Heavy goods vehicle (HGV) fires cause more serious fire safety problems than other vehicle fires in road tunnels due to the large fire size. The fire size is a critical parameter in road tunnel fire safety design and this parameter varies considerably under different environmental conditions. It is impractical to experimentally measure heat release rate (HRR) for HGV fires under different tunnel conditions because of the large experimental cost. There is a desire to use a cost-effective

more »

... ional fluid dynamics (CFD) modelling method to study tunnel fires, such as fire dynamics simulator (FDS). The pyrolysis model in FDS can predict HRR based on fuel properties and environmental conditions. Therefore, the FDS pyrolysis model is adopted in this research to simulate a large-scale tunnel simulated HGV cargo experiment, which was carried out on behalf of the Land Transport Authority (LTA), Singapore. There are three major objectives in this research: to understand fuel properties for the application of the pyrolysis model; to understand influence of forced ventilation on the HRR of tunnel fires; and to assess the predictive capability of the pyrolysis model in FDS to simulate tunnel fires. The material properties of the fuels (plastic and wood) adopted in the LTA experiment are investigated. A simple hand calculation method using multiple-component schemes is proposed in this research to analyse the kinetic properties for the LTA materials through a series of material-scale experiments. Favourable FDS predictions of decomposition behaviour are obtained based on the derived kinetic properties. Following the studies of the kinetic properties, a manual optimisation process is used to determine other thermal properties for the application of the FDS pyrolysis model. The results from FDS simulations for a series of cone calorimeter experiments reveal that the use of component schemes and thermal property settings are critical in accurately predicting burning behaviour in FDS. A series of small-scale tunnel experiments [...]