Date/Time
Date(s) - 11/03/2024
11:30 am - 12:30 pm

Location: HH-207

Speaker: Avesta Ahmadi

Title: Characterization of Lithium Plating in Li-ion Cells via Optimal Reconstruction of Constitutive Relations

Abstract: Lithium-ion batteries stand as one of the most promising energy storage solutions, promising advancements in sustainable energy systems. Addressing challenges related to the performance and reliability of these cells is crucial for their wide deployment. Various factors contribute to inefficiencies within these cells, leading to degradation over time. This study focuses on understanding, predicting, and managing one of the primary degradation mechanisms known as Lithium Plating. Employing a data-driven approach in conjunction with a suitable mathematical model, we aim to identify the dynamics of lithium plating within the cell. Experimental measurements are acquired through Lithium Nuclear Magnetic Resonance spectroscopy, providing crucial insights into the system’s behaviour. Initially, we reduce the well-established Doyle-Fuller-Newman (DFN) physical model to a single-particle (SP) model using asymptotic reduction techniques. The SP model is then spatially averaged to track the temporal evolution of key concentrations in the system, eliminating the model’s spatial resolution. However, this simplified physical model still contains unknown parameters and constitutive relations, which are functions of the system’s state variables. In this study, our objective is to determine an optimal form for these constitutive relations, without making any a priori assumptions about their functional form. To achieve this, we employ a specially designed inverse model and utilize techniques such as adjoint analysis and gradient descent optimization. The resulting optimized model will serve as a powerful tool for prediction and control, advancing our understanding and management of lithium plating in lithium-ion cells.