The ESTiMatE project, composed of seven public institutions from three countries and Rolls Royce as Topic Leader, aims to contribute to the decarbonisation objectives envisioned by the European Union for next decades by providing new insights in the soot formation for conditions representative of aero-engines by the synergic joint of both experiments and simulations. The integral approach followed in the project comprises the formulation of a new mechanism for a jet-A1 surrogate, the implementation of advanced soot models in CFD (Computational Fluid Dynamics) codes and the measurement of laminar and turbulent flames and is a unique chance to step forward in a research that has an increasing impact in terms of health and environment.
ESTiMatE is about to start its third year and, therefore, has entered in its maturity stage in which the most sounded results are expected to be generated. During this period the main efforts have been dedicated to the development of solid tools to face the incoming challenges. To this end, a detailed mechanism for jet-A1 surrogate has been developed by the Universität Stuttgart (IVLR), in which special attention has been drawn to polycyclic aromatic hydrocarbons (PAHs) whose reliable prediction is of paramount importance to reproduce the first stages of soot formation, while in parallel two different advanced soot models have been implemented by the Technische Universität Darmstadt (TUD) in OpenFOAM and by the Barcelona Supercomputing Center (BSC) and the Technische Universiteit Eindhoven (TUE) in the CFD code Alya from BSC. Moreover, a primary breakup model for air blast atomisers is being developed by the Universitat Politècnica de València (UPV) as the atomisation process plays a fundamental role in pollutant formation.
These models have been largely tested by the simulation of sooting laminar flames measured by the Karlsruher Institut für Technologie (KIT) to give validity to next steps. Such steps correspond to the Large Eddy Simulations (LES) scheduled for the incoming months of a turbulent counterflow diffusion flame and a swirling flame configuration currently being measured at the Technische Universität Berlin (TUB). Therefore, the aim of these simulations is to provide solid grounds for soot modelling by determining the influence of the combustion model and, more particularly, the Turbulence-Chemistry Interaction (TCI) on the prediction of particle size distributions. This assessment will be conducted as a gradual transition from ethylene to kerosene fuels using dedicated measurements for conditions representative of aero-engines.
The ability of the models to accurately reproduce soot formation from oxidation to particle size distributions has to come at an affordable computational cost. Consequently, these methodologies need to be tested in current supercomputers in order to ensure a high computational performance. To detect possible overheads in the models, the ESTiMatE consortium members have been granted with a Preparatory Access (2010PA5720) from PRACE (Partnership for Advanced Computing in Europe) to test the performance of the models before they are used in practical applications. Results from the strong scaling tests are gathered in Figure 1 where it is shown that speed up factors close to the ideal values are maintained up to several thousands of cores. Since the final simulations of the project will be run with meshes in the order of 100 million cells it is necessary that the codes show very high efficiencies at these high number of cores.
It is expected that the great effort invested up to this moment in ESTiMatE, where a consistent and seamless approach has served as a guide in the development of the project, will allow to successfully cope with the challenging simulations for sooting turbulent flames at aero-engine relevant conditions.
Acknowledgment:
We acknowledge PRACE for awarding us access to MareNostrum hosted at Barcelona Supercomputing Center (BSC), Spain, as part of PRACE Preparatory Access Type A from 12/4/2021 to 12/6/2021.