Inverse estimation of fire properties under constant external heat flux

초록

Fire properties of many organic matters under chemical reactions are very difficult to be measured directly. In this study, we inversely estimated convective heat transfer coefficient, surface emissivity, specific heat and flame heat flux from cone calorimeter experimental data by applying the repulsive particle swarm optimization (RPSO) algorithm. The standard cone calorimeter test procedure described in ISO 5660 is used to measure the surface temperature, mass loss rate and ignition time. The experiment is carried out using 21[mm] thick Douglas fir and the external heat flux from the cone heater into the specimen is set to 30, 40, 50, 60, 70 and 80[kW/m2] respectively. A computer code is developed to determine the critical heat flux and flame heat flux from the experimental data which are then used to estimate the specific heat, convective heat transfer coefficient and surface emissivity of the reacting specimen by using the RPSO. From this study we conclude that the fire properties of Douglas fir can be predicted successfully from results of the cone calorimeter test by applying the RPSO algorithm.