Modeling Volume Change of Large Format Pouch Cells Due to Intercalation

Abstract

Most current production electric vehicles (EVs) contain cells within a battery pack or module in order to maintain electrical conductivity, prevent fatigue due to vibrations, and provide efficient cooling. Modules may contain cooling fins, thermistors, foam separators, and repeating frame elements to hold the cells. As the cells expand and contract during cycling, the stresses generated can cause the materials in the battery module to deform and crack. This fatigue can result in a loss of electrical or thermal contact and therefore, decreased range, reduced battery life or loss of function, or a loss of heat transfer fluid which could result in electrical shorts. Therefore, it is critical to properly design modules and packs to properly account for the volume change inherent in these pouch cells. Currently, empirical mechanical data is used to predict the thickness changes of cells with varying electrode porosities, chemistries, and thicknesses. The ability to relate electrode expansion to electrochemical operation and accurately predict cell volume change would save significant time in experimental efforts as well as material costs. Previously, our group developed a coupled electrochemical and mechanical model that accurately predicts the split between electrode porosity and dimensional changes [1-3]. Here, we employ our model along with lithiation-based particle expansion data for graphite [4], lithium nickel-manganese-cobalt oxide (NMC) [5], and lithium manganese oxide (LMO) [5], to predict the thickness change of a large-format pouch cell used in automotive battery packs. Then, we measured the thickness of the cell as the cell was discharged from .99 to .01 state-of-charge. Figure 1 shows the predicted anode and cathode strain compared to the measured cell displacement divided by the sum of the anode and cathode thicknesses. This method allows battery pack designers to appropriately account for new cell types using electrochemical performance and individual particle-level expansion parameters as inputs. Design considerations, as well as cell level volume changes will be discussed.