| dc.description.abstract | The greatest challenge in the production of future generation electric and hybrid vehicle (EV and HEV) technology is the control and management of operating temperatures and heat generation. Vehicle performance, reliability and ultimately consumer market adoption are dependent on the successful design of the thermal management system. In addition, accurate battery thermal models capable of predicting the behavior of lithium-ion batteries under various operating conditions are necessary. Therefore, this work presents the thermal characterization of a prismatic lithium-ion battery cell and pack comprised of LiFePO4 electrode material. Thermal characterization is performed via experiments that enable the development of an empirical battery thermal model.
This work starts with the design and development of an apparatus to measure the surface temperature profiles, heat flux, and heat generation from a lithium-ion battery cell and pack at different discharge rates of 1C, 2C, 3C, and 4C and varying operating temperature/boundary conditions (BCs) of 5ºC, 15°C, 25°C, and 35°C for water cooling and ~22°C for air cooling. For this, a large sized prismatic LiFePO4 battery is cooled by two cold plates and nineteen thermocouples and three heat flux sensors are applied to the battery at distributed locations. The experimental results show that the temperature distribution is greatly affected by both the discharge rate and BCs. The developed experimental facility can be used for the measurement of heat generation from any prismatic battery, regardless of chemistry. In addition, thermal images are obtained at different discharge rates to enable visualization of the temperature distribution.
In the second part of the research, an empirical battery thermal model is developed at the above mentioned discharge rates and varying BCs based on the acquired data using a neural network approach. The simulated data from the developed model is validated with experimental data in terms of the discharge temperature, discharge voltage, heat flux profiles, and the rate of heat generation profile. It is noted that the lowest temperature is 7.11°C observed for 1C-5°C and the highest temperature is observed to be 41.11°C at the end of discharge for 4C-35°C for cell level testing. The proposed battery thermal model can be used for any kind of Lithium-ion battery. An example of this use is demonstrated by validating the thermal performance of a realistic drive cycle collected from an EV at different environment temperatures.
In the third part of the research, an electrochemical battery thermal model is developed for a large sized prismatic lithium-ion battery under different C-rates. This model is based on the principles of transport phenomena, electrochemistry, and thermodynamics presented by coupled nonlinear partial differential equations (PDEs) in x, r, and t. The developed model is validated with an experimental data and IR imaging obtained for this particular battery. It is seen that the surface temperature increases faster at a higher discharge rate and a higher temperature distribution is noted near electrodes.
In the fourth part of the research, temperature and velocity contours are studied using a computational approach for mini-channel cold plates used for a water cooled large sized prismatic lithium-ion battery at different C-rates and BCs. Computationally, a high-fidelity turbulence model is also developed using ANSYS Fluent for a mini-channel cold plate, and the simulated data are then validated with the experimental data for temperature profiles. The present results show that increased discharge rates and increased operating temperature results in increased temperature at the cold plates.
In the last part of this research, a battery degradation model of a lithium-ion battery, using real world drive cycles collected from an EV, is presented. For this, a data logger is installed in the EV and real world drive cycle data are collected. The vehicle is driven in the province of Ontario, Canada, and several drive cycles were recorded over a three-month period. A Thevenin battery model is developed in MATLAB along with an empirical degradation model. The model is validated in terms of voltage and state of charge (SOC) for all collected drive cycles. The presented model closely estimates the profiles observed in the experimental data. Data collected from the drive cycles show that a 4.60% capacity fade occurred over 3 months of driving. | en |
