Monitoring wildfire behavior has recently emerged as a key public policy issue due to the occurrence of extreme events, in particular in the Euro-Mediterranean area that is exposed to more frequent and more severe wildfires under climate change. Key to this monitoring is the development of an event-scale numerical simulation capability as a means to understand and predict the interactions between the atmosphere and the wildfire that drive its behavior. FIREFLY tackles this challenge by designing and evaluating a unique wildland fire reanalysis capability, which combines information coming from a coupled atmosphere/fire model and from airborne thermal infrared images in order to reconstruct as best as possible wildland fire progression. The key idea in FIREFLY is to design a novel data assimilation approach able to reduce fire front and plume uncertainties in the ensemble simulations by taking advantage of available observations. FIREFLY builds up on the Meso-NH/BLAZE coupled model, which will be extended to fire simulation in forest environment. A sensitivity analysis will be performed to spot the most influential fuel parameters to infer through the data assimilation process. A front shape similarity measure along with novel shape-based statistics will be a fundamental aspect to adequately measure the discrepancies between observed and simulated fire front positions. Test cases of increasing complexity, from experimental fires to actual wildfires, will be considered to assess the data assimilation performance. FIREFLY will bring a paradigm shift in the way we anticipate wildfire behavior, in particular fire-induced atmospheric processes possibly exacerbated under climate change.

This project is focused on the fundamental problem of combustion dynamics that has many practical implications. It specifically considers annular systems like those used in land based gas turbines and aeroengines. The corresponding combustors feature new architectures that reduce pollutant NOx emissions to comply with increasingly stringent regulations. However, these new designs and the essentially lean premixed mode they employ also promote a resonant coupling between combustion and acoustic modes. These combustion instabilities have many detrimental effects leading in extreme cases to mechanical failure. Despite the large number of investigations concerning instabilities arising with premixed flames yet few are the studies dealing with cases where the fuel is injected as a spray. Furthermore, the limited availability of engine data is a further obstacle for these studies in a real combustor. The present project aims to fill this gap of knowledge by proposing combined numerical and experimental investigations on spray flame dynamics and spray flame coupling with azimuthal modes in annular systems. The project first focuses on the analysis of the nonlinear response of a single spray flame to incoming acoustic perturbations via Large Eddy Simulations. The study is then extended to a multiple swirl spray injector injector system to investigate the impact on the flame response of the interaction between multiple adjacent flames. Numerical results are compared with experiments carried out in unique facilities available at EM2C Laboratory including the annular MICCA-Spray combustor. Theoretical and numerical procedures involving advanced coupling of the combustor acoustics with the obtained flame responses are developed to predict the limit cycles observed during experiments. Finally, the combustion dynamics observed experimentally in a Safran Aircraft Engine combustor will be investigated to verify the transferability of the developed techniques to a full-scale system.

The application of an external electric field has proven to be a very effective way of controlling combustion. Laboratory test cases have shown that this technology is capable of controlling many aspects of premixed and non-premixed combustion in small burners as well as in industrial-scale systems. However, the fundamental aspects of this interaction between a reacting flow and an electric field remain mostly unknown and this prevents the application of this technology in industrial applications. The complexity of the required models and the large computational cost associated with the simulation of the electric response of a flame to an electric field have so far limited the numerical analysis of this type fo flows to one-dimensional and two-dimensional configurations taking into account only laminar flames. Considering that most of the industrial burners are operated in the turbulent regime, it appears necessary to shed light on the behavior of electrified flames at high Reynolds number. In this context, the present project proposes to study the behavior of a turbulent reacting mixing layer impinged by an external electric field using a detailed description of the reacting and transport phenomena involved in this flow. The calculation of this complex configuration will be allowed by the use of the parallel computing paradigm proposed with the "Regent" programming language recently developed at Stanford University. The result of these calculations will allow me to extensively define the modification induced by the imposed electric field on the thermochemical fields produced in the turbulent reacting layer as well as the impact of the electric force on the statistical properties of the turbulence. Moreover, the database obtained with these calculations will be used as a benchmark solution for the development and validation of a reduced-order model necessary for the calculation of electrified reacting flows in industrial applications.

The enhancement of an engine operability involves increasing the compressor surge margin which is still an open challenge. The project FLORA (FLOw control in RAdial compressor) is twofold. First, it proposes to achieve a comprehensive understanding of the transient behavior of the radial compressor delivered by the Topic Leader through a precise characterization of the instabilities which develop at various rotation speeds and at different IGV (Inlet Guide Vanes) stagger angles. Detailed experimental investigations are planned providing an improved and time-resolved description of the path to surge. Then, it proposes to apply passive flow control strategies in order to push back the compressor surge line towards low mass flow which will consequently enhance the compressor stability, hence the engine operability. The proposal particularly aims at evaluating the benefits from the boundary layer aspiration in radial geometries in terms of performance (gain in pressure ratio and efficiency) and surge margin. Besides experiments, calculations will help for the understanding of the internal flow structures which develop from stable operating points up to surge. High-fidelity Large Eddy Simulations (LES) will be used to get an in-depth comprehension of the impact of the flow control on the internal flow. The project FLORA will contribute to the development of stable and efficient radial compressors with extended operating range.

The ubiquity of H2 as a solution to decarbonize our industries raises major safety concerns related to its deployment: storage, transportation and use. Leakage scenarios from tanks may jeopardise the development of this solution to the energy transition. In confined areas, a laminar flame front, resulting from flammable mixtures of H2/O2/N2 exposed to an ignition source, often goes through an acceleration phase, which may trigger an event often associated with an abrupt escalation of this hazardous scenario: deflagration-detonation transition (DDT). This rare and intricate process is induced by several factors: presence of obstacles, development of wall-boundary layers, shock waves, etc. The controlling parameters of DDT remain an open problem. The TRACKDEMO project is a combined experimental and numerical investigation of DDT aiming to lay the foundations for its modelling in large-scale configurations. Laboratory scale investigations have highlighted that the acceleration phase creates the conditions to trigger DDT: coupling between the focusing of shock waves ahead of the flame front and the presence of reactivity gradients. This process is considerably amplified by the presence of obstacles periodically placed along the path of the reactive front. Even if realising that DDT often takes the form of a shock-to-detonation transition (SDT), studies with weak shocks devoted to this canonical configuration remain scarce. The objectives of TRACKDEMO are to: (i) study the SDT for weak shocks propagating in an obstructed channel filled with a reactive premixture, (ii) study the mechanism of formation of explosion centres following the amplification of shock waves likely to occur close to the obstacles, and (iii) propose a proper modelling of these abrupt events in an LES context, a subject very poorly addressed in the literature. The impact of the obstacle spacing and their associated blockage ratio, key parameters for the propensity to transition, have to be addressed. TRACKDEMO is therefore a combined experimental and numerical investigation of SDT in H2/O2/N2 mixtures with the following outcomes: (1) A highly instrumented experimental database of SDT results made available to the community: The database will cover a large range of shock, obstacle arrangement and mixture parameters. (2) A fundamental understanding of the controlling parameters of SDT using the DNS approach: the details of the physics governing SDT, including the combustion regimes at play, will be unveiled. (3) A carefully designed LES approach, validated against the experimental database and DNS results produced in TRACKDEMO, and suitable for the numerical investigation of SDT/DDT in large scale and complex configurations. (4) An identification of the thermo-chemical mechanisms governing quasi-detonation propagation and the sources of instabilities and velocity losses. All these expected results will be of utmost help for the transportation industry willing to implement the hydrogen solution in a safe manner.