Spray cooling is used in an increasing numbers of industrial applications. In the steel industry, during heat treatment of alloys, very high quenching rates are required to obtain strong and hard bonding in the metallurgical microstructure. The main purpose of this project is to deliver new models of spray cooling of solid surfaces at very high temperatures, typically around the Leidenfrost temperature and beyond. Compared to other cooling methods like jet impingement and pool boiling, spray cooling can potentially achieve very high heat dissipation rates while enforcing uniform and controlled cooling and reducing water consumption. However, despite its advantages, integration of spray cooling is still limited because of incomplete understanding of the complex fluid flow and heat transfer characteristics of sprays interacting with the hot surface. In particular, at high surface temperatures, an insulating vapour film quickly develops at the contact between a drop and the solid surface. The presence of this vapour layer modifies the dynamic of the impinging drops and reduces heat transfer considerably. When considering a practical spray, cooling is the most important beneath the centre of the spray, where droplet concentration is generally the highest, which results in a rewetting of the solid surface. A liquid film is then created and supplied with the incoming droplets; the heat flux extracted from the wall is also considerably enhanced. The present project has the ambition to combine advanced experiments, direct numerical simulation, Euler-Lagrange simulation and a large scale industrial experiment. To infer new and efficient models of spray cooling, experiments will be performed at different scales with a graded complexity. In this purpose, this project will rely on two distinct experiments. The first one will be devoted to the study of the impingement of a monodisperse and well-controlled droplet chain or of an array of parallel droplets chains in order to feature the space and time interaction phenomena occurring in sprays. These impingements will occur on a plate heated by induction with a constant flux. Significant efforts will be made to achieve experiments with a high-performance instrumentation. Non-intrusive innovative diagnostics will be specially developed and adapted in this project. Experiments dedicated to droplets chains will benefit from one of the largest range of experimental techniques never implemented yet in such a configuration. For instance, temperature droplets (before and after impingement on the wall) will be measured by two colors planar laser induced fluorescence, their sizes by forward scattering interferometry and their velocities by laser Doppler velocimetry. The heat flux removed from the hot solid surface will be inferred from infrared thermography combined with inverse conduction. In close interactions with the experimental accomplishments on droplet chains, direct numerical simulations (DNS) with interface tracking using the level-set technique will be achieved. This task implies taking into account phase change, Marangoni effect and radiative heating. DNS will help to assess the influence of many physical parameters especially when the experiment is not possible. The impingement of a single droplet chain or of an array of droplets chains on a heated wall will be studied experimentally and also numerically, using DNS, which will highlight the effect of successive impingements or of the spatial proximity of the impingements. Data coming from both experiments and DNS will be used to develop sub-models of droplet-heated wall interactions, taking into account heat and mass transfers. Euler-Lagrange numerical simulations will also be implemented: the sub-models developed for the bouncing and splashing regimes with inclusion of thermal transfer at the wall will be integrated in the code. A liquid film model will be also developed and integrated in the Euler-Lagrange code, which will be combined to a heat transfer solver to account for the heat conduction within the solid wall. The second experiment will focus on polydisperse sprays cooling with well controlled laboratory conditions. Here investigations will rely on a fine characterization of the statistical distribution of physical parameters like droplets diameter, velocity and number concentration through Phase Doppler Anemometry (PDA). Relevant PDA data will be supplied to initiate the numerical simulations of the spray/solid surface interaction that will be compared to the heat flux extracted from the wall, measured by IR thermography associated with inverse conduction. In a last step, an industrial cooling pilot unit of steel products at unit scale will be used in order to provide the extracted fluxes for previously characterized spraying conditions. The data coming from these experiments will be compared to the results of the Euler-Lagrange simulation code.