In the current investigation, a fully coupled electrochemical and thermal model for lithium-ion batteries is developed to investigate the effects of different thermal management strategies on battery performance. This work represents the first ever study of these coupled electrochemical-thermal phenomena in batteries from the electrochemical heat generation all the way to the dynamic heat removal in actual hybrid electric vehicles (HEV) drive cycles. In addition, a novel, passive internal cooling system that uses heat removal through liquid-vapor phase change is developed. The proposed cooling system passively removes heat almost isothermally with negligible thermal resistances between the heat source and cooling fluid, thereby allowing battery performance to improve unimpeded by thermal limitations.

For the battery model, local electrochemical reaction rates are predicted using temperature-dependent data on a commercially available battery designed for high rates (C/LiFePO₄) in a computationally efficient manner. Data were collected on this small battery (∼1 Ah) over a wide range of temperatures (10°C to 60°C), depths of discharge (0.15 Ah < DOD < 0.95 Ah), and rates (−5 A to 5 A) using two separate test facilities to maintain sufficient temperature fidelity and to discern the relative influence of reversible and irreversible heating. The results show that total volumetric heat generation is a primarily a function of current and DOD, and secondarily a function of temperature. The results also show that reversible heating is significant compared to irreversible heating, with a minimum of 7.5% of the total heat generation attributable to reversible heating at 5 A and 15°C. Additional tests show that these constant current data can be used to simulate the response of the battery to dynamic loading, which serves as the basis for the electrochemical-thermal model development. This model is then used to compare the effects of external and internal cooling on battery performance.

The proposed internal cooling system utilizes microchannels inserted into the interior of the cell that contain a liquid-vapor phase change fluid for heat removal at the source of heat generation. Although there have been prior investigations of phase change at the microscales, fluid flow for pure refrigerants at low mass fluxes (G < 120 kg m⁻² s⁻¹) experienced in the passive internal cooling system is not well understood. Therefore, passive, thermally driven refrigerant (R134a) flow in a representative test section geometry (3.175 mm × 160 mm) is investigated using a surrogate heat source. Heat inputs were varied over a wide range of values representative of battery operating conditions (120 < Q˙^m < 6500 W L⁻¹ ). The measured mass flow rate and test section outlet quality from these experiments are utilized to accurately calculate the two-phase frictional pressure drop in the test section, which is the dominant flow loss in the passive system in most cases.

The two-phase frictional pressure drop model is used to predict the performance of a simplified passive internal cooling system. This thermal-hydraulic performance model is coupled to the electrochemical-thermal model for performance assessment of two-scaled up HEV battery packs (9.6 kWh based on 8 Ah and 20 Ah cells) subjected to an aggressive highway dynamic simulation. This assessment is used to compare the impact of air, liquid, and edge external cooling on battery performance. The results show that edge cooling causes large thermal gradients inside the cells, leading to non-uniform cycling. Air cooling also causes unacceptable temperature rise, while liquid cooling is sufficient only for the pack based on the thinner 8 Ah cell. In contrast, internally cooled cells reduce peak temperature without imposing significant thermal gradients. As a result, packs with internal cooling can be cycled more aggressively, leading to higher charge and discharge energy extraction densities in spite of the volume increase due to 160 µm channels inserted into the 284.5 µm unit cell. Furthermore, the saturation temperature of the phase change fluid can be optimized to balance capacity fade and energy extraction at elevated temperatures. At a saturation temperature of 34°C, the energy extraction density was 80.2% and 66.7% greater than for the best externally cooled system (liquid) even when the pack volume increased due to incorporation of the channels. (Abstract shortened by UMI.)