Statistical variation of the burning rate and extinction characteristics of engineered timber products

Abstract

Understanding the burning rate (mass loss rate, MLR) of timber products is essential in characterizing the ignition characteristics, fire size, and extinction phenomena experienced by timber. Key parameters for timber, such as self-extinction criteria, are presented throughout literature. These parameters are often determined using bench-scale experiments in relatively low trial quantities (e.g., n = 3). This study investigates the influence of trial quantity on the observed statistical variation in key burning rate metrics for timber products (e.g., peak MLR, MLR at extinction). Experiments were performed using a conical heater to conduct 100 repeat trials at incident heat fluxes of 20 kW/m², 40 kW/m², and 50 kW/m² on 12.7 mm (0.5 in.) thick ACX cross laminated plywood samples. Significant variability was observed in trials conducted at 40 kW/m² due to bimodal behavior where 39% of samples experienced self-extinction and the remaining 61% of samples sustained combustion until burnout (i.e., complete pyrolysis of the material). The MLR at extinction for the trials at 40 kW/m² displayed nearly double the magnitude compared to trials conducted at 20 kW/m² due to the breakdown of the semi-infinite solid condition. The results from this work illustrate the significance of large trial quantities when investigating complex phenomena.