His research develops advanced combustion technologies for the flexible, efficient and clean conversion of energy: how turbulence and chemistry interact in reacting flows, reduced-order models that make simulations faster, alternative fuels such as hydrogen and ammonia, and the uncertainty in the results.
Moderate or Intense Low-oxygen Dilution (MILD) combustion is a highly efficient, low-emission technology. His group combines data analysis, numerical simulation and new modelling approaches to show that MILD systems can run on alternative fuels, such as hydrogen and ammonia blends, in industrial furnaces and gas turbines.
Using principal component analysis, empirical low-dimensional manifolds identify the variables that control a reacting flow simulation and lead to accurate reduced-order models, applied from zero-dimensional reactors to three-dimensional simulations. This work was the basis of the ERC Starting Grant VADEMECOM and the ERC Proof of Concept grant INVENT.
His group builds digital twins and cyber-physical infrastructures for combustion systems, using physics-based reduced-order models trained on CFD simulations and experimental data. This led to the first simulation-based digital twin of a furnace operating in flameless combustion.
A comprehensive method for simulating atmospheric boundary layer flows over complex terrain and around obstacles, now used by other research groups for pollutant dispersion and wind farm optimisation.
