Prediction of Aerodynamic Flow Fields Using Convolutional Neural Networks

Abstract

An approximation model based on convolutional neural networks (CNNs) is proposed for flow field predictions. The CNN is used to predict the velocity and pressure field in unseen flow conditions and geometries given the pixelated shape of the object. In particular, we consider Reynolds Averaged Navier-Stokes (RANS) flow solutions over airfoil shapes. The CNN can automatically detect essential features with minimal human supervision and shown to effectively estimate the velocity and pressure field orders of magnitude faster than the RANS solver, making it possible to study the impact of the airfoil shape and operating conditions on the aerodynamic forces and the flow field in near-real time. The use of specific convolution operations, parameter sharing, and robustness to noise are shown to enhance the predictive capabilities of CNN. We explore the network architecture and its effectiveness in predicting the flow field for different airfoil shapes, angles of attack, and Reynolds numbers.

References in corpus (3)

• Deep Learning in Neural Networks: An Overview
• Turbulence Modeling in the Age of Data
• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations

Cited by in corpus (61)

• PhyGeoNet: Physics-Informed Geometry-Adaptive Convolutional Neural Networks for Solving Parameterized Steady-State PDEs on Irregular Domain
• A physics-informed variational DeepONet for predicting the crack path in brittle materials
• Machine Learning in Aerodynamic Shape Optimization
• Physics-informed graph neural Galerkin networks: A unified framework for solving PDE-governed forward and inverse problems
• A Point-Cloud Deep Learning Framework for Prediction of Fluid Flow Fields on Irregular Geometries
• PhyCRNet: Physics-informed Convolutional-Recurrent Network for Solving Spatiotemporal PDEs
• DeepM&Mnet: Inferring the electroconvection multiphysics fields based on operator approximation by neural networks
• Deep Learning in Deterministic Computational Mechanics
• Physics-informed PointNet: A deep learning solver for steady-state incompressible flows and thermal fields on multiple sets of irregular geometries
• Physics guided machine learning using simplified theories
• Convolutional neural networks for fluid flow analysis: toward effective metamodeling and low-dimensionalization
• Learning Deep Implicit Fourier Neural Operators (IFNOs) with Applications to Heterogeneous Material Modeling
• Assessment of unsteady flow predictions using hybrid deep learning based reduced order models
• Learned Turbulence Modelling with Differentiable Fluid Solvers: Physics-based Loss-functions and Optimisation Horizons
• Graph Convolutional Neural Networks for Body Force Prediction
• Multi-fidelity prediction of fluid flow and temperature field based on transfer learning using Fourier Neural Operator
• Three-dimensional deep learning-based reduced order model for unsteady flow dynamics with variable Reynolds number
• Mosaic Flows: A Transferable Deep Learning Framework for Solving PDEs on Unseen Domains
• Nonlinear proper orthogonal decomposition for convection-dominated flows
• Machine Learning for Physics-Informed Generation of Dispersed Multiphase Flow Using Generative Adversarial Networks
• A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence
• Nonlocal Kernel Network (NKN): a Stable and Resolution-Independent Deep Neural Network
• Point-Cloud Deep Learning of Porous Media for Permeability Prediction
• Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression
• Numerical investigation of minimum drag profiles in laminar flow using deep learning surrogates
• Finite volume method network for acceleration of unsteady computational fluid dynamics: non-reacting and reacting flows
• Flow Completion Network: Inferring the Fluid Dynamics from Incomplete Flow Information using Graph Neural Networks
• Airfoil's Aerodynamic Coefficients Prediction using Artificial Neural Network
• Reinforcement-learning-based actuator selection method for active flow control
• Wasserstein Generative Adversarial Uncertainty Quantification in Physics-Informed Neural Networks
• On the Use of Neural Networks for Full Waveform Inversion
• Deep Learning for Real-Time Aerodynamic Evaluations of Arbitrary Vehicle Shapes
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Neural Network Training Using ${\ell }_1$-Regularization and Bi-fidelity Data
• Machine Learning to Predict Aerodynamic Stall
• Sample-Efficient and Surrogate-Based Design Optimization of Underwater Vehicle Hulls
• Deep encoder-decoder hierarchical convolutional neural networks for conjugate heat transfer surrogate modeling
• Adaptive physics-informed neural operator for coarse-grained non-equilibrium flows
• Flowfield prediction of airfoil off-design conditions based on a modified variational autoencoder
• Mapping of coherent structures in parameterized flows by learning optimal transportation with Gaussian models
• An accuracy-enhanced transonic flow prediction method fusing deep learning and reduced-order model
• Data-Driven Multi-Grid Solver for Accelerated Pressure Projection
• Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes
• Deep learning for surrogate modelling of 2D mantle convection
• Multi-scale Time-stepping of Partial Differential Equations with Transformers
• opPINN: Physics-Informed Neural Network with operator learning to approximate solutions to the Fokker-Planck-Landau equation
• A Machine Learning Approach to Classify Vortex Wakes of Energy Harvesting Oscillating Foils
• A Deep Learning algorithm to accelerate Algebraic Multigrid methods in Finite Element solvers of 3D elliptic PDEs
• LordNet: An Efficient Neural Network for Learning to Solve Parametric Partial Differential Equations without Simulated Data
• Physics Informed Neural Networks for Modeling of 3D Flow-Thermal Problems with Sparse Domain Data
• Local neural operator for solving transient partial differential equations on varied domains
• Benchmarking Convolutional Neural Network and Graph Neural Network based Surrogate Models on a Real-World Car External Aerodynamics Dataset
• Quantum DeepONet: Neural operators accelerated by quantum computing
• Seeing moiré: convolutional network learning applied to twistronics
• Bridging the Gap: Machine Learning to Resolve Improperly Modeled Dynamics
• Learning phase field mean curvature flows with neural networks
• Polynomial Ridge Flowfield Estimation
• TripOptimizer: Generative 3D Shape Optimization and Drag Prediction using Triplane VAE Networks
• Real-time simulation of parameter-dependent fluid flows through deep learning-based reduced order models
• A robust and stable hybrid neural network/finite element method for 2D flows that generalizes to different geometries
• CompNO: A Novel Foundation Model approach for solving Partial Differential Equations