Understanding and PredictingHydrogen Combustion

30/11/2022 – Day 1: Role of hydrogen in the energy and transportation green transition

This session introduces the context and market opportunities of hydrogen across different sectors from production to utilization. External speakers from industry and academia will give lectures on the application of hydrogen for power generation and transportation.

Part I: The Hydrogen Economy: A Background

Part II: Utilization of hydrogen by industrial sector

01/12/2022 – Day 2: From theory to simulation of hydrogen flames

This session introduces the fundamentals of hydrogen combustion from theory to simulations using high-fidelity methods. It includes lectures and hands on activities where the students can utilize the physical models and numerical methods described in the lectures.

This lecture aims to highlight the effects of hydrogen thermo-physical properties on the combustion process. First, the hydrogen combustion chemistry will be discussed in the context of the high burning velocities. Then, the focus will be on hydrodynamic and thermo-diffusive instabilities including the physical mechanisms behind them, the impact on laminar premixed flames and on lean turbulent flames, and the possibility of modelling such non-linear phenomena.

In this lecture, all the necessary numerical and modelling aspects for the simulation of hydrogen flames will be reviewed. This will include chemical models, transport models, as well as turbulence and flame / turbulence interaction models in the context of Large Eddy Simulation (LES). The various flame combustion regimes will be described, and some particular flame behaviours will be highlighted, like flame-wall interaction, and flame stabilization. The lecture will end with the numerical study of some technical challenges raised using hydrogen as a fuel in practical systems.

This part provides an overview and perspective on the computational requirements to simulate turbulent flames in the context of DNS and LES. It will introduce different levels of modelling like RANS, LES, and DNS, solution approaches low Mach or fully compressible, discretization strategies, and high-order methods, with an overview of different algorithms and strategies that can be used to accelerate the computations. The last part of the lecture will introduce some practical applications of LES for hydrogen combustion with a focus on the numerical and modelling aspects described in previous lectures.

This hands-on session is focused on the computation and characterization of hydrogen flames using the Cantera software. The effects of chemistry and transport will be discussed in the context of 1D flames with emphasis on the chemistry description and the flame structure.

02/12/2022 - Day 3: Understanding and predicting hydrogen flames

This hands-on session is focused on the study of numerical methods applied to the solution of reactive flow simulations. The students will have the opportunity to select different numerical methods and explore their influence on the accuracy and cost of the simulations in the context of hydrogen flames. Laminar flames and weakly turbulent flames will be used as examples.

The use of machine learning algorithms to predict the behaviors of complex systems is booming. The present talk reviews some open opportunities for applying data-driven, reduced-order modelling to tackle some of the challenges associated with combustion modelling: the large dimensionality of detailed chemical mechanisms and the complex turbulence-chemistry interactions. Examples of feature extraction in turbulent combustion data, empirical low dimensional manifold identification, classification, regression, and reduced-order modelling are provided.

The last hands-on session is focused on the study of the methodologies used for the analysis of massive databases. These methodologies are a common feature to be able to automatically extract non-linear dependency among the quantities of interest. The students will have the possibility to learn more about the structure of data-driven algorithms, explore the effects of tuning parameters on the models extracted and their accuracy. Laminar flames and weakly turbulent flames will be used as examples.

R​EGISTRATION IS NOW CLOSED

REGISTRATION FEES: