Key points are not available for this paper at this time.
The performance of lithium-ion batteries is strongly affected by temperature. This study presents a coupled electro-thermal model of a lithium nickel manganese cobalt oxides pouch cell battery (Li-NMC), via a 2D and 3D approach for both electrical and thermal problems, respectively, validated through an experimental analysis. In particular, the model accounts for dispersive bars, which cause heat losses in both experiments and conventional use. Different fitting equations have been formulated for closure parameters, included a piecewise fitting for entropy coefficient. Convection was modelled using heat transfer correlations, providing a reproducible model with all parameters explicitly reported. Three configurations were analyzed: (i) excluding the connection bars, (ii) including them, and (iii) considering their thermal effects via a lumped capacitance method, by using an equivalent boundary condition on the battery tabs. An experimental validation was performed via thermal imaging. The model excluding bars overestimated the maximum temperature by 8.08 °C root mean square error (RMSE) for a 5C discharge, whereas the model including bars reduced this to 2.16 °C. Lumped capacitance model accurately reproduced average and maximum temperature trends, obtaining a maximum difference between the models of 0.18 °C for 5C discharge on the average temperature. The methodology was extended to a multi-cell battery module. Both maximum and average temperatures were very similar when replacing the busbar domain with equivalent boundary conditions, with deviations lower than 0.85 °C. This confirms the validity of the proposed approach in reducing computational effort while maintaining predictive accuracy at both cell and module level. • Coupled battery electro-thermal model validated through experimental analysis. • Impact of connection bars effects on battery thermal behavior is analyzed. • RMSE of 0.3/2.1 °C for average/maximum temperature at 5C discharge. • Tab boundary condition reduces computational domain and maintains accuracy. • Tab equivalent boundary condition validated also at a module level.
Bianco et al. (Sat,) studied this question.