Towards Next-Generation Electric Flight: Battery Pack Design and Power Electronics Interface

Abstract

Electric aircraft (EA) are an important pathway toward zero-emission transportation. Their deployment is constrained by technological and regulatory challenges, including standards and certification, battery development, powertrain hardware, and grid integration. Among these, the battery is the most critical, as it serves as the primary energy source in all-electric aircraft (AEA). Across flight profiles and propulsion configurations, AEA batteries require significant improvements in specific energy, specific power, and cycle life. This thesis defines the performance requirements of battery systems for AEA and develops methods to address them from two perspectives: battery pack design and power electronics interfaces connecting the pack to the powertrain.

This thesis begins with an overview of current EA development, including AEA, hybrid-electric aircraft (HEA), and electric vertical takeoff and landing (eVTOL) aircraft. Key battery performance metrics for aviation are examined, including specific energy, specific power, cycle life, safety, and pack design. Future trends, challenges, and suitable battery chemistries for different stages of EA deployment are also discussed.

To determine battery pack size, a methodology is proposed to map EA performance requirements to the corresponding energy and power demands across mission phases. These include takeoff, climb, cruise, and descent for all-electric conventional takeoff and landing (eCTOL) aircraft, and hover, climb, cruise, descent, and landing for eVTOL aircraft. Since existing cell chemistries typically involve a trade-off between specific energy and specific power, battery packs based on a single cell type may be oversized to satisfy either energy or peak power demands, increasing mass unnecessarily. To address this limitation, a hybrid battery pack is developed. High-specific-energy cells form the primary pack for energy-intensive cruise operation, while high-specific-power cells in a secondary pack support power-intensive phases such as takeoff and climb. The design is formulated as an optimization problem that minimizes pack mass under energy and power constraints.

Powertrain integration of the hybrid battery pack is then investigated. In conventional hybrid energy storage systems (HESS), a semi-active topology is commonly used, where a DC/DC converter regulates power flow from the secondary pack. This requires the converter to handle peak secondary-pack power, increasing system mass and reducing efficiency. To overcome these limitations, a partial power processing (PPP) architecture is adopted. In this structure, only part of the total power is processed by the converter, while the remainder follows a direct path to the bus. This reduces converter rating, losses, and thermal stress, improving overall efficiency.

To further improve efficiency, a partition variable-frequency triple-phase-shift (TPS) modulation strategy is developed for the dual-active-bridge (DAB) converter in the PPP architecture. This enables zero-voltage switching (ZVS) across all switches and reduces inductor current stress. In addition, the optimal secondary battery pack voltage is determined using a detailed converter loss model to minimize total integration mass, including converter mass and additional battery mass required to compensate for conversion losses.

Finally, battery thermal management is optimized at the pack level. Since pack-level energy density is lower than cell-level energy density due to packaging and the battery thermal management system (BTMS), improving pack design is essential. An electrical-thermal-aging (ETA) model is developed for a liquid-cooled battery pack, accounting for cell C-rate, state of charge, temperature, aging state, electrical topology, and BTMS design parameters. This model enables prediction and optimization of pack performance under EA operating conditions, resulting in improved energy density while maintaining favorable thermal and aging performance.