System design and analysis methods for optimal electric vehicle thermal management

Abstract

The thermal management system (TMS) in an electric vehicle (EV) encounters many challenges due to the stringent thermal requirements of EV components and concurrent range reduction in cold conditions. Efficient systems require thermal architectures with highly interconnected components to satisfy a wide range of operating conditions. This work addresses the lack of a design methodology for clean sheet EV TMS designs by introducing a decision tree to enable analysis-driven design space exploration using simulation and modeling tools. A versatile simulation framework is developed inside MATLAB-Simulink using Simscape to capture dynamic physical interactions of coolant and refrigerant thermal systems and is validated at both the component and system-levels. Direct and indirect configurations for cabin conditioning are analyzed to compare relative performance. The indirect configuration is found to have a 1.6–1.8x longer conditioning time and a coefficient of performance decrease of 18–31% and 31–41% for heating and cooling, respectively. A previously unexplored general integrated loop architecture is created for concept-level analysis of various EV TMS configurations. Operating modes are formulated for all possible driving conditions and are switched with a control strategy. A detailed analysis is done for an idealized system to study the system-level performance, and critical modes are identified for different driving conditions. The range increase of the heat pump relative to positive temperature coefficient heating varies from 4 to 33%, with an extra 1–4.4% possible by idealized waste heat recovery. Waste heat recovery also increases the heating capacity by 28%, making heat pump operation feasible at low temperatures. The applicability of the decision tree in the context of various EV TMS designs of leading manufacturers and existing literature is also discussed.