A systematic methodological approach is proposed for the selection and dimensioning of an appropriate on-board energy supply concept for any given specific propulsion configurations. Its development comprises a detailed model-description of energy consumers and suppliers in the system. From these, physically-motivated reduced order models are derived. An adequate super-structure like framework is used to interlink the single components. Uncertainties are incorporated by stochastic analysis and surrogate models, enabled by applying ideas from the arbitrary Polynomial Chaos expansion. This at hand, global sensitivities of the system can be analyzed and robustness and reliability measures can be computed. This comprehensive and fundamental rethinking of energy supply concepts and technologies shall lift the potential for a considerable increase in efficiency, specific energy and power, and sustainability. Not only for the energy supply system itself, but also for the single components, which is possible due to the strong interactions between the projects unit in ICA-C. Additionally, the energy supply system model is made available for overall aircraft design studies in ICA-B and on-ground supply concepts in ICA-A and helps to identify promising concepts and to properly select the energy system components and organize them in an appropriate structure.
The result from this work package provide the interlinking framework for the component models and the design optimization. The energy flows between the different components will be balanced by means of an energy management strategy. The system model must represent the electric, enthalpic and thermal energy flows of the propulsion architecture as well as of the required subsystems, as e.g. heat management and cooling systems. An indispensable prerequisite is to include all consumers, such as the wheel braking system, the producers, as e.g. the fuel cell or turbine and the storage components. Therefore, the framework must be set-up in a way, which is suitable for the development of a super-structure and capable to cover a flexible topological design.
Uncertainties are ubiquitous when exploring new design spaces for aeronautical energy supply systems. They may stem from simplified descriptions on various levels, missing data or more generally lack of knowledge when new or future system components are considered. A key challenge in uncertainty quantification, understood as a computational discipline, is the high complexity required to propagate probability distributions through models. This can be accounted for by replacing the original model with numerical surrogates that can be cheaply evaluated and hence, are well-suited in a multi-query context.
In a first step uncertainties arising in the energy system model are identified and modelled as random input parameters. The distributions and possible correlations of these parameters are defined in collaboration with the component expert groups. In addition to input parameter uncertainties model-form uncertainties have to be considered. These model-form uncertainties give rise to the fact that surrogate models are used in the superstructure system model.
With a probabilistic description of the input uncertainties at hand, dedicated methods are used to propagate these uncertainties through the energy system model. Surrogate modelling will be applied to obtain a tractable computational complexity for uncertainty propagation. In a post-processing step of the uncertainty propagation a sensitivity analysis is performed. Based on these results key-drivers of uncertainty are identified and reliable and robust development targets can be provided for component development in ICA C.
Prof. Dr.-Ing. Richard Hanke-Rauschenbach
Institut für Elektrische Energiesysteme
+49 511 762 14401