An inverse methodology to estimate the thermal properties and heat generation of a Li-ion battery

Abstract

Accurate estimation of temperature-dependent orthotropic thermal properties and volumetric heat generation of a Li-ion battery is crucial for thermal modeling, thermal safety, and the design of thermal management systems for electric vehicles. Though various studies are available on estimating thermophysical properties and heat generation, simple and easily applicable methods are rare. Moreover, these studies failed to report temperature sensitivity and standard deviation of the estimates. In this study, the temperature-dependent orthotropic thermal conductivities (kr,kθ,kz), specific heat (cp), and volumetric heat generation (qv) of Panasonic NCR18650BD cylindrical battery are estimated using an inverse approach (Metropolis Hastings-Markov Chain Monte Carlo algorithm based Bayesian method) with the help of experimentally obtained surface temperatures measured at convenient locations on the battery. From the estimation, the average values of kr,kθ,kz and cv are observed to be 3.18 ± 0.19, 20.34 ± 1.26, 19.89 ± 1.29 (W/mK), and 3180 ± 202 (J/kgK), respectively. The average heat generation rates from the same battery obtained using the same methodology are 0.1 ± 0.005, 0.34 ± 0.012, and 1.51 ± 0.026 W for 0.5, 1, and 2C discharge rates, respectively. The estimated thermophysical properties and heat generation rates are in good agreement with the results obtained from both in-house experiments and literature. In addition to the estimation of thermophysical properties and heat generation, the proposed methodology opens vistas to predict the strength and location of hotspots in the battery domain, which helps in designing appropriate and effective thermal management systems for battery packs.