Search

High temperature gases released through the safety vent of a lithium-ion cell during a thermal runaway event contain flammable components that, if ignited, can increase the risk of thermal runaway propagation to other cells in a multi-cell pack configuration. Computational fluid dynamics (CFD) simulations of flow through detailed geometric models of four vent-activated commercial 18650 lithium-ion cell caps were conducted using two turbulence modeling approaches: Reynolds-averaged Navier-Stokes (RANS) and scale-resolving simulations (SRS). The RANS method was compared with independent experiments of discharge coefficient through the cap across a range of pressure ratios and then used to investigate the ensemble-averaged flow field for the four caps. At high pressure ratios, choked flow occurs either at the current collector plate when flow through the current collector plate is more restrictive or the positive terminal vent holes when flow through the current collector plate is less restrictive. Turbulent mixing occurred within the vent cap assembly, in the jets emerging from the vent holes, and in recirculating zones directly above the vent cap assembly. The global maximum turbulent viscosity ratio (μT/μ) of the MTI, LG MJ1, K2, and LG M36 caps at pressure ratio of P1/P2 = 7 were 4,575, 3,360, 3,855, and 2,993, respectively. SRS and RANS simulations showed that both velocity magnitude and fluctuating velocity magnitude were lower for vent holes which are obstructed by the burst disk. SRS showed high levels of fluctuating velocity in the jets, up to 48.5% of the global maximum velocity. The present CFD models and the resulting insights provide the groundwork for future studies to investigate how jet structure and turbulence levels influence combustion and heat transfer in propagating thermal runaway scenarios.

Thermal runaway of lithium-ion batteries is particularly dangerous in systems where runaway can propagate through many cells. To understand thermal runaway propagation in these systems, it is important to understand the differences between failure characteristics of the initial cell that fails due to an abuse event and subsequent cells that fail due to thermal runaway propagation. This work compares thermal runaway events for single cells failed using a heater with cells failed due to propagation in an array. Most of the tests were conducted in an inert environment within a pressure vessel. Cell clamping stress, vessel gas pressure and temperature data show that runaway in cells within an array traverses the cells more slowly and releases gas at a slower rate than singly failed cells. For heater-initiated cell failure, post-experiment cell teardowns suggest that thermal runaway originates at a single point and creates a gas flow stream that causes heavy damage and exits from one side of the cell. Teardowns of cells which failed due to array propagation show more uniform damage and gas release from a larger area. [ABSTRACT FROM AUTHOR]