Periodic Reporting for period 1 - MYTHOS (Medium-range hybrid low-pollution flexi-fuel/hydrogen sustainable engine)
Okres sprawozdawczy: 2023-01-01 do 2024-06-30
MYTHOS proposes to develop a demonstrated innovative and disruptive design methodology for future short/medium range civil engines capable of using a wide range of liquid and gaseous fuels including SAFs and, ultimately, pure hydrogen, thus aiming at fulfilling the objective of decarbonize civil aviation as fore-seen by the ACARE SRIA short, mid and long-term Goals by 2050.
The MYTHOS consortium develops and adopts a multidisciplinary multi-fidelity modelling approach for the characterization of the relevant engine components deploying the full power of the method of machine learning. The latter will lead through hidden-physics discovery to advance data-driven reduced models which will be embedded in a holistic tool for the prediction of the environmen-tal footprint of the civil aviation of all speeds. Through this approach the projct will contribute to reduce time-to-market for engines designed and engineered to burn various types of environmentally friendly fuels, such as SAF, in the short and medium term, and hydrogen, in the long term. The project follows a strongly holistic approach, supported by a hierarchic and targeted usage of low- to high-fidelity simulation methods combined to experimental validation, to the design and assessment of future, flexi-fuel propulsion systems. The final outcome of this effort shall be a low-order, fast design and assessment tool, equipped with high-fidelity pollutant emission prediction capabilities.
The MYTHOS consortium develops and adopts a multidisciplinary multi-fidelity modelling approach for the characterization of the relevant engine components deploying the full power of the method of machine learning. The latter will lead through hidden-physics discovery to advance data-driven reduced models which will be embedded in a holistic tool for the prediction of the environmen-tal footprint of the civil aviation of all speeds. Through this approach the projct will contribute to reduce time-to-market for engines designed and engineered to burn various types of environmentally friendly fuels, such as SAF, in the short and medium term, and hydrogen, in the long term. The project follows a strongly holistic approach, supported by a hierarchic and targeted usage of low- to high-fidelity simulation methods combined to experimental validation, to the design and assessment of future, flexi-fuel propulsion systems. The final outcome of this effort shall be a low-order, fast design and assessment tool, equipped with high-fidelity pollutant emission prediction capabilities.
A first, minimal-fidelity model of the complete propulsion system and generic airframe has been generated. The necessary level of detail has been achieved for high-fidelity, reduced-order models to be embedded, accounting for the properties of the fuel and hence allowing an accurate and yet fast assessment of the engine environmental impact and performance at any arbitrary point along the mission’s envelope.
An extensive review of drop-in SAFs according to their properties and production processes was conducted, aiming at the selection of appropriate fuels for numerical and experimental testing. On one hand, candidates among synthetic kerosene were sought, showing a high degree of similarity with conventional jet fuel (category A fuels). On the other hand, compounds belonging to category “C”, representative of fuels being at the viscosity limit or fuels whose composition is outside of the typical composition of conventional jet fuels of category A, have been identified as posing the biggest challenges for the conception of a flexible injection and combustion system.
Great progress has been made in the development and assessment of an efficient chemistry-turbulence interaction model, suitable for the high-fidelity simulations representing one of the central activities to be addressed. In particular, the Multidimensional Chemistry Coordinate Mapping (MCCM) approach has been successfully applied to turbulent spray combustion and results validated against experiments in a realistic swirl-stabilised burner, thus exploiting valuable existing know-how and synergies with previous research carried out in the consortium. The configuration investigated in these preliminary studies will serve as basis for the analysis of flame stabilisation mechanisms and for the successive optimisation steps leading to the flexi-fuel design, which is the main objective of the project. The numerical validation framework for the chemical reaction mechanisms as well as for the chemistry-turbulence interaction modelling has been set and also verified against literature data and prior experiments, respectively, also setting the foundation for the design and analysis methods hierarchy at the heart of the project’s methodology. The 0D-engine design has been augmented by means of the first improved reduced-order combustor model based on the Chemical Reaction Network concept, in the first instance consisting of only three (perfectly stirred) reactors, representing the three main (primary, secondary and dilution) combustor areas. A first validation of the reaction mechanism in this context against the ICAO emission database has produced very encouraging results and allowed the identification of aspects which must be addressed in deeper details (e.g. some aspect of surrogate fuel definition and fuel break-down reaction paths as well as degree of mixing). High-fidelity Large Eddy Simulations of the model swirled injector (TARS) to be used as a starting point for the future design analysis have been successfully conducted (LUND) highlighting sensible differences in the flame placement and dynamics observable when burning standard jet fuel (JetA) and the C-class fuel discussed above.
A central part of the activities carried out during the first reporting period is the generation of a virtual test-rig reproducing the core components of a realistic commercial engine to be deployed on short/medium range flights. The first design of a virtual core engine for the verification of the low-order physical models was developedleading to a unique set of component geometries (high-pressure compressor, combustor and high-pressure turbine) which define the baseline for further optimisation steps.
An extensive review of drop-in SAFs according to their properties and production processes was conducted, aiming at the selection of appropriate fuels for numerical and experimental testing. On one hand, candidates among synthetic kerosene were sought, showing a high degree of similarity with conventional jet fuel (category A fuels). On the other hand, compounds belonging to category “C”, representative of fuels being at the viscosity limit or fuels whose composition is outside of the typical composition of conventional jet fuels of category A, have been identified as posing the biggest challenges for the conception of a flexible injection and combustion system.
Great progress has been made in the development and assessment of an efficient chemistry-turbulence interaction model, suitable for the high-fidelity simulations representing one of the central activities to be addressed. In particular, the Multidimensional Chemistry Coordinate Mapping (MCCM) approach has been successfully applied to turbulent spray combustion and results validated against experiments in a realistic swirl-stabilised burner, thus exploiting valuable existing know-how and synergies with previous research carried out in the consortium. The configuration investigated in these preliminary studies will serve as basis for the analysis of flame stabilisation mechanisms and for the successive optimisation steps leading to the flexi-fuel design, which is the main objective of the project. The numerical validation framework for the chemical reaction mechanisms as well as for the chemistry-turbulence interaction modelling has been set and also verified against literature data and prior experiments, respectively, also setting the foundation for the design and analysis methods hierarchy at the heart of the project’s methodology. The 0D-engine design has been augmented by means of the first improved reduced-order combustor model based on the Chemical Reaction Network concept, in the first instance consisting of only three (perfectly stirred) reactors, representing the three main (primary, secondary and dilution) combustor areas. A first validation of the reaction mechanism in this context against the ICAO emission database has produced very encouraging results and allowed the identification of aspects which must be addressed in deeper details (e.g. some aspect of surrogate fuel definition and fuel break-down reaction paths as well as degree of mixing). High-fidelity Large Eddy Simulations of the model swirled injector (TARS) to be used as a starting point for the future design analysis have been successfully conducted (LUND) highlighting sensible differences in the flame placement and dynamics observable when burning standard jet fuel (JetA) and the C-class fuel discussed above.
A central part of the activities carried out during the first reporting period is the generation of a virtual test-rig reproducing the core components of a realistic commercial engine to be deployed on short/medium range flights. The first design of a virtual core engine for the verification of the low-order physical models was developedleading to a unique set of component geometries (high-pressure compressor, combustor and high-pressure turbine) which define the baseline for further optimisation steps.
The progress made in the past 8 months on the generation of a virtual test rig geometry for the core engine components (HP compressor, combustor and HP turbine) starting from the preliminary aerorthermal deisgn is particularly noteworthy and beyond the standard modelling capabilities of standard approached. The configuration generated by partners RUB and DREAM, is based on an RQL-combustor architecture and will host the new burner design generated by beneficiary LUND. A novel, two-way coupling strategies for arbitrary flow solver has been devised by partner RUB, such that an all-Mach approach can be realised for seamless simulations of component interactions which is of paramount importance for the data-mining and model order reduction activities planned in the project. This approach has been attempted in the past by non-EU research initiatives (see Accelerated Strategic Computing Initiative [ASCI] project at Stanford University, USA), however not considering a full deployment of scale-resolving methods (LES, DES, DNS) and a bi-directional exchange of flow information for a truly seamless aerothermal coupling of engine components.