PREDICT: PREDIctive numerical simulation of novel Combustion Technologies through validation and uncertainty quantification

flameless MILD combustion represents a very attractive solution for combustion systems as it can provide high combustion efficiency with low pollutant emissions. The increasing interest in MILD combustion is also motivated by its large fuel flexibility, representing a valuable technology for low-calorific value fuels and high-calorific industrial wastes. MILD combustion still appears worthy of investigations and attention. In particular, the fundamental mechanism of the interaction between turbulent mixing and chemical kinetics needs to be elucidated, due to the very strong coupling between chemical and mixing time-scales. Considering the complex nature of these non-conventional regimes, the use of Computational Fluid Dynamics (CFD) is acknowledged to be essential for the development and implementation of novel combustion technologies. Indeed, CFD calculations may be applied directly at the industrial scale of interest, thus avoiding scaling-up the results from lab- scale experiments. This appears very significant for combustion processes, for which scale-up procedures are generally complicated by the strong interaction between turbulence, reaction kinetics, heat release and radiation. However, users and developers of computational tools face today the critical issue of assessing the confidence in modelling and simulations, to effectively help decision making in new design and regulation. This appears particularly relevant in the context of MILD combustion. Indeed, the available sub-models (e.g. kinetic schemes, turbulence/chemistry interaction models) have been derived for conventional combustion systems and need to be validated and revised for novel applications, to be used with “confidence”.

The present project aims at developing predictive simulation tools for non-conventional combustion regimes including MILD combustion, based on new-generation models able to reduce the dependence on sub- models and increase the fidelity of numerical simulations. For this purpose, high fidelity experimental data will be collected under different operating conditions and with various fuel blends. An experimental apparatus will be designed and built to allow investigation of the global performances of the combustion system as well as the visualization of the main features of the combustion process. Such information will constitute the ideal basis for the development and validation of new-generation modelling approaches based on the concept of Empirical Manifolds.