No abstract

This paper presents a 3D electro-thermal model approach for large-format lithium ion cells. The model development is focused on the prediction of 3D behavior of large-format lithium ion cells up to high discharge currents. The distribution and inhomogeneity of internal state values in the cell like temperature, current, power loss, SOC as well as voltage can be predicted reliably with this kind of models. The demand for calculating these state values on battery management systems can be met with a model order reduction introduced in this paper. Moreover, the interdependency of the temperature inhomogeneities and the inhomogeneities of inner state values in the cell can be studied. These dependencies have major influence on lifetime and performance of large-format lithium ion cells during common traction applications. The paper presents the configuration of a 3D equivalent circuit model considering full cell geometry and internal electrical resistances. In order to consider 3D thermal as well as 3D electrical effects, the direct coupling of the 3D electrical model with a 3D thermal FEM model is introduced. Furthermore, in this paper a convergence study on elementary cell resolution is performed. Additionally, the model is comprehensively validated with a thermal characterization over a wide range of temperature and current up to 300 A.

Thermal management of lithium ion batteries is a key-task in order to meet performance, lifetime and safety requirements for automotive application. Especially for large format lithium ion cells, which are commonly used in hybrid and full electric vehicles, a proper thermal management can extend significant lifetime and range. Therefore, thermal modeling is an integral part of battery development for traction applications in order to shorten development time and to reduce cost of development. In this contribution a 3D electro-thermal modeling approach is shown in order to predict three dimensional behavior of large format lithium ion cells up to high discharge currents. The model is able to predict reliable the distribution and inhomogeneities of inner cell parameters like temperature, power loss, state of charges, current and potential. The development of a 3D electrical model using an impedance-based cell model is presented showing its direct linking with a 3D thermal FEM model. Besides for this model also a convergence test in sub-cell resolution has to be done. Each sub-cell is treated as elementary cell linked with electrical resistances due to the current collectors. Moreover a comprehensive validation of this model is shown with data from a thermal characterization of the large lithium ion cells at different temperatures and up to 300 A discharge currents. These measurements where done with a thermal camera on cell level and with a fiber optical measurement system on battery pack level. The validation is done regarding absolute cell temperatures. In addition with this kind of models one is also able to predict temperature gradients during high current discharge in total agreement with measurement. One can also predict gradients over cell geometry of heat dissipation, state of charge, current and potential, which have crucial influence on performance and lifetime. For state of charge distribution an indirect validation is presented due to its impact on time development of temperature gradients during constant current discharges. The model is used in order to study the effect of geometrical cell design on internal cell parameters. This study is dedicated to inhomogeneities in distribution of temperature, heat dissipation, state of charge, current and voltage. Furthermore the study points out the effect of self-intensifying and damping of these gradients depending on the state of operation. It is also shown the prediction of cell gradients and local hot and cold spots during different operation state in electrical vehicles, addressing the self-intensifying of external thermal gradients due to the parallel circuit character of elementary cells inside a lithium ion cell. Especially the external gradients due to different cooling strategies are investigated. For an effective operation strategy it is essential to know during all states of operation the cold and hot spots in addition to the gradient over the cells. Due to the fact that the position of min and max temperature differs because of operation conditions and state of health it is not trivial to measure these values during operation. Nevertheless one should also know the internal cell parameters like current density to address lifetime issues like lithium-plating, which strongly depends on temperature and current density. Therefore there is also the need to run this kind of models in real time on battery management systems. Here also an approach of model order reduction is shown on system level simulation in order to reduce significant computing time.