• Corpus ID: 211010687

Embedding Hard Physical Constraints in Neural Network Coarse-Graining of 3D Turbulence

@article{Mohan2020EmbeddingHP,
  title={Embedding Hard Physical Constraints in Neural Network Coarse-Graining of 3D Turbulence},
  author={Arvind T. Mohan and Nick Lubbers and Daniel Livescu and Michael Chertkov},
  journal={arXiv: Computational Physics},
  year={2020}
}
In the recent years, deep learning approaches have shown much promise in modeling complex systems in the physical sciences. A major challenge in deep learning of PDEs is enforcing physical constraints and boundary conditions. In this work, we propose a general framework to directly embed the notion of an incompressible fluid into Convolutional Neural Networks, and apply this to coarse-graining of turbulent flow. These physics-embedded neural networks leverage interpretable strategies from… 

Figures from this paper

Effects of boundary conditions in fully convolutional networks for learning spatio-temporal dynamics
TLDR
Padding, improved spatial context, and explicit encoding of physical boundaries are investigated in the context of fully convolutional networks applied to recurrent tasks, revealing a high sensitivity of both accuracy and stability on the boundary implementation in such recurrent tasks.
Learning Incompressible Fluid Dynamics from Scratch - Towards Fast, Differentiable Fluid Models that Generalize
TLDR
A novel physics-constrained training approach that generalizes to new fluid domains, requires no fluid simulation data, and allows convolutional neural networks to map a fluid state from time-point t to a subsequent state at time t + dt in a single forward pass is proposed, which simplifies the pipeline to train and evaluate neural fluid models.
Teaching Solid Mechanics to Artificial Intelligence: a fast solver for heterogeneous solids
TLDR
A completely new and highly efficient approach to solve non-linear mechanical boundary value problems and / or augment existing solution methods in corresponding hybrid variants, with an acceleration up to factor of 8300 for heterogeneous elastic-plastic materials in comparison to the currently fastest available solvers.
SCRATCH-TOWARDS FAST, DIFFERENTIABLE FLUID MODELS
TLDR
A novel physics-constrained training approach that generalizes to new fluid domains, requires no fluid simulation data, and allows convolutional neural networks to map a fluid state from time-point t to a subsequent state at time t + dt in a single forward pass is proposed, which simplifies the pipeline to train and evaluate neural fluid models.
Computationally Efficient Multiscale Neural Networks Applied to Fluid Flow in Complex 3D Porous Media
TLDR
A general multiscale deep learning model that is able to learn from porous media simulation data and able to perform accurate prediction on domains exhibiting a large degree of heterogeneity is addressed.
PhyGeoNet: Physics-Informed Geometry-Adaptive Convolutional Neural Networks for Solving Parametric PDEs on Irregular Domain
TLDR
A novel physics-constrained CNN learning architecture is proposed, aiming to learn solutions of parametric PDEs on irregular domains without any labeled data, and elliptic coordinate mapping is introduced to enable coordinate transforms between the irregular physical domain and regular reference domain.
A Tailored Convolutional Neural Network for Nonlinear Manifold Learning of Computational Physics Data Using Unstructured Spatial Discretizations
TLDR
A nonlinear manifold learning technique based on deep convolutional autoencoders that is appropriate for model order reduction of physical systems in complex geometries and shows better than an order of magnitude improvement in accuracy over linear methods.
Learning physics-constrained subgrid-scale closures in the small-data regime for stable and accurate LES
TLDR
GCNN+EnCon, which combines these two physics constraints, shows the best a posteriori performance in this regime, and these results illustrate the power of physics-constrained learning in the small-data regime for accurate and stable LES.
Teaching solid mechanics to artificial intelligence—a fast solver for heterogeneous materials
TLDR
The DNN model proves suited for reproducing the stress distribution in geometries different from those used for training, with an efficient approach to solve non-linear mechanical problems, with a acceleration up to a factor of 8300 for elastic-plastic materials compared to typical solvers.
...
...

References

SHOWING 1-10 OF 50 REFERENCES
Towards Physics-informed Deep Learning for Turbulent Flow Prediction
TLDR
This paper proposes a hybrid approach to predict turbulent flow by learning its highly nonlinear dynamics from spatiotemporal velocity fields of large-scale fluid flow simulations of relevance to turbulence modeling and climate modeling by marrying two well-established turbulent flow simulation techniques with deep learning.
Compressed Convolutional LSTM: An Efficient Deep Learning framework to Model High Fidelity 3D Turbulence
High-fidelity modeling of turbulent flows is one of the major challenges in computational physics, with diverse applications in engineering, earth sciences and astrophysics, among many others. The
A Deep Learning based Approach to Reduced Order Modeling for Turbulent Flow Control using LSTM Neural Networks
TLDR
A deep learning based approach is demonstrated to build a ROM using the POD basis of canonical DNS datasets, for turbulent flow control applications and finds that a type of Recurrent Neural Network, the Long Short Term Memory (LSTM) shows attractive potential in modeling temporal dynamics of turbulence.
Latent Space Physics: Towards Learning the Temporal Evolution of Fluid Flow
TLDR
It is demonstrated for the first time that dense 3D+time functions of physics system can be predicted within the latent spaces of neural networks, and the method arrives at a neural‐network based simulation algorithm with significant practical speed‐ups.
Reynolds averaged turbulence modelling using deep neural networks with embedded invariance
TLDR
This paper presents a method of using deep neural networks to learn a model for the Reynolds stress anisotropy tensor from high-fidelity simulation data and proposes a novel neural network architecture which uses a multiplicative layer with an invariant tensor basis to embed Galilean invariance into the predicted anisotropic tensor.
From Deep to Physics-Informed Learning of Turbulence: Diagnostics
TLDR
Tests validating progress made toward acceleration and automation of hydrodynamic codes in the regime of developed turbulence by three Deep Learning (DL) Neural Network (NN) schemes trained on Direct Numerical Simulations of turbulence suggest a path forward towards improving reproducibility of the large-scale geometry of turbulence with NN.
Enforcing statistical constraints in generative adversarial networks for modeling chaotic dynamical systems
Lat-Net: Compressing Lattice Boltzmann Flow Simulations using Deep Neural Networks
TLDR
It is shown that once Lat-Net is trained, it can generalize to large grid sizes and complex geometries while maintaining accuracy and is a general method for compressing other Lattice Boltzmann based simulations such as Electromagnetism.
Physics-informed Autoencoders for Lyapunov-stable Fluid Flow Prediction
TLDR
This work investigates whether it is possible to include physics-informed prior knowledge for improving the model quality, and focuses on the stability of an equilibrium, one of the most basic properties a dynamic system can have, via the lens of Lyapunov analysis.
...
...