Exploring Batteries & Energy Mobility Systems
This study focuses on the comprehensive analysis of commercial lithium-ion batteries (LIBs) through both electrochemical characterization and internal material investigations. In particular, degraded cells subjected to charge–discharge cycling were examined using advanced analytical techniques such as scanning electron microscopy (SEM), X-ray diffraction (XRD), transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS). These methods enabled systematic identification of microstructural and compositional changes in electrodes and materials, thereby providing critical insights into the degradation mechanisms of LIBs. By correlating electrochemical performance deterioration with physicochemical property changes, reliable diagnostic indicators of cell degradation are established. The outcomes of this research are expected to serve as a valuable foundation for lifetime prediction and performance improvement strategies, ultimately contributing to the development of safe and high-reliability energy storage systems.
Lead by Dain Kim
This study analyzes the correlation between mechanical behavior and electrochemical performance of Li-ion batteries (Pouch-type, 900 mAh), focusing on the effect of the number of cells. Mechanical measurements were conducted on both single- and multi-cell systems to evaluate thickness variation and inter-cell pressure distribution during operation. The results revealed that in multi-cell configurations, nonlinear thickness growth occurred, which did not scale proportionally with the number of cells. To interpret this phenomenon, a theoretical prediction model considering SEI (Solid Electrolyte Interphase) layer growth and Li plating was developed to quantitatively predict irreversible thickness expansion. In addition, the nonlinear compressive characteristics according to the number of cells were incorporated to analyze deformation behavior under realistic operating conditions. This study elucidates how increased cell stacking density influences mechanical stress and performance degradation, providing valuable insights for the design and performance prediction of high-reliability battery systems.
Lead by Yeonghwan Jang
This study develops an AI-based framework for the accurate estimation of the state of health (SoH) in lithium-ion batteries (LIBs). A comprehensive database reflecting real-world operating conditions, including variations in temperature, charge/discharge profiles, and usage history, was constructed. The dataset was divided into training and validation sets, and machine learning algorithms were employed to capture nonlinear degradation patterns. The proposed model effectively accounts for current, voltage, and state-of-charge (SoC) conditions, and its predictive performance was evaluated using root mean square error (RMSE) metrics. Moreover, full-cycle SoH estimation was performed and validated against experimental ground truths, confirming the reliability of the model. The findings demonstrate that the proposed approach can contribute to the development of precise diagnostic tools for enhancing battery performance and safety, and can be further applied to the design of high-reliability battery management systems (BMS) and next-generation energy storage technologies.
Lead by Yeonghwan Jang
We develop accelerated degradation modeling frameworks for lithium-ion batteries that couple electrochemical and side reaction mechanisms. These frameworks capture SEI growth, transition-metal dissolution, solvent oxidation, and salt decomposition, among other mechanisms, enabling the prediction of long-term aging with reduced computational cost. Through advanced acceleration strategies, our approach maintains accuracy while significantly reducing simulation time, allowing efficient evaluation of cell- and system-level performance.
Lead by Yeonghwan Jang
Lithium(Li) metal batteries (LMBs) are considered promising candidates for next-generation energy storage systems owing to their exceptionally high theoretical capacity and low electrochemical potential. However, the uncontrolled growth of Li dendrites remains a critical challenge, leading to severe safety risks and shortened cycle life. In this study, a simulation-based model was developed to observe and investigate dendrite growth in lithium metal batteries. The model enables analysis of dendrite behavior under various operating conditions, such as current and voltage, and is employed to identify strategies for suppressing Li dendrite formation. The findings provide fundamental insights into enhancing the performance and safety of LMBs and can serve as a basis for establishing design and control strategies for next-generation high–energy-density batteries.
Lead by Yeonghwan Jang
Lithium metal anodes provide high energy density but inevitably suffer from the growth of inactive lithium layers during cycling. This inactive Li (dead Li) formation, often accompanied by dendrite growth, causes voltage polarization, capacity fade, and mechanical swelling in the cell. The generation of dead Li and dendrites is driven by non-uniform ion transport, local current density fluctuations, and electrolyte wetting inhomogeneity. To address this, we developed an electrochemical–mechanical coupled model that accounts for inactive Li layer porosity, thickness evolution, and SEI formation, linking them directly to stress and pressure development. Ultimately, by identifying and quantifying the factors that accelerate or slow down inactive Li growth, our research aims to improve the Coulombic efficiency and long-term stability of lithium metal batteries.
Lead by Dain Kim
The aim is to deeply elucidate and predict the phenomenon of thermal runaway (TR)—a critical safety issue in lithium-ion batteries—and its propagation mechanisms. A high-precision simulation model will be established to observe the entire process, from initiation to propagation of thermal runaway, within a virtual environment. This allows for the quantitative analysis of the physical and chemical characteristics at each stage. Ultimately, this study seeks to provide essential predictive technologies for battery system design and the establishment of safety standards, thereby contributing to a significant improvement in battery safety.
Lead by Dongwook Kim
This study primarily focuses on developing a generalized optimization framework aimed at enhancing the energy density of battery modules while satisfying mechanical and thermal safety requirements. In addition, a Reliability-Based Design Optimization (RBDO) framework is established to account for system uncertainties, thereby mitigating the likelihood of safety failures and ensuring a reliable module design. The mechanical behavior of the module is analyzed to illustrate how factors such as swelling, breathing, material type, and material properties influence the final module length and internal stress. Moreover, the effects of C-rate and design parameters on the maximum temperature and temperature distribution are investigated.
Paper: Optimization of module structure considering mechanical and thermal safety of pouch cell lithium-ion batteries using a reliability-based design optimization approach
This study develops an efficient multi-scale design optimization framework for electric vehicles (EVs) that integrates four hierarchical levels, ranging from the battery cell level to the marketing level. In the engineering domain, three physics-based models are constructed to describe the system’s performance: an electrochemical cell model, a thermal module model, and an EV dynamics model. In the marketing domain, a utility model is employed to maximize profit while considering customer preferences and cost constraints. Furthermore, the vehicle model, which represents the highest level within the engineering hierarchy, is interlinked with the marketing model, enabling integrated decision-making. As a result, all four model levels are hierarchically connected, allowing bidirectional information exchange across engineering and marketing domains for comprehensive optimization.
Paper: Multi-scale design optimization of electric vehicles by analytical target cascading: From battery cell level to marketing level
We develop multi-scale optimization frameworks that integrate cell-level physics with vehicle and operational-level performance. These frameworks incorporate degradation-coupled models at the cell scale, link them into vehicle-level constraints, and extend to fleet and charging strategies at the operation scale. By integrating design variables and engineering constraints across all levels, we enable holistic optimization of battery and mobility systems.
Paper: Shared autonomous electric vehicle system design and optimization under dynamic battery degradation considering varying load conditions
We develop control strategies and BMS algorithms for estimation, balancing, and fault diagnosis. These methods are demonstrated and validated at the module, pack, and system levels under real-world operating conditions. Our field studies extend across EVs, ESS, UAM, and advanced mobility platforms
Lead by Donghae Kim
This study aims to develop a high-reliability battery cycling emulator that replicates real-world driving scenarios. To this end, we are developing a module that converts the driving scenarios (real-time profiles) of a coupled simulator into battery current. In this module, the current required to meet the demanded driving load is calculated and then corrected based on the battery’s state (temperature and voltage). The driving load generated at the wheel passes through Model 1 and Model 2, where it is converted into battery current and subsequently applied. Finally, the converted current is applied to the battery module/pack installed in the actual battery cycler, enabling real-time monitoring of the battery state. As shown in the following figure, this system enables validation of a battery system model derived from real-world data by reproducing battery responses in the simulation environment.
Lead by Dongmin Kim
We study a BMS topology that embeds COMSOL’s physics-based battery model directly into a Simulink circuit, so that electrochemical and thermal state changes (e.g., lithium stoichiometry, side-reaction heat, SEI growth, parameter fade) feed the controller in real time. The co-simulation lets the BMS make decisions—balancing, derating, protection—using physico-chemical truth rather than lumped surrogates, enabling precise prediction of pack heat generation and cell degradation under realistic drive cycles and control actions.
Lead by Donghae Kim
