Numerical models to predict fire behaviour of flax-polypropylene composites with intumescent flame-retardant

• Thermal characterization to derive thermal parameters validated by cone calorimeter tests. • Thermal decomposition model of flax-PP-APP composite using first-order reactions. • Detailed multi-fuel combustion model integrating CO and soot emissions. • Coupled solid and gas-phase to simulate thermal response in flax-PP-APP composites. This study investigates the fire-retardant properties of flax fibre-reinforced polypropylene (PP) composites with ammonium polyphosphate (APP) through computational fluid dynamics (CFD) simulations and thermal analysis. The complex chemical interaction between PP and APP, which forms an intumescent char layer, presents significant challenges for accurate modelling. To address this, the effect of intumescent char formation is modelled in both the solid phase and gas phase within the numerical framework of Fire Dynamics Simulator (FDS), accounting for the thermal decomposition of the composite. Multi-fuel gas-phase reactions, incorporating CO and soot emissions, are included to enhance accuracy in simulating combustion behaviour. Two approaches are employed to formulate the boundary conditions in FDS, illustrating their influence on the composite’s burning behaviour. The heat capacities of flax, APP, and char are determined using temperature-modulated differential scanning calorimetry (TMDSC), while the char residue of the composite and its constituents is obtained from thermogravimetric (TG) analysis. The thermal decomposition of flax-PP-APP composites is modelled using three consecutive first-order reactions, with kinetic parameters derived from TGA profiles and refined through manual iteration in FDS to achieve a coefficient of determination (R 2 ) above 0.85. The CFD simulations, validated against experimental cone calorimeter data, demonstrate a deviation of less than 3% in peak heat release rates and provide accurate predictions of key fire behaviour parameters, including ignition time, flameout time, mass loss rate, and multi-fuel gas-phase emissions. These results confirm the model’s capability to simulate the combustion behaviour of natural fibre composites and to elucidate the mechanisms underlying their fire-retardant performance.