Methodological development of a digital twin for Li-ion cells simulation: An experimental and numerical assessment

Lithium-ion batteries are essential for the energy transition as they can be used both in the automotive field and as a stationary storage solution. They are the first choice in these sectors as they feature high energy densities, low self-discharge property, nearly zero-memory effect and long lifespan. The purpose of the thesis is to develop a procedure for the characterization of lithium-ion batteries using the GT-AutoLion tool included in the worldwide spread CAE simulation software GT SUITE. CAE simulation is a time- and cost-effective means that can support the design and testing of Li-ion cells. This methodology was defined and validated based on experimental data from coin cells assembled and tested in laboratory. In particular, charge and discharge tests at different C-RATEs and external temperatures were carried out, as well as cycle aging tests conducted exploring the impact of a different Depth of Discharge. Geometrical measurements taken on the real component and materials specifications were used to set-up a physical virtual model of the cell, calibrated and correlated against experiments. The predictivity and accuracy of this digital twin was thoroughly investigated and assessed: it was possible to model the discharge trends of the coin-cells by calibrating some fundamental parameters for cell balancing and mass transport and to estimate the degradation during constant current cycles, obtaining a very satisfactory accuracy and fitting with the experiments despite the use of non-commercial batteries (RRMSE<2.0% and R²>0.95). This demonstrate the potential of CAE simulation in supporting OEMs and cells manufactures in the development of their products. This work was carried out with the collaboration of the company POWERTECH Engineering S.r.l. (PWT).