A Methodology For Dynamic Sizing of Electric Power Generation and Distribution Architectures


Electric and hybrid electric aircraft pose a significant architecture challenge, as these concepts not only deal with considerably high electrical loads, but also are extremely weight-sensitive. Ideally, a design space exploration should be conducted for each aircraft design including the transient electric propulsion and generation subsystem models for different propulsion architectures. However, such transient and detailed models require significantly small time steps (generally in the order of microseconds) during simulations, compared to the time steps required for aircraft mission analysis (generally in the order of minutes). Hence, the inclusion of such models bring enormous computational burden to the early design phases, and therefore are usually neglected in the aircraft conceptual design stage. Combined with the lack of historical data, the uncertainty in the design and performance estimation of these subsystems can have a cascading impact on the vehicle design and mission performance, which results in non-optimal designs with weight and performance penalties. The over-arching objective of this thesis is to develop a methodology to perform the sizing, integration and performance evaluation of electric power generation and distribution subsystems (EPGDS) and architectures within electric and hybrid electric aircraft concepts. To this end, this dissertation presents the creation of a novel methodological framework, called Electric Propulsion Sizing and Synthesis (E-PASS), which integrates EPGDS considerations into the aircraft sizing and synthesis process to enable quantitative and adequate comparisons between different types of electric and hybrid electric propulsion architectures. E-PASS has three main capabilities to overcome the aforementioned limitations. First, the traditional sizing and synthesis approach is modified to incorporate a modular weight estimation technique along with an energy-based mission analysis approach which stems from the conservation laws. The new, generalized approach enables the design and performance evaluation of any vehicle configuration, including the electric and hybrid electric aircraft. Second, a power split schedule optimization algorithm is wrapped around the sizing and synthesis capability to ensure that the candidate architectures at their optimum performance. Third, the dynamic nature of the EPGDS is taken into account by developing bi-level, physics-based and parametric models in addition to the adaptive step sizing capability which enables performing transient analysis at the conceptual design stage without sacrificing valuable computational resources. As a result, the transient analysis are performed only when required so that the knowledge about the subsystem design is maximized while minimizing the computational burden. Consequently, E-PASS incorporates these elements and provides a capability to integrate subsystem performance and dynamics of novel architectures to the aircraft sizing process at early design phases, enabling adequate comparisons between competing architectures.

Georgia Institute of Technology