The urgent need to decarbonize the transportation sector—particularly aviation, shipping, and road transport—is driven by environmental challenges, fossil fuel depletion, and the EU's 2050 roadmap targeting an 80–95% emissions reduction. A central solution involves integrating Sustainable Aviation Fuels (SAFs), e-fuels, and blends derived from green energy sources like solar and wind. These synthetic, drop-in fuels can be used in existing combustion systems with minimal redesign. However, a robust understanding of their chemical composition, combustion behavior, and pollutant emissions (NOx, CO₂, soot) is essential.
CERFACS is strategically positioned to address these challenges, leveraging its E&S team's expertise in turbulent combustion, chemistry, multiphase flows, and thermodynamics. The objective is to deliver high-fidelity simulations, particularly Large Eddy Simulations (LES), to assess SAF behavior in realistic industrial contexts, focusing on engine operability, pollutant output, and climate impact (e.g., contrail formation).
Three primary challenges guide the research roadmap:
- Chemical complexity – SAFs consist of diverse hydrocarbon blends requiring advanced surrogate models and reduced chemical schemes, demanding high-performance computing (HPC) tools like ARCANE and CANTERA.
- Multiphysics coupling – Accurate simulation must integrate spray atomization, evaporation, combustion, and turbine flows using Eulerian-Lagrangian methods within AVBP.
- Computational scale – The large range of spatial and temporal scales, combined with stiff chemical kinetics, necessitates efficient solvers and explicit time integration to maintain industrial relevance.
The corresponding software includes ARCANE for chemistry reduction, CANTERA for canonical flames, and AVBP for LES in HPC environments.
Strategic axes supporting this work are the following:
- “Sustainable programming“: improving computational performance for complex, coupled, and multiphase phenomena.
- “Numerical algorithms“: adapting solvers to better handle stiff chemistry, thermal radiation, and small-scale physics and enhancing solvers for radiation and chemical kinetics, particularly in parallelized settings.
- “Data driven modelling“: applying machine learning for sub-grid scale modeling and chemistry tabulation, using DNS and LES data to improve simulation fidelity.
This applied axis aligns with the priorities of CERFACS shareholders—Airbus, Safran, ONERA, EDF, Total, and CNES—who are actively involved in the transition to clean propulsion and energy systems. The work aims to provide predictive, efficient, and scalable tools for evaluating SAF and other alternative fuels in next-generation transportation technologies.
