Comprehensive Numerical Framework for Simulating Cell Venting and Gas-Phase Reactions in 18650 Lithium-Ion Batteries During Thermal Runaway: Validation and Comparison of Integration Schemes

A comprehensive numerical framework is developed to investigate the complex behavior of 18650 lithium-ion cells during thermal runaway, focusing on gas-phase flow, internal pressure evolution, and venting dynamics. External turbulent gas flow during venting is modeled using a k-ε Reynolds-Averaged Navier–Stokes (RANS) approach, while internal porous flow within the cell is described by the Darcy–Forchheimer equation. Thermal abuse reactions and gas generation are represented by a single-step lumped reaction model. To efficiently handle the stiff kinetics characteristic of thermal runaway, five numerical integration schemes are implemented and compared: fourth-order Runge-Kutta (RK4), Backward Euler, Crank-Nicolson, and Modified Patankar-Runge-Kutta methods (MPRK2, MPRK3). The model is validated against comprehensive experimental data from 18650 NCA cells at 100% state of charge. Results demonstrate that Crank-Nicolson achieves the best accuracy with 5.7% gas mass error and 28.6% temperature error, while RK4 provides a reliable alternative with 9.7% and 26.1% errors respectively. MPRK methods offer computational efficiency but with elevated errors. Backward Euler exhibits catastrophic failure with errors exceeding 21,000%. The validated framework provides practical computational tools for battery safety design and establishes clear guidelines for numerical scheme selection in thermal runaway simulations.