The prediction of damage caused by blast waves, generated by large scale explosions or industrial hazards (accidents in industrial systems or storage units) is an important technological and research problem that needs further understanding from disaster prevention point of view. The ERANE project has two main important goals: (1) to understand the complex physical phenomena induced by blast and shock waves that pass through complex media (buildings, industrial plants, topographic reliefs, slopes, river valley, etc.) and (2) to help designing new devices for protection against shock and blast loading in an urban agglomeration.

The recent use of slippery surfaces such as super hydrophobic and liquid-infused surfaces have demonstrated remarkable properties in reducing flow separation around bluff and slender bodies, modifying the separation of the boundary layer and the dynamics of the recirculation region. However, designing strategies to reduce the noise radiated by the separated region remains an open question. When considering the noise radiated by propulsion systems, the origin of sound generation is generated by the incipient separation whose precursor events lead to massively separated flow regions. The latter has a solid unsteady nature and is at the origin of the radiated noise in the near wake. The ambition of the LOTUS project is to understand how slippery surfaces can allow for attenuating the noise radiated using a multidisciplinary multi-scale approach coupling laboratory experiments with high-fidelity numerical simulations. In particular, we propose a novel strategy for the design and optimisation of superhydrophobic and liquid-infused surfaces over a blade profile with the aim of controlling incipient separation to attenuate the sources of noise. The tasks of the project will be jointly performed by PRISME laboratory and ENSTA Paris whose experimental facilities will allow for exploring a large range of physical parameters associated with (i) incoming turbulence, (ii) slippery-interfaces' deformation, and (iii) cavitation regimes. The LOTUS project aims at designing a methodology capable of engineering and optimizing turbulent flows in complex geometries, by means of textured surfaces which either repel water or use different fluids at the interface. The goal of this work aims at demonstrating the feasibility of a passive-control method, based on a homogenization technique for the coupling between the slippery surface and the outer fluid, to attenuate the noise generated by turbulence as well as cavitation for submarine propellers. The project combines high-fidelity numerical simulations for the characterisation of the turbulent flow and an adjoint method to the simulation for the design of the slippery surfaces. In addition, an experimental characterisation in a cavitating tunnel will allow for assessing the performances and the durability of the newly engineered surfaces with an application on a propeller blade. This project, therefore, tackles several open scientific questions regarding the analysis and the design of slippery surfaces towards the passive control of hydroacoustics. The radiated noise will be modelled using Ligthill's analogy and the slippery surfaces' design will be dedicated to the optimal attenuation of the mechanisms leading to noise generation. In addition, the best designs will be tested in a cavitating water channel assessing the validity and durability of both super-hydrophobic and liquid-infused surfaces. The project is built around four tasks: (i) High-fidelity numerical simulations of turbulent flows in the presence of a pressure gradient over a deformable wall, representing a submarine propeller blade. (ii) The design of a slippery wall based on an adjoint procedure with the aim of attenuating noise generation mechanisms. (iii) The analysis of the sound generated over different optimised deformable slippery surfaces. (iv) Testing the best designs in a cavitating tunnel and applying the same principles to an actual propeller-blade model assessing both the performances and the durability of the approach.

The project aims to develop resilient- and secure-by-design solutions for the real-time monitoring and control of networked control and cyber-physical systems in the framework of critical infrastructures monitoring. To realize this goal, the project aims to train several PhD students in the fields of cyber-security, control theory, and autonomous systems to shape future resilient- and secure-by-design cyber-physical systems (CPS). The PhD topics will investigate open research questions about the use of systems and control theory to build secure and resilient CPS. The project will also promote industrial excellence by offering opportunities to the researchers for testing their tools and frameworks in real-world scenarios, provided by the industrial partners. In this line, industrial partners will provide scenarios focusing on water, energy and transportation services, inspired by a real deployment using state-of-the-art and innovative IoT components. The research outcomes that will be validated in the context of this use cases will be reusable to other smart energy or smart cities scenarios.

This project proposes a significant contribution for the improvement of explosive and propulsive energy systems. Major advances will be proposed on the two following axes: - Optimization of materials and designed systems, - Prediction capacity of composition of energetic materials and system design under current and future environmental constraints. To do so, the three following items will be combined: - The scientific expertise of the project academic partners, expertise acquired during several decades in the modelling of reactive media based on solid fuels, - The design capability of safe propulsive and explosive energy systems with solid-phase or condensed materials, know-how mastered by the project's industrial partners, - The multidisciplinarity of the leading laboratory, particularly in energy systems, but also in nonlinear and hybrid dynamic systems in the context of spatial (porous materials, heterogeneity) and temporal (event and probabilistic) discontinuities. Optimization concerns the composition of energetic materials, as well as the induced industrial systems, at the level of ignition or combustion but also at small scale at the level of the organization of the heterogeneities. Indeed, the reactive medium leading to combustion or explosion must be perfectly defined from a physic-chemical point of view. For example, in the case of energetic materials such as pyrotechnic compositions, the grains stacking mode, in addition to the knowledge of their sizes and shapes, plays an important role at local level. In the case of low-vulnerability propellants, the energies and environmental conditions (pressure, nature of the pressurizing gases) must be well defined to obtain a safe and reproducible ignition despite the loss of sensitivity of the propellant. In addition to optimization, there is also a need to adapt to the new environmental conditions more and more present (e.g. REACH Regulation). This adaptability induces a continuous possibility of anticipation. This implies a better control of sensitivities of energetic materials and their relationships with the induced energy conversion system. To minimize the number of trials and errors, we propose, in this project, a multidisciplinary method to "capture" the relevant parameters, best translators of "local" to "global" and vice versa. It is based on a fine description, at temporal and spatial scales, of physic-chemical properties with the help of cellular automata model, inverse methods for parametric identification and images processing. It will then be possible to follow a flame front instantaneously, but also to define the pyrotechnic composition under anticipated constraints more and more severe. New materials are currently being developed for the replacement of pyrotechnic compositions in thermal batteries (consequence of the application of REACH Regulation). Combustion process of these substances may be affected, or even totally changed, in case of a partial or complete containment. We will therefore be interested in the evacuation of the propellant combustion products (gun propellant in interior ballistic applications) through a non-adiabatic tortuous channel with the search of a flow optimum in order to preserve the limit pressure necessary to maintain the quasi-stationary combustion. We will also optimize the transition of the combustion in a pyrotechnic environment modelled in 1D: transition from a first medium to a second one, through a discontinuity interface of the substance.

The aim of the SONATE project is to understand the aerodynamics, performance and losses of a wind farm at different scales and under inlet flows that correspond to realistic conditions of a real wind farm. Understanding the collective aerodynamics of wind turbines is crucial to achieving, improving and optimising the performance and lifetime of individual wind turbines operating in wind farms. The complex physics of farm aerodynamics is known to be multi-scale, due to the non-linear interaction between the unsteady flow generated by the rotating blades and the tower, the interaction of the wakes of individual turbines placed close together, the wake generated by the farm itself, and the large scales of atmospheric turbulent flow affecting the farm sites. The wake interaction in particular significantly increases the power losses of the farm, up to 30-40% of production compared to its nominal operation. However, for practical applications, most industrial codes use simple analytical models of the wind farm wake, mostly based on the mean velocity deficit. Realistic flow conditions upstream of the farm, such as the atmospheric boundary layer, blockage and wake interaction dynamics, are not always included in these codes. Therefore, what is missing in the literature, and what is crucial to achieve efficient industrial codes predicting wake farm performance, is an equivalent simplified model of the wind farm to generate the same blockages, losses and dynamic phenomena, based on realistic experiments under physical flow conditions corresponding to a full-scale wind farm. Therefore, data-driven reduced order models (POD-ROM) are known to be efficient to have surrogate reduced models with dynamical phenomena and to be simulated quickly. The project will focus on building representative reduced-order models from experiments, capable of predicting the performance of the wind farms with different inputs depending on the flow conditions (wind angle, turbulence, angular momentum added by the blades, etc.). A wind farm model will be designed and characterised with large-scale measurements in the environmental section of the PRISME laboratory wind tunnel. Porous discs will be used as surrogates for the wind turbine, which produces similar far-field wakes and for which the team is particularly renowned. The effect of the inflow parameters on the aerodynamics of the wind farm will be studied, focusing on the turbulence (using turbulence grids), the wind angle (by placing the farm in a wind rose) and the angular momentum generated by the blades rotation. Three physically relevant scales will be studied: (i) the turbine scale (WP1); (ii) a spanwise/longitudinal line of turbines scale (WP2); (iii) the whole farm (WP3). From all the scales studied in these WPs, reduced order models of the farm aerodynamics will be designed in the final WP4. Two different strategies will be tried here: (i) to model the mean blockage of the farm using a porous medium, (ii) to model the flow properties and dynamics using POD-ROM. The outcomes of the project will be proposed to improve the industrial models (such as FAST.Farm) for a better representation of the dynamic phenomena. The results will be available for collaborations in France and Europe, for comparison with high-fidelity numerical work and field measurements. The project will allow me to take on supervision and management responsibilities and to promote environmental aerodynamics research in the aerodynamics team of the PRISME laboratory.