Meeting the demand for clean, renewable energy in the future will require the use of batteries to meet power demand at times of low supply. Having batteries with a large enough capacity and power output to meet this requirement is imperative, prompting the need for research into estimating the state of health of batteries. This thesis takes two approaches to state of health monitoring: experimental and model driven. The experimental approach consisted of taking displacement measurements of an LMN-8790140-1C pouch cell using 3D DIC technology. It was determined that there is a strong linear relationship between the displacement of a completely discharged battery and the battery’s state of health. It was also determined that there is a potential relationship between displacement, voltage, and state of health. More work needs to be done to verify this relationship. The points that best represented the average displacement were in the middle of the cell or closer to the long sides. The model driven approach consisted of creating an equivalent hydraulic model to simulate a silicon-graphite composite anode battery. An LG-MJ1 18650 cell was cycled to collect current and voltage data at several different state of health stages. The particle swarm algorithm in MATLAB was used to identify key parameters of the model. Using identified parameters, the model could accurately simulate voltage given a simple current input. The model struggled with simulating a UDDS cycle, but that could be due to poor parameter identification. A relationship was identified between the diffusive time constant of silicon and state of health. More work needs to be done to determine if other state of health indicating parameters, like estimated resistance or the diffusive time constant of the cathode or graphite, can be used in composite anode batteries.

KEYWORDS: State of Health Estimation, GOM Aramis, DIC, Equivalent Hydraulic Model, Composite Anode Batteries