We present the latest developments of our High-Order Spectral Element Solver (), an open source high-order discontinuous Galerkin framework, capable solving a variety flow applications, including compressible flows (with or without shocks), incompressible flows, various RANS and LES turbulence models, particle dynamics, multiphase aeroacoustics. provide overview spatial discretisation (including energy/entropy stable schemes) anisotropic p-adaptation capabilities. The solver is parallelised...

Aeroacoustic noise is a major concern in wind turbine design that can be minimized by optimizing the airfoils shape rotating blades. To this end, we present framework for airfoil optimization to reduce trailing edge of Far-field evaluated using Amiet's theory coupled with TNO-Blake model calculate wall pressure spectrum and fast turn-around XFOIL simulations evaluate boundary layer parameters. The computational first validated NACA0012 at 0° angle attack. Particle swarm used find optimized...

We propose an invariant feature space for the detection of viscous dominated and turbulent regions (i.e., boundary layers wakes). The developed methodology uses principal invariants strain rotational rate tensors as input to unsupervised Machine Learning Gaussian mixture model. selected is independent coordinate frame used generate processed data, it relies on rate, which are Galilean invariants. This allows us identify two distinct flow regions: a dominated, region (boundary layer wake...

Abstract. The growing demand for offshore wind energy has led to a significant increase in turbine size and the development of large-scale farms, often comprising 100 150 turbines. However, environmental impact underwater noise emissions remains largely unaddressed. This paper quantifies, first time, aerodynamic footprint three large turbines (5 MW, 10 22 MW) farms composed these We propose novel methodology that integrates validated prediction techniques with plane wave propagation theory...

Aerodynamic noise is a limitation for further exploitation of wind energy resources. As this type caused by the interaction turbulent flow with airframe, detailed resolution necessary to obtain an accurate prediction far-field noise. Computational fluid dynamic (CFD) solvers simulate field but only at high computational cost, which much increased when acoustic resolved. Therefore, turbine predictions using numerical approaches remain challenge. This paper presents methodology that couples...

Jump penalty stabilization techniques for under-resolved turbulence have been recently proposed continuous and discontinuous high order Galerkin schemes [1], [2], [3]. The relies on the gradient or solution discontinuity at element interfaces to incorporate localised numerical diffusion in scheme. This acts as an implicit subgrid model stablizes turbulent simulations. paper investigates effect of jump methods (penalising solution) improvement high-order regime. We analyze these using...

This paper compares two actuator line methodologies for modelling wind turbines employing high-order h/p solvers and large-eddy simulations. The methods combine the accuracy of (in this work maximum order is 6) with computational efficiency lines to capture aerodynamic effects turbine blades. Comparisons experiments validate methodologies. We explore polynomial smoothing parameter associated Gaussian regularization function, use them blend forcing in mesh, show that both parameters influence...

We present the latest developments of our High-Order Spectral Element Solver (HORSES3D), an open source high-order discontinuous Galerkin framework, capable solving a variety flow applications, including compressible flows (with or without shocks), incompressible flows, various RANS and LES turbulence models, particle dynamics, multiphase aeroacoustics. provide overview spatial discretisation (including energy/entropy stable schemes) anisotropic p-adaptation capabilities. The solver is...

With the aim of minimising losses produced by fire accidents, engineering applies physics and principles to preserve integrity people, environment infrastructure. Fire modelling is complex due interaction between chemistry, heat transfer fluid dynamics. Commercially available simulation tools necessarily simplify this complexity, excluding less fundamental processes, such as soot production. By not including compound in simulations, interactions radiation transfer, propagation toxicity must...

View Video Presentation: https://doi.org/10.2514/6.2022-0413.vid We use a implicit Large Eddy Simulations methodology on under-resolved meshes to compute the aeroacoustic generated by Siemens-Gamesa airfoil. The solver employ high order discontiuous Galerkin spectral element scheme supplemented with stabilising split-form entropy-stable formulation. aim is compare direct computations of far field pressure fluctuations those obtained Ffowcs Williams and Hawkings predictions. Specifically, we...

We present a framework for airfoil shape optimization to reduce the trailing edge noise design of wind turbine blades. Far-field is evaluated using Amiet's theory coupled with TNO-Blake model calculate wall pressure spectrum and fast turn-around XFOIL simulations evaluate boundary layer parameters. The computational first validated NACA0012 at 0 angle attack. Particle swarm used find optimized configuration. multi-objective minimizes A-weighted overall sound level various angles attack,...

Aerodynamic noise is a limitation for further exploitation of wind energy resources. As this type caused by the interaction turbulent flow with airframe, detailed resolution necessary to obtain an accurate prediction far-field noise. Computational fluid dynamic (CFD) solvers simulate field but only at high computational cost, which much increased when acoustic resolved. Therefore, turbine predictions using numerical approaches remain challenge. This paper presents methodology that couples...

This paper explores two Large Eddy Simulation (LES) approaches within the framework of high-order discontinuous Galerkin solver, Horses3D. The investigation focuses on an Inverted Multi-element Wing in Ground Effect (i.e. 2.5D Imperial Front section) representing a Formula 1 front wing, and compares strengths limitations LES methods. explicit formulation relies Vreman model, that adapts to laminar, transitional turbulent regimes. numerical uses nodal basis functions Gauss points. implicit...

We propose a reinforcement learning strategy to control wind turbine energy generation by actively changing the rotor speed, yaw angle and blade pitch angle. A double deep Q-learning with prioritized experience replay agent is coupled element momentum model trained allow for winds. The decide best (speed, yaw, pitch) simple steady winds subsequently challenged real dynamic turbulent winds, showing good performance. Q- compared classic value iteration both strategies outperform PID in all...

We present a novel approach to automate and optimize anisotropic p-adaptation in high-order h/p solvers using Reinforcement Learning (RL). The dynamic RL adaptation uses the evolving solution adjust polynomials. develop an offline training approach, decoupled from main solver, which shows minimal overcost when performing simulations. In addition, we derive RL-based error estimation that enables quantification of local discretization errors. proposed methodology is agnostic both computational...

High-order solvers are accurate but computationally expensive as they require small time steps to advance the solution in time. In this work, we include a corrective forcing low-order achieve high accuracy while advancing with larger and achieving fast computations. This work is continuation of our previous research [Manrique de Lara Ferrer, “Accelerating order discontinuous Galerkin using neural networks: 1D Burgers' equation,” Comput. Fluids 235, 105274 (2022) F. Manrique E. 3D...

We develop a torque-pitch control framework using deep reinforcement learning for wind turbines to optimize the generation of turbine energy while minimizing operational noise. employ double Q-learning, coupled blade element momentum solver, enable precise over parameters. In addition momentum, we use acoustic model Brooks Pope and Marcolini. Through training with simple winds, agent learns optimal policies that allow efficient complex turbulent winds. Our experiments demonstrate is able...

Jump penalty stabilisation techniques have been recently proposed for continuous and discontinuous high order Galerkin schemes [1,2,3]. The relies on the gradient or solution discontinuity at element interfaces to incorporate localised numerical diffusion in scheme. This acts as an implicit subgrid model stablises under-resolved turbulent simulations. paper investigates effect of jump methods (penalising solution) improvement high-order regime. We analyse these using eigensolution analysis,...

We present a framework for airfoil shape optimization to reduce the trailing edge noise design of wind turbine blades. Far-field is evaluated using Amiet's theory coupled with TNO-Blake model calculate wall pressure spectrum and fast turn-around XFOIL simulations evaluate boundary layer parameters. The computational first validated NACA0012 at zero angle attack. Particle swarm used find optimized configuration. multi-objective minimizes A-weighted overall sound level various angles attack,...

We propose a reinforcement learning strategy to control wind turbine energy generation, by actively changing the rotor speed, yaw angle and blade pitch angle. A double deep Q-Learning with prioritized experience replay agent is coupled element momentum model trained allow for winds. The decide best (speed, yaw, pitch) simple steady winds subsequently challenged real dynamic turbulent winds, showing good performance. shown outperform classic PID in all environments shows be well suited great...