Reactive Flow

Reactive flow modeling aims to accurately and efficiently simulate complex chemical reactions within fluid systems, crucial for applications ranging from combustion engines to environmental remediation. Current research heavily emphasizes data-driven approaches, employing machine learning techniques like neural networks (including Bayesian and physics-informed variants), and generative adversarial networks (GANs) to reduce computational costs and improve predictive accuracy, often incorporating transfer learning to leverage existing datasets. These advancements focus on enhancing the efficiency and reliability of simulations, particularly for under-resolved flows and scenarios with limited experimental data, ultimately improving the design and optimization of various technological systems.

Papers

October 31, 2024

Understanding Optimization in Deep Learning with Central Flows
Jeremy M. Cohen, Alex Damian, Ameet Talwalkar, Zico Kolter, Jason D. Lee
Deep Learning Optimization Purpose Adaptive Optimizers Minimum Time Trajectory Reactive Flow

October 22, 2024

EnKode: Active Learning of Unknown Flows with Koopman Operators
Alice Kate Li, Thales C. Silva, M. Ani Hsieh
Active Learning Koopman Operator Adaptive Sampling Active Sampling Flow Based Model Flow Estimation Reactive Flow

May 17, 2024

Probabilistic transfer learning methodology to expedite high fidelity simulation of reactive flows
Bruno S. Soriano, Ki Sung Jung, Tarek Echekki, Jacqueline H. Chen, Mohammad Khalil
NCD Method High Fidelity Reduced Order Principal Component Bayesian Transfer Reactive Flow

December 1, 2023

Transfer learning for predicting source terms of principal component transport in chemically reactive flow
Ki Sung Jung, Tarek Echekki, Jacqueline H. Chen, Mohammad Khalil
Transfer Learning Formality Transfer Model Reduction Data Driven Reduced Order Laser Induced Reactive Flow

November 1, 2022

Transfer Learning with Physics-Informed Neural Networks for Efficient Simulation of Branched Flows
Raphaël Pellegrin, Blake Bullwinkel, Marios Mattheakis, Pavlos Protopapas
Neural Network Deep Learning Transfer Learning Physic Informed Neural Network Multi Head Fast Simulation Random Dynamical System Reactive Flow

October 28, 2022

Applying Physics-Informed Enhanced Super-Resolution Generative Adversarial Networks to Finite-Rate-Chemistry Flows and Predicting Lean Premixed Gas Turbine Combustors
Mathis Bode
Super Resolution Generative Adversarial Network Finite Rate Chemistry Reactive Flow

April 25, 2022

On the Performance of Machine Learning Methods for Breakthrough Curve Prediction
Daria Fokina, Oleg Iliev, Pavel Toktaliev, Ivan Oseledets, Felix Schindler
Machine Learning System Performance Selective Focus Porous Medium Reactive Flow

Reactive Flow

Papers

Understanding Optimization in Deep Learning with Central Flows

EnKode: Active Learning of Unknown Flows with Koopman Operators

Probabilistic transfer learning methodology to expedite high fidelity simulation of reactive flows

Transfer learning for predicting source terms of principal component transport in chemically reactive flow

Transfer Learning with Physics-Informed Neural Networks for Efficient Simulation of Branched Flows

Applying Physics-Informed Enhanced Super-Resolution Generative Adversarial Networks to Finite-Rate-Chemistry Flows and Predicting Lean Premixed Gas Turbine Combustors

On the Performance of Machine Learning Methods for Breakthrough Curve Prediction