Direct mapping from LES resolved scales to filtered-flame generated manifolds using convolutional neural networks

Seltz, Andréa; Domingo, Pascale; Vervisch, Luc; Nikolaou, Zacharias M.

doi:10.1016/j.combustflame.2019.08.014

Cited by 58 publications

(26 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(2020) suggested method used an ANN approach to improve CFD code computational time for combustion simulations. Seltz et al. (2019) developed stoichiometric predictions for turbulent premixed jet flames of methane and air, based on ANN approach.…”

Section: Introductionmentioning

confidence: 99%

Analysis of deep learning neural network combined with experiments to develop predictive models for a propane vertical jet fire

Mashhadimoslem

Ghaemi

Palacios

2020

Heliyon

View full text Add to dashboard Cite

Fires are important responsible factors to cause catastrophic events in the process industries, whose consequences usually initiate domino effects. The artificial neural network has been shown to be one of the rapid methods to simulate processes in the risk analysis field. In the present work, experimental data points on jet fire shape ratios, defined by the 800 K isotherm, have been applied for ANN development. The mass flow rates and the nozzle diameters of these jet flames have been considered as input dataset; while, the jet flame lengths and widths have been collected as output dataset by the ANN models. A Bayesian Regularization algorithm has been chosen as the three-layer backpropagation training from Multi-layer perceptron algorithm. Then it has been compared with a Radial based functions algorithm, based on single hidden layer. The optimized number of neurons in the first and second hidden layers of the MLP algorithm, and in the single hidden layer of the RBF algorithm has been found to be twenty and fifteen, respectively. The best MSE validation performance of MLP and RBF networks has been found to be 0.00286 and 0.00426 at 100 and 20 epochs, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Analysis of deep learning neural network combined with experiments to develop predictive models for a propane vertical jet fire

Mashhadimoslem

Ghaemi

Palacios

2020

Heliyon

View full text Add to dashboard Cite

show abstract

“…The complete structure of the sequential CNNs is summarized in Table 1. The CNN architecture is inspired from [23], with two additional dense layers to secure the performance of the regression task required here. In total, the neural network contains 237,579 weights, which need to be adjusted.…”

Section: Turbulent Flame Configuration and Numericsmentioning

confidence: 99%

“…Artificial neural networks (ANNs) were found accurate, and both CPU and memory efficient [18,19]. Along similar lines, convolutional neural networks (CNNs), which were originally developed for analyzing visual representations [20], have been introduced in turbulent combustion modeling for direct deconvolution of the filtered progress variable [21], for modeling the unresolved flame surface wrinkling [22] and for the direct reconstruction of unresolved sources and fluxes from mesh-resolved quantities in large-eddy simulation (LES) [23]. Compared to ANNs, CNNs reduce the number of connections per layer to stack more of them efficiently, thus increasing the depth of the neural network.…”

Section: Introductionmentioning

confidence: 99%

Machine learning for detailed chemistry reduction in DNS of a syngas turbulent oxy-flame with side-wall effects

Wan

Barnaud

Domingo

2021

Proceedings of the Combustion Institute

Self Cite

View full text Add to dashboard Cite

A novel chemistry reduction strategy based on convolutional neural networks (CNNs) is developed and applied to direct numerical simulation (DNS) of a turbulent non-premixed flame interacting with a cooled wall. The fuel syngas mixture is burning in pure oxygen. The training and the subsequent application of the CNN rely on the processing of two-dimensional (2D) images built from species mass fractions and temperature (CNN input), to predict the corresponding chemical sources at the center of the image (CNN output). This image-type treatment of chemistry is found to efficiently capture intermediate radicals species highly sensitive to the local flame topology. To reduce the CPU cost, a simplified 2D DNS database with detailed chemistry serves as reference and is used for training and testing the neural network. Comparisons are also made a posteriori against the same 2D DNS with a reduced chemical scheme specialized for syngas. Then, three-dimensional (3D) DNS are conducted either with CNN or the reduced chemistry for more a posteriori tests. The CNN reduced chemistry outperforms the reduced Arrhenius based mechanism in the prediction of radical species, such as monoatomic hydrogen, and also in terms of CPU cost.

show abstract

“…ω + |c from DNS vsc for filter sizes ×: 0.3 mm, : 0.6 mm, •: 0.9 mm, (δ L = 0.4 mm). Reprinted with permission [39].…”

Section: Normalised Sourcementioning

confidence: 99%

From Discrete and Iterative Deconvolution Operators to Machine Learning for Premixed Turbulent Combustion Modeling

Domingo

Nikolaou

Seltz

2020

Data Analysis for Direct Numerical Simulations of Turbulent Combustion

View full text Add to dashboard Cite

Following the rapid and continuous progress of computing power allowing for increasing the mesh resolution in large eddy simulation (LES), new modeling strategies appear which are based on a direct treatment of the now well-resolved, but still not fully-resolved scalar signals. Along this line, deconvolution or inverse filtering, either based on discrete or iterative operators, is first discussed. Recent results obtained from a direct numerical simulation (DNS) database and LES of a premixed turbulent jet flame are presented. The analysis confirms the potential of deconvolution to approximate the unclosed non-linear terms and the SGS fluxes. Then, the introduction of machine learning in turbulent combustion modeling is illustrated in the context of convolutional neural networks.

show abstract

Direct mapping from LES resolved scales to filtered-flame generated manifolds using convolutional neural networks

Cited by 58 publications

References 56 publications

Analysis of deep learning neural network combined with experiments to develop predictive models for a propane vertical jet fire

Analysis of deep learning neural network combined with experiments to develop predictive models for a propane vertical jet fire

Machine learning for detailed chemistry reduction in DNS of a syngas turbulent oxy-flame with side-wall effects

From Discrete and Iterative Deconvolution Operators to Machine Learning for Premixed Turbulent Combustion Modeling

Contact Info

Product

Resources

About