An extended hybrid chemistry framework for complex hydrocarbon fuels

Ranade, Rishikesh; Alqahtani, Sultan; Farooq, Aamir; Echekki, Tarek

doi:10.1016/j.fuel.2019.04.053

Cited by 24 publications

(18 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, the maximum concentration of 𝑦 2 is higher than 𝑦 16 by 16 orders of magnitude. We then applied QSSA to ten intermediate species of [3,5,6,10,11,13,14,16,19,20], i.e., assuming that the net production rates of these species are zero: After substituting (A.20) into (A.12), (A.13), (A.14), (A.15), we can get the algebraic expression of each variable to be solved.…”

Section: Appendix a Details Of The Pollution Modelmentioning

confidence: 99%

“…Depending on the applications, many different neural network architectures have been developed, including Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Graph Neural Network (GNN). Some of them have also been employed for datadriven physics modeling [1][2][3][4][5][6][7][8], including turbulent flow modeling [9] and chemical kinetic modeling [10]. Those different neural network architectures introduce specific regularization to the neural network based on the nature of the task such as the scale and rotation invariant of the convolutional kernel in CNN.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stiff-PINN: Physics-Informed Neural Network for Stiff Chemical Kinetics

et al. 2021

View full text Add to dashboard Cite

Recently developed physics-informed neural network (PINN) has achieved success in many science and engineering disciplines by encoding physics laws into the loss functions of the neural network, such that the network not only conforms to the measurements, initial and boundary conditions but also satisfies the governing equations. This work first investigates the performance of PINN in solving stiff chemical kinetic problems with governing equations of stiff ordinary differential equations (ODEs). The results elucidate the challenges of utilizing PINN in stiff ODE systems. Consequently, we employ Quasi-Steady-State-Assumptions (QSSA) to reduce the stiffness of the ODE systems, and the PINN then can be successfully applied to the converted non/mild-stiff systems. Therefore, the results suggest that stiffness could be the major reason for the failure of the regular PINN in the studied stiff chemical kinetic systems. The developed Stiff-PINN approach that utilizes QSSA to enable PINN to solve stiff chemical kinetics shall open the possibility of applying PINN to various reaction-diffusion systems involving stiff dynamics.

show abstract

Section: Appendix a Details Of The Pollution Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Stiff-PINN: Physics-Informed Neural Network for Stiff Chemical Kinetics

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Emerging new methods to develop HyChem models have been recently proposed that use machine learning approaches to fit models for reference fuels against pre-published shock tube experimental data [18,19]. The current study extends these concepts by providing new measurements of a previously uncharacterized fuel and computationally constructing a HyChem model.…”

Section: Introductionmentioning

confidence: 93%

A streamlined approach to hybrid-chemistry modeling for a low cetane-number alternative jet fuel

Pinkowski

Wang

Cassady

et al. 2019

Combustion and Flame

View full text Add to dashboard Cite

The development of renewable, alternative jet fuels presents an exigent challenge to the aviation community. In this work, a streamlined methodology for building computationally efficient kinetic models of real fuels from shock tube experiments is developed and applied to a low cetane-number, broad-boiling alternative jet fuel (termed C-4). A multi-wavelength laser absorption spectroscopy technique was used to determine species time-histories during the high-temperature pyrolysis of C-4, and a batch gradient descent optimization routine built a hybrid-chemistry (HyChem) kinetic model from the measured data. The model was evaluated using combustor-relevant, high-pressure ignition delay time measurements with satisfactory agreement. The present model enables predictive simulations of C-4 in practical environments, while the underlying methodology described here can be readily extended to build kinetic models for a broad range of real fuels of interest.

show abstract

“…In [16], some chemically reasonable requirements were considered such as the mass conservation and the principle of detailed balance. The deep neural networks (DNNs) were applied to extract the chemical reaction rate information in [17,18], but the weights are difficult to interpret physically. In [19], the authors adapted the sparse identification of nonlinear dynamics (SINDy) method [20,21] to the present problem.…”

Section: Introductionmentioning

confidence: 99%

Data-driven discovery of multiscale chemical reactions governed by the law of mass action

Huang,

Zhou,

Yong

2021

Preprint

View full text Add to dashboard Cite

In this paper, we propose a data-driven method to discover multiscale chemical reactions governed by the law of mass action. First, we use a single matrix to represent the stoichiometric coefficients for both the reactants and products in a system without catalysis reactions. The negative entries in the matrix denote the stoichiometric coefficients for the reactants and the positive ones for the products. Second, we find that the conventional optimization methods usually get stuck in the local minima and could not find the true solution in learning the multiscale chemical reactions. To overcome this difficulty, we propose a partial-parameters-freezing (PPF) technique to progressively determine the network parameters by using the fact that the stoichiometric coefficients are integers. With such a technique, the dimension of the searching space is gradually reduced in the training process and the global mimina can be eventually obtained. Several numerical experiments including the classical Michaelis-Menten kinetics and the hydrogen oxidation reactions verify the good performance of our algorithm in learning the multiscale chemical reactions. The code is available at https://github.com/JuntaoHuang/multiscale-chemical-reaction.

show abstract

An extended hybrid chemistry framework for complex hydrocarbon fuels

Cited by 24 publications

References 12 publications

Stiff-PINN: Physics-Informed Neural Network for Stiff Chemical Kinetics

Stiff-PINN: Physics-Informed Neural Network for Stiff Chemical Kinetics

A streamlined approach to hybrid-chemistry modeling for a low cetane-number alternative jet fuel

Data-driven discovery of multiscale chemical reactions governed by the law of mass action

Contact Info

Product

Resources

About