Inverse modeling of thermal decomposition of flame-retardant flexible PVC with model-free coupled with model-fitting method

Highlights: • Thermal decomposition modeling on flame-retardant flexible PVC. • The model-free method is coupled with model-fitting method. • The Ant lion Optimization algorithm is applied for inverse modeling process. • The performance of different optimization algorithm is compared. The thermal decomposition model of flame-retardant flexible PVC is essential for relevant waste management and predicting its fire behavior. In this work, thermal decomposition kinetics of this material was investigated via thermogravimetric analysis in non-isothermal conditions. The kinetic analysis was performed using model-free (including Kissinger, Friedman, and advanced Vyazovkin methods) coupled with model-fitting method (Ant Lion Optimization algorithm, ALO). The kinetic parameters obtained from model-free method provide guides for initial guess and search range for ALO. Two different decomposition models are developed and the validations show that second two-step parallel model could better predict experimental data and explain the decomposition process. The performance of ALO is compared with Shuffled Complex (SCE) Evolution algorithm. The results show that the ALO shows better performance because all solutions from SCE are trapped in local minima. In this way, the ALO could be used as an alternate optimization tool for inverse modeling.