Direction on Flow Over a Realistic Urban Area: A Large-Eddy Simulation Study

We conducted high-resolution large-eddy simulations over a real urban district in Barcelona to examine the impact of wind direction on near-ground flow. The computational mesh resolves over 500 million degrees of freedom, with a spatial resolution on the order of 1 m at pedestrian level. This allows a detailed analysis of mean velocity and turbulence patterns within the canopy layer. Although instantaneous flow fields differ significantly between cases, double-averaged profiles of velocity and turbulence intensity remain remarkably consistent across all wind directions. The results reveal a shear-driven mixing layer below the average building height and turbulence maxima near the tallest buildings, highlighting the influence of urban morphology on the development of flow and turbulence.

physics.flu-dyn↗

Projection-based stabilization for high-order incompressible flow solvers

This work presents a novel stabilization strategy for the Galerkin formulation of the incompressible Navier-Stokes equations, developed to achieve high accuracy while ensuring convergence and compatibility with high-order elements on unstructured meshes. The numerical algorithm employs a fractional step method with carefully defined boundary conditions to obtain a consistent pressure field, enabling high-order temporal accuracy. The proposed stabilization is seamlessly integrated into the algorithm and shares the same underlying principle as the natural stabilization inherent in the fractional step method, both rely on the difference between the gradient operator and its projection. The numerical dissipation associated to the stabilization term is found to diminish with increasing polynomial order of the elements. Numerical test cases confirm the effectiveness of the method, demonstrating convergence under mesh refinement and increasing polynomial order.

math.NA↗

SmartFlow: A CFD-solver-agnostic deep reinforcement learning framework for computational fluid dynamics on HPC platforms

Deep reinforcement learning (DRL) is emerging as a powerful tool for fluid-dynamics research, encompassing active flow control, autonomous navigation, turbulence modeling and discovery of novel numerical schemes. We introduce SmartFlow, a CFD-solver-agnostic framework for both single- and multi-agent DRL algorithms that can easily integrate with MPI-parallel CPU and GPU-accelerated solvers. Built on Relexi and SmartSOD2D, SmartFlow uses the SmartSim infrastructure library and our newly developed SmartRedis-MPI library to enable asynchronous, low-latency, in-memory communication between CFD solvers and Python-based DRL algorithms. SmartFlow leverages PyTorch's Stable-Baselines3 for training, which provides a modular, Gym-like environment API. We demonstrate its versatility via three case studies: single-agent synthetic-jet control for drag reduction in a cylinder flow simulated by the high-order FLEXI solver, multi-agent cylinder wake control using the GPU-accelerated spectral-element code SOD2D, and multi-agent wall-model learning for large-eddy simulation with the finite-difference solver CaLES. SmartFlow's CFD-solver-agnostic design and seamless HPC integration is promising to accelerate RL-driven fluid-mechanics studies.

physics.flu-dyn↗

Deep-reinforcement-learning-based separation control in a two-dimensional airfoil

The aim of this study is to discover new active-flow-control (AFC) techniques for separation mitigation in a two-dimensional NACA 0012 airfoil at a Reynolds number of 3000. To find these AFC strategies, a framework consisting of a deep-reinforcement-learning (DRL) agent has been used to determine the action strategies to apply to the flow. The actions involve blowing and suction through jets at the airfoil surface. The flow is simulated with the numerical code Alya, which is a low-dissipation finite-element code, on a high-performance computing system. Various control strategies obtained through DRL led to 43.9% drag reduction, while others yielded an increase in aerodynamic efficiency of 58.6%. In comparison, periodic-control strategies demonstrated lower energy efficiency while failing to achieve the same level of aerodynamic improvements as the DRL-based approach. These gains have been attained through the implementation of a dynamic, closed-loop, time-dependent, active control mechanism over the airfoil.

physics.flu-dyn↗

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

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. The latter mimics a typical central European urban geometry. In both cases, the incident wind angle is $0^\circ$. The Reynolds number for each case are $Re_H = 5.0 \times 10^6$ and $8.0 \times 10^6$, respectively ($Re_H = U_{ref} H/ν$ with $U_{ref}$ and $H$ denoting the reference velocity and building height, respectively, and $ν$ the kinematic viscosity). The LES employs a high-order, low-dissipation numerical scheme with a spatial resolution of 0.75m within the urban canopy. An online precursor simulation ensures realistic turbulent inflow conditions. The simulations performed successfully captures 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. 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.

physics.flu-dyn↗

On Deep-Learning-Based Closures for Algebraic Surrogate Models of Turbulent Flows

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 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 [2.5,5,7.5,10,12.5], with 7.5 as a test case. Key coherent modes are identified by clustering them based on dominant frequency dynamics using Hotelling T2 on the spectral properties of temporal coefficients. These coherent modes account for nearly 60% of the total energy while comprising less than 10% of all modes. A common POD basis is created by concatenating coherent modes from training angles and orthonormalizing the set, reducing the basis vectors from 142 to 90 without losing information. Transformers with different size on the attention layer, (64, 128 and 256), are trained to model the missing fluctuations. Larger attention sizes always improve predictions for the training set, but the transformer with an attention layer of size 256 overshoots the fluctuations predictions in the test set because they have lower intensity than in the training cases. Adding the predicted fluctuations closes the energy gap between the reconstruction and the original flow field, improving predictions for energy, root-mean-square velocity fluctuations, and instantaneous flow fields. The deepest architecture reduces mean energy error from 37% to 12% and decreases the Kullback--Leibler divergence of velocity distributions from KL=0.2 to below KL=0.026.

physics.flu-dyn↗

Differentially heated turbulent channel flow two-point correlations

This study analyzes the behavior of a differentially heated channel flow by means of a direct numerical simulations (DNS) with variable thermophysical properties under low-speed conditions focusing on the impact of the temperature gradient on the turbulence structures near the channel walls. The simulations were conducted at a mean friction Reynolds number of Reτm = 400 with a temperature ratio between the walls of Thot/Tcold = 2. Results show significant differences between the hot and cold walls that lead to an increased heat transfer at the hot wall and a higher turbulent production in the cold wall.

physics.flu-dyn↗

Towards Active Flow Control Strategies Through Deep Reinforcement Learning

This paper presents a deep reinforcement learning (DRL) framework for active flow control (AFC) to reduce drag in aerodynamic bodies. Tested on a 3D cylinder at Re = 100, the DRL approach achieved a 9.32% drag reduction and a 78.4% decrease in lift oscillations by learning advanced actuation strategies. The methodology integrates a CFD solver with a DRL model using an in-memory database for efficient communication between

cs.LG↗

Turbulent Boundary Layer in a 3-Element High-LiftWing: Coherent Structures Identification

A wall-resolved large-eddy simulation (LES) of the fluid flow around a 30P30N airfoil is conducted at a Reynolds number of Rec=750,000 and an angle of attack (AoA) of 9 degrees. The simulation results are validated against experimental data from previous studies and further analyzed, focusing on the suction side of the wing main element. The boundary layer development is investigated, showing characteristics typical of a zero-pressure-gradient turbulent boundary layer (ZPG TBL). In particular, the boundary layer exhibits limited growth, and the outer peak of the streamwise Reynolds stresses is virtually absent, distinguishing it from an adverse-pressure-gradient turbulent boundary layer (APG TBL). A proper orthogonal decomposition (POD) analysis is performed on a portion of the turbulent boundary layer, revealing a significant energy spread across higher-order modes. Despite this, TBL streaks are identified, and the locations of the most energetic structures correspond to the peaks in the Reynolds stresses.

physics.flu-dyn↗

pyLOM: A HPC open source reduced order model suite for fluid dynamics applications

This paper describes the numerical implementation in a high-performance computing environment of an open-source library for model order reduction in fluid dynamics. This library, called pyLOM, contains the algorithms of proper orthogonal decomposition (POD), dynamic mode decomposition (DMD) and spectral proper orthogonal decomposition (SPOD), as well as, efficient SVD and matrix-matrix multiplication, all of them tailored for supercomputers. The library is profiled in detail under the MareNostrum IV supercomputer. The bottleneck is found to be in the QR factorization, which has been solved by an efficient binary tree communications pattern. Strong and weak scalability benchmarks reveal that the serial part (i.e., the part of the code that cannot be parallelized) of these algorithms is under 10% for the strong scaling and under 0.7% for the weak scaling. Using pyLOM, a POD of a dataset containing 1.14 x 108 gridpoints and 1808 snapshots that takes 6.3Tb of memory can be computed in 81.08 seconds using 10368 CPUs. Additioally, the algorithms are validated using the datasets of a flow around a circular cylinder at ReD = 100 and ReD = 1 x 104, as well as the flow in the Stanford diffuser at Reh = 1 x 104.

physics.flu-dyn↗